Post on 25-Mar-2020
1
Nucleotide binding domain and leucine-rich repeat pyrin domain-
containing protein 12: characterization of its binding to
hematopoietic cell kinase
Yue Zhang* and Curtis T. Okamoto 1
2
Department of Pharmacology and Pharmaceutical Sciences, School of Pharmacy, University of 3
Southern California, USA 90089-9121 4
5
*Correspondence: 6
Yue Zhang 7
zhan576@usc.edu 8
9
Running title: NlrpNLRP12 binds to HckHCK 10
11
Key words: yeast two-hybrid, innate immunity, protein‐protein interaction, NlrpNLRP12, 12
HckHCK, acute myeloid leukemia13
14
15
Abstract 16
Protein-protein interactions are key to defining the function of nucleotide binding domain (NBD) 17
and leucine-rich repeat (LRR), pyrin domain (PYD)-containing family of proteins 12 18
(NlrpNLRP12). cDNA encoding the human PYD + NBD of NlrpNLRP12 was used as bait in a 19
NLRPlrp12 binds to HCKck
2
yeast two-hybrid screen with a human leukocyte cDNA library as prey. Hematopoiesis cell 20
kinase (HckHCK), a member of the Sc-SRCrc family of non-receptor tyrosine kinases, was 21
among the top hits. The C-terminal 40 amino acids of HckHCK specifically bound to 22
NlrpNLRP12’s PYD + NBD, but not to that of NlrpNLRP3 and NlrpNLRP8. Amino acids F503, 23
I506, Q507, L510, and D511 of HckHCK are critical for the binding of HckHCK’s C-terminal 24
40 amino acids to NlrpNLRP12’s PYD + NBD. Additionally, the C-terminal 30 amino acids of 25
HckHCK are sufficient to bind to NlrpNLRP12’s PYD + NBD, but not to its PYD alone nor to 26
its NBD alone. In cell lines that express HckHCK endogenously, it was co-immunoprecipitated 27
with stably expressed exogenous NlrpNLRP12. Also, NlrpNLRP12 co-immunoprecipitated and 28
co-localized with HckHCK when both were overexpressed in 293T cells. In addition, in this 29
overexpression system, steady-state NlrpNLRP12 protein expression levels significantly 30
decreased when HckHCK was co-expressed. Bioinformatic analysis showed that HckHCK 31
mRNA co-occurred with NlrpNLRP12 mRNA, but not with other NlrpNLRP proteinsmRNAs, in 32
blood and marrow samples from acute myeloid leukemia (AML) patients. The mRNA of 33
NlrpNLRP12 is also co-expressed with HckHCK in AML patient samples, and the levels of 34
mRNA expression of each are correlated. Together these data suggest that NlrpNLRP12, 35
through its binding to HckHCK, may have an effect on the progression of AML. 36
37
Introduction 38
Pattern recognition receptors (PRR) are the first major line of cellular defense against damage-39
associated molecular patterns (DAMP) and pathogen-associated molecular patterns (PAMP) (1). 40
Among PRR, one class of proteins is the nucleotide-binding domain (NBD) and leucine-rich 41
repeat (LRR) receptor family (NLR). NBD and LRR PYD protein 12 (NlrpNLRP12) is one 42
NLRPlrp12 binds to HCKck
3
protein that belongs to the PYD subfamily of the NLR class and is a multi-functional protein (2). 43
NlrpNLRP12’s functions likely all depend on defined protein-protein interactions. For example, 44
in anti-inflammatory functions of NlrpNLRP12, NlrpNLRP12 interacts with tumor necrosis 45
factor receptor associated factor 3 (TRAF3) and nuclear factor kappa-light-chain-enhancer of 46
activated B cells (NF-κB)-inducing kinase (NIK) in the non-canonical NF-κB pathway, when 47
toll-like receptors (TLR) are stimulated with Pam3Cys4 triacylated lipopeptides followed by 48
addition of cluster of differentiation 40 (CD40) (3). Dr. Kanneganti’s group also showed that 49
NLRP12 expression downregulated the Nlrpcanonical NF-kB pathway during the Salmonella 50
infection (4,5). NLRP12’s inhibition of the non-canonical NF-κB pathway, thereby effecting an 51
anti-inflammatory function, has been reported in many publications (7-11). For example, in 52
colitis-driven colon cancer, NlrpNlrp12 knock-out (KO) mice experience a higher frequency of 53
death, weigh less, have a shorter colon length, and have more severe disease progression than 54
wild type (WT) mice, all of which may occur through NlrpNLRP12’s inhibition of the non-55
canonical NF-κB pathway (3,4). In addition, NlrpNlrpLRP12 KO mice have been reported to 56
exhibit higher innate immunity when it is infected by the virus because NlrpNLRP12 inhibits the 57
non-canonical NF-κB pathway; thus, in a KO mice, a more significant increase of NF- κB 58
activity is observed compared to WT mice when it is infected by the virus (6). And in the case 59
of NlrpNLRP12 protecting the diversity of intestinal microbiota (7), thereby preventing obesity 60
(8), it has also been reported that NlrpNlrpLRP12 KO mice display significantly decreased 61
intestinal microbial diversity due to increased innate immunity, resulting from the lack of 62
inhibition of the non-canonical NF-κB pathway. 63
64
NLRPlrp12 binds to HCKck
4
In addition, NlrpNLRP12 also interacts with interleukin (IL)-1 receptor-associated kinase 1 65
(IRAK1) in the canonical NF-κB pathway when TLRs are stimulated, also to inhibit the 66
inflammatory response (9). In contrast, for human studies of NlrpNLRP12-associated 67
autoinflammatory disorders, no protein-protein interactions have been reported (10,11). 68
Functionally, however, NlrpNlrpLRP12 KO mice had a higher immunological response in the 69
autoimmune model of contact hypersensitivity (12); thus, these mice are more prone to develop 70
contact hypersensitivity. This results most likely from NlrpNlrpLRP12 KO mice having 71
defective migration of their dendritic cells (DC). In addition, Luken et al. demonstrated that 72
NLRP12 regulatesimpacts the eExperimental aautoimmune eencephalomyelitis (EAE) model via 73
regulating T- cell responses (13). But, in these mouse studies(Lukens et al., 2015)y, no protein-74
protein interactions had been reported either (12). And NlrpNLRP12 is protective against 75
Yersinia pestis infection through NlrpNLRP12’s activation of the inflammasome which results in 76
secretion of IL-18 and IL-1β. But again, no protein-protein interactions of NlrpNLRP12 within 77
inflammasomes had been characterized (14). Also, there is also a study showing that NLRP12 78
also regulates the c-Jun N-terminal kinase (JNK) pathway in hepatocytes (15);. bBut, again, no 79
protein-protein interactions related to howwithin the NLRP12 activates- JNK pathway hads not 80
been showeddemonstrated. Thus, there is a further need to find and to characterize previously 81
unreported NlrpNLRP12-binding proteins. Knowing NlrpNLRP12’s protein binding partners 82
may help to gain mechanistic insight into pathways previously characterized as being regulated 83
by NlrpNLRP12. Moreover, identifying and validating potentially novel binding partners of 84
NlrpNLRP12 may suggest novel functions for NlrpNLRP12. 85
86
NLRPlrp12 binds to HCKck
5
In humans, NlrpNLRP12 is mainly expressed in macrophages, DCs, and neutrophils. In mice, 87
NlrpNLRP12 is primarily expressed in DCs and neutrophils (6,12,16). The NlrpNLRP12 protein 88
has three domains, from N- to C-terminus: PYD, NBD, and LRR domains (2,17,18). It is 89
generally accepted that the LRR domain, like NlrpNLRP1’s LRR domain and NlrpNLRP3’s 90
LRR domain, is responsible for sensing ligands or other types of activators of NlrpNLRPs in 91
modulating their role in an immunologic response (19-22). However, interestingly, it has been 92
reported that without the LRR domain, NlrpNLRP3 still can be activated by triggers of the 93
canonical NF-κB pathway (23). The NBD domain is responsible for self-oligomerization when 94
NlrpNLRP is activated, and the PYD domain is responsible for regulating downstream signaling 95
pathways (and thus, it is called the “interaction domain”) (16,23). Here we used a truncated 96
form of NlrpNLRP12 (PYD + NBD) to identify possible protein-protein interactions that may 97
regulate signaling pathways in which NlrpNLRP12 is involved. In addition, to our knowledge, 98
there are currently no protein-protein interaction studies that have been performed for this 99
truncated form of NlrpNLRP12 and may thus represent a new general approach to screening for 100
protein-protein interactions involved in signaling complexes with NlrpNLRPs. 101
102
A yeast two-hybrid screen was performed to identify potential binding partners of 103
NlrpNLRP12’s PYD + NBD. HckHCK, a member of the Sc-SRCrc family of non-receptor 104
tyrosine kinases, was one of the top positive clones (“hits”) among all of the proteins that 105
appeared to interact with NlrpNLRP12. Further experiments were done to confirm the novel 106
binding of NlrpNLRP12 to HckHCK and to characterize key structural features of the binding 107
interaction. In addition, co-expression of HckHCK with NlrpNLRP12 resulted in a significant 108
decrease of steady-state NlrpNLRP12 protein expression levels. In consideration tFinally, 109
NLRPlrp12 binds to HCKck
6
together with a bioinformatics analysis included here, a potential interaction between 110
NlrpNLRP12 and HckHCK may be involved in modulating the role of HckHCK in the 111
pathogenesis of AML (24-26). 112
113
Results 114
NlrpNLRP12 interacts with HckHCK in the yeast two-hybrid assay 115
A human leukocyte cDNA library was used as prey in the screening of binding partners for 116
NlrpNLRP12 using the yeast two-hybrid assay. The basic principle of the yeast two-hybrid 117
assay is that if the cDNA library contains clones that express proteins, or parts thereof, that 118
interact with the co-expressed bait, the NlrpNLRP12 PYD + NBD (Figure 1A) in this case, then 119
the co-transformed yeast clones will grow. The cDNA clones from the library that supported 120
growth would then be isolated and characterize, to identify what proteins interact with 121
NlrpNLRP12’s PYD + NBD. The leukocyte library was used because NlrpNLRP12 is mainly 122
expressed in leukocytes in humans (2,16,27,28), and thus putative binding partners for 123
NlrpNLRP12 are also likely to be expressed there. 124
125
The screening resulted in 119 positive clones from the leukocyte cDNA library (“hits”) 126
expressed from 48 distinct genes (Supplementary materialsSupporting information, Table S1) 127
that were subsequently identified as encoding for proteins (or portions thereof) that potentially 128
interact with NlrpNLRP12’s PYD + NBD. The clones with the most hits are listed in Table 1. 129
Among the 119 hits, which included repeated hits of the same genes, 9 were individual clones 130
that encoded fragments of a member of the Sc-SRCrc family of non-receptor tyrosine kinases, 131
HckHCK, which was among the highest number of hits for a single gene (Figure 1B and Table 1). 132
NLRPlrp12 binds to HCKck
7
Moreover, of the 9 positive HckHCK clones, 5 were encoded distinct HckHCK fragments 133
(Figure 1C). Full-length HckHCK p61 was also obtained as a hit in another independent yeast 134
two-hybrid screen (data not shown). HckHCK has two isoforms, p59 and P61, in Homo sapiens. 135
P61 and p59 share the same transcript, but p61 uses an uncommon start codon CTG, and p59 136
uses the common one, ATG (29), to give rise to two distinct transcripts. Endogenously 137
expressed in phagocytes, HckHCK p59 is mainly localized to the plasma membrane, while 138
HckHCK p61 is primarily localized to lysosomes, and both of them are also found in the Golgi 139
apparatus (30). It has been suggested that overexpression of p59 promotes the formation of 140
membrane protrusions such as pseudopodia and overexpression of p61 induces podosome 141
rosettes and triggers actin polymerization around lysosomes (30,31). However, there are no 142
differences between p61 and p59 in the region of HckHCK that appears to bind to 143
NlrpNLRP12’s PYD + NBD (see below), the C-terminal part of HckHCK. Thus, detecting a 144
full-length clone of only p61 as a hit, opposed to p59, appeared to be coincidental. 145
146
In directed screens using the yeast two-hybrid assay, NlrpNLRP12’s PYD + NBD domain 147
binding to HckHCK C-terminal 40 amino acids appears to be selective for NlrpNLRP12, 148
and HckHCK binding to NlrpNLRP12 appears to be selective for HckHCK 149
Initially, the HckHCK C-terminal 40 amino acids (“HckHCK C1 fragment”) was the shortest 150
HckHCK fragment from a positive clone that interacted with the bait, NlrpNLRP12’s PYD + 151
NBD (Figure 2A). NlrpNLRP 3 and NlrpNLRP8 were chosen based on the phylogenetic trees of 152
the PYD + NBD domains of members of the NlrpNLRP family: phylogenetically, NlrpNLRP3’s 153
PYD + NBD is the most closely related NlrpNLRP PYD + NBD to that of NlrpNLRP12, while 154
NlrpNLRP8’s PYD + NBD is among the furthest related PYD + NBD to that of NlrpNLRP12 155
NLRPlrp12 binds to HCKck
8
(Figure 2B). In Figure 2C, yeast clones grew on the 4 drop out (DO) plates (synthetic dropout 156
(SD) plus amino acids supplements minus tryptophan, leucine, histidine, and adenine (-Trp-Leu-157
His-Ade)) and 4DO + 3-Amino-1,2,4-triazole (3-AT) high stringency plates when the yeast 158
were co-transformed with NlrpNLRP12’s PYD + NBD and HckHCK C1 fragment; and thus, this 159
result is indicative of an interaction between the two constructs. However, yeast clones were not 160
observed on 4DO and 4DO + 3-AT plates when NlrpNLRP3’s PYD + NBD and HckHCK C1 161
fragment were co-transformed, nor when NlrpNLRP8’s PYD + NBD and HckHCK C1 fragment 162
were co-transformed. These results suggest that the HckHCK C1 fragment binds to 163
NlrpNLRP12’s PYD + NBD, but not to those from either NlrpNLRP3 or NlrpNLRP8, results 164
consistent with a selective interaction between NlrpNLRP12’s PYD + NBD and the HckHCK C1 165
fragment. 166
167
Similarly, yeast clones again grew on the high stringency 4DO and 4DO + 3-AT plates when the 168
HckHCK C1 fragment was co-transformed with NlrpNLRP12’s PYD + NBD. However, clones 169
did not grow on the 4DO and 4DO + 3-AT plates when the C1 fragment of other members of the 170
Sc-SRCrc family of non-receptor tyrosine kinases were co-transformed with NlrpNLRP12’s PYD 171
+ NBD (Figure 2D). This experiment suggests that the interaction of HckHCK C1 fragment 172
with NlrpNLRP12 is selective for HckHCK; the C1 fragment of other members of the Sc-SRCrc 173
family of non-receptor tyrosine kinases did not bind to NlrpNLRP12’s PYD + NBD (Figure 2D). 174
175
In directed screens using the yeast two-hybrid assay, amino acids F503, I506, Q507, L510, 176
and D511 of HckHCK are critical for the binding between NlrpNLRP12’s PYD + NBD and 177
HckHCK C-terminal 40 amino acids 178
NLRPlrp12 binds to HCKck
9
The HckHCK C1 fragment were divided into four helical regions according to its atomic 179
structure (32) (protein data bank (PDB) accession number: 1AD5): R1, R2, R3, and R4 180
(designated by us) (Figure 3A). As shown in the sequence alignments in Figure 3B, the primary 181
amino acid sequences of R2, R3, and R4 are very well conserved among members of the Sc-182
SRCrc family of non-receptor tyrosine kinases, while R1 is not. To characterize further the 183
important structural features of HckHCK’s binding to NlrpNLRP12’s PYD + NBD, all of the 184
amino acids comprising each of the R2, R3, and R4 domains were mutated to alanine, but only 185
mutated one section at a time was mutated, and these were denoted as the HckHCK M2 fragment, 186
HckHCK M3 fragment, and HckHCK M4 fragment, respectively. The amino acids comprising 187
the R1 region were mutated to those amino acids in the R1 region of the Sc-SRCrc family non-188
receptor tyrosine kinase LYNyn, which has the closest phylogenetic relationship to HckHCK 189
(Figure 3C). All of the mutated constructs are illustrated in Figure 3D. Yeast still grew on the 190
high stringency 4DO + 3AT plates when NlrpNLRP12’s PYD + NBD was co-transformed with 191
the HckHCK M1 fragment, HckHCK M2 fragment, and HckHCK M4 fragment. But, yeast could 192
not grow when NlrpNLRP12’s PYD + NBD was co-transformed with HckHCK M3 fragment 193
(Figure 3E). These experiments suggested that the R1, R2, and R4 regions are dispensable for 194
binding, while R3 may be important for binding. Further analysis was conducted to identify 195
individual critical amino acids of HckHCK’s R3 region binding to NlrpNLRP12’s PYD + NBD 196
using alanine scanning mutagenesis. However, the intervening sequences between R1, R2, R3, 197
and R4 were not examined for binding. 198
199
To identify the key amino acid residues involved in binding of R3 to NlrpNLRP12’s PYD + 200
NBD, each amino acid in R3 was individually mutated to alanine in an alanine scanning 201
NLRPlrp12 binds to HCKck
10
mutagenesis analysis, on the background of the HckHCK C1 fragment. As shown by the yeast 202
two-hybrid assay, the mutations of F503A, I506A, Q507A, L510A, and D511A abolished the 203
growth of yeast clones on the high stringency 4DO when these yeast strains were co-transformed 204
with NlrpNLRP12’s PYD + NBD (Figure 3F). These experiments suggest that these four amino 205
acids are critical for binding. On the other hand, yeast clones still grew on the high stringency 206
4DO plates when the HckHCK C1 fragments with the E504A, Y505A, S508A, or V509A 207
mutations were co-transformed with NlrpNLRP12’s PYD + NBD (Figure 3F). These 208
experiments suggest that E504A, Y505, S508, and V509 are dispensable for binding of HckHCK 209
C1 fragment to NlrpNLRP12’s PYD + NBD. Whether E504 and I506 are critical for binding is 210
unfortunately not clear as the yeast did not grow on the control plates (SD/-Trp-Leu). Overall, 211
however, these results suggest that the amino acids F503, I506, Q507, L510, and D511 in 212
HckHCK are critical for the binding of HckHCK C1 fragment to NlrpNLRP12’s PYD + NBD. 213
214
On the other hand, YES also hasve these fiveour amino acid residues, are all conserved in all 215
members of the Src family of non-receptor tyrosine kinases (Figure 3A), suggesting that, while 216
they are important for binding, they cannot account for the specificity or selectivity of binding 217
between NlrpNLRP12 and HckHCK. In addition, these four critical residues, as shown in Figure 218
3C, appear to be in the same plane along a helix, which may then form an interface that is 219
important for binding to NlrpNLRP12. However, it is unclear whether the structure of the 220
isolated HckHCK C-terminus used in the screens here will retain theits structural features that it 221
possesses within the context of in full-length HckHCK, due to the potential loss of 222
intermolecular interactions for the C-terminal region. Moreover, for example, In addition, the 223
amino acids R500, P501, T502, D512, T515, Q520, and Y521, are 100% conserved among all 224
NLRPlrp12 binds to HCKck
11
other members of the Sc-SRCrc family of non-receptor tyrosine kinases and are therefore 225
unlikely to play a role in specificity and selectivity of binding of NlrpNLRP12 with HckHCK. 226
But, it is possible that two or more amino acids in the intervening sections between the R1-R4 227
regions, which were not analyzed by mutagenesis, may then actually account for the specificity 228
and selectivity of binding of HckHCK to NlrpNLRP12. One example is that the combination of 229
R496, P497, and Q522 is unique to HckHCK compared to other members of the cS-SRCrc 230
family of non-receptor tyrosine kinases. 231
232
In directed screens using the yeast two-hybrid assay, the shortest fragment of HckHCK 233
that binds to NlrpNLRP12’s PYD + NBD is its C-terminal 30 amino acids 234
The shortest form of HckHCK that bound to NlrpNLRP12’s PYD + NBD was determined by 235
observing whether the yeast clones grew on high stringency 4DO or 4DO +3AT plates when 236
NlrpNLRP12’s PYD + NBD was co-transformed with HckHCK fragments C1 to C7 and N1 to 237
N5 (Figure 4) (data of NlrpNLRP12’s PYD + NBD co-transformed with HckHCK fragment C7 238
did not shown). HckHCK’s C1 to C7 fragments were gradually truncated from the N-terminus 239
of the HckHCK C-terminal 40 amino acid fragment (Figure 4A and 4B). HckHCK N1 to N5 240
fragments were gradually truncated from the C-terminus of the HckHCK C-terminal 40 amino 241
acid fragment (Figure 4BC and 4D). Yeast strains grew on high stringency 4DO or 4DO +3AT 242
plates when either the HckHCK C1, C2, or C3 fragment was co-transformed with NlrpNLRP12’s 243
PYD + NBD. Thus, of all of the constructs screened, the shortest fragment of HckHCK that can 244
apparently bind to NlrpNLRP12’s PYD + NBD is the C-terminal 30 amino acids, as indicated by 245
the HckHCK C3 fragment (Figure 4B and 4D). In addition, the last C-terminal 5 amino acids of 246
the HCK C-terminus also appear to be important for binding, as clones co-transformed with 247
NLRPlrp12 binds to HCKck
12
NlrpNLRP12’s PYD + NBD with N1, with a deletion of the C-terminal 5 amino acids of 248
HckHCK, did not grow on the high stringency 4DO and 4DO + 3AT plates (Figure 4BD). The 249
structures of both inactive HckHCK (PDB accession number: 1AD5) and active c-SSRCrc (PDB 250
accession number: 1Y57), the prototypical member of this family of nonreceptor tyrosine kinases 251
related to HckHCK, show that the last C-terminal 4 amino acids of c-SSRCrc are surface-252
exposed, indicating that these four amino acids that are critical for binding are available to 253
interact with NlrpNLRP12 (33,34). 254
255
NlrpNLRP12 co-immunoprecipitates with HckHCK from mammalian cells, and 256
NlrpNLRP12 is co-localized with HckHCK in mammalian cells 257
NlrpNLRP12 and HckHCK could be co-immunoprecipitated when they were both 258
heterologously co-expressed in 293T cells (Figure 5A). Also, semi-endogenous conditions were 259
used for co-immunoprecipitations from cells in which HckHCK is endogenously expressed, and 260
NlrpNLRP12 was heterologously stably expressed in epitope-tagged form, such as in 261
macrophage-like RAW 264.7 cells (Figure 5B), monocyte-like THP-1 cells (Figure 5C), and 262
lymphocyte -like U937 cells (Figure 5D). Semi-endogenous conditions have been used in 263
previous reports where NlrpNLRP12 was shown to interact with TRAF3 (3), NIK (35), and 264
IRAK-1(9). In all cases, NlrpNLRP12 co-immunoprecipitated with the p59 and p61 isoforms of 265
HckHCK, but from these experiments, it is not clear that whether NlrpNLRP12 interacts with 266
p59, p61, or both isoforms of HckHCK. In U937 cells and THP-1 cells, since heterologous 267
NlrpNLRP12 expression level was low, adding phorbol 12-myristate 13-acetate (PMA) (1 µM 268
for 24 hr) increased NlrpNLRP12 expression level by stimulating the PMA-sensitive 269
cytomegalovirus promoter of NlrpNLRP12’s expression plasmid (Figure 5C and 5D). Taken 270
NLRPlrp12 binds to HCKck
13
together, these data suggest that NlrpNLRP12 and HckHCK (either p59, p61, or both isoforms) 271
can interact with each other in a mammalian cell system, a result that in part validates the 272
findings from the yeast two-hybrid assay. 273
274
After heterologous co-expression of NlrpNLRP12 with either the p61 or the p59 isoform of 275
HckHCK in 293T cells, immunofluorescent staining of NlrpNLRP12 and HckHCK showed that 276
NlrpNLRP12 can be co-localized with either p61 or p59 isoforms of HckHCK in most of the 277
293T cells (Figure 5E and Supplementary materials, Supporting Information Figure S1). The co-278
localization of NlrpNLRP12 and HckHCK is consistent with an intracellular interaction between 279
the two proteins. In addition, the lack of selectivity of NlrpNLRP12 co-localizing with either 280
p61 or p59 isoform of HckHCK is consistent with the results from the screening assay, where the 281
putative binding region of HckHCK to NlrpNLRP12 is identical in both the p61 and p59 282
isoforms. 283
284
The steady-state level of NlrpNLRP12 protein may is significantly decreased when 285
HckHCK protein is co-expressed in an increasing amount 286
Figure 6 shows that in 293T cells, the co-expression of increasing amount of HckHCK may led 287
to a significant decrease in steady-state protein expression levels of NlrpNLRP12. Although, 288
alternatively, the apparent decrease in expression level of NlrpNLRP12 might be due actually to 289
a decrease in the epitope-tag itself or a decrease in the expression of the epitope-tagged protein 290
only. Interestingly, HckHCK protein levels did not show a linearlyn increasing level of 291
expression upon transfection with increasing amounts of plasmid DNA, possibly due to 292
HckHCK’s maximum expression levels in these cells being achieved with the lowest amounts of 293
NLRPlrp12 binds to HCKck
14
plasmid DNA, which may be resolved if the transfections are performed with lower amounts of 294
plasmid DNA. This lack of linearity in protein expression It could also be due to saturation of the 295
signal for HCK on the the fact that blot at relatively low levels of HCK expression is saturated 296
or a combination of HCK’s maximum expression levels has been reached as well as saturation 297
of the and blot, as mentioned above. is saturated. Alternatively, reduction of NLRP12 expression 298
could be due to plasmid competition; that is, since the expression of both genes from either 299
plasmid are driven from the same promoter, an increase in the amount of one plasmid may 300
competitively decrease expression of the gene from the other plasmid due to their competition 301
for transcription factors.. 302
303
In blood and marrow samples from patients with AML, NlrpNLRP12 and HckHCK co-304
occur and are co-expressed 305
The odds ratio test (a statistical test that defines the likelihood of association of two events) 306
determines whether two proteins either co-occur or are mutually exclusive in their expression. 307
The Fisher exact test is a statistical test that uses P-values to determine if two events, such as co-308
occurrence or mutual exclusivity, are quantitatively significantly associated. Applying these 309
tests to The Cancer Genome Atlas (TCGA) provisional data sets for AML patients (36-38), 310
NlrpNLRP12 and NlrpNLRP1 are the only two NlrpNLRP family members that show a 311
statistically significant co-occurrence with HckHCK (Table 2). However, NlrpNLRP1’s co-312
occurrence has a P value of 0.048, while NlrpNLRP12’s co-occurrence has a P value of <0.001. 313
Therefore, NlrpNLRP12 is most likely the only NlrpNLRP family protein mRNAs that shows co-314
occurrence with HckHCK in samples from patients with AML. 315
316
NLRPlrp12 binds to HCKck
15
Moreover, in this analysis, NlrpNLRP12 is significantly co-expressed with HckHCK (Pearson 317
coefficient = 0.80 and Spearman coefficient = 0.73, Pearson and Spearman coefficients show the 318
possibility of a linear relationship of two factors, Figure 7). This analysis is also consistent with 319
NlrpNLRP12 and HckHCK co-occurring in the same patients with AML. However, 320
NlrpNLRP12 was not the only NlrpNLRP protein mRNA that shows significant co-expression 321
with HckHCK. NlrpNLRP1 and NlrpNLRP3 also showed significant co-expression with 322
HckHCK. This can be explained by the fact that co-expression is less stringent than co-323
occurrence. In summary, NlrpNLRP12 is the only NlrpNLRP protein that shows co-occurrence 324
with HckHCK, and it also shows co-expression with HckHCK, although it is not the only 325
member of the NlrpNLRP family to do so. Other data sets, including Hausera data, Metzeler 1 326
data, Metzeler2 data, and Raponi data, were not available to evaluate the co-expression of 327
NlrpNLRP12 versus HckHCK, as these other data sets did not have information incorporate data 328
regarding NlrpNLRP12 expression. 329
330
Other potential binding partners of NlrpNLRP12 from the yeast two-hybrid screen 331
Other potentially interesting and unique genes that whose products may interact with 332
NlrpNLRP12’s PYD + NBD were identified in the yeast two-hybrid screen (Table 1 and 333
Supplementary materials, Supporting information Table S1). Table 1 only shows those hits with 334
frequency larger than 2, and these hits were analyzed through an Ingenuity Pathway Analysis 335
(IPA) diagram (Figure 8). In the IPA diagram, protein-protein interactions found by our yeast 336
two-hybrid screen are connected by solid lines and interactions between protein binding partners 337
collated by IPA are connected by dashed lines. In addition, those molecules that are associated 338
with AML are colored purple. Interestingly, NlrpNLRP12 is also associated with AML from the 339
NLRPlrp12 binds to HCKck
16
database Catalogue of somatic mutations in cancer (COSMIC). Although the mutation occurs in 340
the NBD of NlrpNLRP12, only 2 samples of these mutations are observed within 175 samples 341
and may be worthy of follow-up study. 342
343
The interaction of one of the putative binding partners to NlrpNLRP12 identified in our screen, 344
TRAF3 interacting protein 3 (TRAF3IP3), was validated in a co-immunoprecipitation 345
experiment where the FLAG-tagged form of TRAF3IP3 and V5-tagged form of NlrpNLRP12 346
were overexpressed in 293T cells. TRAF3IP3 was coimmunoprecipitated by adding mouse 347
monoclonal anti-TRAF3IP3 antibody, and a band that showed the same molecular weight size of 348
NlrpNLRP12 was detected by rabbit monoclonal anti-V5 antibody (Figure 8). However, further 349
validation of these other putative interactors detected by the yeast two-hybrid needs to be done. 350
351
Discussion 352
In this paper, the yeast two-hybrid assay was used to screen for NlrpNLRP12 binding partners. 353
This approach identified 48 genes as positive hits encoding proteins that could possibly bind to 354
NlrpNLRP12. Among the hits, the cS-SRCrc non-receptor tyrosine kinase family member 355
HckHCK was selected to focus upon as a potentially interesting target, particularly given its 356
association with hematopoiesis and AML. Further direct screens found that NlrpNLRP12 357
specifically binds to HckHCK, but not to NlrpNLRP8 or NlrpNLRP3. Conversely, HckHCK, 358
but not other members of the SRCc-SRCrc non-receptor tyrosine kinase family, specifically 359
binds to NlrpNLRP12. HckHCK’s C-terminal 30 amino acids are sufficient for binding to 360
NlrpNLRP12. And, amino acids F503, I506, Q507, L510, and D511 of HckHCK are critical for 361
binding to NlrpNLRP12. Co-immunoprecipitation experiments from mammalian cells 362
NLRPlrp12 binds to HCKck
17
confirmed that NlrpNLRP12 and HckHCK interact. Co-expression of NlrpNLRP12 and 363
HckHCK results in a significantly lower steady-state expression level of NlrpNLRP12; thus, the 364
interaction may affect NlrpNLRP12 protein stability. Finally, a bioinformatics analysis shows 365
that NlrpNLRP12 and HckHCK essentially exclusively co-occur, and mRNA levels for both are 366
correlated in blood samples from AML patients. Together, these data suggested that 367
NlrpNLRP12 and HckHCK protein and mRNA co-expression may have an effect on the AML. 368
369
The binding data from yeast two-hybrid (Figure 3 and Figure 4) showed that the R2 and R3 370
regions are both important for binding, but we suspect that R3 region is the major region that is 371
critical for binding, while R2 region may be an “auxiliary” region that is also critical for binding, 372
because when we delete the R2 region but maintain the R3 region, the R3 region itself cannot 373
maintain stable binding of HckHCK’s C-terminal 40 amino acids to NlrpNLRP12’s PYD + NBD. 374
Thus, binding of NlrpNLRP12’s PYD + NBD to this fragment of HckHCK may occur initially 375
through binding to R3, followed by stabilization of the interaction by secondarily binding to R2. 376
This model is also consistent with the finding that when we mutated the R3 region, the binding 377
of HckHCK’s C-terminal 40 amino acid fragment to the NlrpNLRP12’s PYD + NBD was lost 378
despite the R2 region being present. The above evidences are consistent with the model that R2 379
and R3 regions together are both necessary for binding, but the stabilization of binding to the R3 380
region may requires the presence of R2. 381
382
The yeast two-hybrid screen with NlrpNLRP12 also identified products from 47 other genes as 383
potential interactors. According to the Search Tool for Recurring Instances of Neighbouring 384
Genes (STRING) protein network and previous publications, NlrpNLRP12 interactors include 385
NLRPlrp12 binds to HCKck
18
IRAK1 (39), NIK (35), TRAF3(3), heat shock protein 90 (HSP90) (40), Fas associated factor 1 386
(FAF1) (41), and apoptosis-associated speck-like protein containing a caspase activation and 387
recruitment domain (ASC) proteins. IRAK1, NIK, TRAF3, and Heat shock protein (HSP) 90 388
were only found in experiments involving semi-endogenous protein co-immunoprecipitations 389
with NlrpNLRP12. FAF1 was found to be an interacting protein in NMR experiments by 390
showing a difference in the chemical shift perturbation of FAF1’s ubiquitin-associated domains 391
(UBA) domain alone and NlrpNLRP12’s PYD domain bound to FAF1’s UBA domain. And 392
FAF1 was also found to interact with NlrpNLRP12 in a yeast two-hybrid study where the PYD 393
domain of different NlrpNLRP proteins were screened (42). However, in our study, IRAK1 (43), 394
NIK (44), TRAF3 (45), HSP90 (46), FAF1 (47), and ASC (48) proteins were not found in our 395
results of interacting proteins, although they are all expressed in leukocytes. Therefore, given 396
these differences, the results from this yeast two-hybrid screen may provide a unique database 397
for future investigations on the proteins that interact with NlrpNLRP12, and the functional 398
outcomes of these interactions. 399
400
How HckHCK co-expression affects NlrpNLRP12 expression can be further investigated to 401
determine the functional outcomes of the NlrpNLRP12-HckHCK interaction relative to protein 402
stability or degradation. For example, MG-132 and chloroquine, which are a proteasomal 403
inhibitor and an inhibitor of lysosomal degradation, respectively, can be added to cells in these 404
experiments to determine whether the decreasing protein expression level of NlrpNLRP12 in the 405
presence of HckHCK is due to proteasomal degradation or lysosomal degradation. In addition, 406
an NlrpNLRP12 stable cell line, with a typically lower level of NlrpNLRP12 expression, 407
compared to transiently transfected cells, can be used to test formally whether NlrpNLRP12 408
NLRPlrp12 binds to HCKck
19
protein expression levels are increasingly decreased when the HckHCK’s is expression of HCK 409
is increased. Moreover, the rate of NlrpNLRP12 degradation, in the absence or presence of 410
HckHCK, can be studied and compared in pulse-chase assays (35). 411
412
Given the interaction that was characterized between Nlrp12NLRP12 and HckHCK, we sought 413
to identify diseases in which their interaction may be relevant. AML is a type of blood and bone 414
marrow cancer that is characterized by an unusually high level of circulating immature white 415
blood cells (49). HckHCK is found to be highly expressed in hematopoietic stem cells of AML 416
patients (24). The overexpression of HckHCK is linked with the development of AML (24,50). 417
Therefore, HckHCK may be potentially used as a blood biomarker to identify those patients with 418
potential to develop of AML disease. In addition, activation of HckHCK and its signaling 419
pathways often causes the development of AML (25,51). HckHCK has been reported to be 420
involved in the feline McDonough sarcoma (fms)-like tyrosine kinase 3 (FLT3)-HCKHCK- 421
cyclin D-dependent kinase (CDK6) pathway. The mutation in FLT3-internal tandem 422
duplications (FLT3-ITD) causes the activation of HckHCK, and subsequently results in 423
overexpression of CDK6 in MV4-11 cells, a human AML cell line with the FLT3-ITD mutation. 424
CDK6 siRNA resulted in MV4-11 cells with the FLT3-ITD mutation having a lower 425
proliferation rate compared to MV4-11 control cells (25). Also, it has been reported that in AML 426
patients, HckHCK, LynLYN, and Fgr FGR phosphorylation levels are high (51). In summary, 427
the above evidence suggests that either high expression and/or high activation of HckHCK could 428
occur in AML patients as well. 429
430
NLRPlrp12 binds to HCKck
20
Our bioinformatics analysis revealed that Nlrp12NLRP12 is likely the only member of the 431
NlrpNLRP family that shows significant co-occurrence with HckHCK at the level of mRNA 432
expression. However, our cellular functional study revealed that NlrpNLRP12 protein 433
expression levels decreased when it is co-expressed with HckHCK. This discrepancy might be 434
due NlrpNLRP12 mRNA expression level not correlating with protein expression level (52-54). 435
Nevertheless, these data are still consistent with the possibility that NlrpNLRP12 may have an 436
effect on AML through its binding to HckHCK. NlrpNLRP12 and HckHCK co-expression and 437
co-occurrence may then also be relevant to the prognosis of AML. However, further analyses 438
need to be performed to evaluate the ability of NlrpNLRP12 and HckHCK to serve as 439
biomarkers in the prognosis of AML. 440
441
NlrpNLRP12 and HckHCK co-expression and co-occurrence may indicate that NlrpNLRP12 and 442
HckHCK mRNA or protein expression can impact the same pathway or both be regulated by the 443
same molecule. For example, in the case two molecules exhibiting co-occurrence, they may 444
interact in the same signaling pathway, for example, Ki-ras2 Kirsten rat sarcoma viral oncogene 445
homolog (KRAS) and glycogen synthase kinase 3 beta (GSK3β) showed co-occurrence in the 446
AML provisional data, and they are in the same signaling pathway (55). However, further 447
analyses need to be performed to evaluate the ability of NlrpNLRP12 and HckHCK to serve as 448
biomarkers in the prognosis of AML. 449
450
Although many HckHCK inhibitors have been developed, they often lack specificity, with 451
HckHCK inhibitors also inhibiting other members of the Sc-SRCrc non-receptor tyrosine kinase 452
family (24,56-5960). If the potential for the development of AML can be decreased when 453
NLRPlrp12 binds to HCKck
21
NlrpNLRP12 does not bind to HckHCK, then an inhibitor to the interaction between 454
NlrpNLRP12 and HckHCK can be designed. On the other hand, if the potential for AML is 455
increased when NlrpNLRP12 and HckHCK does not bind, then a drug that facilitates 456
NlrpNLRP12 and HckHCK interaction could be developed. By this approach, the non-457
specificity problem with NlrpNLRP inhibitors may still exist but perhaps to a smaller extent, as 458
NlrpNLRP proteins are less conserved overall compared to the members of the Sc-SRCrc non-459
receptor tyrosine kinase family. For example, the amino acid identity between NlrpNLRP12’s 460
PYD + NBD versus another NlrpNLRP PYD + NBD most closely related to NlrpNLRP12, 461
NlrpNLRP3’s PYD + NBD, is 54%. But, the protein identity between the HckHCK C1 fragment 462
versus the one with the closest relationship, the LYNyn C1 fragment, is 78%. 463
464
In conclusion, this paper reports the preliminary findings of NlrpNLRP12 interacting with 465
HckHCK, and NlrpNLRP12 co-occurs with HckHCK in AML patients. The binding data 466
presented here will expand the database regarding the interactors of NlrpNLRP12, and additional 467
bioinformatic and functional in vitro and in vivo analyses may shed light on the functional 468
consequence of the interaction of NlrpNLRP12 with HckHCK. 469
470
Material and Methods 471
Reagents and cells 472
U937, THP-1, RAW 264.7, and 293T cells were all purchased from American Type Culture 473
Collection (ATCC) (Manassas, VA). pCDNA-NlrpNLRP12 plasmid and pCDNA-NlrpNLRP3 474
plasmid were gifts from Dr. Jenny P. Y. Ting (University of North Carolina-Chapel Hill). 475
pCDNA-NlrpNLRP8 plasmid was purchased from Genescript (Piscataway, NJ). All other 476
NLRPlrp12 binds to HCKck
22
chemicals were reagent grade. The mouse monoclonal anti-FLAG antibody and the rabbit 477
monoclonal anti-FLAG antibody were purchased from Sigma-Aldrich (St. Louis, MO). The 478
mouse monoclonal anti-HckHCK antibody, the mouse monoclonal anti-TRAF3IP3 antibody, and 479
the mouse IgGκ binding protein-horseradish peroxidase (HRP) for detecting the anti-HckHCK 480
antibody were purchased from Santa Cruz Biotechnology (Dallas, TX). The rabbit monoclonal 481
anti-V5 antibody and polyclonal rabbit HRP-conjugated anti-mouse IgG secondary 482
antibodies were purchased from Cell Signaling (Beverly, MA). The goat anti-rabbit IgG HRP-483
linked secondary antibody was purchased from Cell Signaling. The Clean-Blot™ IP Detection 484
Reagent was purchased from Thermo Fisher Scientific (Grant Island, NY). Goat anti-mouse IgG 485
(H+L) cross-adsorbed secondary antibody, conjugated to Alexa Fluor 488 and goat anti-rabbit 486
IgG (H+L) cross-adsorbed secondary antibody, conjugated to Alexa Fluor 555 were purchased 487
from Thermo Fisher Scientific (Grant Island, NY). The 4',6-diamidino-2-phenylindole, 488
dihydrochloride (DAPI) was purchased from Thermo Fisher Scientific. 489
490
Cloning 491
NlrpNLRP12’s PYD + NBD (corresponding to a.a. 1-531) was amplified by polymerase chain 492
reaction (PCR) (primers see Table 3), from a pCDNA-NlrpNLRP12 plasmid, was digested with 493
enzymes NdeI and SalI, and subcloned into the pGBKT7 vector (Takara). The sequence 494
encoding the PYD + NBD of NlrpNLRP3 (corresponding to a.a. 1-536) was amplified by PCR 495
(see table 3 for primers), digested with enzymes NdeI and BamHI, and subcloned in the 496
pGBKT7 vector. The sequence encoding the PYD + NBD of NlrpNLRP8 (corresponding to a.a. 497
1- 527) was amplified by PCR (see Table 3 for primers), digested with enzymes EcoRI and 498
BamHI, and subcloned into the pGBKT7 vector. The entire sequences corresponding to of the 499
NLRPlrp12 binds to HCKck
23
C-terminal 42 amino acids of FGRgr, Sc-SRCrc, YESes, and FYNyn, and 40 amino acids of 500
BLKlk, LCKck, LYNyn, and HckHCK (The C terminal 40 (or 42) amino acids is called “C1”), 501
and HckHCK constructs C2, C3, C4, C5, C6, C7, N1, N2, N3, N4, N5, M1, M2, M3, M4, and 502
HckHCK C1 mutations for alanine scanning with F503A, E504A, Y505A, I506A, Q507A, 503
S508A, V509A, L510A, and D511A, were all synthesized from IDT DNA (Coralville, IA) and 504
cloned into the pACT2 vector (Takara). NlrpNLRP12 was amplified by PCR (see Table 3 for 505
primers), digested with AflII and XbaI, and subcloned into pEFIRES-P-3FLAG (601) and 506
pEFIRES -P-V5 (601). 3FLAG and V5 are homemade, i.e., added later into pEFIRES-P. The 507
peptide sequence of the 3FLAG epitope is DYKDDDDKDYKDDDDKDYKDDDDK. The 508
peptide sequence of the V5 epitope is GKPIPNPLLGLDST. The multiple cloning sites of 509
pEFIRES-P were modified according to our needs. Sequences encoding HckHCK p61 and p59 510
isoforms were amplified by PCR (see Table 3 for primers), digested with AflII and XbaI, and 511
subcloned into pEFIRES -3FLAG vector. Final plasmids were sequenced to confirm the 512
constructs. 513
514
Yeast two-hybrid assay 515
Yeast transformed with the plasmid PGBKT7- NlrpNLRP12’s PYD + NBD 516
Yeast strain AH109 (Takara, Mountain View, CA) was stored frozen at -80°C. Upon use, it was 517
streaked onto an agar plate supplemented with yeast extract, peptone, dextrose, and adenine 518
hemisulfate (YPDA) (Takara), which was then placed in a 30°C incubator overnight. The 519
growing colonies were picked and placed into 10 ml of YPDA media, incubated overnight with 520
shaking at 30°C. Then they were transformed with NlrpNLRP12’s PYD + NBD plasmid, 521
following recommendations in the Takara manual (MatchmakerTM Gal4 Two-hybrid System 3 522
NLRPlrp12 binds to HCKck
24
and Libraries User Manual). A 4ml of 1 M of 3-AT was added into 800 mL of the YPDA 523
medium that were used for making the plates due to a leakage problem. 3-AT is a competitive 524
inhibitor for the product of the HIS3 reporter gene. Yeast cells were lysed using a trichloroacetic 525
acid (TCA) method, and western blots were performed to check NlrpNLRP12’s PYD + NBD 526
protein expression (data not shown). 527
528
Library scale transformation of cDNA library into yeast that have been transformed with the 529
PGBKT7- NlrpNLRP12 PYD + NBD plasmid 530
A human leukocyte MatchmakerTM cDNA library was purchased from Takara. The library was 531
titered according to the MatchmakerTM Gal4 Two-hybrid System 3 and Libraries User Manual. 532
The colony-forming units (CFU) for the library were calculated to be about 3.1 x 109 CFU/ml. 533
Four large scale transformations (equal to one library scale) were done at the same time 534
following the MatchmakerTM Gal4 Two-hybrid System 3 and Libraries User’s Manual. To 535
calculate the efficiency of yeast two-hybrid (i.e., the number of colons obtained per µg of library 536
plasmid DNA) and to check the expressions of “bait” and “prey” plasmids, small amounts of the 537
solution, diluted 10X, 100X, 1000X, and 10,000X were plated in SD/-Trp-Leu 2DO (drop out) 538
plates. The yeast two-hybrid efficiency obtained was 3 x 106 CFU per µg of library plasmid 539
DNA. In addition, if the yeast grew on the SD/-Trp-Leu-His-Ade + 3-AT plates, the “bait” and 540
“prey” plasmids were considered to be interacting with each other. 541
542
Small scale transformation 543
The yeast two-hybrid co-transformation on a small scale for all of the direct screenings was 544
performed following the small-scale transformation protocol in the MatchmakerTM Gal4 Two-545
NLRPlrp12 binds to HCKck
25
hybrid System 3 & Libraries User’s Manual. The direct screens were to test whether 546
NlrpNLRP12 PYD + NBD and HckHCK specifically interact; identification of HckHCK F503, 547
Q507, L510, and D511 being critical for binding with NlrpNLRP12 PYD + NBD; and the 548
characterization of the HckHCK C-terminal 30 amino acid fragment binding to NlrpNLRP12 549
PYD + NBD. 550
551
Cell transfection 552
293T cells, THP-1 cells, U937 cells, and RAW 264.7 cells were seeded at a density of 5 x 105 553
per well in a 6-well plate. Transfection followed the Lipofectamine® 3000 reagent (Thermo 554
Fisher Scientific) protocol with the following modifications. Briefly, 3.75 µl of Lipofectamine® 555
3000 was added into 125 µl of Opti-MEM® medium (Thermo Fisher Scientific). 2.5 µg of 556
plasmid DNA (prepared by isolation through cesium chloride (CsCl) density gradient 557
purification (612) and 5 µl of P3000TM reagent were added into another aliquot of 125 µl of 558
Opti-MEM® medium. Then these two tubes were mixed together, and the mixed tube was 559
incubated at room temperature for 15 min. 250 µl of this mixture was added to the cells. The 560
cells were incubated at 37 °C with 5% CO2 for two days until analysis. 561
562
NlrpNLRP12 stable cell lines 563
THP-1 cells, U937 cells, and RAW 264.7 cells were transfected with pEFIRES-P-NlrpNLRP12-564
3FLAG plasmid, and 2.5 μg/ml puromycin (InvitrogenTM by Thermo Fisher Scientifc) was added. 565
After two weeks, the cells were considered to be stably expressing NlrpNLRP12 after screening 566
for expression. The stable cells were “pooled” clones, and they are maintained by adding 2.5 567
μg/ml puromycin. 568
NLRPlrp12 binds to HCKck
26
569
Cell lysis and immunoprecipitation 570
After trypsinization, cells were lysed in 500 μl of lysis buffer (1% NP-40, 4mM Tris-HCl, pH 8.0, 571
150 mM NaCl). Cells were then treated with freeze-thaw cycles, i.e., frozen in the -80C freezer 572
for 5 min and thawed at 37C in a water bath for 3 min, for a total of five times. Cells were then 573
centrifuged at 12,000 x g for 10 min at 4C. After centrifugation, the supernatant was saved at -574
20C until further use. 575
576
For immunoprecipitations, cell lysates were pre-cleared by adding 40 μl of protein A/G agarose 577
beads (Thermo Fisher Scientific) from a 50% slurry, and the sample was rotated for 1 hr at 4C. 578
The sample was spun at 1000 x g for 1min. Then, 30 μl of pre-cleared cell lysate were removed 579
and run on immunoblotting as a control. The remaining cell lysate was incubated with 1 μg anti-580
HckHCK mouse monoclonal primary antibody overnight. On the second day, 30 μl of protein 581
A/G agarose, from a 50% slurry and previously washed with lysis buffer three times, were added 582
to the tube, and the tube was incubated for 3 hrs at 4C. The beads were then washed with lysis 583
buffer, but with the NaCl concentration increased to 500 mM. The tube was briefly centrifuged 584
at 1000 x g for 1min., and the supernatant was removed. This step was performed for a total of 5 585
times. The proteins were eluted from the beads by addition of SDS-PAGE sample buffer. 586
587
Immunoblotting 588
Protein concentrations were determined by the PierceTM BCA protein assay (Thermo Fisher 589
Scientific). About 20µg of cell lysate samples and immunoprecipitated samples were loaded 590
onto the 8% acrylamide gel. After overnight transfer of the proteins from an SDS-gel to a 0.45 591
NLRPlrp12 binds to HCKck
27
μm pore size polyvinylidene difluoride (PVDF) membrane, the membrane was blocked in for 1 592
hr at room temperature 5% non-fat dry milk dissolved in (Tris-buffered saline with Tween 20, 593
pH=8.0) (TBST) (Sigma-Aldrich). Mouse monoclonal anti-FLAG antibody (diluted 1:1000) and 594
mouse monoclonal anti-HckHCK antibody (diluted 1:1000) were added and the membrane 595
incubated overnight. For detection of bands in the cell lysate, rabbit anti-mouse IgG HRP-linked 596
secondary antibody (diluted 1:2000) (for anti-FLAG antibody) and mouse IgGκ binding protein-597
HRP secondary antibody (for anti-HckHCK antibody) (diluted 1:1000) were added, and the 598
membrane was incubated for 1 hr at room temperature. For immunoprecipitations, Clean-Blot™ 599
IP detection reagent (HRP) secondary antibody (1:300) (Thermo Fisher Scientific) was used. 600
The detection signal was visualized by using enhanced chemiluminescence (ECL) reagent 601
(Thomas Scientific, Swedesboro, NJ). Immoblots were analyzed by chemiluminescence on a 602
Bio-Rad ChemiDocTM Touch Gel imaging system (Bio-Rad, Hercules, CA). 603
604
Immunofluorescence 605
The A protocol for immunofluorescence was followed 606
(https://www.cellsignal.com/contents/resources-protocols/immunofluorescence-general-607
protocol/if), with the following modifications. cCoverslips were autoclaved and coated with 0.1 608
mg/ml of poly-D-lysine hydrobromide (mol wt 70,000-150,000) (Sigma-Aldrich). Cells were 609
plated onto the coverslips. The cells were fixed in 4% paraformaldehyde (Alfa Aesar, Ward Hill, 610
MA) in phosphate buffered saline (PBS) for 15 mins. And fFixation was stopped by adding 50 611
mM NH4Cl in PBS for 15 min at room temperature. After raising with PBS, the fFixed cells 612
were permeabilized by adding 0.5% Triton®-X-100 in PBS for 10 min. The cells were then 613
washed with PBS three times and blocked with 0.1% of BSA in PBS at room temperature for 1 614
NLRPlrp12 binds to HCKck
28
hr. The rabbit monoclonal anti-FLAG antibody (diluted 1:100) and mouse monoclonal anti-615
HckHCK antibody (diluted 1:100) were added to each cover slip and incubated overnight. Goat 616
anti-mouse IgG (H+L) cross-adsorbed secondary antibody, conjugated to Alexa Fluor 488 617
(diluted 1:100) and goat anti-rabbit IgG (H+L) cross-adsorbed secondary antibody, conjugated to 618
Alexa Fluor 555 (1: 100) were added. After incubation with secondary antibodies for 1 hr, the 619
cells were washed five times with PBS. During the first step, DAPI (InvitrogenTM by Thermo 620
Fisher Scientific) was also added at 1:1000 in PBS for 2 min. Finally, the coverslips were 621
mounted in ProLong™ Gold Antifade mounting medium (Life Technologies Corporation by 622
Thermo Fisher Scientific, Eugene, OR). The stained slides were viewed on a ZEISS LSM 880 623
(Pleasanton, CA) with Airyscan. The images were viewed and processed using Fiji software 624
(623). 625
626
IPA diagram 627
Yeast two-hybrid data were analyzed by IPA software (QIAGEN Inc., 628
https://www.qiagenbioinformatics.com/products/ingenuity- pathway-analysis). 629
630
Statistical analysis 631
The Spearman coefficient and Pearson coefficient used for co-expression was generated by 632
cBioPortal website (37,38,634). Clinical samples from the Oncomine website 633
(https://www.oncomine.org) were downloaded and checked for NlrpNLRP12 and Hck mRNA 634
expression levels and HckHCK levels. The Fisher exact test and the odds ratio test used for co-635
occurrence were generated by the cBioPortal website. Analysis of variance (ANOVA) was 636
performed to check whether a significant difference of the treatment exists among all the groups. 637
NLRPlrp12 binds to HCKck
29
A p-value < 0.05 was considered statistically significant. If P < 0.05, then the Tukey test was 638
used to determine whether a pair-wise difference exists. Quantification of the western blot bands 639
was performed by FIJI software (623) and made into graphs by Prism 6.0 (GraphPad Software, 640
Inc., San Diego, CA).641
642
Acknowledgement 643
Thank you to Dr. Houda Alachkar for providing valuable suggestions to this paper. We also 644
thank Ms. Meng Li and Dr. Yibu Chen at the University of Southern California’s Norris Medical 645
Library for their suggestions for bioinformatic analysis. Appreciation is given to Mr. Noam 646
Morningstar-Kywi and Dr. Ian Haworth for their helps on structural analysis. 647
648
Abbreviations 649
3-AT: , 3-Amino-1,2,4-triazole; AML: , acute myeloid leukemia; ANOVA: , analysis of variance; 650
ASC: , apoptosis-associated speck-like protein containing a caspase activation and 651
recruitment domain; ATCC: , American Type Culture Collection; C1:, self-named truncations of 652
C-terminal 40 (or 41) amino acids fragment of Srcc-SRC family non-receptor tyrosine kinases; 653
C2-C7 and N1-N5:, self-named truncations HckHCK C1 fragment; CD40:, cluster of 654
differentiation 40; CDK6:, cyclin D-dependent kinase; CFU;, colony-forming units; COSMIC, 655
catalogueCOSMIC: catalogue of somatic mutations in cancer; CsCl:, cesium chloride; DAMP: , 656
damage-associated molecular patterns; DAPI:, 4',6-diamidino-2-phenylindole, dihydrochloride; 657
DC: , dendritic cells; DO:, drop out (2DO: SD/-Trp-Leu; 4DO: SD/-Trp-Leu-His-Ade); ECL:, 658
enhanced chemiluminescence; FAf1: Fas associated factor 1; FLT3-ITD: FLT3-internal tandem 659
duplications; FLT3: , feline McDonough sarcoma (fms)-like tyrosine kinase 3; GSK3b:, 660
NLRPlrp12 binds to HCKck
30
glycogen synthase kinase 3 beta; HCK C1 fragment:, HckHCK C-terminal 40 amino acids; 661
HCK: , hHematopoiesis cell kinase; HRP: , horseradish peroxidase; HSP90:, heat shock protein 662
90; IL: interleukin; IPA:, ingenuity pathway analysis; IRAK1:, IL-1 receptor-associated kinase 1; 663
KO:, knock out; KRAS; Ki-ras2 Kirsten rat sarcoma viral oncogene homolog; LRR: leucine rich 664
repeat domain; M1:, the amino acids in the R1 region were mutated to the amino acids identical 665
to the R1 region in Lyn; M2: , the amino acids in the R2 region were all mutated to Alanine; 666
M3: , the amino acids in the R3 region were all mutated to Alanine; M4: , the amino acids in the 667
R4 region were all mutated to Alanine; NBD:, nucleotide binding domain; NF-κB:, nuclear 668
factor kappa-light-chain-enhancer of activated B cells; NIK:, NF-κB-inducing kinase; NLR: , 669
nucleotide binding domain and leucine-rich repeat; NLRP: , NLR pyrin domain-containing 670
protein; PAMP:, pathogen-associated molecular patterns; PBS: phosphate buffered saline; PCR:, 671
polymerase chain reaction; PDB: , protein data bank; PMA: , phorbol 12-myristate 13-acetate; 672
PRR: , pattern recognition receptors; PVDF: , polyvinylidene difluoride; PYD:, pyrin domain; 673
R1, R2, R3, and R4:, self-named four regions in the HckHCK C terminal 40 amino acids 674
fragment; SD:, synthetic dropout; SD/-Trp-Leu:, SD base plus amino acids supplements minus 675
tryptophan and leucine; SD/-Trp-Leu-His-Ade:, SD base plus amino acids supplements minus 676
tryptophan, leucine, and histidine, and adenine; STRING: search tool for recurring instances of 677
neighboring genes; TBST: tris-buffered saline with Tween 20, pH=8.0; TCA:, trichloroacetic 678
acid; TCGA:, the cancer genome atlas; TLR:, toll-like receptor; TRAF3: , tumor necrosis factor 679
receptor associated factor 3; TRAF3IP3:, TRAF3 interacting protein 3; UBA:, ubiquitin-680
associated domains; WT: wild type; YPDA:, yeast extract, peptone, dextrose, and adenine 681
hemisulfate. 682
683
NLRPlrp12 binds to HCKck
31
Conflict ofCompeting Interest 684
The authors have declared that no competing conflict exists. of interest regarding publication of 685
this article. 686
687
Author Contributions Statement 688
YZ design the experiment and wrote and revised the manuscript. CO helped to design the 689
experiments and to revise the manuscript. 690
691
References 692
1. Takeuchi, O., and Akira, S. (2010) Pattern recognition receptors and inflammation. Cell. 693
2010; 140:, 805-820. 694
2. Tuncer, S., Fiorillo, M. T., and Sorrentino, R. (2014) The multifaceted nature of NLRP12. 695
J Leukoc Biol. 2014; 96:, 991-1000. 696
3. Allen, I. C., Wilson, J. E., Schneider, M., Lich, J. D., Roberts, R. A., Arthur, J. C., 697
Woodford , R. M., Davis, B. K., Uronis, J. M,., Herfarth, H. H., Jobin, C., Rogers, A. B., 698
and Ting , J. P. (2012) NLRP12 suppresses colon inflammation and tumorigenesis 699
through the negative regulation of noncanonical NF-kappaB signaling. Immunity. 2012; 700
36:, 742-754. 701
4. Zaki, M. H., Vogel, P., Malireddi, R. K., Body-Malapel, M., Anand, P. K., Bertin, J., 702
Green, D. R., Lamkanfi, M., and Kanneganti, T. D. (2011) The NOD-like receptor 703
NLRP12 attenuates colon inflammation and tumorigenesis. Cancer Cell. 2011; 20:, 649-704
660. 705
5. Zaki, M. H., Man, S. M., Vogel, P., Lamkanfi, M., and Kanneganti, T. D. (2014) 706
Salmonella exploits NLRP12-dependent innate immune signaling to suppress host 707
defenses during infection. Proc Natl Acad Sci U S A. 2014; 111:, 385-390. 708
6. Chen, S. T., Chen, L., Lin, D. S., Chen, S. Y., Tsao, Y. P., Guo, H., Li, F. J., Tseng, W. 709
T., Tam, J. W., Chao, C. W., Brickey , W. J,., Dzhagalov, I., Song, M. J., Kang, H. R., 710
Jung, J. U., and Ting, J. P. (2019) NLRP12 rRegulates aAnti-viral RIG-I aActivation via 711
iInteraction with TRIM25. Cell Host Microbe. 2019; 25:, 602-616.e607. 712
7. Chen, L., Wilson, J. E., Koenigsknecht, M. J., Chou, W. C., Montgomery, S. A., Truax, 713
A. D., Brickey, W. J., Packey, C. D., Maharshak, N,., Matsushima, G. K., Plevy, S. E., 714
Young, V. B., Sartor, R. B., and Ting, J. P. (2017) NLRP12 attenuates colon 715
inflammation by maintaining colonic microbial diversity and promoting protective 716
commensal bacterial growth. Nat Immunol. 2017; 18:, 541-551. 717
8. Truax, A. D., Chen, L., Tam, J. W., Cheng, N., Guo, H., Koblansky, A. A., Chou, W. C., 718
Wilson, J. E., Brickey, W. J., Petrucelli, A., Liu, R., Cooper, D. E., Koenigsknecht, M. J,., 719
Young, V. B., Netea, M. G., Stienstra, R., Sartor, R. B., Montgomery, S. A., Coleman, R. 720
NLRPlrp12 binds to HCKck
32
A., and Ting, J. P. (2018) The Inhibitory Innate Immune Sensor NLRP12 mMaintains a 721
tThreshold against oObesity by rRegulating gGut mMicrobiota hHomeostasis. Cell Host 722
Microbe. 2018; 24:, 364-378.e366. 723
9. Ye, Z., Lich, J. D., Moore, C. B., Duncan, J. A., Williams, K. L., and Ting, J. P. (2008) 724
ATP binding by monarch-1/NLRP12 is critical for its inhibitory function. Mol Cell Biol. 725
2008; 28:, 1841-1850. 726
10. Jeru, I., Le Borgne, G., Cochet, E., Hayrapetyan, H., Duquesnoy, P., Grateau , G., Morali, 727
A., Sarkisian, T., and Amselem, S. (2011) Identification and functional consequences of a 728
recurrent NLRP12 missense mutation in periodic fever syndromes. Arthritis Rheum. 729
2011; 63: , 1459-1464. 730
11. Jeru, I., Duquesnoy, P., Fernandes-Alnemri, T., Cochet, E., Yu, J. W., Lackmy-Port-Lis, 731
M., Grimprel, E., Landman-Parker, J., Hentgen, V., Marlin, S., McElreavey, K., Sarkisian, 732
T., Grateau, G., Alnemri, E. ES., and Amselem, S. (2008) Mutations in NALP12 cause 733
hereditary periodic fever syndromes. Proc Natl Acad Sci U S A. 2008; 105: , 1614-1619. 734
12. Arthur, J. C., Lich, J. D., Ye, Z., Allen, I. C., Gris, D., Wilson, J. E., Schneider, M., 735
Roney, K. E., O'Connor, B. P., Moore, C. B., Morrison, A., Sutterwala, F. S., Bertin, J., 736
Koller, B. H., Liu, Z., and Ting, J. P. (2010) Cutting edge: NLRP12 controls dendritic 737
and myeloid cell migration to affect contact hypersensitivity. J Immunol. 2010; 185:, 738
4515-4519. 739
13. Lukens, J. R., Gurung, P., Shaw, P. J., Barr, M. J., Zaki, M. H., Brown, S. A., Vogel, P., 740
Chi, H., and Kanneganti, T. D. (2015) The NLRP12 sSensor nNegatively rRegulates 741
aAutoinflammatory dDisease by mModulating iInterleukin-4 pProduction in T cCells. 742
Immunity. 2015; 42:, 654-664. 743
14. Vladimer, G. I., Weng, D., Paquette, S. W. M., Vanaja, S. K., Rathinam, V. A., Aune, M. 744
H., Conlon, J. E., Burbage, J. J., Proulx, M. K., and Liu, Q. (2012) The NLRP12 745
inflammasome recognizes Yersinia pestis. Immunity. 2012; 37:, 96-107. 746
15. Udden, S. N., Kwak, Y. T., Godfrey, V., Khan, M. A. W., Khan, S., Loof, N., Peng, L., 747
Zhu, H., and Zaki, H.. (2019) NLRP12 suppresses hepatocellular carcinoma via 748
downregulation of cJun N-terminal kinase activation in the hepatocyte. Elife. 2019; 8: 749
e40396. 750
16. Hornick, E. E., Banoth, B., Miller, A. M., Zacharias, Z. R., Jain, N., Wilson, M. E., 751
Gibson-Corley, K. N., Legge, K. L., Bishop, G. A., Sutterwala, F. S., and Cassel, S. L. 752
(2018) NLRP12Nlrp12 mMediates aAdverse nNeutrophil rRecruitment during iInfluenza 753
vVirus iInfection. J Immunol. 2018; 200:, 1188-1197. 754
17. Damiano, J. S., Oliveira, V., Welsh, K., and Reed, J. C. (2004) Heterotypic interactions 755
among NACHT domains: implications for regulation of innate immune responses. The 756
Biochemical journal. 2004; 381:, 213-219. 757
18. Janowski, A. M., Kolb, R., Zhang, W., and Sutterwala, F.F S. (2013) Beneficial and 758
dDetrimental rRoles of NLRs in cCarcinogenesis. Frontiers in immunology. 2013; 4:, 759
370-370. 760
19. Faustin, B., Lartigue, L., Bruey, J. M., Luciano, F., Sergienko, E., Bailly-Maitre, B., 761
Volkmann , N., Hanein, D., Rouiller, I., and Reed, J. C. (2007) Reconstituted NALP1 762
inflammasome reveals two-step mechanism of caspase-1 activation. Mol Cell. 200; 25:, 763
713-724. 764
NLRPlrp12 binds to HCKck
33
20. Martinon, F., Agostini, L., Meylan, E., and Tschopp, J. (2004) Identification of bacterial 765
muramyl dipeptide as activator of the NALP3/cryopyrin inflammasome. Curr Biol. 2004; 766
14:, 1929-1934. 767
21. Kanneganti, T. D., Lamkanfi, M., and Nunez, G. (2007) Intracellular NOD-like receptors 768
in host defense and disease. Immunity. 2007; 27:, 549-559. 769
22. Sharif, H., Wang, L., Wang, W. L., Magupalli, V. G., Andreeva, L., Qiao, Q., Hauenstein, 770
A. V., Wu, Z., Núñez, G., Mao, Y., and Wu, H. (2019) Structural mechanism for NEK7-771
licensed activation of NLRP3 inflammasome. Nature. 2019; 570:, 338-343. 772
23. Hafner-Bratkovič, I., Sušjan, P., Lainšček, D., Tapia-Abellán, A., Cerović, K., Kadunc, 773
L., Angosto-Bazarra, D., Pelegrin, P., and Jerala, R. (2018) NLRP3 lacking the leucine-774
rich repeat domain can be fully activated via the canonical inflammasome pathway. 775
Nature Communications. 2018; 9:, 5182. 776
24. Roversi, F. M., Pericole, F. V., Machado-Neto, J. A., da Silva Santos Duarte, A., 777
Longhini, A. L., Corrocher, F. A., Palodetto, B., Ferro, K. P., Rosa, R. G., Baratti, M. O., 778
Verjovski-Almeida, S., Traina, F., Molinari, A., Botta, M., and Saad, S. T. (2017) 779
Hematopoietic cell kinase (HCK) is a potential therapeutic target for dysplastic and 780
leukemic cells due to integration of erythropoietin/PI3K pathway and regulation of 781
erythropoiesis: HCK in erythropoietin/PI3K pathway. Biochim Biophys Acta Mol Basis 782
Dis. 2017; 1863:, 450-461. 783
25. Lopez, S., Voisset, E., Tisserand, J. C., Mosca, C., Prebet, T., Santamaria, D., Dubreuil, 784
P., and De Sepulveda, P. (2016) An essential pathway links FLT3-ITD, HCK and CDK6 785
in acute myeloid leukemia. Oncotarget. 2016; 7:, 51163-51173. 786
26. Roversi, F. M., Pericole, F. V., Duarte, A. d. S. S., Ferro, K. P., Corrocher, F. A., 787
Palodetto, B., Longhini , A. L., Botta, M., and Saad, S. T. O. (2015) Knockdown of HCK 788
rReduces cCell dDeath and eErythroid dDifferentiation in hHuman CD34<sup>+</sup> 789
hHematopoietic pProgenitor cCells. Blood. 2015; 126:, 2860. 790
27. Normand, S., Waldschmitt, N., Neerincx, A., Martinez-Torres, R. J., Chauvin, C., 791
Couturier-Maillard, A., Boulard, O., Cobret, L., Awad, F., Huot, L., Ribeiro-Ribeiro, A., 792
Lautz, K., Ruez, R., Delacre, M., Bondu, C., Guilliams, M., Scott, C., Segal, A., 793
Amselem, S., Hot, D., Karabina, S., Bohn, E., Ryffel, B., Poulin, L. F., Kufer, T. A., and 794
Chamaillard, M. (2018) Proteasomal degradation of NOD2 by NLRP12 in monocytes 795
promotes bacterial tolerance and colonization by enteropathogens. Nature 796
Communications. 2018; 9:, 5338. 797
28. Linz, B. M., Neely, C. J., Kartchner , L. B., Mendoza, A. E., Khoury, A. L., Truax, A., 798
Sempowski, G., Eitas, T., Brickey, J., Ting , J. P., Cairns, B. A., and Maile, R. (2017) 799
Innate iImmune cCell rRecovery iIs pPositively rRegulated by NLRP12 during 800
eEmergency hHematopoiesis. J Immunol. 2017; 198:, 2426-2433. 801
29. Robbins, S. M., Quintrell, N. A., and Bishop, J. M. (1995) Myristoylation and differential 802
palmitoylation of the HCK protein-tyrosine kinases govern their attachment to 803
membranes and association with caveolae. Mol Cell Biol. 1995; 15:, 3507-3515. 804
30. Carreno, S., Gouze, M. E., Schaak, S., Emorine, L. J., and Maridonneau-Parini, I. (2000) 805
Lack of palmitoylation redirects p59HCKHck from the plasma membrane to 806
p61HCKHck-positive lysosomes. J Biol Chem. 2000; 275:, 36223-36229. 807
31. Guiet, R., Poincloux, R., Castandet, J., Marois, L., Labrousse, A., Le Cabec, V., and 808
Maridonneau-Parini, I. (2008) Hematopoietic cell kinase (HCKHck) isoforms and 809
NLRPlrp12 binds to HCKck
34
phagocyte duties - from signaling and actin reorganization to migration and phagocytosis. 810
Eur J Cell Biol. 2008; 87:, 527-542. 811
32. Sicheri, F., Moarefi, I., and Kuriyan, J. (1997) Crystal structure of the c-SRCSrc family 812
tyrosine kinase HCKHck. Nature. 1997; 385:, 602-609. 813
33. Ma , B., Elkayam, T., Wolfson, H., and Nussinov, R. (2003) Protein–protein interactions: 814
Structurally conserved residues distinguish between binding sites and exposed protein 815
surfaces. Proceedings of the National Academy of Sciences Proc Natl Acad Sci U S A. 816
2003; 100:, 5772. 817
34. Knappe, M., Bodevin, S., Selinka, H. C., Spillmann, D., Streeck, R. E., Chen, X. S,., 818
Lindahl, U., and Sapp, M. (2007) Surface-exposed amino acid residues of HPV16 L1 819
protein mediating interaction with cell surface heparan sulfate. J Biol Chem. 2007; 282:, 820
27913-27922. 821
35. Lich, J. D., Williams, K. L., Moore, C. B., Arthur, J. C., Davis, B. K., Taxman, D. J., and 822
Ting, J. P. (2007) Monarch-1 suppresses non-canonical NF-kappaB activation and p52-823
dependent chemokine expression in monocytes. J Immunol. 2007; 178:, 1256-1260. 824
36. Cerami, E., Gao, J,., Dogrusoz, U., Gross, B. E., Sumer, S. O., Aksoy, B. A., Jacobsen, 825
A., Byrne, C. J., Heuer, M. L., Larsson, E., Antipin, Y., Reva, B., Goldberg, A. P., 826
Sander, C., and Schultz, N. (2012) The cBio Cancer Genomics Portal: An oOpen 827
pPlatform for eExploring mMultidimensional cCancer gGenomics dData. Cancer 828
Discovery. 2012; 2:, 401. 829
37. Gao, J., Aksoy, B. A., Dogrusoz, U., Dresdner, G., Gross, B., Sumer, S. O., Sun, Y., 830
Jacobsen, A., Sinha, R., Larsson, E., Cerami, E., Sander, C., and Schultz, N. (2013) 831
Integrative analysis of complex cancer genomics and clinical profiles using the 832
cBioPortal. Sci Signal. 2013; 6:, pl1. 833
38. Cancer Genome Atlas Research Network. (2013) Genomic and eEpigenomic lLandscapes 834
of aAdult dDe nNovo aAcute mMyeloid lLeukemia. NEJMNew England Journal of 835
Medicine. 2013, 368:, 2059-2074. 836
39. Williams, K. L., Lich, J. D., Duncan, J. A., Reed, W., Rallabhandi, P., Moore, C., Kurtz, 837
S., Coffield , V. M., Accavitti-Loper, M. A., Su, L., Vogel, S. N., Braunstein, M., and 838
Ting, J. P. (2005) The CATERPILLER protein monarch-1 is an antagonist of toll-like 839
receptor-, tumor necrosis factor alpha-, and Mycobacterium tuberculosis-induced pro-840
inflammatory signals. J Biol Chem. 2005; 280: , 39914-39924. 841
40. Arthur, J. C., Lich, J. D., Aziz, R. K., Kotb, M., and Ting, J. P. (2007) Heat shock protein 842
90 associates with monarch-1 and regulates its ability to promote degradation of NF-843
kappaB-inducing kinase. J Immunol. 2007; 179:, 6291-6296. 844
41. Pinheiro, A. S., Eibl, C., Ekman-Vural, Z., Schwarzenbacher, R., and Peti, W. (2011) The 845
NLRP12 pyrin domain: structure, dynamics, and functional insights. Journal of molecular 846
biology. 2011; 413:, 790-803. 847
42. Kinoshita, T., Kondoh, C., Hasegawa, M., Imamura, R., and Suda, T. (2006) Fas-848
associated factor 1 is a negative regulator of PYRIN-containing Apaf-1-like protein 1. Int 849
Immunol. 2006; 18:, 1701-1706. 850
43. Shi , C., Miley , J., Nottingham, A., Morooka, T., Prosdocimo, D. A., and Simon, D. I. 851
(2019) Leukocyte integrin signaling regulates FOXP1 gene expression via FOXP1-IT1 852
long non-coding RNA-mediated IRAK1 pathway. Biochim Biophys Acta Gene Regul 853
Mech. 2019; 1862:, 493-508. 854
NLRPlrp12 binds to HCKck
35
44. Fu, G., Xu , Q., Qiu, Y., Jin, X., Xu, T., Dong, S., Wang, J., Ke, Y., Hu, H., Cao, X., 855
Wang, D., Cantor , H., Gao, X., and Lu, L. (2017) Suppression of Th17 cell 856
differentiation by misshapen/NIK-related kinase MINK1. J Exp Med. 2017; 214:, 1453-857
1469. 858
45. Wang, Q., Liu, F., Zhang, M., Zhou, P., Xu, C., Li, Y., Bian, L., Liu, Y., Yao, Y., Wang, 859
F., Fang, Y., and Li, D. (2018) NLRP12 pPromotes mMouse nNeutrophil 860
dDifferentiation through rRegulation of nNon-canonical NF-kappaB and MAPK(ERK1/2) 861
sSignaling. Int J Biol Sci. 2018; 14:, 147-155. 862
46. Kinnaird, J. H., Singh, M., Gillan, V., Weir, W., Calder, E. D., Hostettler, I., Tatu, U., 863
Devaney, E,., and Shiels, B. R. (2017) Characterization of HSP90 isoforms in 864
transformed bovine leukocytes infected with Theileria annulata. Cell Microbiol. 2017; 19: 865
e12669. 866
47. Ryu, S. W., Lee, S. J., Park, M. Y., Jun, J. I., Jung, Y. K., and Kim, E. (2003) Fas-867
associated factor 1, FAF1, is a member of Fas death-inducing signaling complex. J Biol 868
Chem. 2003; 278:, 24003-24010. 869
48. Martin, B. N., Wang, C., Zhang, C. J., Kang, Z., Gulen, M. F., Zepp, J. A., Zhao, J., Bian, 870
G., Do, J. S., Min, B., Pavicic, P. G., Jr., El-Sanadi, C., Fox, P. L., Akitsu, A., Iwakura , 871
Y., Sarkar, A., Wewers, M. D., Kaiser, W. J., Mocarski, E. S., Rothenberg, M. E., Hise, 872
A. G., Dubyak, G. R., Ransohoff, R. M., and Li, X. (2016) T cell-intrinsic ASC critically 873
promotes T(H)17-mediated experimental autoimmune encephalomyelitis. Nat Immunol. 874
2016; 17:, 583-592. 875
49. Prada-Arismendy, J., Arroyave, J. C., and Rothlisberger, S. (2017) Molecular biomarkers 876
in acute myeloid leukemia. Blood Rev. 2017; 31:, 63-76. 877
50. Poh, A. R., O'Donoghue, R. J., and Ernst, M. (2015) Hematopoietic cell kinase (HCK) as 878
a therapeutic target in immune and cancer cells. Oncotarget. 2015; 6:, 15752-15771. 879
51. Dos Santos, C., Demur, C., Bardet, V., Prade-Houdellier, N., Payrastre, B., and Récher, C. 880
(2008) A critical role for LYNLyn in acute myeloid leukemia. Blood. 2008; 111:, 2269. 881
52. Greenbaum, D., Colangelo, C,., Williams, K., and Gerstein, M. (2003) Comparing 882
protein abundance and mRNA expression levels on a genomic scale. Genome Biol. 2003; 883
4:, 117. 884
53. Vogel, C., and Marcotte, E. M. (2012) Insights into the regulation of protein abundance 885
from proteomic and transcriptomic analyses. Nat Rev Genet. 2012; 13:, 227-232. 886
54. Liu, Y., Beyer , A., and Aebersold, R. (2016) On the dDependency of cCellular pProtein 887
lLevels on mRNA aAbundance. Cell. 2016; 165:, 535-550. 888
55. Prochnicki, T., and Latz, E. (2017) Inflammasomes on the cCrossroads of iInnate 889
iImmune rRecognition and mMetabolic cControl. Cell Metab. 2017; 26:, 71-93. 890
56. Saito, Y., Yuki, H., Kuratani, M., Hashizume, Y., Takagi, S., Honma , T., Tanaka, A., 891
Shirouzu , M., Mikuni, J., Handa, N., Ogahara, I., Sone, A., Najima, Y., Tomabechi, Y., 892
Wakiyama, M., Uchida, N., Tomizawa-Murasawa, M., Kaneko, A., Tanaka, S,., Suzuki, 893
N., Kajita , H., Aoki , Y., Ohara, O., Shultz, L. D., Fukami, T., Goto, T., Taniguchi, S., 894
Yokoyama, S., and Ishikawa, F. (2013) A pyrrolo-pyrimidine derivative targets human 895
primary AML stem cells in vivo. Sci Transl Med. 2013; 5:, 181ra152. 896
57. Patel, R. K., Weir, M., Hellwig, S., Dorman, H., and Smithgall, T. E. (2017) Abstract 897
2363: Targeting Src-family kinases to combat acquired inhibitor resistance in FLT3-898
ITD<sup>+</sup> AML. Cancer Research 77, 2363-2363. 899
NLRPlrp12 binds to HCKck
36
5758. Pene-Dumitrescu, T., Peterson, L. F., Donato, N. J., and Smithgall, T. E. (2008) An 900
inhibitor-resistant mutant of HCKHck protects CML cells against the antiproliferative 901
and apoptotic effects of the broad-spectrum c-SRCSrc family kinase inhibitor A-419259. 902
Oncogene. 2008; 27:, 7055-7069. 903
5859. Naganna , N., Opoku-Temeng, C., Choi, E. Y., Larocque, E., Chang, E. T., Carter-904
Cooper, B. A., Wang, M., Torregrosa-Allen, S. E., Elzey, B. D., Lapidus, R. G., and 905
Sintim, H. O. (2019) Amino alkynylisoquinoline and alkynylnaphthyridine compounds 906
potently inhibit acute myeloid leukemia proliferation in mice. EBioMedicine. 2019; 40:, 907
231-239. 908
5960. Hu , Y., Liu , Y., Pelletier, S., Buchdunger, E., Warmuth, M., Fabbro, D., Hallek, M., 909
Van Etten, R. A., and Li, S. (2004) Requirement of c-SRCSrc kinases LYNLyn, 910
HCKHck and FGRFgr for BCR-ABL1-induced B-lymphoblastic leukemia but not 911
chronic myeloid leukemia. Nat Genet. 2004; 36:, 453-461. 912
601. Hobbs, S., Jitrapakdee, S., and Wallace, J. C. (1998) Development of a bBicistronic 913
vVector dDriven by the hHuman pPolypeptide cChain eElongation fFactor 1α pPromoter 914
for cCreation of sStable mMammalian cCell lLines tThat eExpress vVery hHigh lLevels 915
of rRecombinant pProteins. Biochem Biophys Res Commun. Biochemical and 916
Biophysical Research Communications1998; 252:, 368-372. 917
612. Brakke, M. K. (1951) Density Gradient Centrifugation: A New Separation Technique1. 918
Journal of the American Chemical SocietyJACS. 1951; 73:, 1847-1848. 919
623. Schindelin, J., Arganda-Carreras, I., Frise, E., Kaynig, V., Longair, M., Pietzsch, T., 920
Preibisch, S., Rueden, C., Saalfeld, S., Schmid, B., Tinevez, J. Y., White, D. J., 921
Hartenstein, V., Eliceiri, K., Tomancak, P., and Cardona, A. (2012) Fiji: an open-source 922
platform for biological-image analysis. Nat Methods. 2012; 9:, 676-682. 923
634. Cerami, E., Gao, J., Dogrusoz, U., Gross, B. E., Sumer, S. O., Aksoy, B. A., Jacobsen, A., 924
Byrne, C. J,., Heuer, M. L., Larsson, E., Antipin, Y., Reva, B., Goldberg, A. P., Sander, 925
C., and Schultz, N. (2012) The cBio cancer genomics portal: an open platform for 926
exploring multidimensional cancer genomics data. Cancer Discov. 2012; 2: , 401-404. 927
928
Table 1: yeast two hybrid results (hit frequency >= 3) 929
Name Frequency
Deleted in azoospermia (DAZ) associated protein 2 (DAZAP2) 10
Filamin A, alpha (FLNA) 9
Hematopoietic cell kinase (HCK) proto-oncogene, Srcc-SRC family tyrosine kinase
(HCK) 9
Chaperone DnaJ (a.k.a., heat shock protein 40 kD(Hsp40)) homolog, subfamily B,
member 1 (DNAJB1) 9
NLRPlrp12 binds to HCKck
37
Metallothionein 2A (MT2A) 7
Cluster of differentiation 14 molecule (CD14) 6
DEAD (Asp-Glu-Ala-Asp) box helicase 5 (DDX5) 6
Talin 1 (TLN1) 5
Calcium modulating ligand (CAMLG) 4
Squamous cell carcinoma antigen recognized by T cells (SART1) 4
Ras-related nuclear protein (RAN) binding protein 9 (RANBP9) 3
Tumor necrosis factor (TNF) receptor associated factor 3 (TRAF3) interacting protein 3
(TRAF3IP3) 3
Thyroid hormone receptor interactor 4 (TRIP4) 3
930
Table 2: Co-occurrence of mRNA expression level of HckHCK and all of the NlrpNLRP 931
family proteins (from the provisional data, assessed on July 11 2019) 932
HckHCK +
NlrpNLRP family
genesproteins
Co-occurrence (C) or
mutual exclusivity (M)
P-value
NlrpNLRP1 C 0.048
NlrpNLRP2 M 0.546
NlrpNLRP3 C 0.415
NlrpNLRP4 M 0.936
NlrpNLRP5 M 0.877
NlrpNLRP6 M 0.670
NLRPlrp12 binds to HCKck
38
NlrpNLRP7 M 0.475
NlrpNLRP8 M 0.877
NlrpNLRP9 M 0.491
NlrpNLRP10 M 0.626
NlrpNLRP11 M 0.233
NlrpNLRP12 C <0.001
NlrpNLRP13 C 1.000
NlrpNLRP14 M 0.509
933
934
935
Table 3: Primers used in cloning (The underline shows the restriction enzymes) 936
Name Primers
NlrpNLRP12 5’ primer
(cloned into PGBKT7)
GATCTACATATGCTACGAACCGCAGGCAG
NlrpNLRP12 3’ primer
(cloned into PGBKT7)
GATGTAGTCGACTCACCCCTCGTCCAGGATATAGTACATAGCT
NlrpNLRP3 5’ primer
(cloned into pGBKT7)
GATCTACATATGAAGATGGCAAGCACCCGCTGCA
NlrpNLRP3 3’ primer
(cloned into PGBKT7)
GATGTAGGATCCTCACAGCAGGTAGTACATGGCGGCAAAGA
NlrpNLRP8 5’ primer GATCTAGAATTCATGAGTGACGTGAATCCACCCTCTG
NLRPlrp12 binds to HCKck
39
(cloned into pGBKT7)
NlrpNLRP8 3’ primer
(cloned into PGBKT7)
GATGTAGGATCCTCAGAGTCTTTGTGGGAAACAGAGAACA
NlrpNLRP12 5’ (cloned
into pEFIRES-P-3FLAG
and pIRES-V5)
GTCCAGCTTAAGATGCTACGAACCGCAGGCAG
NlrpNLRP12 3’ (cloned
into PIRES-3FLAG and
pEFIRES-P-V5)
CTGAATTCTAGAGCAGCCAATGTCCAAATAAGG
HckHCK p61 5’ (cloned
into pEFIRES-P-3FLAG)
GATCTACTTAAGGCCACCATGGGGGGGCGCTCAAGCTGCGAGG
HckHCK p59 5’ (cloned
into pEFIRES-P-3FLAG)
GATCTACTCGAGGCCACCATGGGGTGCATGAAGTCCAAGTTCC
HckHCK p61/p59 3’
(cloned into pEFIRES-P-
3FLAG)
GTCGATTCTAGATGGCTGCTGTTGGTACTGGCTCTCT
937
Figure legends: 938
Figure 1: (A) Schematic drawing of the domain structure of full-length NlrpNLRP12 (top) and 939
the domains used as the construct for the yeast two-hybrid screen (bottom). PYD: Pyrin domain; 940
NBD: Nucleotide binding domain; and LRR: Leucine rich repeat domain. (B) The domain structures 941
of the two alternatively spliced isoforms of HckHCK, HckHCK p59 and HckHCK p61. (C) 942
Identification of the clones of HckHCK interacting with NlrpNLRP12 (“hits”) obtained from the 943
NLRPlrp12 binds to HCKck
40
yeast two-hybrid screen. Among the nine of the clones positive for HCKHck, one clone encoded 944
for the smallest interacting fragment of HCKck, the C-terminal 40 amino acids of HCKck. The 945
numbers in parentheses indicate the total number of hits obtained in the screen. 946
947
Figure 2: Specificity of binding of NlrpNLRP12 with HCKck. (A) Schematic drawing of 948
HCKck (p61 isoform, top) and, for comparison, the region of HCKck (“HCKck C1 fragment,” 949
lower) used for testing the specificity of binding of HCKck to other members of the NlrpNLRP 950
family of proteins. The numbers represent the amino acid residues of the protein or protein 951
fragment (B) Phylogenetic tree showing the relationships among the PYD + NBD region of all of 952
the members of the human NlrpNLRP family of proteins. The phylogenetic tree was produced 953
through Clustal Omega (https://www.ebi.ac.uk/Tools/msa/clustalo/). For each NlrpNLRP family 954
member, the PYD + NBD of the longest isoform was selected for generating the phylogenetic 955
tree. NlrpNLRP3 is phylogenetically the closest NlrpNLRP to NlrpNLRP12, while NlrpNLRP8 956
is phylogenetically among the most distant from NlrpNLRP12. NlrpNLRP3 and NlrpNLRP8 957
arewere highlighted in red color. (C) When NlrpNLRP12’s PYD + NBD and HCKck C1 958
fragment were co-transformed into yeast, they grew on the highest stringency plates (4DO + 959
3AT), consistent with an interaction between the two constructs. On the other hand, neither 960
NlrpNLRP3’s PYD + NBD co-transformed with the C1 fragment nor NlrpNLRP8’s PYD + NBD 961
co-transformed the C1 fragment grew on the highest stringency plates. These results are 962
consistent with a specific interaction between NlrpNLRP12 and the HCKck C1 fragment, but not 963
with the other NlrpNLRPs. (D) Yeast clones grew on the high stringency plates 4DO + 3AT 964
when HCKck’s C1 fragment and NlrpNLRP12’s PYD + NBD were co-transformed, but they did 965
not grow on the high stringency 4DO + 3AT plates when the C1 fragments of other members of 966
NLRPlrp12 binds to HCKck
41
the Sc-SRCrc family of non-receptor tyrosine kinases were co-transformed with NlrpNLRP12’s 967
PYD + NBD. These results are again consistent with a specific interaction between 968
NlrpNLRP12 and HCKck. 969
970
Figure 3: (A) Structure of the HCKck C1 fragment generated by Chimera from the University 971
of California, San Francisco (version 1.13.1), extracted from its structure within the entire 972
HCKck protein. HCKck C1 fragment was thus divided into four helical domains: R1, R2, R3, 973
and R4. The side chains of amino acids F503, I506, Q507, L510, and D511 are shown. However, 974
it is not clear whether the C1 fragment retains the structural features shown here when it 975
expressed on its own. (B) Primary amino acid sequence alignments showing of the C1 fragments 976
of all members of the Sc-SRCrc non-receptor tyrosine kinase family (40 amino acids for HCKck, 977
BLKlk, LCKck, and LYNyn; 42 amino acids for FGRgr, Sc-SRCrc, YESes, and FYNyn). The 978
black lines indicate the R1, R2, R3, or R4 regions. The arrows denote the four critical amino 979
acids for binding of HCKck C1 fragment to NlrpNLRP12’s PYD + NBD. (C) Phylogenetic tree 980
showing the relationship among all of the C1 fragments of the members of the Sc-SRCrc family 981
of non-receptor tyrosine kinases. The phylogenetic tree was produced through Clustal Omega 982
(https://www.ebi.ac.uk/Tools/msa/ clustalo/). HCKck iswas highlighted in red color. (D) Four 983
sets of distinct mutations were made within the HCKck C1 fragment. M1: the amino acids in the 984
R1 region were mutated to the amino acids identical to the R1 region in LYNyn (highlighted in 985
red color); M2: the amino acids in the R2 region were all mutated to alanine (Ala); M3: the 986
amino acids in the R3 region were all mutated to Ala; and M4: the amino acids in the R4 region 987
were all mutated to Ala. (E) Yeast did not grow on high stringency 4DO + 3AT plates when 988
HCKck C1 fragment with the M3 mutations was co-transformed with NlrpNLRP12’s PYD + 989
NLRPlrp12 binds to HCKck
42
NBD. But yeast can grow on the high stringency 4DO + 3AT plates when the HCKck C1 990
fragment contained either the M1, M2, or M4 mutations, suggesting that the R3 region within the 991
C1 fragment was necessary for binding to NlrpNLRP12 NBD + PYD. (E) Yeast did not grow on 992
high stringency 4DO plates when NlrpNLRP12’s PYD + NBD was co-transformed with HCKck 993
C1 fragment with single Ala substitutions at either F503, I506, Q507, L510, or D511. But yeast 994
grew on high stringency 4DO plates when NlrpNLRP12’s PYD + NBD was co-transformed with 995
HCKck’s C1 fragment in which the following amino acids in the R3 region were individually 996
mutated to Ala: E504, Y505, S508, or V509. Yeast did not grow on either 2DO nor 4DO plates 997
when NlrpNLRP12’s PYD + NBD was co-transformed with Hck C1 fragment in which either 998
E504 or I506 was individually mutated to Ala, making the results with these particular mutants 999
inconclusive. 1000
1001
Figure 4: (A) Top: sSchematic drawing of HCKck truncations comprising the C1 to C7 1002
fragments. Bottom: . (B) yYeast grew on the high stringency 4DO plates and 4DO +3AT plates 1003
when HCKck’s C-terminal 40 amino acid fragment (C1), HCKck’s C-terminal 35 amino acid 1004
fragment (C2), and HCKck’s C-terminal 30 amino acid fragment (C3), individually, were co-1005
transformed with NlrpNLRP12’s PYD + NBD. But yeast did not grow on the high stringency 1006
4DO plates nor 4DO+ 3AT plates when C4 to C7 fragments were co-transformed to with 1007
NlrpNLRP12’s PYD + NBD. The lack of yeast growth upon co-transformation of the C7 1008
fragment and NlrpNLRP12 are not shown. (BC) Top: sSchematic drawing of Hck truncations 1009
comprising the N1 to N5 fragments. Bottom:(D) yYeast grew on the high stringency 4DO plates 1010
and 4DO +3AT plates when HCKck’s C-terminal 40 amino acid fragment (C1) was co-1011
transformed with NlrpNLRP12’s PYD + NBD (control). But yeast did not grow on the high 1012
NLRPlrp12 binds to HCKck
43
stringency 4DO plates nor 4DO+ 3AT plates when N1 and N5 fragments were co-transformed 1013
to with NlrpNLRP12’s PYD + NBD, suggesting that the very C-terminal 5 amino acids are also 1014
necessary for binding of NlrpNLRP12. The colored boxes represent different fragments that 1015
were deleted each time. 1016
1017
Figure 5: NlrpNLRP12 co-immunoprecipitates with HCKck and co-localizes with HCKck by 1018
immunofluorescence. (A) NlrpNLRP12 co-immunoprecipitated with HCKck when both proteins 1019
were transiently exogenously co-expressed in 293T cells, with NlrpNLRP12 having to be 1020
expressed as epitope-tagged forms. Cells were lysed using a non-denaturing cell lysis buffer (see 1021
Material and Methods). Co-immunoprecipitations were done using a mouse monoclonal anti-1022
HCKck antibody followed by protein A/G agarose. The immunoprecipitated proteins were 1023
analyzed by immunoblotting. The same method was used for experiments shown in (B), (C), 1024
and (D). (B) NlrpNLRP12 co-immunoprecipitated with endogenous HCKck in macrophage-like 1025
RAW 264.7 cells, when NlrpNLRP12 is exogenously and stably expressed. Arrows show the 1026
two isoforms of HCKck p59 and p61 that were immunoprecipitated by the anti-HCKck antibody, 1027
although it is not clear whether NlrpNLRP12 co-immunopreciptitated with one form or the other, 1028
or both. (C) Exogenously and stably expressed NlrpNLRP12 co-immunoprecipitated with 1029
endogenously expressed HCKck in monocyte-like THP-1 cells. In these cells, phorbol 12-1030
myristate 13-acetate (PMA) (1µM) was added to enhance the expression of NlrpNLRP12 under 1031
the PMA-sensitive cytomegalovirus promoter of the NlrpNLRP12 plasmid. (D) Left panel 1032
shows that NlrpNLRP12 was exogenously and stably expressed in lymphoblast-like U937 cells. 1033
Right panel shows that NlrpNLRP12 co-immunoprecipitated with endogenously expressed 1034
HCKck in U937 cells after PMA was added to 1µM, as was done for THP-1 cells, to enhance 1035
NLRPlrp12 binds to HCKck
44
expression of NlrpNLRP12. (E) Immunofluorescent images showing the colocalization of a 1036
cohort of NlrpNLRP12 with either HCKck p59 or p61. HCKck was immunolabeled with a 1037
mouse monoclonal anti-HCKck antibody, followed by a secondary goat anti-mouse IgG (H+L) 1038
cross-adsorbed, conjugated to Alexa Fluor 488. For NlrpNLRP12, the primary antibody was a 1039
rabbit monoclonal anti-FLAG antibody, and the secondary antibody was a goat anti-rabbit IgG 1040
(H+L) cross-adsorbed secondary antibody, conjugated to Alexa Fluor 555. Nuclei were 1041
counterstained by DAPI. 1042
1043
Figure 6: The co-expression of HCKck protein significantly decreased the steady-state protein 1044
expression levels of NlrpNLRP12 (A) Epitope-tagged NlrpNLRP12-V5 and HCKck p61 were 1045
co-transfected into 293T cells. Cell lysates were immunoblotted with rabbit monoclonal anti-V5 1046
antibody for NlrpNLRP12 and mouse monoclonal anti-HCKck antibody. Comparing lane 2 1047
(basal level of NlrpNLRP12 expression) with lanes 5, 6, and 7, co-expression of HCK ck protein 1048
leads to a significant decrease of protein expression levels of NlrpNLRP12 over three times they 1049
are they co-expressed. Increasing protein expression of HCKck, relative to the plasmid DNA 1050
added (ramp: lanes 5, 6, and 7; 2.5-20 μg of DNA) was not observed possibly because the 1051
amount of plasmid for HCKHck transfection were already too high, resulting in the maximum 1052
expression of HCKck’s at all plasmid DNA amounts. The figure is a representative of one of 1053
threethe results from three experiments. (B) Quantification of the NlrpNLRP12 protein bands 1054
(mean + standard error) to show that NlrpNLRP12 protein expression levels decreased in the 1055
presence of co-expressed HCKck, relative to equal amounts of cell lysate being assayed. The 1056
NlrpNLRP12 protein expression levels are indicated as arbitrary units. (C) Quantification of the 1057
HCKck protein bands (mean + standard error) confirming that HCKck protein expression levels 1058
NLRPlrp12 binds to HCKck
45
did not change although the amount of plasmid DNA for HCK Hck increased. Figure (B) and 1059
(C) were generated by Prism 6.0 (GraphPad Software, Inc., San Diego, CA). Twenty μg of cell 1060
lysate proteins were loaded for each lane. ANOVA and Tukey tests were used to compare the 1061
NlrpNLRP12 and HCKck protein bands. 1062
1063
Figure 7: NlrpNLRP12 is co-expressed with Hck in acute myeloid leukemia (AML) patient 1064
samples. Both NlrpNLRP12 expression and Hck expression data are shown as log2 values. The 1065
data are RNA seq data obtained from cBioportal (provisional) database assessed on July 11, 1066
2019. The graph was taken from the cBioportal website. Spearman coefficient, indicating the 1067
association of NlrpNLRP12 and Hck mRNA expression level (if NlrpNLRP12 and Hck mRNA 1068
expression levels are non-parametric) = 0.80. Pearson coefficient, indicating the degree of the 1069
linear relationship between the NlrpNLRP12 mRNA expression level and Hck mRNA expression 1070
level (if NlrpNLRP12 and Hck mRNA expression levels are parametric) = 0.73. The p-values for 1071
Spearman coefficient and Pearson coefficient are both < 0.001. 1072
1073
Figure 8: (A) Ingenuity Pathway Analysis (IPA) displaying the 13 distinctive hits from the total 1074
results of the yeast two-hybrid screen for proteins that interacted with NlrpNLRP12. The figure 1075
only shows the NlrpNLRP12-interacting proteins with hits that have a frequency larger than 2. 1076
The solid lines indicate interacting proteins found through our yeast two-hybrid screen. The 1077
dashed lines are the NlrpNLRP12-interacting proteins previously reported in published data. 1078
The purple-colored molecules are those associated with AML. (B) To validate an interaction 1079
shown in the IPA diagram, tumor necrosis factor receptor-associated factor 3 interacting protein 1080
3 (TRAF3IP3)-3FLAG and NlrpNLRP12-V5 were co-immunoprecipitated from co-transfected 1081
NLRPlrp12 binds to HCKck
46
293T cells. A non-denaturing cell lysate buffer (see Material and Methods) was used to lyse the 1082
cell, and co-immunoprecipitations were performed with a mouse monoclonal anti-TRAF3IP3 1083
antibody followed by protein A/G agarose. The immunoprecipitates were analyzed for 1084
NlrpNLRP12 by immunoblotting with rabbit monoclonal anti-V5 antibody. The results from the 1085
co-immunoprecipitation are consistent with a confirmation of an interaction between TRAF3IP3 1086
and NlrpNLRP12. 1087
1088
1089
1090
1091
1092
NLRPlrp12 binds to HCKck
47
1093
1094
NLRPlrp12 binds to HCKck
48
1095
Figure 1 1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
NLRPlrp12 binds to HCKck
49
1106
NLRPlrp12 binds to HCKck
50
1107
Figure 2 1108
1109
1110
1111
NLRPlrp12 binds to HCKck
51
1112
1113
Figure 3 1114
NLRPlrp12 binds to HCKck
52
1115
Figure 3 1116
NLRPlrp12 binds to HCKck
53
1117
1118
1119
NLRPlrp12 binds to HCKck
54
1120
Figure 4 1121
1122
NLRPlrp12 binds to HCKck
55
1123
NLRPlrp12 binds to HCKck
56
1124
Figure 5 1125
1126
NLRPlrp12 binds to HCKck
57
1127
NLRPlrp12 binds to HCKck
58
1128
Figure 6 1129
NLRPlrp12 binds to HCKck
59
1130
NLRPlrp12 binds to HCKck
60
1131
Figure 7 1132
1133
1134
1135
1136
1137
1138
1139
1140
NLRPlrp12 binds to HCKck
61
1141
NLRPlrp12 binds to HCKck
62
1142
Figure 8 1143
1144
1
Nucleotide binding domain and leucine-rich repeat pyrin domain-
containing protein 12: characterization of its binding to
hematopoietic cell kinase
Yue Zhang* and Curtis T. Okamoto 1
2
Department of Pharmacology and Pharmaceutical Sciences, School of Pharmacy, University of 3
Southern California, USA 90089-9121 4
5
*Correspondence: 6
Yue Zhang 7
zhan576@usc.edu 8
9
Running title: NLRP12 binds to HCK 10
11
Key words: yeast two-hybrid, innate immunity, protein‐protein interaction, NLRP12, HCK, 12
acute myeloid leukemia13
14
15
Abstract 16
Protein-protein interactions are key to defining the function of nucleotide binding domain (NBD) 17
and leucine-rich repeat (LRR), pyrin domain (PYD)-containing family of proteins 12 (NLRP12). 18
cDNA encoding the human PYD + NBD of NLRP12 was used as bait in a yeast two-hybrid 19
NLRP12 binds to HCK
2
screen with a human leukocyte cDNA library as prey. Hematopoiesis cell kinase (HCK), a 20
member of the c-SRC family of non-receptor tyrosine kinases, was among the top hits. The C-21
terminal 40 amino acids of HCK specifically bound to NLRP12’s PYD + NBD, but not to that of 22
NLRP3 and NLRP8. Amino acids F503, I506, Q507, L510, and D511 of HCK are critical for the 23
binding of HCK’s C-terminal 40 amino acids to NLRP12’s PYD + NBD. Additionally, the C-24
terminal 30 amino acids of HCK are sufficient to bind to NLRP12’s PYD + NBD, but not to its 25
PYD alone nor to its NBD alone. In cell lines that express HCK endogenously, it was co-26
immunoprecipitated with stably expressed exogenous NLRP12. Also, NLRP12 co-27
immunoprecipitated and co-localized with HCK when both were overexpressed in 293T cells. In 28
addition, in this overexpression system, steady-state NLRP12 protein expression levels 29
significantly decreased when HCK was co-expressed. Bioinformatic analysis showed that HCK 30
mRNA co-occurred with NLRP12 mRNA, but not with other NLRP mRNAs, in blood and 31
marrow samples from acute myeloid leukemia (AML) patients. The mRNA of NLRP12 is also 32
co-expressed with HCK in AML patient samples, and the levels of mRNA expression of each are 33
correlated. Together these data suggest that NLRP12, through its binding to HCK, may have an 34
effect on the progression of AML. 35
36
Introduction 37
Pattern recognition receptors (PRR) are the first major line of cellular defense against damage-38
associated molecular patterns (DAMP) and pathogen-associated molecular patterns (PAMP) (1). 39
Among PRR, one class of proteins is the nucleotide-binding domain (NBD) and leucine-rich 40
repeat (LRR) receptor family (NLR). NBD and LRR PYD protein 12 (NLRP12) is one protein 41
that belongs to the PYD subfamily of the NLR class and is a multi-functional protein (2). 42
NLRP12 binds to HCK
3
NLRP12’s functions likely all depend on defined protein-protein interactions. For example, in 43
anti-inflammatory functions of NLRP12, NLRP12 interacts with tumor necrosis factor receptor 44
associated factor 3 (TRAF3) and nuclear factor kappa-light-chain-enhancer of activated B cells 45
(NF-κB)-inducing kinase (NIK) in the non-canonical NF-κB pathway, when toll-like receptors 46
(TLR) are stimulated with Pam3Cys4 triacylated lipopeptides followed by addition of cluster of 47
differentiation 40 (CD40) (3). Kanneganti’s group also showed that NLRP12 expression 48
downregulated the canonical NF-kB pathway during Salmonella infection (4,5). NLRP12’s 49
inhibition of the non-canonical NF-κB pathway, thereby effecting an anti-inflammatory function, 50
has been reported in many publications (7-11). For example, in colitis-driven colon cancer, 51
Nlrp12 knock-out (KO) mice experience a higher frequency of death, weigh less, have a shorter 52
colon length, and have more severe disease progression than wild type (WT) mice, all of which 53
may occur through NLRP12’s inhibition of the non-canonical NF-κB pathway (3,4). In addition, 54
Nlrp12 KO mice have been reported to exhibit higher innate immunity when it is infected by the 55
virus because NLRP12 inhibits the non-canonical NF-κB pathway; thus, in a KO mice, a more 56
significant increase of NF- κB activity is observed compared to WT mice when it is infected by 57
the virus (6). And in the case of NLRP12 protecting the diversity of intestinal microbiota (7), 58
thereby preventing obesity (8), it has also been reported that Nlrp12 KO mice display 59
significantly decreased intestinal microbial diversity due to increased innate immunity, resulting 60
from the lack of inhibition of the non-canonical NF-κB pathway. 61
62
In addition, NLRP12 also interacts with interleukin (IL)-1 receptor-associated kinase 1 (IRAK1) 63
in the canonical NF-κB pathway when TLRs are stimulated, also to inhibit the inflammatory 64
response (9). In contrast, for human studies of NLRP12-associated autoinflammatory disorders, 65
NLRP12 binds to HCK
4
no protein-protein interactions have been reported (10,11). Functionally, however, Nlrp12 KO 66
mice had a higher immunological response in the autoimmune model of contact hypersensitivity 67
(12); thus, these mice are more prone to develop contact hypersensitivity. This results most likely 68
from Nlrp12 KO mice having defective migration of their dendritic cells (DC). In addition, 69
Luken et al. demonstrated that NLRP12 impacts the experimental autoimmune 70
encephalomyelitis (EAE) model via regulating T-cell responses (13). But, in these mouse studies, 71
no protein-protein interactions had been reported either (12). And NLRP12 is protective against 72
Yersinia pestis infection through NLRP12’s activation of the inflammasome which results in 73
secretion of IL-18 and IL-1β. But, no protein-protein interactions of NLRP12 within 74
inflammasomes had been characterized (14). Also, NLRP12 also regulates the c-Jun N-terminal 75
kinase (JNK) pathway in hepatocytes (15); but, again, no protein-protein interactions within the 76
NLRP12-JNK pathway had been demonstrated. Thus, there is a further need to find and to 77
characterize previously unreported NLRP12-binding proteins. Knowing NLRP12’s protein 78
binding partners may help to gain mechanistic insight into pathways previously characterized as 79
being regulated by NLRP12. Moreover, identifying and validating potentially novel binding 80
partners of NLRP12 may suggest novel functions for NLRP12. 81
82
In humans, NLRP12 is mainly expressed in macrophages, DCs, and neutrophils. In mice, 83
NLRP12 is primarily expressed in DCs and neutrophils (6,12,16). The NLRP12 protein has 84
three domains, from N- to C-terminus: PYD, NBD, and LRR domains (2,17,18). It is generally 85
accepted that the LRR domain, like NLRP1’s LRR domain and NLRP3’s LRR domain, is 86
responsible for sensing ligands or other types of activators of NLRPs in modulating their role in 87
an immunologic response (19-22). However, interestingly, it has been reported that without the 88
NLRP12 binds to HCK
5
LRR domain, NLRP3 still can be activated by triggers of the canonical NF-κB pathway (23). 89
The NBD domain is responsible for self-oligomerization when NLRP is activated, and the PYD 90
domain is responsible for regulating downstream signaling pathways (and thus, it is called the 91
“interaction domain”) (16,23). Here we used a truncated form of NLRP12 (PYD + NBD) to 92
identify possible protein-protein interactions that may regulate signaling pathways in which 93
NLRP12 is involved. In addition, to our knowledge, there are currently no protein-protein 94
interaction studies that have been performed for this truncated form of NLRP12 and may thus 95
represent a new general approach to screening for protein-protein interactions involved in 96
signaling complexes with NLRPs. 97
98
A yeast two-hybrid screen was performed to identify potential binding partners of NLRP12’s 99
PYD + NBD. HCK, a member of the c-SRC family of non-receptor tyrosine kinases, was one of 100
the top positive clones (“hits”) among all of the proteins that appeared to interact with NLRP12. 101
Further experiments were done to confirm the novel binding of NLRP12 to HCK and to 102
characterize key structural features of the binding interaction. In addition, co-expression of HCK 103
with NLRP12 resulted in a significant decrease of steady-state NLRP12 protein expression levels. 104
Finally, together with a bioinformatics analysis included here, a potential interaction between 105
NLRP12 and HCK may be involved in modulating the role of HCK in the pathogenesis of AML 106
(24-26). 107
108
Results 109
NLRP12 interacts with HCK in the yeast two-hybrid assay 110
NLRP12 binds to HCK
6
A human leukocyte cDNA library was used as prey in the screening of binding partners for 111
NLRP12 using the yeast two-hybrid assay. The basic principle of the yeast two-hybrid assay is 112
that if the cDNA library contains clones that express proteins, or parts thereof, that interact with 113
the co-expressed bait, the NLRP12 PYD + NBD (Figure 1A) in this case, then the co-114
transformed yeast clones will grow. The cDNA clones from the library that supported growth 115
would then be isolated and characterize, to identify what proteins interact with NLRP12’s PYD + 116
NBD. The leukocyte library was used because NLRP12 is mainly expressed in leukocytes in 117
humans (2,16,27,28), and thus putative binding partners for NLRP12 are also likely to be 118
expressed there. The screening resulted in 119 positive clones from the leukocyte cDNA library 119
(“hits”) expressed from 48 distinct genes (Supplementary materials, Table S1) that were 120
subsequently identified as encoding for proteins (or portions thereof) that potentially interact 121
with NLRP12’s PYD + NBD. The clones with the most hits are listed in Table 1. Among the 122
119 hits, which included repeated hits of the same genes, 9 were individual clones that encoded 123
fragments of a member of the c-SRC family of non-receptor tyrosine kinases, HCK, which was 124
among the highest number of hits for a single gene (Figure 1B and Table 1). Moreover, of the 9 125
positive HCK clones, 5 encoded distinct HCK fragments (Figure 1C). Full-length HCK p61 was 126
also obtained as a hit in another independent yeast two-hybrid screen (data not shown). HCK 127
has two isoforms, p59 and P61, in Homo sapiens. P61 and p59 share the same transcript, but p61 128
uses an uncommon start codon CTG, and p59 uses the common one, ATG (29), to give rise to 129
two distinct transcripts. Endogenously expressed in phagocytes, HCK p59 is mainly localized to 130
the plasma membrane, while HCK p61 is primarily localized to lysosomes, and both of them are 131
also found in the Golgi apparatus (30). It has been suggested that overexpression of p59 132
promotes the formation of membrane protrusions such as pseudopodia and overexpression of p61 133
NLRP12 binds to HCK
7
induces podosome rosettes and triggers actin polymerization around lysosomes (30,31). 134
However, there are no differences between p61 and p59 in the region of HCK that appears to 135
bind to NLRP12’s PYD + NBD (see below), the C-terminal part of HCK. Thus, detecting a full-136
length clone of only p61 as a hit, opposed to p59, appeared to be coincidental. 137
138
In directed screens using the yeast two-hybrid assay, NLRP12’s PYD + NBD domain 139
binding to HCK C-terminal 40 amino acids appears to be selective for NLRP12, and HCK 140
binding to NLRP12 appears to be selective for HCK 141
Initially, the HCK C-terminal 40 amino acids (“HCK C1 fragment”) was the shortest HCK 142
fragment from a positive clone that interacted with the bait, NLRP12’s PYD + NBD (Figure 2A). 143
NLRP 3 and NLRP8 were chosen based on the phylogenetic trees of the PYD + NBD domains 144
of members of the NLRP family: phylogenetically, NLRP3’s PYD + NBD is the most closely 145
related NLRP PYD + NBD to that of NLRP12, while NLRP8’s PYD + NBD is among the 146
furthest related PYD + NBD to that of NLRP12 (Figure 2B). In Figure 2C, yeast clones grew on 147
the 4 drop out (DO) (synthetic dropout (SD) plus amino acids supplements minus tryptophan, 148
leucine, histidine, and adenine (-Trp-Leu-His-Ade)) and 4DO + 3-Amino-1,2,4-triazole (3-AT) 149
high stringency plates when the yeast were co-transformed with NLRP12’s PYD + NBD and 150
HCK C1 fragment; and thus, this result is indicative of an interaction between the two constructs. 151
However, yeast clones were not observed on 4DO and 4DO + 3-AT plates when NLRP3’s PYD 152
+ NBD and HCK C1 fragment were co-transformed, nor when NLRP8’s PYD + NBD and HCK 153
C1 fragment were co-transformed. These results suggest that the HCK C1 fragment binds to 154
NLRP12’s PYD + NBD, but not to those from either NLRP3 or NLRP8, results consistent with a 155
selective interaction between NLRP12’s PYD + NBD and the HCK C1 fragment. 156
NLRP12 binds to HCK
8
157
Similarly, yeast clones again grew on the high stringency 4DO and 4DO + 3-AT plates when the 158
HCK C1 fragment was co-transformed with NLRP12’s PYD + NBD. However, clones did not 159
grow on the 4DO and 4DO + 3-AT plates when the C1 fragment of other members of the c-SRC 160
family of non-receptor tyrosine kinases were co-transformed with NLRP12’s PYD + NBD 161
(Figure 2D). This experiment suggests that the interaction of HCK C1 fragment with NLRP12 is 162
selective for HCK; the C1 fragment of other members of the c-SRC family of non-receptor 163
tyrosine kinases did not bind to NLRP12’s PYD + NBD (Figure 2D). 164
165
In directed screens using the yeast two-hybrid assay, amino acids F503, I506, Q507, L510, 166
and D511 of HCK are critical for the binding between NLRP12’s PYD + NBD and HCK C-167
terminal 40 amino acids 168
The HCK C1 fragment were divided into four helical regions according to its atomic structure 169
(32) (protein data bank (PDB) accession number: 1AD5): R1, R2, R3, and R4 (designated by us) 170
(Figure 3A). As shown in the sequence alignments in Figure 3B, the primary amino acid 171
sequences of R2, R3, and R4 are very well conserved among members of the c-SRC family of 172
non-receptor tyrosine kinases, while R1 is not. To characterize further the important structural 173
features of HCK’s binding to NLRP12’s PYD + NBD, all of the amino acids comprising each of 174
the R2, R3, and R4 domains were mutated to alanine, but only one section at a time was mutated, 175
and these were denoted as the HCK M2 fragment, HCK M3 fragment, and HCK M4 fragment, 176
respectively. The amino acids comprising the R1 region were mutated to those amino acids in 177
the R1 region of the c-SRC family non-receptor tyrosine kinase LYN, which has the closest 178
phylogenetic relationship to HCK (Figure 3C). All of the mutated constructs are illustrated in 179
NLRP12 binds to HCK
9
Figure 3D. Yeast still grew on the high stringency 4DO + 3AT plates when NLRP12’s PYD + 180
NBD was co-transformed with the HCK M1 fragment, HCK M2 fragment, and HCK M4 181
fragment. But, yeast could not grow when NLRP12’s PYD + NBD was co-transformed with 182
HCK M3 fragment (Figure 3E). These experiments suggested that the R1, R2, and R4 regions 183
are dispensable for binding, while R3 may be important for binding. Further analysis was 184
conducted to identify individual critical amino acids of HCK’s R3 region binding to NLRP12’s 185
PYD + NBD using alanine scanning mutagenesis. However, the intervening sequences between 186
R1, R2, R3, and R4 were not examined for binding. 187
188
To identify the key amino acid residues involved in binding of R3 to NLRP12’s PYD + NBD, 189
each amino acid in R3 was individually mutated to alanine in an alanine scanning mutagenesis 190
analysis, on the background of the HCK C1 fragment. As shown by the yeast two-hybrid assay, 191
the mutations of F503A, I506A, Q507A, L510A, and D511A abolished the growth of yeast 192
clones on the high stringency 4DO when these yeast strains were co-transformed with NLRP12’s 193
PYD + NBD (Figure 3F). These experiments suggest that these four amino acids are critical for 194
binding. On the other hand, yeast clones still grew on the high stringency 4DO plates when the 195
HCK C1 fragments with the E504A, Y505A, S508A, or V509A mutations were co-transformed 196
with NLRP12’s PYD + NBD (Figure 3F). These experiments suggest that E504A, Y505, S508, 197
and V509 are dispensable for binding of HCK C1 fragment to NLRP12’s PYD + NBD. Overall, 198
however, these results suggest that the amino acids F503, I506, Q507, L510, and D511 in HCK 199
are critical for the binding of HCK C1 fragment to NLRP12’s PYD + NBD. 200
201
NLRP12 binds to HCK
10
On the other hand, YES also has these five amino acid residues,(Figure 3A), suggesting that, 202
while they are important for binding, they cannot account for the specificity or selectivity of 203
binding between NLRP12 and HCK. In addition, the amino acids R500, P501, T502, D512, 204
T515, Q520, and Y521, are 100% conserved among all other members of the c-SRC family of 205
non-receptor tyrosine kinases and are therefore unlikely to play a role in specificity and 206
selectivity of binding of NLRP12 with HCK. But, it is possible that two or more amino acids in 207
the intervening sections between the R1-R4 regions, which were not analyzed by mutagenesis, 208
may then actually account for the specificity and selectivity of binding of HCK to NLRP12. One 209
example is that the combination of R496, P497, and Q522 is unique to HCK compared to other 210
members of the c-SRC family of non-receptor tyrosine kinases. 211
212
In directed screens using the yeast two-hybrid assay, the shortest fragment of HCK that 213
binds to NLRP12’s PYD + NBD is its C-terminal 30 amino acids 214
The shortest form of HCK that bound to NLRP12’s PYD + NBD was determined by observing 215
whether the yeast clones grew on high stringency 4DO or 4DO +3AT plates when NLRP12’s 216
PYD + NBD was co-transformed with HCK fragments C1 to C7 and N1 to N5 (Figure 4) (data of 217
NLRP12’s PYD + NBD co-transformed with HCK fragment C7 not shown). HCK’s C1 to C7 218
fragments were gradually truncated from the N-terminus of the HCK C-terminal 40 amino acid 219
fragment (Figure 4A). HCK N1 to N5 fragments were gradually truncated from the C-terminus 220
of the HCK C-terminal 40 amino acid fragment (Figure 4B). Yeast strains grew on high 221
stringency 4DO or 4DO +3AT plates when either the HCK C1, C2, or C3 fragment was co-222
transformed with NLRP12’s PYD + NBD. Thus, of all of the constructs screened, the shortest 223
fragment of HCK that can apparently bind to NLRP12’s PYD + NBD is the C-terminal 30 amino 224
NLRP12 binds to HCK
11
acids, as indicated by the HCK C3 fragment (Figure 4). In addition, the last C-terminal 5 amino 225
acids of the HCK C-terminus also appear to be important for binding, as clones co-transformed 226
with NLRP12’s PYD + NBD with N1, with a deletion of the C-terminal 5 amino acids of HCK, 227
did not grow on the high stringency 4DO and 4DO + 3AT plates (Figure 4B). The structures of 228
both inactive HCK (PDB accession number: 1AD5) and active c-SRC (PDB accession number: 229
1Y57), the prototypical member of this family of nonreceptor tyrosine kinases related to HCK, 230
show that the last C-terminal 4 amino acids of c-SRC are surface-exposed, indicating that these 231
four amino acids that are critical for binding are available to interact with NLRP12 (33,34). 232
233
NLRP12 co-immunoprecipitates with HCK from mammalian cells, and NLRP12 is co-234
localized with HCK in mammalian cells 235
NLRP12 and HCK could be co-immunoprecipitated when they were both heterologously co-236
expressed in 293T cells (Figure 5A). Also, semi-endogenous conditions were used for co-237
immunoprecipitations from cells in which HCK is endogenously expressed, and NLRP12 was 238
heterologously stably expressed in epitope-tagged form, such as in macrophage-like RAW 264.7 239
cells (Figure 5B), monocyte-like THP-1 cells (Figure 5C), and lymphocyte -like U937 cells 240
(Figure 5D). Semi-endogenous conditions have been used in previous reports where NLRP12 241
was shown to interact with TRAF3 (3), NIK (35), and IRAK1(9). In all cases, NLRP12 co-242
immunoprecipitated with the p59 and p61 isoforms of HCK, but from these experiments, it is not 243
clear that whether NLRP12 interacts with p59, p61, or both isoforms of HCK. In U937 cells and 244
THP-1 cells, since heterologous NLRP12 expression level was low, adding phorbol 12-myristate 245
13-acetate (PMA) (1 µM for 24 hr) increased NLRP12 expression level by stimulating the PMA-246
sensitive cytomegalovirus promoter of NLRP12’s expression plasmid (Figure 5C and 5D). 247
NLRP12 binds to HCK
12
Taken together, these data suggest that NLRP12 and HCK (either p59, p61, or both isoforms) 248
can interact with each other in a mammalian cell system, a result that in part validates the 249
findings from the yeast two-hybrid assay. 250
251
After heterologous co-expression of NLRP12 with either the p61 or the p59 isoform of HCK in 252
293T cells, immunofluorescent staining of NLRP12 and HCK showed that NLRP12 can be co-253
localized with either p61 or p59 isoforms of HCK in most of the 293T cells (Figure 5E and 254
Supplementary materials, Figure S1). The co-localization of NLRP12 and HCK is consistent 255
with an intracellular interaction between the two proteins. In addition, the lack of selectivity of 256
NLRP12 co-localizing with either p61 or p59 isoform of HCK is consistent with the results from 257
the screening assay, where the putative binding region of HCK to NLRP12 is identical in both 258
the p61 and p59 isoforms. 259
260
The steady-state level of NLRP12 protein may decrease when HCK protein is co-expressed 261
Figure 6 shows that in 293T cells, the co-expression of increasing amount of HCK may led to a 262
decrease in steady-state protein expression levels of NLRP12. Although, alternatively, the 263
apparent decrease in expression level of NLRP12 might be due actually to a decrease in the 264
epitope-tag itself or a decrease in the expression of the epitope-tagged protein only. Interestingly, 265
HCK protein levels did not show a linearly increasing level of expression upon transfection with 266
increasing amounts of plasmid DNA, possibly due to HCK’s maximum expression levels in 267
these cells being achieved with the lowest amounts of plasmid DNA, which may be resolved if 268
the transfections are performed with lower amounts of plasmid DNA. This lack of linearity in 269
protein expression could also be due to saturation of the signal for HCK on the blot at relatively 270
NLRP12 binds to HCK
13
low levels of HCK expression or a combination of HCK’s maximum expression levels has been 271
reached as well as saturation of the blot, as mentioned above. Alternatively, reduction of 272
NLRP12 expression could be due to plasmid competition; that is, since the expression of both 273
genes from either plasmid are driven from the same promoter, an increase in the amount of one 274
plasmid may competitively decrease expression of the gene from the other plasmid due to their 275
competition for transcription factors. 276
277
In blood and marrow samples from patients with AML, NLRP12 and HCK co-occur and 278
are co-expressed 279
The odds ratio test (a statistical test that defines the likelihood of association of two events) 280
determines whether two proteins either co-occur or are mutually exclusive in their expression. 281
The Fisher exact test is a statistical test that uses P-values to determine if two events, such as co-282
occurrence or mutual exclusivity, are quantitatively significantly associated. Applying these 283
tests to The Cancer Genome Atlas (TCGA) provisional data sets for AML patients (36-38), 284
NLRP12 and NLRP1 are the only two NLRP family members that show a statistically significant 285
co-occurrence with HCK (Table 2). However, NLRP1’s co-occurrence has a P value of 0.048, 286
while NLRP12’s co-occurrence has a P value of <0.001. Therefore, NLRP12 is most likely the 287
only NLRP family mRNAs that shows co-occurrence with HCK in samples from patients with 288
AML. 289
290
Moreover, in this analysis, NLRP12 is significantly co-expressed with HCK (Pearson coefficient 291
= 0.80 and Spearman coefficient = 0.73, Pearson and Spearman coefficients show the possibility 292
of a linear relationship of two factors, Figure 7). This analysis is also consistent with NLRP12 293
NLRP12 binds to HCK
14
and HCK co-occurring in the same patients with AML. However, NLRP12 was not the only 294
NLRP mRNA that shows significant co-expression with HCK. NLRP1 and NLRP3 also showed 295
significant co-expression with HCK. This can be explained by the fact that co-expression is less 296
stringent than co-occurrence. In summary, NLRP12 is the only NLRP protein that shows co-297
occurrence with HCK, and it also shows co-expression with HCK, although it is not the only 298
member of the NLRP family to do so. Other data sets, including Hausera data, Metzeler 1 data, 299
Metzeler2 data, and Raponi data, were not available to evaluate the co-expression of NLRP12 300
versus HCK, as these other data sets did not incorporate data regarding NLRP12 expression. 301
302
Other potential binding partners of NLRP12 from the yeast two-hybrid screen 303
Other potentially interesting and unique genes that whose products may interact with NLRP12’s 304
PYD + NBD were identified in the yeast two-hybrid screen (Table 1 and Supplementary 305
materials, Table S1). Table 1 only shows those hits with frequency larger than 2, and these hits 306
were analyzed through an Ingenuity Pathway Analysis (IPA) diagram (Figure 8). In the IPA 307
diagram, protein-protein interactions found by our yeast two-hybrid screen are connected by 308
solid lines and interactions between protein binding partners collated by IPA are connected by 309
dashed lines. In addition, those molecules that are associated with AML are colored purple. 310
Interestingly, NLRP12 is also associated with AML from the database Catalogue of somatic 311
mutations in cancer (COSMIC). Although the mutation occurs in the NBD of NLRP12, only 2 312
samples of these mutations are observed within 175 samples and may be worthy of follow-up 313
study. 314
315
NLRP12 binds to HCK
15
The interaction of one of the putative binding partners to NLRP12 identified in our screen, 316
TRAF3 interacting protein 3 (TRAF3IP3), was validated in a co-immunoprecipitation 317
experiment where the FLAG-tagged form of TRAF3IP3 and V5-tagged form of NLRP12 were 318
overexpressed in 293T cells. TRAF3IP3 was coimmunoprecipitated by adding mouse 319
monoclonal anti-TRAF3IP3 antibody, and a band that showed the same molecular weight size of 320
NLRP12 was detected by rabbit monoclonal anti-V5 antibody (Figure 8). However, further 321
validation of these other putative interactors detected by the yeast two-hybrid needs to be done. 322
323
Discussion 324
In this paper, the yeast two-hybrid assay was used to screen for NLRP12 binding partners. This 325
approach identified 48 genes as positive hits encoding proteins that could possibly bind to 326
NLRP12. Among the hits, the c-SRC non-receptor tyrosine kinase family member HCK was 327
selected to focus upon as a potentially interesting target, particularly given its association with 328
hematopoiesis and AML. Further direct screens found that NLRP12 specifically binds to HCK, 329
but not to NLRP8 or NLRP3. Conversely, HCK, but not other members of the c-SRC non-330
receptor tyrosine kinase family, specifically binds to NLRP12. HCK’s C-terminal 30 amino 331
acids are sufficient for binding to NLRP12. And, amino acids F503, I506, Q507, L510, and 332
D511 of HCK are critical for binding to NLRP12. Co-immunoprecipitation experiments from 333
mammalian cells confirmed that NLRP12 and HCK interact. Co-expression of NLRP12 and 334
HCK results in a significantly lower steady-state expression level of NLRP12; thus, the 335
interaction may affect NLRP12 protein stability. Finally, a bioinformatics analysis shows that 336
NLRP12 and HCK essentially exclusively co-occur, and mRNA levels for both are correlated in 337
NLRP12 binds to HCK
16
blood samples from AML patients. Together, these data suggested that NLRP12 and HCK 338
protein and mRNA co-expression may have an effect on the AML. 339
340
The binding data from yeast two-hybrid (Figure 3 and Figure 4) showed that the R2 and R3 341
regions are both important for binding, but we suspect that R3 region is the major region that is 342
critical for binding, while R2 region may be an “auxiliary” region that is also critical for binding, 343
because when we delete the R2 region but maintain the R3 region, the R3 region itself cannot 344
maintain stable binding of HCK’s C-terminal 40 amino acids to NLRP12’s PYD + NBD. Thus, 345
binding of NLRP12’s PYD + NBD to this fragment of HCK may occur initially through binding 346
to R3, followed by stabilization of the interaction by secondarily binding to R2. This model is 347
also consistent with the finding that when we mutated the R3 region, the binding of HCK’s C-348
terminal 40 amino acid fragment to the NLRP12’s PYD + NBD was lost despite the R2 region 349
being present. The above evidences are consistent with the model that R2 and R3 regions 350
together are both necessary for binding, but the stabilization of binding to the R3 region may 351
require the presence of R2. 352
353
The yeast two-hybrid screen with NLRP12 also identified products from 47 other genes as 354
potential interactors. According to the Search Tool for Recurring Instances of Neighbouring 355
Genes (STRING) protein network and previous publications, NLRP12 interactors include 356
IRAK1 (39), NIK (35), TRAF3(3), heat shock protein 90 (HSP90) (40), Fas associated factor 1 357
(FAF1) (41), and apoptosis-associated speck-like protein containing a caspase activation and 358
recruitment domain (ASC) proteins. IRAK1, NIK, TRAF3, and Heat shock protein (HSP) 90 359
were only found in experiments involving semi-endogenous protein co-immunoprecipitations 360
NLRP12 binds to HCK
17
with NLRP12. FAF1 was found to be an interacting protein in NMR experiments by showing a 361
difference in the chemical shift perturbation of FAF1’s ubiquitin-associated domains (UBA) 362
domain alone and NLRP12’s PYD domain bound to FAF1’s UBA domain. And FAF1 was also 363
found to interact with NLRP12 in a yeast two-hybrid study where the PYD domain of different 364
NLRP proteins were screened (42). However, in our study, IRAK1 (43), NIK (44), TRAF3 (45), 365
HSP90 (46), FAF1 (47), and ASC (48) proteins were not found in our results of interacting 366
proteins, although they are all expressed in leukocytes. Therefore, given these differences, the 367
results from this yeast two-hybrid screen may provide a unique database for future investigations 368
on the proteins that interact with NLRP12, and the functional outcomes of these interactions. 369
370
How HCK co-expression affects NLRP12 expression can be further investigated to determine the 371
functional outcomes of the NLRP12-HCK interaction relative to protein stability or degradation. 372
For example, MG-132 and chloroquine, which are a proteasomal inhibitor and an inhibitor of 373
lysosomal degradation, respectively, can be added to cells in these experiments to determine 374
whether the decreasing protein expression level of NLRP12 in the presence of HCK is due to 375
proteasomal degradation or lysosomal degradation. In addition, an NLRP12 stable cell line, with 376
a typically lower level of NLRP12 expression, compared to transiently transfected cells, can be 377
used to test formally whether NLRP12 protein expression levels are increasingly decreased when 378
the expression of HCK is increased. Moreover, the rate of NLRP12 degradation, in the absence 379
or presence of HCK, can be studied and compared in pulse-chase assays (35). 380
381
Given the interaction that was characterized between NLRP12 and HCK, we sought to identify 382
diseases in which their interaction may be relevant. AML is a type of blood and bone marrow 383
NLRP12 binds to HCK
18
cancer that is characterized by an unusually high level of circulating immature white blood cells 384
(49). HCK is found to be highly expressed in hematopoietic stem cells of AML patients (24). 385
The overexpression of HCK is linked with the development of AML (24,50). Therefore, HCK 386
may be potentially used as a blood biomarker to identify those patients with potential to develop 387
of AML disease. In addition, activation of HCK and its signaling pathways often causes the 388
development of AML (25,51). HCK has been reported to be involved in the feline McDonough 389
sarcoma (fms)-like tyrosine kinase 3 (FLT3)-HCK- cyclin D-dependent kinase (CDK6) pathway. 390
The mutation in FLT3-internal tandem duplications (FLT3-ITD) causes the activation of HCK, 391
and subsequently results in overexpression of CDK6 in MV4-11 cells, a human AML cell line 392
with the FLT3-ITD mutation. CDK6 siRNA resulted in MV4-11 cells with the FLT3-ITD 393
mutation having a lower proliferation rate compared to MV4-11 control cells (25). Also, it has 394
been reported that in AML patients, HCK, LYN, and FGR phosphorylation levels are high (51). 395
In summary, the above evidence suggests that either high expression and/or high activation of 396
HCK could occur in AML patients as well. 397
398
Our bioinformatics analysis revealed that NLRP12 is likely the only member of the NLRP family 399
that shows significant co-occurrence with HCK at the level of mRNA expression. However, our 400
cellular functional study revealed that NLRP12 protein expression levels decreased when it is co-401
expressed with HCK. This discrepancy might be due NLRP12 mRNA expression level not 402
correlating with protein expression level (52-54). Nevertheless, these data are still consistent 403
with the possibility that NLRP12 may have an effect on AML through its binding to HCK. 404
NLRP12 and HCK co-expression and co-occurrence may then also be relevant to the prognosis 405
NLRP12 binds to HCK
19
of AML. However, further analyses need to be performed to evaluate the ability of NLRP12 and 406
HCK to serve as biomarkers in the prognosis of AML. 407
408
NLRP12 and HCK co-expression and co-occurrence may indicate that NLRP12 and HCK mRNA 409
or protein expression can impact the same pathway or both be regulated by the same molecule. 410
For example, in the case two molecules exhibiting co-occurrence, they may interact in the same 411
signaling pathway, for example, Ki-ras2 Kirsten rat sarcoma viral oncogene homolog (KRAS) 412
and glycogen synthase kinase 3 beta (GSK3β) showed co-occurrence in the AML provisional 413
data, and they are in the same signaling pathway (55). 414
415
Although many HCK inhibitors have been developed, they often lack specificity, with HCK 416
inhibitors also inhibiting other members of the c-SRC non-receptor tyrosine kinase family 417
(24,56-59). If the potential for the development of AML can be decreased when NLRP12 does 418
not bind to HCK, then an inhibitor to the interaction between NLRP12 and HCK can be designed. 419
On the other hand, if the potential for AML is increased when NLRP12 and HCK does not bind, 420
then a drug that facilitates NLRP12 and HCK interaction could be developed. By this approach, 421
the non-specificity problem with NLRP inhibitors may still exist but perhaps to a smaller extent, 422
as NLRP proteins are less conserved overall compared to the members of the c-SRC non-423
receptor tyrosine kinase family. For example, the amino acid identity between NLRP12’s PYD 424
+ NBD versus another NLRP PYD + NBD most closely related to NLRP12, NLRP3’s PYD + 425
NBD, is 54%. But, the protein identity between the HCK C1 fragment versus the one with the 426
closest relationship, the LYN C1 fragment, is 78%. 427
428
NLRP12 binds to HCK
20
In conclusion, this paper reports the preliminary findings of NLRP12 interacting with HCK, and 429
NLRP12 co-occurs with HCK in AML patients. The binding data presented here will expand the 430
database regarding the interactors of NLRP12, and additional bioinformatic and functional in 431
vitro and in vivo analyses may shed light on the functional consequence of the interaction of 432
NLRP12 with HCK. 433
434
Material and Methods 435
Reagents and cells 436
U937, THP-1, RAW 264.7, and 293T cells were all purchased from American Type Culture 437
Collection (ATCC) (Manassas, VA). pCDNA-NLRP12 plasmid and pCDNA-NLRP3 plasmid 438
were gifts from Dr. Jenny P. Y. Ting (University of North Carolina-Chapel Hill). pCDNA-439
NLRP8 plasmid was purchased from Genescript (Piscataway, NJ). All other chemicals were 440
reagent grade. The mouse monoclonal anti-FLAG antibody and the rabbit monoclonal anti-441
FLAG antibody were purchased from Sigma-Aldrich (St. Louis, MO). The mouse monoclonal 442
anti-HCK antibody, the mouse monoclonal anti-TRAF3IP3 antibody, and the mouse IgGκ 443
binding protein-horseradish peroxidase (HRP) for detecting the anti-HCK antibody were 444
purchased from Santa Cruz Biotechnology (Dallas, TX). The rabbit monoclonal anti-V5 445
antibody and polyclonal rabbit HRP-conjugated anti-mouse IgG secondary antibodies were 446
purchased from Cell Signaling (Beverly, MA). The goat anti-rabbit IgG HRP-linked secondary 447
antibody was purchased from Cell Signaling. The Clean-Blot™ IP Detection Reagent was 448
purchased from Thermo Fisher Scientific (Grant Island, NY). Goat anti-mouse IgG (H+L) cross-449
adsorbed secondary antibody, conjugated to Alexa Fluor 488 and goat anti-rabbit IgG (H+L) 450
cross-adsorbed secondary antibody, conjugated to Alexa Fluor 555 were purchased from Thermo 451
NLRP12 binds to HCK
21
Fisher Scientific (Grant Island, NY). The 4',6-diamidino-2-phenylindole, dihydrochloride (DAPI) 452
was purchased from Thermo Fisher Scientific. 453
454
Cloning 455
NLRP12’s PYD + NBD (corresponding to a.a. 1-531) was amplified by polymerase chain 456
reaction (PCR) (primers see Table 3), from a pCDNA-NLRP12 plasmid, was digested with 457
enzymes NdeI and SalI, and subcloned into the pGBKT7 vector (Takara). The sequence 458
encoding the PYD + NBD of NLRP3 (corresponding to a.a. 1-536) was amplified by PCR (see 459
table 3 for primers), digested with enzymes NdeI and BamHI, and subcloned in the pGBKT7 460
vector. The sequence encoding the PYD + NBD of NLRP8 (corresponding to a.a. 1- 527) was 461
amplified by PCR (see Table 3 for primers), digested with enzymes EcoRI and BamHI, and 462
subcloned into the pGBKT7 vector. The entire sequences corresponding to the C-terminal 42 463
amino acids of FGR, c-SRC, YES, and FYN, and 40 amino acids of BLK, LCK, LYN, and HCK 464
(The C terminal 40 (or 42) amino acids is called “C1”), and HCK constructs C2, C3, C4, C5, C6, 465
C7, N1, N2, N3, N4, N5, M1, M2, M3, M4, and HCK C1 mutations for alanine scanning with 466
F503A, E504A, Y505A, I506A, Q507A, S508A, V509A, L510A, and D511A, were all 467
synthesized from IDT DNA (Coralville, IA) and cloned into the pACT2 vector (Takara). 468
NLRP12 was amplified by PCR (see Table 3 for primers), digested with AflII and XbaI, and 469
subcloned into pEFIRES-P-3FLAG (60) and pEFIRES -P-V5 (60). 3FLAG and V5 are 470
homemade, i.e., added later into pEFIRES-P. The peptide sequence of the 3FLAG epitope is 471
DYKDDDDKDYKDDDDKDYKDDDDK. The peptide sequence of the V5 epitope is 472
GKPIPNPLLGLDST. The multiple cloning sites of pEFIRES-P were modified according to our 473
needs. Sequences encoding HCK p61 and p59 isoforms were amplified by PCR (see Table 3 for 474
NLRP12 binds to HCK
22
primers), digested with AflII and XbaI, and subcloned into pEFIRES -3FLAG vector. Final 475
plasmids were sequenced to confirm the constructs. 476
477
Yeast two-hybrid assay 478
Yeast transformed with the plasmid PGBKT7- NLRP12’s PYD + NBD 479
Yeast strain AH109 (Takara, Mountain View, CA) was stored frozen at -80°C. Upon use, it was 480
streaked onto an agar plate supplemented with yeast extract, peptone, dextrose, and adenine 481
hemisulfate (YPDA) (Takara), which was then placed in a 30°C incubator overnight. The 482
growing colonies were picked and placed into 10 ml of YPDA media, incubated overnight with 483
shaking at 30°C. Then they were transformed with NLRP12’s PYD + NBD plasmid, following 484
recommendations in the Takara manual (MatchmakerTM Gal4 Two-hybrid System 3 and 485
Libraries User Manual). A 4ml of 1 M of 3-AT was added into 800 mL of the YPDA medium 486
that were used for making the plates due to a leakage problem. 3-AT is a competitive inhibitor 487
for the product of the HIS3 reporter gene. Yeast cells were lysed using a trichloroacetic acid 488
(TCA) method, and western blots were performed to check NLRP12’s PYD + NBD protein 489
expression (data not shown). 490
491
Library scale transformation of cDNA library into yeast that have been transformed with the 492
PGBKT7- NLRP12 PYD + NBD plasmid 493
A human leukocyte MatchmakerTM cDNA library was purchased from Takara. The library was 494
titered according to the MatchmakerTM Gal4 Two-hybrid System 3 and Libraries User Manual. 495
The colony-forming units (CFU) for the library were calculated to be about 3.1 x 109 CFU/ml. 496
Four large scale transformations (equal to one library scale) were done at the same time 497
NLRP12 binds to HCK
23
following the MatchmakerTM Gal4 Two-hybrid System 3 and Libraries User’s Manual. To 498
calculate the efficiency of yeast two-hybrid (i.e., the number of colons obtained per µg of library 499
plasmid DNA) and to check the expressions of “bait” and “prey” plasmids, small amounts of the 500
solution, diluted 10X, 100X, 1000X, and 10,000X were plated in SD/-Trp-Leu 2DO (drop out) 501
plates. The yeast two-hybrid efficiency obtained was 3 x 106 CFU per µg of library plasmid 502
DNA. In addition, if the yeast grew on the SD/-Trp-Leu-His-Ade + 3-AT plates, the “bait” and 503
“prey” plasmids were considered to be interacting with each other. 504
505
Small scale transformation 506
The yeast two-hybrid co-transformation on a small scale for all of the direct screenings was 507
performed following the small-scale transformation protocol in the MatchmakerTM Gal4 Two-508
hybrid System 3 & Libraries User’s Manual. The direct screens were to test whether NLRP12 509
PYD + NBD and HCK specifically interact; identification of HCK F503, Q507, L510, and D511 510
being critical for binding with NLRP12 PYD + NBD; and the characterization of the HCK C-511
terminal 30 amino acid fragment binding to NLRP12 PYD + NBD. 512
513
Cell transfection 514
293T cells, THP-1 cells, U937 cells, and RAW 264.7 cells were seeded at a density of 5 x 105 515
per well in a 6-well plate. Transfection followed the Lipofectamine® 3000 reagent (Thermo 516
Fisher Scientific) protocol with the following modifications. Briefly, 3.75 µl of Lipofectamine® 517
3000 was added into 125 µl of Opti-MEM® medium (Thermo Fisher Scientific). 2.5 µg of 518
plasmid DNA (prepared by isolation through cesium chloride (CsCl) density gradient 519
purification (61) and 5 µl of P3000TM reagent were added into another aliquot of 125 µl of Opti-520
NLRP12 binds to HCK
24
MEM® medium. Then these two tubes were mixed together, and the mixed tube was incubated 521
at room temperature for 15 min. 250 µl of this mixture was added to the cells. The cells were 522
incubated at 37 °C with 5% CO2 for two days until analysis. 523
524
NLRP12 stable cell lines 525
THP-1 cells, U937 cells, and RAW 264.7 cells were transfected with pEFIRES-P-NLRP12-526
3FLAG plasmid, and 2.5 μg/ml puromycin (InvitrogenTM by Thermo Fisher Scientifc) was added. 527
After two weeks, the cells were considered to be stably expressing NLRP12 after screening for 528
expression. The stable cells were “pooled” clones, and they are maintained by adding 2.5 μg/ml 529
puromycin. 530
531
Cell lysis and immunoprecipitation 532
After trypsinization, cells were lysed in 500 μl of lysis buffer (1% NP-40, 4mM Tris-HCl, pH 8.0, 533
150 mM NaCl). Cells were then treated with freeze-thaw cycles, i.e., frozen in the -80C freezer 534
for 5 min and thawed at 37C in a water bath for 3 min, for a total of five times. Cells were then 535
centrifuged at 12,000 x g for 10 min at 4C. After centrifugation, the supernatant was saved at -536
20C until further use. 537
538
For immunoprecipitations, cell lysates were pre-cleared by adding 40 μl of protein A/G agarose 539
beads (Thermo Fisher Scientific) from a 50% slurry, and the sample was rotated for 1 hr at 4C. 540
The sample was spun at 1000 x g for 1min. Then, 30 μl of pre-cleared cell lysate were removed 541
and run on immunoblotting as a control. The remaining cell lysate was incubated with 1 μg anti-542
HCK mouse monoclonal primary antibody overnight. On the second day, 30 μl of protein A/G 543
NLRP12 binds to HCK
25
agarose, from a 50% slurry and previously washed with lysis buffer three times, were added to 544
the tube, and the tube was incubated for 3 hrs at 4C. The beads were then washed with lysis 545
buffer, but with the NaCl concentration increased to 500 mM. The tube was briefly centrifuged 546
at 1000 x g for 1min., and the supernatant was removed. This step was performed for a total of 5 547
times. The proteins were eluted from the beads by addition of SDS-PAGE sample buffer. 548
549
Immunoblotting 550
Protein concentrations were determined by the PierceTM BCA protein assay (Thermo Fisher 551
Scientific). About 20µg of cell lysate samples and immunoprecipitated samples were loaded 552
onto the 8% acrylamide gel. After overnight transfer of the proteins from an SDS-gel to a 0.45 553
μm pore size polyvinylidene difluoride (PVDF) membrane, the membrane was blocked in for 1 554
hr at room temperature 5% non-fat dry milk dissolved in (Tris-buffered saline with Tween 20, 555
pH=8.0) (TBST) (Sigma-Aldrich). Mouse monoclonal anti-FLAG antibody (diluted 1:1000) and 556
mouse monoclonal anti-HCK antibody (diluted 1:1000) were added and the membrane incubated 557
overnight. For detection of bands in the cell lysate, rabbit anti-mouse IgG HRP-linked secondary 558
antibody (diluted 1:2000) (for anti-FLAG antibody) and mouse IgGκ binding protein-HRP 559
secondary antibody (for anti-HCK antibody) (diluted 1:1000) were added, and the membrane 560
was incubated for 1 hr at room temperature. For immunoprecipitations, Clean-Blot™ IP 561
detection reagent (HRP) secondary antibody (1:300) (Thermo Fisher Scientific) was used. The 562
detection signal was visualized by using enhanced chemiluminescence (ECL) reagent (Thomas 563
Scientific, Swedesboro, NJ). Immoblots were analyzed by chemiluminescence on a Bio-Rad 564
ChemiDocTM Touch Gel imaging system (Bio-Rad, Hercules, CA). 565
566
NLRP12 binds to HCK
26
Immunofluorescence 567
The coverslips were autoclaved and coated with 0.1 mg/ml of poly-D-lysine hydrobromide (mol 568
wt 70,000-150,000) (Sigma-Aldrich). Cells were plated onto the coverslips. The cells were 569
fixed in 4% paraformaldehyde (Alfa Aesar, Ward Hill, MA) in phosphate buffered saline (PBS) 570
for 15 mins. And fixation was stopped by adding 50 mM NH4Cl in PBS for 15 min at room 571
temperature. After raising with PBS, the fixed cells were permeabilized by adding 0.5% 572
Triton®-X-100 in PBS for 10 min. The cells were then washed with PBS three times and 573
blocked with 0.1% of BSA in PBS at room temperature for 1 hr. The rabbit monoclonal anti-574
FLAG antibody (diluted 1:100) and mouse monoclonal anti-HCK antibody (diluted 1:100) were 575
added to each cover slip and incubated overnight. Goat anti-mouse IgG (H+L) cross-adsorbed 576
secondary antibody, conjugated to Alexa Fluor 488 (diluted 1:100) and goat anti-rabbit IgG 577
(H+L) cross-adsorbed secondary antibody, conjugated to Alexa Fluor 555 (1: 100) were added. 578
After incubation with secondary antibodies for 1 hr, the cells were washed five times with PBS. 579
During the first step, DAPI (InvitrogenTM by Thermo Fisher Scientific) was also added at 1:1000 580
in PBS for 2 min. Finally, the coverslips were mounted in ProLong™ Gold Antifade mounting 581
medium (Life Technologies Corporation by Thermo Fisher Scientific, Eugene, OR). The stained 582
slides were viewed on a ZEISS LSM 880 (Pleasanton, CA) with Airyscan. The images were 583
viewed and processed using Fiji software (62). 584
585
IPA diagram 586
Yeast two-hybrid data were analyzed by IPA software (QIAGEN Inc., 587
https://www.qiagenbioinformatics.com/products/ingenuity- pathway-analysis). 588
589
NLRP12 binds to HCK
27
Statistical analysis 590
The Spearman coefficient and Pearson coefficient used for co-expression was generated by 591
cBioPortal website (37,38,63). Clinical samples from the Oncomine website 592
(https://www.oncomine.org) were downloaded and checked for NLRP12 and Hck mRNA 593
expression levels. The Fisher exact test and the odds ratio test used for co-occurrence were 594
generated by the cBioPortal website. Analysis of variance (ANOVA) was performed to check 595
whether a significant difference of the treatment exists among all the groups. A p-value < 0.05 596
was considered statistically significant. If P < 0.05, then the Tukey test was used to determine 597
whether a pair-wise difference exists. Quantification of the western blot bands was performed 598
by FIJI software (62) and made into graphs by Prism 6.0 (GraphPad Software, Inc., San Diego, 599
CA).600
601
Acknowledgement 602
Thank you to Dr. Houda Alachkar for providing valuable suggestions to this paper. We also 603
thank Ms. Meng Li and Dr. Yibu Chen at the University of Southern California’s Norris Medical 604
Library for their suggestions for bioinformatic analysis. Appreciation is given to Mr. Noam 605
Morningstar-Kywi and Dr. Ian Haworth for their helps on structural analysis. 606
607
Abbreviations 608
3-AT: 3-Amino-1,2,4-triazole; AML: acute myeloid leukemia; ANOVA: analysis of variance; 609
ASC: apoptosis-associated speck-like protein containing a caspase activation and 610
recruitment domain; ATCC: American Type Culture Collection; C1: self-named truncations of 611
C-terminal 40 (or 41) amino acids fragment of c-SRC family non-receptor tyrosine kinases; C2-612
NLRP12 binds to HCK
28
C7 and N1-N5: self-named truncations HCK C1 fragment; CD40: cluster of differentiation 40; 613
CDK6: cyclin D-dependent kinase; CFU; colony-forming units; COSMIC: catalogue of somatic 614
mutations in cancer; CsCl: cesium chloride; DAMP: damage-associated molecular patterns; 615
DAPI: 4',6-diamidino-2-phenylindole, dihydrochloride; DC: dendritic cells; DO: drop out (2DO: 616
SD/-Trp-Leu; 4DO: SD/-Trp-Leu-His-Ade); ECL: enhanced chemiluminescence; FAf1: Fas 617
associated factor 1; FLT3-ITD: FLT3-internal tandem duplications; FLT3: feline McDonough 618
sarcoma (fms)-like tyrosine kinase 3; GSK3b: glycogen synthase kinase 3 beta; HCK C1 619
fragment: HCK C-terminal 40 amino acids; HCK: hematopoiesis cell kinase; HRP: horseradish 620
peroxidase; HSP90: heat shock protein 90; IL: interleukin; IPA: ingenuity pathway analysis; 621
IRAK1: IL-1 receptor-associated kinase 1; KO: knock out; KRAS; Ki-ras2 Kirsten rat sarcoma 622
viral oncogene homolog; LRR: leucine rich repeat domain; M1: the amino acids in the R1 region 623
were mutated to the amino acids identical to the R1 region in Lyn; M2: the amino acids in the 624
R2 region were all mutated to Alanine; M3: the amino acids in the R3 region were all mutated to 625
Alanine; M4: the amino acids in the R4 region were all mutated to Alanine; NBD: nucleotide 626
binding domain; NF-κB: nuclear factor kappa-light-chain-enhancer of activated B cells; NIK: 627
NF-κB-inducing kinase; NLR: nucleotide binding domain and leucine-rich repeat; NLRP: NLR 628
pyrin domain-containing protein; PAMP: pathogen-associated molecular patterns; PBS: 629
phosphate buffered saline; PCR: polymerase chain reaction; PDB: protein data bank; PMA: 630
phorbol 12-myristate 13-acetate; PRR: pattern recognition receptors; PVDF: polyvinylidene 631
difluoride; PYD: pyrin domain; R1, R2, R3, and R4: self-named four regions in the HCK C 632
terminal 40 amino acids fragment; SD: synthetic dropout; SD/-Trp-Leu: SD base plus amino 633
acids supplements minus tryptophan and leucine; SD/-Trp-Leu-His-Ade: SD base plus amino 634
acids supplements minus tryptophan, leucine, and histidine, and adenine; STRING: search tool 635
NLRP12 binds to HCK
29
for recurring instances of neighboring genes; TBST: tris-buffered saline with Tween 20, pH=8.0; 636
TCA: trichloroacetic acid; TCGA: the cancer genome atlas; TLR: toll-like receptor; TRAF3: 637
tumor necrosis factor receptor associated factor 3; TRAF3IP3: TRAF3 interacting protein 3; 638
UBA: ubiquitin-associated domains; WT: wild type; YPDA: yeast extract, peptone, dextrose, and 639
adenine hemisulfate. 640
641
Competing Interest 642
The authors have declared that no competing conflict exists. 643
644
Author Contributions Statement 645
YZ design the experiment and wrote and revised the manuscript. CO helped to design the 646
experiments and to revise the manuscript. 647
648
References 649
1. Takeuchi O, Akira S. Pattern recognition receptors and inflammation. Cell. 2010; 140: 650
805-820. 651
2. Tuncer S, Fiorillo MT, Sorrentino R. The multifaceted nature of NLRP12. J Leukoc Biol. 652
2014; 96: 991-1000. 653
3. Allen IC, Wilson JE, Schneider M, Lich JD, Roberts RA, Arthur JC, Woodford RM, 654
Davis BK, Uronis JM, Herfarth HH, Jobin C, Rogers AB, Ting JP. NLRP12 suppresses 655
colon inflammation and tumorigenesis through the negative regulation of noncanonical 656
NF-kappaB signaling. Immunity. 2012; 36: 742-754. 657
4. Zaki MH, Vogel P, Malireddi RK, Body-Malapel M, Anand PK, Bertin J, Green DR, 658
Lamkanfi M, Kanneganti TD. The NOD-like receptor NLRP12 attenuates colon 659
inflammation and tumorigenesis. Cancer Cell. 2011; 20: 649-660. 660
5. Zaki MH, Man SM, Vogel P, Lamkanfi M, Kanneganti TD. Salmonella exploits 661
NLRP12-dependent innate immune signaling to suppress host defenses during infection. 662
Proc Natl Acad Sci U S A. 2014; 111: 385-390. 663
6. Chen ST, Chen L, Lin DS, Chen SY, Tsao YP, Guo H, Li FJ, Tseng WT, Tam JW, Chao 664
CW, Brickey WJ, Dzhagalov I, Song MJ, Kang HR, Jung JU, Ting JP. NLRP12 regulates 665
anti-viral RIG-I activation via interaction with TRIM25. Cell Host Microbe. 2019; 25: 666
602-616.e607. 667
NLRP12 binds to HCK
30
7. Chen L, Wilson JE, Koenigsknecht MJ, Chou WC, Montgomery SA, Truax AD, Brickey 668
WJ, Packey CD, Maharshak N, Matsushima GK, Plevy SE, Young VB, Sartor RB, Ting 669
JP. NLRP12 attenuates colon inflammation by maintaining colonic microbial diversity 670
and promoting protective commensal bacterial growth. Nat Immunol. 2017; 18: 541-551. 671
8. Truax AD, Chen L, Tam JW, Cheng N, Guo H, Koblansky AA, Chou WC, Wilson JE, 672
Brickey WJ, Petrucelli A, Liu R, Cooper DE, Koenigsknecht MJ, Young VB, Netea MG, 673
Stienstra R, Sartor RB, Montgomery SA, Coleman RA, Ting JP. The Inhibitory Innate 674
Immune Sensor NLRP12 maintains a threshold against obesity by regulating gut 675
microbiota homeostasis. Cell Host Microbe. 2018; 24: 364-378.e366. 676
9. Ye, Z, Lich JD, Moore CB, Duncan JA, Williams KL, Ting JP. ATP binding by 677
monarch-1/NLRP12 is critical for its inhibitory function. Mol Cell Biol. 2008; 28: 1841-678
1850. 679
10. Jeru I, Le Borgne G, Cochet E, Hayrapetyan H, Duquesnoy P, Grateau G, Morali A, 680
Sarkisian T, Amselem S. Identification and functional consequences of a recurrent 681
NLRP12 missense mutation in periodic fever syndromes. Arthritis Rheum. 2011; 63: 682
1459-1464. 683
11. Jeru I, Duquesnoy P, Fernandes-Alnemri T, Cochet E, Yu JW, Lackmy-Port-Lis M, 684
Grimprel E, Landman-Parker J, Hentgen V, Marlin S, McElreavey K, Sarkisian T, 685
Grateau G, Alnemri ES, Amselem S. Mutations in NALP12 cause hereditary periodic 686
fever syndromes. Proc Natl Acad Sci U S A. 2008; 105: 1614-1619. 687
12. Arthur J C, Lich JD, Ye Z, Allen IC, Gris D, Wilson JE, Schneider M, Roney KE, 688
O'Connor BP, Moore CB, Morrison A, Sutterwala FS, Bertin J, Koller BH, Liu Z, Ting 689
JP. Cutting edge: NLRP12 controls dendritic and myeloid cell migration to affect contact 690
hypersensitivity. J Immunol. 2010; 185: 4515-4519. 691
13. Lukens JR, Gurung P, Shaw PJ, Barr MJ, Zaki MH, Brown SA, Vogel P, Chi H, 692
Kanneganti TD. The NLRP12 sensor negatively regulates autoinflammatory disease by 693
modulating interleukin-4 production in T cells. Immunity. 2015; 42: 654-664. 694
14. Vladimer GI, Weng D, Paquette SWM, Vanaja SK, Rathinam VA, Aune MH, Conlon J E, 695
Burbage JJ, Proulx MK, Liu Q. The NLRP12 inflammasome recognizes Yersinia pestis. 696
Immunity. 2012; 37: 96-107. 697
15. Udden SN, Kwak YT, Godfrey V, Khan MAW, Khan S, Loof N, Peng L, Zhu H, Zaki H. 698
NLRP12 suppresses hepatocellular carcinoma via downregulation of cJun N-terminal 699
kinase activation in the hepatocyte. Elife. 2019; 8: e40396. 700
16. Hornick EE, Banoth B, Miller AM, Zacharias ZR, Jain N, Wilson ME, Gibson-Corley 701
KN, Legge KL, Bishop GA, Sutterwala FS, Cassel SL. NLRP12 mediates adverse 702
neutrophil recruitment during influenza virus infection. J Immunol. 2018; 200: 1188-703
1197. 704
17. Damiano JS, Oliveira V, Welsh K, Reed JC. Heterotypic interactions among NACHT 705
domains: implications for regulation of innate immune responses. The Biochemical 706
journal. 2004; 381: 213-219. 707
18. Janowski AM, Kolb R, Zhang W, Sutterwala FS. Beneficial and detrimental roles of 708
NLRs in carcinogenesis. Frontiers in immunology. 2013; 4: 370-370. 709
19. Faustin B, Lartigue L, Bruey JM, Luciano F, Sergienko E, Bailly-Maitre B, Volkmann N, 710
Hanein D, Rouiller I, Reed JC. Reconstituted NALP1 inflammasome reveals two-step 711
mechanism of caspase-1 activation. Mol Cell. 200; 25: 713-724. 712
NLRP12 binds to HCK
31
20. Martinon F, Agostini L, Meylan E, Tschopp J. Identification of bacterial muramyl 713
dipeptide as activator of the NALP3/cryopyrin inflammasome. Curr Biol. 2004; 14: 714
1929-1934. 715
21. Kanneganti TD, Lamkanfi M, Nunez G. Intracellular NOD-like receptors in host defense 716
and disease. Immunity. 2007; 27: 549. 717
22. Sharif H, Wang L, Wang WL, Magupalli VG, Andreeva L, Qiao Q, Hauenstein AV, Wu 718
Z, Núñez G, Mao Y, Wu H. Structural mechanism for NEK7-licensed activation of 719
NLRP3 inflammasome. Nature. 2019; 570: 338-343. 720
23. Hafner-Bratkovič I, Sušjan P, Lainšček D, Tapia-Abellán A, Cerović K, Kadunc L, 721
Angosto-Bazarra D, Pelegrin P, Jerala R. NLRP3 lacking the leucine-rich repeat domain 722
can be fully activated via the canonical inflammasome pathway. Nature Communications. 723
2018; 9: 5182. 724
24. Roversi FM, Pericole FV, Machado-Neto JA, da Silva Santos Duarte A, Longhini AL, 725
Corrocher FA, Palodetto B, Ferro KP, Rosa RG, Baratti MO, Verjovski-Almeida S, 726
Traina F, Molinari A, Botta M, Saad ST. Hematopoietic cell kinase (HCK) is a potential 727
therapeutic target for dysplastic and leukemic cells due to integration of 728
erythropoietin/PI3K pathway and regulation of erythropoiesis: HCK in 729
erythropoietin/PI3K pathway. Biochim Biophys Acta Mol Basis Dis. 2017; 1863: 450-730
461. 731
25. Lopez S, Voisset E, Tisserand JC, Mosca C, Prebet T, Santamaria D, Dubreuil P, De 732
Sepulveda, P. An essential pathway links FLT3-ITD, HCK and CDK6 in acute myeloid 733
leukemia. Oncotarget. 2016; 7: 51163-51173. 734
26. Roversi FM, Pericole FV, Duarte AdSS, Ferro KP, Corrocher FA, Palodetto B, Longhini 735
AL, Botta M, Saad STO. Knockdown of HCK reduces cell death and erythroid 736
differentiation in human CD34<sup>+</sup> hematopoietic progenitor cells. Blood. 737
2015; 126: 2860. 738
27. Normand S, Waldschmitt N, Neerincx A, Martinez-Torres RJ, Chauvin C, Couturier-739
Maillard A, Boulard O, Cobret L, Awad F, Huot L, Ribeiro-Ribeiro A, Lautz K, Ruez R, 740
Delacre M, Bondu C, Guilliams M, Scott C, Segal A, Amselem S, Hot D, Karabina S, 741
Bohn E, Ryffel B, Poulin LF, Kufer TA, Chamaillard M. Proteasomal degradation of 742
NOD2 by NLRP12 in monocytes promotes bacterial tolerance and colonization by 743
enteropathogens. Nature Communications. 2018; 9: 5338. 744
28. Linz BM, Neely CJ, Kartchner LB, Mendoza AE, Khoury AL, Truax A, Sempowski G, 745
Eitas T, Brickey J, Ting JP, Cairns BA, Maile R. Innate immune cell recovery is 746
positively regulated by NLRP12 during emergency hematopoiesis. J Immunol. 2017; 198: 747
2426-2433. 748
29. Robbins SM, Quintrell NA, Bishop, JM. Myristoylation and differential palmitoylation of 749
the HCK protein-tyrosine kinases govern their attachment to membranes and association 750
with caveolae. Mol Cell Biol. 1995; 15: 3507-3515. 751
30. Carreno S, Gouze ME, Schaak S, Emorine LJ, Maridonneau-Parini I. Lack of 752
palmitoylation redirects p59HCK from the plasma membrane to p61HCK-positive 753
lysosomes. J Biol Chem. 2000; 275: 36223-36229. 754
31. Guiet R, Poincloux R, Castandet J, Marois L, Labrousse A, Le Cabec V, Maridonneau-755
Parini I. Hematopoietic cell kinase (HCK) isoforms and phagocyte duties - from 756
signaling and actin reorganization to migration and phagocytosis. Eur J Cell Biol. 2008; 757
87: 527-542. 758
NLRP12 binds to HCK
32
32. Sicheri F, Moarefi I, Kuriyan J. Crystal structure of the c-SRC family tyrosine kinase 759
HCK. Nature. 1997; 385: 602-609. 760
33. Ma B, Elkayam T, Wolfson H, Nussinov R. Protein–protein interactions: Structurally 761
conserved residues distinguish between binding sites and exposed protein surfaces. Proc 762
Natl Acad Sci U S A. 2003; 100: 5772. 763
34. Knappe M, Bodevin S, Selinka HC, Spillmann D, Streeck RE, Chen XS, Lindahl U, Sapp 764
M. Surface-exposed amino acid residues of HPV16 L1 protein mediating interaction with 765
cell surface heparan sulfate. J Biol Chem. 2007; 282: 27913-27922. 766
35. Lich JD, Williams KL, Moore CB, Arthur JC, Davis BK, Taxman DJ, Ting, JP. 767
Monarch-1 suppresses non-canonical NF-kappaB activation and p52-dependent 768
chemokine expression in monocytes. J Immunol. 2007; 178: 1256-1260. 769
36. Cerami E, Gao J, Dogrusoz U, Gross BE, Sumer SO, Aksoy BA, Jacobsen A, Byrne CJ, 770
Heuer ML, Larsson E, Antipin Y, Reva B, Goldberg AP, Sander C, Schultz N. The cBio 771
Cancer Genomics Portal: An open platform for exploring multidimensional cancer 772
genomics data. Cancer Discovery. 2012; 2: 401. 773
37. Gao J, Aksoy BA, Dogrusoz U, Dresdner G, Gross B, Sumer SO, Sun Y, Jacobsen A, 774
Sinha R, Larsson E, Cerami E, Sander C, Schultz N. Integrative analysis of complex 775
cancer genomics and clinical profiles using the cBioPortal. Sci Signal. 2013; 6: pl1. 776
38. Cancer Genome Atlas Research Network. Genomic and epigenomic landscapes of adult 777
de novo acute myeloid leukemia. NEJM. 2013, 368: 2059-2074. 778
39. Williams KL, Lich JD, Duncan JA, Reed W, Rallabhandi P, Moore C, Kurtz S, Coffield 779
VM, Accavitti-Loper MA, Su L, Vogel SN, Braunstein M, Ting JP. The CATERPILLER 780
protein monarch-1 is an antagonist of toll-like receptor-, tumor necrosis factor alpha-, and 781
Mycobacterium tuberculosis-induced pro-inflammatory signals. J Biol Chem. 2005; 280: 782
39914-39924. 783
40. Arthur JC, Lich JD, Aziz RK, Kotb M, Ting JP. Heat shock protein 90 associates with 784
monarch-1 and regulates its ability to promote degradation of NF-kappaB-inducing 785
kinase. J Immunol. 2007; 179: 6291-6296. 786
41. Pinheiro AS, Eibl C, Ekman-Vural Z, Schwarzenbacher R, Peti W. The NLRP12 pyrin 787
domain: structure, dynamics, and functional insights. Journal of molecular biology. 2011; 788
413: 790-803. 789
42. Kinoshita T, Kondoh C, Hasegawa M, Imamura R, Suda, T. Fas-associated factor 1 is a 790
negative regulator of PYRIN-containing Apaf-1-like protein 1. Int Immunol. 2006; 18: 791
1701-1706. 792
43. Shi C, Miley J, Nottingham A, Morooka T, Prosdocimo DA, Simon DI. Leukocyte 793
integrin signaling regulates FOXP1 gene expression via FOXP1-IT1 long non-coding 794
RNA-mediated IRAK1 pathway. Biochim Biophys Acta Gene Regul Mech. 2019; 1862: 795
493-508. 796
44. Fu G, Xu Q, Qiu Y, Jin X, Xu T, Dong S, Wang J, Ke Y, Hu H, Cao X, Wang D, Cantor 797
H, Gao X, Lu L. Suppression of Th17 cell differentiation by misshapen/NIK-related 798
kinase MINK1. J Exp Med. 2017; 214: 1453-1469. 799
45. Wang Q, Liu F, Zhang M, Zhou P, Xu C, Li Y, Bian L, Liu Y, Yao Y, Wang F, Fang Y, 800
Li D. NLRP12 promotes mouse neutrophil differentiation through regulation of non-801
canonical NF-kappaB and MAPK(ERK1/2) signaling. Int J Biol Sci. 2018; 14: 147-155. 802
NLRP12 binds to HCK
33
46. Kinnaird JH, Singh M, Gillan V, Weir W, Calder ED, Hostettler I, Tatu U, Devaney E, 803
Shiels BR. Characterization of HSP90 isoforms in transformed bovine leukocytes 804
infected with Theileria annulata. Cell Microbiol. 2017; 19: e12669. 805
47. Ryu SW, Lee SJ, Park MY, Jun JI, Jung YK, Kim E. Fas-associated factor 1, FAF1, is a 806
member of Fas death-inducing signaling complex. J Biol Chem. 2003; 278: 24003-24010. 807
48. Martin BN, Wang C, Zhang CJ, Kang Z, Gulen MF, Zepp JA, Zhao J, Bian G, Do JS, 808
Min B, Pavicic PG Jr, El-Sanadi C, Fox PL, Akitsu A, Iwakura Y, Sarkar A, Wewers MD, 809
Kaiser WJ, Mocarski ES, Rothenberg ME, Hise AG, Dubyak GR, Ransohoff RM, Li X. 810
T cell-intrinsic ASC critically promotes T(H)17-mediated experimental autoimmune 811
encephalomyelitis. Nat Immunol. 2016; 17: 583-592. 812
49. Prada-Arismendy J, Arroyave JC, Rothlisberger S. Molecular biomarkers in acute 813
myeloid leukemia. Blood Rev. 2017; 31: 63-76. 814
50. Poh AR, O'Donoghue RJ, Ernst M. Hematopoietic cell kinase (HCK) as a therapeutic 815
target in immune and cancer cells. Oncotarget. 2015; 6: 15752-15771. 816
51. Dos Santos C, Demur C, Bardet V, Prade-Houdellier N, Payrastre B, Récher C. A critical 817
role for LYN in acute myeloid leukemia. Blood. 2008; 111: 2269. 818
52. Greenbaum D, Colangelo C, Williams K, Gerstein M. Comparing protein abundance and 819
mRNA expression levels on a genomic scale. Genome Biol. 2003; 4: 117. 820
53. Vogel C, Marcotte EM. Insights into the regulation of protein abundance from proteomic 821
and transcriptomic analyses. Nat Rev Genet. 2012; 13: 227-232. 822
54. Liu Y, Beyer A, Aebersold R. On the dependency of cellular protein levels on mRNA 823
abundance. Cell. 2016; 165: 535-550. 824
55. Prochnicki T, Latz E. Inflammasomes on the crossroads of innate immune recognition 825
and metabolic control. Cell Metab. 2017; 26: 71-93. 826
56. Saito Y, Yuki H, Kuratani M, Hashizume Y, Takagi S, Honma T, Tanaka A, Shirouzu M, 827
Mikuni J, Handa N, Ogahara I, Sone A, Najima Y, Tomabechi Y, Wakiyama M, Uchida 828
N, Tomizawa-Murasawa M, Kaneko A, Tanaka S, Suzuki N, Kajita H, Aoki Y, Ohara O, 829
Shultz LD, Fukami T, Goto T, Taniguchi S, Yokoyama S, Ishikawa F. A pyrrolo-830
pyrimidine derivative targets human primary AML stem cells in vivo. Sci Transl Med. 831
2013; 5: 181ra152. 832
. 833
57. Pene-Dumitrescu T, Peterson LF, Donato NJ, Smithgall TE. An inhibitor-resistant mutant 834
of HCK protects CML cells against the antiproliferative and apoptotic effects of the 835
broad-spectrum c-SRC family kinase inhibitor A-419259. Oncogene. 2008; 27: 7055-836
7069. 837
58. Naganna N, Opoku-Temeng C, Choi EY, Larocque E, Chang ET, Carter-Cooper BA, 838
Wang M, Torregrosa-Allen SE, Elzey BD, Lapidus RG, Sintim HO. Amino 839
alkynylisoquinoline and alkynylnaphthyridine compounds potently inhibit acute myeloid 840
leukemia proliferation in mice. EBioMedicine. 2019; 40: 231-239. 841
59. Hu Y, Liu Y, Pelletier S, Buchdunger E, Warmuth M, Fabbro D, Hallek M, Van Etten 842
RA, Li S. Requirement of c-SRC kinases LYN, HCK and FGR for BCR-ABL1-induced 843
B-lymphoblastic leukemia but not chronic myeloid leukemia. Nat Genet. 2004; 36: 453-844
461. 845
60. Hobbs S, Jitrapakdee S, Wallace JC. Development of a bicistronic vector driven by the 846
human polypeptide chain elongation factor 1α promoter for creation of stable mammalian 847
NLRP12 binds to HCK
34
cell lines that express very high levels of recombinant proteins. Biochem Biophys Res 848
Commun. 1998; 252: 368-372. 849
61. Brakke MK. Density Gradient Centrifugation: A New Separation Technique1. JACS. 850
1951; 73: 1847-1848. 851
62. Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, Preibisch S, 852
Rueden C, Saalfeld S, Schmid B, Tinevez JY, White DJ, Hartenstein V, Eliceiri K, 853
Tomancak P, Cardona A. Fiji: an open-source platform for biological-image analysis. Nat 854
Methods. 2012; 9: 676-682. 855
63. Cerami E, Gao J, Dogrusoz U, Gross BE, Sumer SO, Aksoy BA, Jacobsen A, Byrne CJ, 856
Heuer ML, Larsson E, Antipin Y, Reva B, Goldberg AP, Sander C, Schultz N. The cBio 857
cancer genomics portal: an open platform for exploring multidimensional cancer 858
genomics data. Cancer Discov. 2012; 2: 401-404. 859
860
Table 1: yeast two hybrid results (hit frequency >= 3) 861
Name Frequency
Deleted in azoospermia (DAZ) associated protein 2 (DAZAP2) 10
Filamin A, alpha (FLNA) 9
Hematopoietic cell kinase (HCK) proto-oncogene, c-SRC family tyrosine kinase (HCK) 9
Chaperone DnaJ (a.k.a., heat shock protein 40 kD(Hsp40)) homolog, subfamily B,
member 1 (DNAJB1) 9
Metallothionein 2A (MT2A) 7
Cluster of differentiation 14 molecule (CD14) 6
DEAD (Asp-Glu-Ala-Asp) box helicase 5 (DDX5) 6
Talin 1 (TLN1) 5
Calcium modulating ligand (CAMLG) 4
Squamous cell carcinoma antigen recognized by T cells (SART1) 4
Ras-related nuclear protein (RAN) binding protein 9 (RANBP9) 3
Tumor necrosis factor (TNF) receptor associated factor 3 (TRAF3) interacting protein 3 3
NLRP12 binds to HCK
35
(TRAF3IP3)
Thyroid hormone receptor interactor 4 (TRIP4) 3
862
Table 2: Co-occurrence of mRNA expression level of HCK and all of the NLRP family 863
proteins (from the provisional data, assessed on July 11 2019) 864
HCK + NLRP
family genes
Co-occurrence (C) or
mutual exclusivity (M)
P-value
NLRP1 C 0.048
NLRP2 M 0.546
NLRP3 C 0.415
NLRP4 M 0.936
NLRP5 M 0.877
NLRP6 M 0.670
NLRP7 M 0.475
NLRP8 M 0.877
NLRP9 M 0.491
NLRP10 M 0.626
NLRP11 M 0.233
NLRP12 C <0.001
NLRP13 C 1.000
NLRP14 M 0.509
865
Table 3: Primers used in cloning (The underline shows the restriction enzymes) 866
NLRP12 binds to HCK
36
Name Primers
NLRP12 5’ primer
(cloned into PGBKT7)
GATCTACATATGCTACGAACCGCAGGCAG
NLRP12 3’ primer
(cloned into PGBKT7)
GATGTAGTCGACTCACCCCTCGTCCAGGATATAGTACATAGCT
NLRP3 5’ primer (cloned
into pGBKT7)
GATCTACATATGAAGATGGCAAGCACCCGCTGCA
NLRP3 3’ primer (cloned
into PGBKT7)
GATGTAGGATCCTCACAGCAGGTAGTACATGGCGGCAAAGA
NLRP8 5’ primer (cloned
into pGBKT7)
GATCTAGAATTCATGAGTGACGTGAATCCACCCTCTG
NLRP8 3’ primer (cloned
into PGBKT7)
GATGTAGGATCCTCAGAGTCTTTGTGGGAAACAGAGAACA
NLRP12 5’ (cloned into
pEFIRES-P-3FLAG and
pIRES-V5)
GTCCAGCTTAAGATGCTACGAACCGCAGGCAG
NLRP12 3’ (cloned into
PIRES-3FLAG and
pEFIRES-P-V5)
CTGAATTCTAGAGCAGCCAATGTCCAAATAAGG
HCK p61 5’ (cloned into
pEFIRES-P-3FLAG)
GATCTACTTAAGGCCACCATGGGGGGGCGCTCAAGCTGCGAGG
HCK p59 5’ (cloned into
pEFIRES-P-3FLAG)
GATCTACTCGAGGCCACCATGGGGTGCATGAAGTCCAAGTTCC
NLRP12 binds to HCK
37
HCK p61/p59 3’ (cloned
into pEFIRES-P-3FLAG)
GTCGATTCTAGATGGCTGCTGTTGGTACTGGCTCTCT
867
Figure legends: 868
Figure 1: (A) Schematic drawing of the domain structure of full-length NLRP12 (top) and the 869
domains used as the construct for the yeast two-hybrid screen (bottom). PYD: Pyrin domain; NBD: 870
Nucleotide binding domain; and LRR: Leucine rich repeat domain. (B) The domain structures of the 871
two alternatively spliced isoforms of HCK, HCK p59 and HCK p61. (C) Identification of the 872
clones of HCK interacting with NLRP12 (“hits”) obtained from the yeast two-hybrid screen. 873
Among the nine of the clones positive for HCK, one clone encoded for the smallest interacting 874
fragment of HCK, the C-terminal 40 amino acids of HCK. The numbers in parentheses indicate 875
the total number of hits obtained in the screen. 876
877
Figure 2: Specificity of binding of NLRP12 with HCK. (A) Schematic drawing of HCK (p61 878
isoform, top) and, for comparison, the region of HCK (“HCK C1 fragment,” lower) used for 879
testing the specificity of binding of HCK to other members of the NLRP family of proteins. The 880
numbers represent the amino acid residues of the protein or protein fragment (B) Phylogenetic 881
tree showing the relationships among the PYD + NBD region of all of the members of the human 882
NLRP family of proteins. The phylogenetic tree was produced through Clustal Omega 883
(https://www.ebi.ac.uk/Tools/msa/clustalo/). For each NLRP family member, the PYD + NBD 884
of the longest isoform was selected for generating the phylogenetic tree. NLRP3 is 885
phylogenetically the closest NLRP to NLRP12, while NLRP8 is phylogenetically among the 886
most distant from NLRP12. NLRP3 and NLRP8 are highlighted in red color. (C) When 887
NLRP12 binds to HCK
38
NLRP12’s PYD + NBD and HCK C1 fragment were co-transformed into yeast, they grew on the 888
highest stringency plates (4DO + 3AT), consistent with an interaction between the two constructs. 889
On the other hand, neither NLRP3’s PYD + NBD co-transformed with the C1 fragment nor 890
NLRP8’s PYD + NBD co-transformed the C1 fragment grew on the highest stringency plates. 891
These results are consistent with a specific interaction between NLRP12 and the HCK C1 892
fragment, but not with the other NLRPs. (D) Yeast clones grew on the high stringency plates 893
4DO + 3AT when HCK’s C1 fragment and NLRP12’s PYD + NBD were co-transformed, but 894
they did not grow on the high stringency 4DO + 3AT plates when the C1 fragments of other 895
members of the c-SRC family of non-receptor tyrosine kinases were co-transformed with 896
NLRP12’s PYD + NBD. These results are again consistent with a specific interaction between 897
NLRP12 and HCK. 898
899
Figure 3: (A) Structure of the HCK C1 fragment generated by Chimera from the University of 900
California, San Francisco (version 1.13.1), extracted from its structure within the entire HCK 901
protein. HCK C1 fragment was thus divided into four helical domains: R1, R2, R3, and R4. The 902
side chains of amino acids F503, I506, Q507, L510, and D511 are shown. However, it is not 903
clear whether the C1 fragment retains the structural features shown here when it expressed on its 904
own. (B) Primary amino acid sequence alignments showing of the C1 fragments of all members 905
of the c-SRC non-receptor tyrosine kinase family (40 amino acids for HCK, BLK, LCK, and 906
LYN; 42 amino acids for FGR, c-SRC, YES, and FYN). The black lines indicate the R1, R2, R3, 907
or R4 regions. The arrows denote the four critical amino acids for binding of HCK C1 fragment 908
to NLRP12’s PYD + NBD. (C) Phylogenetic tree showing the relationship among all of the C1 909
fragments of the members of the c-SRC family of non-receptor tyrosine kinases. The 910
NLRP12 binds to HCK
39
phylogenetic tree was produced through Clustal Omega (https://www.ebi.ac.uk/Tools/msa/ 911
clustalo/). HCK is highlighted in red color. (D) Four sets of distinct mutations were made within 912
the HCK C1 fragment. M1: the amino acids in the R1 region were mutated to the amino acids 913
identical to the R1 region in LYN (highlighted in red color); M2: the amino acids in the R2 914
region were all mutated to alanine (Ala); M3: the amino acids in the R3 region were all mutated 915
to Ala; and M4: the amino acids in the R4 region were all mutated to Ala. (E) Yeast did not 916
grow on high stringency 4DO + 3AT plates when HCK C1 fragment with the M3 mutations was 917
co-transformed with NLRP12’s PYD + NBD. But yeast can grow on the high stringency 4DO + 918
3AT plates when the HCK C1 fragment contained either the M1, M2, or M4 mutations, 919
suggesting that the R3 region within the C1 fragment was necessary for binding to NLRP12 920
NBD + PYD. (E) Yeast did not grow on high stringency 4DO plates when NLRP12’s PYD + 921
NBD was co-transformed with HCK C1 fragment with single Ala substitutions at either F503, 922
I506, Q507, L510, or D511. But yeast grew on high stringency 4DO plates when NLRP12’s 923
PYD + NBD was co-transformed with HCK’s C1 fragment in which the following amino acids in 924
the R3 region were individually mutated to Ala: E504, Y505, S508, or V509. 925
926
Figure 4: (A) Top: schematic drawing of HCK truncations comprising the C1 to C7 fragments. 927
Bottom: yeast grew on the high stringency 4DO plates and 4DO +3AT plates when HCK’s C-928
terminal 40 amino acid fragment (C1), HCK’s C-terminal 35 amino acid fragment (C2), and 929
HCK’s C-terminal 30 amino acid fragment (C3), individually, were co-transformed with 930
NLRP12’s PYD + NBD. But yeast did not grow on the high stringency 4DO plates nor 4DO+ 931
3AT plates when C4 to C7 fragments were co-transformed to with NLRP12’s PYD + NBD. The 932
lack of yeast growth upon co-transformation of the C7 fragment and NLRP12 are not shown. (B) 933
NLRP12 binds to HCK
40
Top: schematic drawing of Hck truncations comprising the N1 to N5 fragments. Bottom: yeast 934
grew on the high stringency 4DO plates and 4DO +3AT plates when HCK’s C-terminal 40 935
amino acid fragment (C1) was co-transformed with NLRP12’s PYD + NBD (control). But yeast 936
did not grow on the high stringency 4DO plates nor 4DO+ 3AT plates when N1 and N5 937
fragments were co-transformed with NLRP12’s PYD + NBD, suggesting that the very C-terminal 938
5 amino acids are also necessary for binding of NLRP12. The colored boxes represent different 939
fragments that were deleted each time. 940
941
Figure 5: NLRP12 co-immunoprecipitates with HCK and co-localizes with HCK by 942
immunofluorescence. (A) NLRP12 co-immunoprecipitated with HCK when both proteins were 943
transiently exogenously co-expressed in 293T cells, with NLRP12 having to be expressed as 944
epitope-tagged forms. Cells were lysed using a non-denaturing cell lysis buffer (see Material 945
and Methods). Co-immunoprecipitations were done using a mouse monoclonal anti-HCK 946
antibody followed by protein A/G agarose. The immunoprecipitated proteins were analyzed by 947
immunoblotting. The same method was used for experiments shown in (B), (C), and (D). (B) 948
NLRP12 co-immunoprecipitated with endogenous HCK in macrophage-like RAW 264.7 cells, 949
when NLRP12 is exogenously and stably expressed. Arrows show the two isoforms of HCK p59 950
and p61 that were immunoprecipitated by the anti-HCK antibody, although it is not clear 951
whether NLRP12 co-immunopreciptitated with one form or the other, or both. (C) Exogenously 952
and stably expressed NLRP12 co-immunoprecipitated with endogenously expressed HCK in 953
monocyte-like THP-1 cells. In these cells, phorbol 12-myristate 13-acetate (PMA) (1µM) was 954
added to enhance the expression of NLRP12 under the PMA-sensitive cytomegalovirus promoter 955
of the NLRP12 plasmid. (D) Left panel shows that NLRP12 was exogenously and stably 956
NLRP12 binds to HCK
41
expressed in lymphoblast-like U937 cells. Right panel shows that NLRP12 co-957
immunoprecipitated with endogenously expressed HCK in U937 cells after PMA was added to 958
1µM, as was done for THP-1 cells, to enhance expression of NLRP12. (E) Immunofluorescent 959
images showing the colocalization of a cohort of NLRP12 with either HCK p59 or p61. HCK 960
was immunolabeled with a mouse monoclonal anti-HCK antibody, followed by a secondary goat 961
anti-mouse IgG (H+L) cross-adsorbed, conjugated to Alexa Fluor 488. For NLRP12, the 962
primary antibody was a rabbit monoclonal anti-FLAG antibody, and the secondary antibody was 963
a goat anti-rabbit IgG (H+L) cross-adsorbed secondary antibody, conjugated to Alexa Fluor 555. 964
Nuclei were counterstained by DAPI. 965
966
Figure 6: The co-expression of HCK protein significantly decreased the steady-state protein 967
expression levels of NLRP12 (A) Epitope-tagged NLRP12-V5 and HCK p61 were co-968
transfected into 293T cells. Cell lysates were immunoblotted with rabbit monoclonal anti-V5 969
antibody for NLRP12 and mouse monoclonal anti-HCK antibody. Comparing lane 2 (basal level 970
of NLRP12 expression) with lanes 5, 6, and 7, co-expression of HCK protein leads to a 971
significant decrease of protein expression levels of NLRP12 over three times they are co-972
expressed. Increasing protein expression of HCK, relative to the plasmid DNA added (ramp: 973
lanes 5, 6, and 7; 2.5-20 μg of DNA) was not observed possibly because the amount of plasmid 974
for HCK transfection were already too high, resulting in the maximum expression of HCK at all 975
plasmid DNA amounts. The figure is representative of the results from three experiments. (B) 976
Quantification of the NLRP12 protein bands (mean + standard error) to show that NLRP12 977
protein expression levels decreased in the presence of co-expressed HCK, relative to equal 978
amounts of cell lysate being assayed. The NLRP12 protein expression levels are indicated as 979
NLRP12 binds to HCK
42
arbitrary units. (C) Quantification of the HCK protein bands (mean + standard error) confirming 980
that HCK protein expression levels did not change although the amount of plasmid DNA for 981
HCK increased. Figure (B) and (C) were generated by Prism 6.0 (GraphPad Software, Inc., San 982
Diego, CA). Twenty μg of cell lysate proteins were loaded for each lane. ANOVA and Tukey 983
tests were used to compare the NLRP12 and HCK protein bands. 984
985
Figure 7: NLRP12 is co-expressed with Hck in acute myeloid leukemia (AML) patient samples. 986
Both NLRP12 expression and Hck expression data are shown as log2 values. The data are RNA 987
seq data obtained from cBioportal (provisional) database assessed on July 11, 2019. The graph 988
was taken from the cBioportal website. Spearman coefficient, indicating the association of 989
NLRP12 and Hck mRNA expression level (if NLRP12 and Hck mRNA expression levels are 990
non-parametric) = 0.80. Pearson coefficient, indicating the degree of the linear relationship 991
between the NLRP12 mRNA expression level and Hck mRNA expression level (if NLRP12 and 992
Hck mRNA expression levels are parametric) = 0.73. The p-values for Spearman coefficient and 993
Pearson coefficient are both < 0.001. 994
995
Figure 8: (A) Ingenuity Pathway Analysis (IPA) displaying the 13 distinctive hits from the total 996
results of the yeast two-hybrid screen for proteins that interacted with NLRP12. The figure only 997
shows the NLRP12-interacting proteins with hits that have a frequency larger than 2. The solid 998
lines indicate interacting proteins found through our yeast two-hybrid screen. The dashed lines 999
are the NLRP12-interacting proteins previously reported in published data. The purple-colored 1000
molecules are those associated with AML. (B) To validate an interaction shown in the IPA 1001
diagram, tumor necrosis factor receptor-associated factor 3 interacting protein 3 (TRAF3IP3)-1002
NLRP12 binds to HCK
43
3FLAG and NLRP12-V5 were co-immunoprecipitated from co-transfected 293T cells. A non-1003
denaturing cell lysate buffer (see Material and Methods) was used to lyse the cell, and co-1004
immunoprecipitations were performed with a mouse monoclonal anti-TRAF3IP3 antibody 1005
followed by protein A/G agarose. The immunoprecipitates were analyzed for NLRP12 by 1006
immunoblotting with rabbit monoclonal anti-V5 antibody. The results from the co-1007
immunoprecipitation are consistent with a confirmation of an interaction between TRAF3IP3 1008
and NLRP12. 1009
1010
1011
1012
1013
1014
1015
1016
NLRP12 binds to HCK
44
1017
Figure 1 1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
NLRP12 binds to HCK
45
1028
1029
Figure 2 1030
1031
1032
NLRP12 binds to HCK
46
1033
1034
1035
1036
NLRP12 binds to HCK
47
1037
Figure 3 1038
1039
NLRP12 binds to HCK
48
1040
1041
1042
NLRP12 binds to HCK
49
Figure 4 1043
1044
1045
NLRP12 binds to HCK
50
1046
Figure 5 1047
1048
NLRP12 binds to HCK
51
1049
1050
Figure 6 1051
1052
NLRP12 binds to HCK
52
1053
Figure 7 1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
NLRP12 binds to HCK
53
1064
Figure 8 1065
1066
1
Nucleotide binding domain and leucine-rich repeat pyrin domain-
containing protein 12: characterization of its binding to
hematopoietic cell kinase
Yue Zhang* and Curtis T. Okamoto 1
2
Department of Pharmacology and Pharmaceutical Sciences, School of Pharmacy, University of 3
Southern California, USA 90089-9121 4
5
*Correspondence: 6
Yue Zhang 7
zhan576@usc.edu 8
9
Running title: Nlrp12 binds to Hck 10
11
Key words: yeast two-hybrid, innate immunity, protein‐protein interaction, Nlrp12, Hck, acute 12
myeloid leukemia13
14
15
Abstract 16
Protein-protein interactions are key to defining the function of nucleotide binding domain (NBD) 17
and leucine-rich repeat (LRR), pyrin domain (PYD)-containing family of proteins 12 (Nlrp12). 18
cDNA encoding the human PYD + NBD of Nlrp12 was used as bait in a yeast two-hybrid screen 19
Nlrp12 binds to Hck
2
with a human leukocyte cDNA library as prey. Hematopoiesis cell kinase (Hck), a member of 20
the Src family of non-receptor tyrosine kinases, was among the top hits. The C-terminal 40 21
amino acids of Hck specifically bound to Nlrp12’s PYD + NBD, but not to that of Nlrp3 and 22
Nlrp8. Amino acids F503, Q507, L510, and D511 of Hck are critical for the binding of Hck’s C-23
terminal 40 amino acids to Nlrp12’s PYD + NBD. Additionally, the C-terminal 30 amino acids 24
of Hck are sufficient to bind to Nlrp12’s PYD + NBD, but not to its PYD alone nor to its NBD 25
alone. In cell lines that express Hck endogenously, it was co-immunoprecipitated with stably 26
expressed exogenous Nlrp12. Also, Nlrp12 co-immunoprecipitated and co-localized with Hck 27
when both were overexpressed in 293T cells. In addition, in this overexpression system, steady-28
state Nlrp12 protein expression levels significantly decreased when Hck was co-expressed. 29
Bioinformatic analysis showed that Hck mRNA co-occurred with Nlrp12 mRNA, but not with 30
other Nlrp proteins, in blood and marrow samples from acute myeloid leukemia (AML) patients. 31
The mRNA of Nlrp12 is also co-expressed with Hck in AML patient samples, and the levels of 32
mRNA expression of each are correlated. Together these data suggest that Nlrp12, through its 33
binding to Hck, may have an effect on the progression of AML. 34
35
Introduction 36
Pattern recognition receptors (PRR) are the first major line of cellular defense against damage-37
associated molecular patterns (DAMP) and pathogen-associated molecular patterns 38
(PAMP)(Takeuchi and Akira, 2010). Among PRR, one class of proteins is the nucleotide-39
binding domain (NBD) and leucine-rich repeat (LRR) receptor family (NLR). NBD and LRR 40
PYD protein 12 (Nlrp12) is one protein that belongs to the PYD subfamily of the NLR class and 41
is a multi-functional protein (Tuncer et al., 2014). Nlrp12’s functions likely all depend on 42
Nlrp12 binds to Hck
3
defined protein-protein interactions. For example, in anti-inflammatory functions of Nlrp12, 43
Nlrp12 interacts with tumor necrosis factor receptor associated factor 3 (TRAF3) and nuclear 44
factor kappa-light-chain-enhancer of activated B cells (NF-κB)-inducing kinase (NIK) in the 45
non-canonical NF-κB pathway, when toll-like receptors (TLR) are stimulated with Pam3Cys4 46
triacylated lipopeptides followed by addition of cluster of differentiation 40 (CD40) (Allen et al., 47
2012). Nlrp12’s inhibition of the non-canonical NF-κB pathway, thereby effecting an anti-48
inflammatory function, has been reported in many publications (7-11). For example, in colitis-49
driven colon cancer, Nlrp12 knock-out (KO) mice experience a higher frequency of death, weigh 50
less, have a shorter colon length, and have more severe disease progression than wild type (WT) 51
mice, all of which may occur through Nlrp12’s inhibition of the non-canonical NF-κB pathway 52
(Zaki et al., 2011; Allen et al., 2012). In addition, Nlrp12 KO mice have been reported to exhibit 53
higher innate immunity when it is infected by the virus because Nlrp12 inhibits the non-54
canonical NF-κB pathway; thus, in a KO mice, a more significant increase of NF- κB activity is 55
observed compared to WT mice when it is infected by the virus (Chen et al., 2019). And in the 56
case of Nlrp12 protecting the diversity of intestinal microbiota (Chen et al., 2017), thereby 57
preventing obesity (Truax et al., 2018), it has also been reported that Nlrp12 KO mice display 58
significantly decreased intestinal microbial diversity due to increased innate immunity, resulting 59
from the lack of inhibition of the non-canonical NF-κB pathway. 60
61
In addition, Nlrp12 also interacts with interleukin (IL)-1 receptor-associated kinase 1 (IRAK1) in 62
the canonical NF-κB pathway when TLRs are stimulated, also to inhibit the inflammatory 63
response (Ye et al., 2008). In contrast, for human studies of Nlrp12-associated autoinflammatory 64
disorders, no protein-protein interactions have been reported (Jeru et al., 2008; Jeru et al., 2011). 65
Nlrp12 binds to Hck
4
Functionally, however, Nlrp12 KO mice had a higher immunological response in the 66
autoimmune model of contact hypersensitivity (Arthur et al., 2010); thus, these mice are more 67
prone to develop contact hypersensitivity. This results most likely from Nlrp12 KO mice having 68
defective migration of their dendritic cells (DC). But, in the mouse study, no protein-protein 69
interactions had been reported either (Arthur et al., 2010). And Nlrp12 is protective against 70
Yersinia pestis infection through Nlrp12’s activation of the inflammasome which results in 71
secretion of IL-18 and IL-1β. But again, no protein-protein interactions of Nlrp12 within 72
inflammasomes had been characterized (Vladimer et al., 2012). Thus, there is a further need to 73
find and to characterize previously unreported Nlrp12-binding proteins. Knowing Nlrp12’s 74
protein binding partners may help to gain mechanistic insight into pathways previously 75
characterized as being regulated by Nlrp12. Moreover, identifying and validating potentially 76
novel binding partners of Nlrp12 may suggest novel functions for Nlrp12. 77
78
In humans, Nlrp12 is mainly expressed in macrophages, DCs, and neutrophils. In mice, Nlrp12 79
is primarily expressed in DCs and neutrophils (Arthur et al., 2010; Hornick et al., 2018; Chen et 80
al., 2019). The Nlrp12 protein has three domains, from N- to C-terminus: PYD, NBD, and LRR 81
domains (Damiano et al., 2004; Janowski et al., 2013; Tuncer et al., 2014). It is generally 82
accepted that the LRR domain, like Nlrp1’s LRR domain and Nlrp3’s LRR domain, is 83
responsible for sensing ligands or other types of activators of Nlrps in modulating their role in an 84
immunologic response (Martinon et al., 2004; Faustin et al., 2007; Kanneganti et al., 2007; 85
Sharif et al., 2019). However, interestingly, it has been reported that without the LRR domain, 86
Nlrp3 still can be activated by triggers of the canonical NF-κB pathway (Hafner-Bratkovič et al., 87
2018). The NBD domain is responsible for self-oligomerization when Nlrp is activated, and the 88
Nlrp12 binds to Hck
5
PYD domain is responsible for regulating downstream signaling pathways (and thus, it is called 89
the “interaction domain”) (Hafner-Bratkovič et al., 2018; Hornick et al., 2018). Here we used a 90
truncated form of Nlrp12 (PYD + NBD) to identify possible protein-protein interactions that may 91
regulate signaling pathways in which Nlrp12 is involved. In addition, to our knowledge, there 92
are currently no protein-protein interaction studies that have been performed for this truncated 93
form of Nlrp12 and may thus represent a new general approach to screening for protein-protein 94
interactions involved in signaling complexes with Nlrps. 95
96
A yeast two-hybrid screen was performed to identify potential binding partners of Nlrp12’s PYD 97
+ NBD. Hck, a member of the Src family of non-receptor tyrosine kinases, was one of the top 98
positive clones (“hits”) among all of the proteins that appeared to interact with Nlrp12. Further 99
experiments were done to confirm the novel binding of Nlrp12 to Hck and to characterize key 100
structural features of the binding interaction. In addition, co-expression of Hck with Nlrp12 101
resulted in a significant decrease of steady-state Nlrp12 protein expression levels. In 102
consideration together with a bioinformatics analysis included here, a potential interaction 103
between Nlrp12 and Hck may be involved in modulating the role of Hck in the pathogenesis of 104
AML (Roversi et al., 2015; Lopez et al., 2016; Roversi et al., 2017). 105
106
Results 107
Nlrp12 interacts with Hck in the yeast two-hybrid assay 108
A human leukocyte cDNA library was used as prey in the screening of binding partners for 109
Nlrp12 using the yeast two-hybrid assay. The basic principle of the yeast two-hybrid assay is 110
that if the cDNA library contains clones that express proteins, or parts thereof, that interact with 111
Nlrp12 binds to Hck
6
the co-expressed bait, the Nlrp12 PYD + NBD (Figure 1A) in this case, then the co-transformed 112
yeast clones will grow. The cDNA clones from the library that supported growth would then be 113
isolated and characterize, to identify what proteins interact with Nlrp12’s PYD + NBD. The 114
leukocyte library was used because Nlrp12 is mainly expressed in leukocytes in humans (Tuncer 115
et al., 2014; Linz et al., 2017; Hornick et al., 2018; Normand et al., 2018), and thus putative 116
binding partners for Nlrp12 are also likely to be expressed there. 117
118
The screening resulted in 119 positive clones from the leukocyte cDNA library (“hits”) 119
expressed from 48 distinct genes (Supporting information, Table S1) that were subsequently 120
identified as encoding for proteins (or portions thereof) that potentially interact with Nlrp12’s 121
PYD + NBD. The clones with the most hits are listed in Table 1. Among the 119 hits, which 122
included repeated hits of the same genes, 9 were individual clones that encoded fragments of a 123
member of the Src family of non-receptor tyrosine kinases, Hck, which was among the highest 124
number of hits for a single gene (Figure 1B and Table 1). Moreover, of the 9 positive Hck 125
clones, 5 were distinct Hck fragments (Figure 1C). Full-length Hck p61 was also obtained as a 126
hit in another independent yeast two-hybrid screen (data not shown). Hck has two isoforms, p59 127
and P61, in Homo sapiens. P61 and p59 share the same transcript, but p61 uses an uncommon 128
start codon CTG, and p59 uses the common one, ATG (Robbins et al., 1995), to give rise to two 129
distinct transcripts. Endogenously expressed in phagocytes, Hck p59 is mainly localized to the 130
plasma membrane, while Hck p61 is primarily localized to lysosomes, and both of them are also 131
found in the Golgi apparatus (Carreno et al., 2000). It has been suggested that overexpression of 132
p59 promotes the formation of membrane protrusions such as pseudopodia and overexpression of 133
p61 induces podosome rosettes and triggers actin polymerization around lysosomes (Carreno et 134
Nlrp12 binds to Hck
7
al., 2000; Guiet et al., 2008). However, there are no differences between p61 and p59 in the 135
region of Hck that appears to bind to Nlrp12’s PYD + NBD (see below), the C-terminal part of 136
Hck. Thus, detecting a full-length clone of only p61 as a hit, opposed to p59, appeared to be 137
coincidental. 138
139
In directed screens using the yeast two-hybrid assay, Nlrp12’s PYD + NBD domain binding 140
to Hck C-terminal 40 amino acids appears to be selective for Nlrp12 and Hck binding to 141
Nlrp12 appears to be selective for Hck 142
Initially, the Hck C-terminal 40 amino acids (“Hck C1 fragment”) was the shortest Hck fragment 143
from a positive clone that interacted with the bait, Nlrp12’s PYD + NBD (Figure 2A). Nlrp 3 144
and Nlrp8 were chosen based on the phylogenetic trees of the PYD + NBD domains of members 145
of the Nlrp family: phylogenetically, Nlrp3’s PYD + NBD is the most closely related Nlrp PYD 146
+ NBD to that of Nlrp12, while Nlrp8’s PYD + NBD is among the furthest related PYD + NBD 147
to that of Nlrp12 (Figure 2B). In Figure 2C, yeast clones grew on the 4 drop out (DO) (synthetic 148
dropout (SD) plus amino acids supplements minus tryptophan, leucine, histidine, and adenine (-149
Trp-Leu-His-Ade) and 4DO + 3-Amino-1,2,4-triazole (3-AT) high stringency plates when the 150
yeast were co-transformed with Nlrp12’s PYD + NBD and Hck C1 fragment; and thus, this 151
result is indicative of an interaction between the two constructs. However, yeast clones were not 152
observed on 4DO and 4DO + 3-AT plates when Nlrp3’s PYD + NBD and Hck C1 fragment were 153
co-transformed, nor when Nlrp8’s PYD + NBD and Hck C1 fragment were co-transformed. 154
These results suggest that the Hck C1 fragment binds to Nlrp12’s PYD + NBD, but not to those 155
from either Nlrp3 or Nlrp8, results consistent with a selective interaction between Nlrp12’s PYD 156
+ NBD and the Hck C1 fragment. 157
Nlrp12 binds to Hck
8
158
Similarly, yeast clones again grew on the high stringency 4DO and 4DO + 3-AT plates when the 159
Hck C1 fragment was co-transformed with Nlrp12’s PYD + NBD. However, clones did not 160
grow on the 4DO and 4DO + 3-AT plates when the C1 fragment of other members of the Src 161
family of non-receptor tyrosine kinases were co-transformed with Nlrp12’s PYD + NBD (Figure 162
2D). This experiment suggests that the interaction of Hck C1 fragment with Nlrp12 is selective 163
for Hck; the C1 fragment of other members of the Src family of non-receptor tyrosine kinases 164
did not bind to Nlrp12’s PYD + NBD (Figure 2D). 165
166
In directed screens using the yeast two-hybrid assay, amino acids F503, Q507, L510, and 167
D511 of Hck are critical for the binding between Nlrp12’s PYD + NBD and Hck C-terminal 168
40 amino acids 169
The Hck C1 fragment were divided into four helical regions according to its atomic structure 170
(Sicheri et al., 1997) (protein data bank (PDB) accession number: 1AD5): R1, R2, R3, and R4 171
(designated by us) (Figure 3A). As shown in the sequence alignments in Figure 3B, the primary 172
amino acid sequences of R2, R3, and R4 are very well conserved among members of the Src 173
family of non-receptor tyrosine kinases, while R1 is not. To characterize further the important 174
structural features of Hck’s binding to Nlrp12’s PYD + NBD, all of the amino acids comprising 175
each of the R2, R3, and R4 domains were mutated to alanine, but mutated one section at a time, 176
and were denoted as the Hck M2 fragment, Hck M3 fragment, and Hck M4 fragment, 177
respectively. The amino acids comprising the R1 region were mutated to those amino acids in 178
the R1 region of the Src family non-receptor tyrosine kinase Lyn, which has the closest 179
phylogenetic relationship to Hck (Figure 3C). All of the mutated constructs are illustrated in 180
Nlrp12 binds to Hck
9
Figure 3D. Yeast still grew on the high stringency 4DO + 3AT plates when Nlrp12’s PYD + 181
NBD was co-transformed with the Hck M1 fragment, Hck M2 fragment, and Hck M4 fragment. 182
But, yeast could not grow when Nlrp12’s PYD + NBD was co-transformed with Hck M3 183
fragment (Figure 3E). These experiments suggested that the R1, R2, and R4 regions are 184
dispensable for binding, while R3 may be important for binding. Further analysis was conducted 185
to identify individual critical amino acids of Hck’s R3 region binding to Nlrp12’s PYD + NBD 186
using alanine scanning mutagenesis. However, the intervening sequences between R1, R2, R3, 187
and R4 were not examined for binding. 188
189
To identify the key amino acid residues involved in binding of R3 to Nlrp12’s PYD + NBD, 190
each amino acid in R3 was individually mutated to alanine in an alanine scanning mutagenesis 191
analysis, on the background of the Hck C1 fragment. As shown by the yeast two-hybrid assay, 192
the mutations of F503A, Q507A, L510A, and D511A abolished the growth of yeast clones on the 193
high stringency 4DO when these yeast strains were co-transformed with Nlrp12’s PYD + NBD 194
(Figure 3F). These experiments suggest that these four amino acids are critical for binding. On 195
the other hand, yeast clones still grew on the high stringency 4DO plates when the Hck C1 196
fragments with the Y505A, S508A, or V509A mutations were co-transformed with Nlrp12’s 197
PYD + NBD (Figure 3F). These experiments suggest that Y505, S508, and V509 are 198
dispensable for binding of Hck C1 fragment to Nlrp12’s PYD + NBD. Whether E504 and I506 199
are critical for binding is unfortunately not clear as the yeast did not grow on the control plates 200
(SD/-Trp-Leu). Overall, however, these results suggest that the amino acids F503, Q507, L510, 201
and D511 in Hck are critical for the binding of Hck C1 fragment to Nlrp12’s PYD + NBD. 202
203
Nlrp12 binds to Hck
10
On the other hand, these four amino acid residues are all conserved in all members of the Src 204
family of non-receptor tyrosine kinases (Figure 3A), suggesting that, while they are important for 205
binding, they cannot account for the specificity or selectivity of binding between Nlrp12 and Hck. 206
In addition, these four critical residues, as shown in Figure 3C, appear to be in the same plane 207
along a helix, which may then form an interface that is important for binding to Nlrp12. 208
However, it is unclear whether the structure of the isolated Hck C-terminus used in the screens 209
here will retain its structural features that it possesses in full-length Hck, due to the loss of 210
intermolecular interactions for the C-terminal region. Moreover, for example, the amino acids 211
R500, P501, T502, D512, T515, Q520, and Y521, are 100% conserved among all other members 212
of the Src family of non-receptor tyrosine kinases and are therefore unlikely to play a role in 213
specificity and selectivity of binding of Nlrp12 with Hck. But, it is possible that two or more 214
amino acids in the intervening sections between the R1-R4 regions, which were not analyzed by 215
mutagenesis, may then actually account for the specificity and selectivity of binding of Hck to 216
Nlrp12. One example is that the combination of R496, P497, and Q522 is unique to Hck 217
compared to other members of the Src family of non-receptor tyrosine kinases. 218
219
In directed screens using the yeast two-hybrid assay, the shortest fragment of Hck that 220
binds to Nlrp12’s PYD + NBD is its C-terminal 30 amino acids 221
The shortest form of Hck that bound to Nlrp12’s PYD + NBD was determined by observing 222
whether the yeast clones grew on high stringency 4DO or 4DO +3AT plates when Nlrp12’s PYD 223
+ NBD was co-transformed with Hck fragments C1 to C7 and N1 to N5 (Figure 4) (data of 224
Nlrp12’s PYD + NBD co-transformed with Hck fragment C7 did not shown). Hck’s C1 to C7 225
fragments were gradually truncated from the N-terminus of the Hck C-terminal 40 amino acid 226
Nlrp12 binds to Hck
11
fragment (Figure 4A and 4B). Hck N1 to N5 fragments were gradually truncated from the C-227
terminus of the Hck C-terminal 40 amino acid fragment (Figure 4C and 4D). Yeast strains grew 228
on high stringency 4DO or 4DO +3AT plates when either the Hck C1, C2, or C3 fragment was 229
co-transformed with Nlrp12’s PYD + NBD. Thus, of all of the constructs screened, the shortest 230
fragment of Hck that can apparently bind to Nlrp12’s PYD + NBD is the C-terminal 30 amino 231
acids, as indicated by the Hck C3 fragment (Figure 4B and 4D). In addition, the C-terminal 5 232
amino acids also appear to be important for binding, as clones co-transformed with Nlrp12’s 233
PYD + NBD with N1, with a deletion of the C-terminal 5 amino acids of Hck, did not grow on 234
the high stringency 4DO and 4DO + 3AT plates (Figure 4D). The structures of both inactive 235
Hck (PDB accession number: 1AD5) and active c-Src (PDB accession number: 1Y57), the 236
prototypical member of this family of nonreceptor tyrosine kinases related to Hck, show that the 237
C-terminal 4 amino acids are surface-exposed, indicating that these four amino acids that are 238
critical for binding are available to interact with Nlrp12. 239
240
Nlrp12 co-immunoprecipitates with Hck from mammalian cells, and Nlrp12 is co-localized 241
with Hck in mammalian cells 242
Nlrp12 and Hck could be co-immunoprecipitated when they were both heterologously co-243
expressed in 293T cells (Figure 5A). Also, semi-endogenous conditions were used for co-244
immunoprecipitations from cells in which Hck is endogenously expressed, and Nlrp12 was 245
heterologously stably expressed in epitope-tagged form, such as in macrophage-like RAW 264.7 246
cells (Figure 5B), monocyte-like THP-1 cells (Figure 5C), and lymphocyte -like U937 cells 247
(Figure 5D). Semi-endogenous conditions have been used in previous reports where Nlrp12 was 248
shown to interact with TRAF3 (Allen et al., 2012), NIK (Lich et al., 2007), and IRAK-1(Ye et al., 249
Nlrp12 binds to Hck
12
2008). In all cases, Nlrp12 co-immunoprecipitated with the p59 and p61 isoforms of Hck, but 250
from these experiments, it is not clear that whether Nlrp12 interacts with p59, p61, or both 251
isoforms of Hck. In U937 cells and THP-1 cells, since heterologous Nlrp12 expression level was 252
low, adding phorbol 12-myristate 13-acetate (PMA) (1 µM for 24 hr) increased Nlrp12 253
expression level by stimulating the PMA-sensitive cytomegalovirus promoter of Nlrp12’s 254
expression plasmid (Figure 5C and 5D). Taken together, these data suggest that Nlrp12 and Hck 255
(either p59, p61, or both isoforms) can interact with each other in a mammalian cell system, a 256
result that in part validates the findings from the yeast two-hybrid assay. 257
258
After heterologous co-expression of Nlrp12 with either the p61 or the p59 isoform of Hck in 259
293T cells, immunofluorescent staining of Nlrp12 and Hck showed that Nlrp12 can be co-260
localized with either p61 or p59 isoforms of Hck in most of the 293T cells (Figure 5E and 261
Supporting Information Figure S1). The co-localization of Nlrp12 and Hck is consistent with an 262
intracellular interaction between the two proteins. In addition, the lack of selectivity of Nlrp12 263
co-localizing with either p61 or p59 isoform of Hck is consistent with the results from the 264
screening assay, where the putative binding region of Hck to Nlrp12 is identical in both the p61 265
and p59 isoforms. 266
267
The steady-state level of Nlrp12 protein is significantly decreased when Hck protein is co-268
expressed 269
Figure 6 shows that in 293T cells, the co-expression of Hck led to a significant decrease in 270
steady-state protein expression levels of Nlrp12. Although, alternatively, the apparent decrease in 271
expression level of Nlrp12 might be due actually to a decrease in the epitope-tag itself or a 272
Nlrp12 binds to Hck
13
decrease in the expression of the epitope-tagged protein only. Interestingly, Hck protein levels 273
did not show an increasing level of expression upon transfection with increasing amounts of 274
plasmid DNA, possibly due to Hck’s maximum expression levels in these cells being achieved 275
with the lowest amounts of plasmid DNA, which may be resolved if the transfections are 276
performed with lower amounts of plasmid DNA. 277
278
In blood and marrow samples from patients with AML, Nlrp12 and Hck co-occur and are 279
co-expressed 280
The odds ratio test (a statistical test that defines the likelihood of association of two events) 281
determines whether two proteins either co-occur or are mutually exclusive in their expression. 282
The Fisher exact test is a statistical test that uses P-values to determine if two events, such as co-283
occurrence or mutual exclusivity, are quantitatively significantly associated. Applying these 284
tests to The Cancer Genome Atlas (TCGA) provisional data sets for AML patients (Cerami et al., 285
2012a; 2013; Gao et al., 2013), Nlrp12 and Nlrp1 are the only two Nlrp family members that 286
show a statistically significant co-occurrence with Hck (Table 2). However, Nlrp1’s co-287
occurrence has a P value of 0.048, while Nlrp12’s co-occurrence has a P value of <0.001. 288
Therefore, Nlrp12 is most likely the only Nlrp family protein that shows co-occurrence with Hck 289
in samples from patients with AML. 290
291
Moreover, in this analysis, Nlrp12 is significantly co-expressed with Hck (Pearson coefficient = 292
0.80 and Spearman coefficient = 0.73, Pearson and Spearman coefficients show the possibility of 293
a linear relationship of two factors, Figure 7). This analysis is also consistent with Nlrp12 and 294
Hck co-occurring in the same patients with AML. However, Nlrp12 was not the only Nlrp 295
Nlrp12 binds to Hck
14
protein that shows significant co-expression with Hck. Nlrp1 and Nlrp3 also showed significant 296
co-expression with Hck. This can be explained by the fact that co-expression is less stringent 297
than co-occurrence. In summary, Nlrp12 is the only Nlrp protein that shows co-occurrence with 298
Hck, and it also shows co-expression with Hck, although it is not the only member of the Nlrp 299
family to do so. Other data sets, including Hausera data, Metzeler 1 data, Metzeler2 data, and 300
Raponi data, were not available to evaluate the co-expression of Nlrp12 versus Hck, as these 301
other data sets did not have information regarding Nlrp12 expression. 302
303
Other potential binding partners of Nlrp12 from the yeast two-hybrid screen 304
Other potentially interesting and unique genes that whose products may interact with Nlrp12’s 305
PYD + NBD were identified in the yeast two-hybrid screen (Table 1 and Supporting information 306
Table S1). Table 1 only shows those hits with frequency larger than 2, and these hits were 307
analyzed through an Ingenuity Pathway Analysis (IPA) diagram (Figure 8). In the IPA diagram, 308
protein-protein interactions found by our yeast two-hybrid screen are connected by solid lines 309
and interactions between protein binding partners collated by IPA are connected by dashed lines. 310
In addition, those molecules that are associated with AML are colored purple. Interestingly, 311
Nlrp12 is also associated with AML from the database Catalogue of somatic mutations in cancer 312
(COSMIC). Although the mutation occurs in the NBD of Nlrp12, only 2 samples of these 313
mutations are observed within 175 samples. 314
315
The interaction of one of the putative binding partners to Nlrp12 identified in our screen, TRAF3 316
interacting protein 3 (TRAF3IP3), was validated in a co-immunoprecipitation experiment where 317
the FLAG-tagged form of TRAF3IP3 and V5-tagged form of Nlrp12 were overexpressed in 318
Nlrp12 binds to Hck
15
293T cells. TRAF3IP3 was coimmunoprecipitated by adding mouse monoclonal anti-319
TRAF3IP3 antibody, and a band that showed the same molecular weight size of Nlrp12 was 320
detected by rabbit monoclonal anti-V5 antibody (Figure 8). However, further validation of these 321
other putative interactors detected by the yeast two-hybrid needs to be done. 322
323
Discussion 324
In this paper, the yeast two-hybrid assay was used to screen for Nlrp12 binding partners. This 325
approach identified 48 genes as positive hits encoding proteins that could possibly bind to 326
Nlrp12. Among the hits, the Src non-receptor tyrosine kinase family member Hck was selected 327
to focus upon as a potentially interesting target, particularly given its association with 328
hematopoiesis and AML. Further direct screens found that Nlrp12 specifically binds to Hck, but 329
not to Nlrp8 or Nlrp3. Conversely, Hck, but not other members of the Src non-receptor tyrosine 330
kinase family, specifically binds to Nlrp12. Hck’s C-terminal 30 amino acids are sufficient for 331
binding to Nlrp12. And, amino acids F503, Q507, L510, and D511 of Hck are critical for 332
binding to Nlrp12. Co-immunoprecipitation experiments from mammalian cells confirmed that 333
Nlrp12 and Hck interact. Co-expression of Nlrp12 and Hck results in a significantly lower 334
steady-state expression level of Nlrp12; thus, the interaction may affect Nlrp12 protein stability. 335
Finally, a bioinformatics analysis shows that Nlrp12 and Hck essentially exclusively co-occur, 336
and mRNA levels for both are correlated in blood samples from AML patients. Together, these 337
data suggested that Nlrp12 and Hck protein and mRNA co-expression may have an effect on the 338
AML. 339
340
Nlrp12 binds to Hck
16
The binding data from yeast two-hybrid (Figure 3 and Figure 4) showed that the R2 and R3 341
regions are both important for binding, but we suspect that R3 region is the major region that is 342
critical for binding, while R2 region may be an “auxiliary” region that is also critical for binding, 343
because when we delete the R2 region but maintain the R3 region, the R3 region itself cannot 344
maintain stable binding of Hck’s C-terminal 40 amino acids to Nlrp12’s PYD + NBD. Thus, 345
binding of Nlrp12’s PYD + NBD to this fragment of Hck may occur initially through binding to 346
R3, followed by stabilization of the interaction by secondarily binding to R2. This model is also 347
consistent with the finding that when we mutated the R3 region, the binding of Hck’s C-terminal 348
40 amino acid fragment to the Nlrp12’s PYD + NBD was lost despite the R2 region being 349
present. The above evidences are consistent with the model that R2 and R3 regions together are 350
both necessary for binding, but the stabilization of binding to the R3 region requires the presence 351
of R2. 352
353
The yeast two-hybrid screen with Nlrp12 also identified 47 other genes as potential interactors. 354
According to the Search Tool for Recurring Instances of Neighbouring Genes (STRING) protein 355
network and previous publications, Nlrp12 interactors include IRAK1 (Williams et al., 2005), 356
NIK (Lich et al., 2007), TRAF3(Allen et al., 2012), heat shock protein 90 (HSP90) (Arthur et al., 357
2007), Fas associated factor 1 (FAF1) (Pinheiro et al., 2011), and apoptosis-associated speck-like 358
protein containing a caspase activation and recruitment domain (ASC) proteins. IRAK1, NIK, 359
TRAF3, and Heat shock protein (HSP) 90 were only found in experiments involving semi-360
endogenous protein co-immunoprecipitations with Nlrp12. FAF1 was found to be an interacting 361
protein in NMR experiments by showing a difference in the chemical shift perturbation of 362
FAF1’s ubiquitin-associated domains (UBA) domain alone and Nlrp12’s PYD domain bound to 363
Nlrp12 binds to Hck
17
FAF1’s UBA domain. And FAF1 was also found to interact with Nlrp12 in a yeast two-hybrid 364
study where the PYD domain of different Nlrp proteins were screened (Kinoshita et al., 2006). 365
However, in our study, IRAK1 (Shi et al., 2019), NIK (Fu et al., 2017), TRAF3 (Wang et al., 366
2018), HSP90 (Kinnaird et al., 2017), FAF1 (Ryu et al., 2003), and ASC (Martin et al., 2016) 367
proteins were not found in our results of interacting proteins, although they are all expressed in 368
leukocytes. Therefore, given these differences, the results from this yeast two-hybrid screen 369
may provide a unique database for future investigations on the proteins that interact with Nlrp12, 370
and the functional outcomes of these interactions. 371
372
How Hck co-expression affects Nlrp12 expression can be further investigated to determine the 373
functional outcomes of the Nlrp12-Hck interaction relative to protein stability or degradation. 374
For example, MG-132 and chloroquine, which are a proteasomal inhibitor and an inhibitor of 375
lysosomal degradation, respectively, can be added to cells in these experiments to determine 376
whether the decreasing protein expression level of Nlrp12 in the presence of Hck is due to 377
proteasomal degradation or lysosomal degradation. In addition, an Nlrp12 stable cell line, with a 378
typically lower level of Nlrp12 expression, compared to transiently transfected cells, can be used 379
to test formally whether Nlrp12 protein expression levels are increasingly decreased when Hck is 380
expression is increased. Moreover, the rate of Nlrp12 degradation, in the absence or presence of 381
Hck, can be studied and compared in pulse-chase assays (Lich et al., 2007). 382
383
Given the interaction that was characterized between Nlrp12 and Hck, we sought to identify 384
diseases in which their interaction may be relevant. AML is a type of blood and bone marrow 385
cancer that is characterized by an unusually high level of circulating immature white blood cells 386
Nlrp12 binds to Hck
18
(Prada-Arismendy et al., 2017). Hck is found to be highly expressed in hematopoietic stem cells 387
of AML patients (Roversi et al., 2017). The overexpression of Hck is linked with the 388
development of AML (Poh et al., 2015; Roversi et al., 2017). Therefore, Hck may be potentially 389
used as a blood biomarker to identify those patients with potential to develop of AML disease. 390
In addition, activation of Hck and its signaling pathways often causes the development of AML 391
(Dos Santos et al., 2008; Lopez et al., 2016). Hck has been reported to be involved in the feline 392
McDonough sarcoma (fms)-like tyrosine kinase 3 (FLT3)-HCK- cyclin D-dependent kinase 393
(CDK6) pathway. The mutation in FLT3-internal tandem duplications (FLT3-ITD) causes the 394
activation of Hck, and subsequently results in overexpression of CDK6 in MV4-11 cells, a 395
human AML cell line with the FLT3-ITD mutation. CDK6 siRNA resulted in MV4-11 cells 396
with the FLT3-ITD mutation having a lower proliferation rate compared to MV4-11 control cells 397
(Lopez et al., 2016). Also, it has been reported that in AML patients, Hck, Lyn, and Fgr 398
phosphorylation levels are high (Dos Santos et al., 2008). In summary, the above evidence 399
suggests that either high expression and/or high activation of Hck could occur in AML patients 400
as well. 401
402
Our bioinformatics analysis revealed that Nlrp12 is likely the only member of the Nlrp family 403
that shows significant co-occurrence with Hck at the level of mRNA expression. However, our 404
cellular functional study revealed that Nlrp12 protein expression levels decreased when it is co-405
expressed with Hck. This discrepancy might be due Nlrp12 mRNA expression level not 406
correlating with protein expression level (Greenbaum et al., 2003; Vogel and Marcotte, 2012; 407
Liu et al., 2016). Nevertheless, these data are still consistent with the possibility that Nlrp12 may 408
have an effect on AML through its binding to Hck. Nlrp12 and Hck co-expression and co-409
Nlrp12 binds to Hck
19
occurrence may then also be relevant to the prognosis of AML. However, further analyses need 410
to be performed to evluate the ability of Nlrp12 and Hck to serve as biomarkers in the prognosis 411
of AML. 412
413
Nlrp12 and Hck co-expression and co-occurrence may indicate that Nlrp12 and Hck mRNA or 414
protein expression can impact the same pathway or both be regulated by the same molecule. For 415
example, in the case two molecules exhibiting co-occurrence, they may interact in the same 416
signaling pathway, for example, Ki-ras2 Kirsten rat sarcoma viral oncogene homolog (KRAS) 417
and glycogen synthase kinase 3 beta (GSK3β) showed co-occurrence in the AML provisional 418
data, and they are in the same signaling pathway (Prochnicki and Latz, 2017). However, further 419
analyses need to be performed to evaluate the ability of Nlrp12 and Hck to serve as biomarkers 420
in the prognosis of AML. 421
422
Although many Hck inhibitors have been developed, they often lack specificity, with Hck 423
inhibitors also inhibiting other members of the Src non-receptor tyrosine kinase family (Hu et al., 424
2004; Pene-Dumitrescu et al., 2008; Saito et al., 2013; Patel et al., 2017; Roversi et al., 2017; 425
Naganna et al., 2019). If the potential for the development of AML can be decreased when 426
Nlrp12 does not bind to Hck, then an inhibitor to the interaction between Nlrp12 and Hck can be 427
designed. On the other hand, if the potential for AML is increased when Nlrp12 and Hck does 428
not bind, then a drug that facilitates Nlrp12 and Hck interaction could be developed. By this 429
approach, the non-specificity problem with Nlrp inhibitors may still exist but perhaps to a 430
smaller extent, as Nlrp proteins are less conserved overall compared to the members of the Src 431
non-receptor tyrosine kinase family. For example, the amino acid identity between Nlrp12’s 432
Nlrp12 binds to Hck
20
PYD + NBD versus another Nlrp PYD + NBD most closely related to Nlrp12, Nlrp3’s PYD + 433
NBD, is 54%. But, the protein identity between the Hck C1 fragment versus the one with the 434
closest relationship, the Lyn C1 fragment, is 78%. 435
436
In conclusion, this paper reports the preliminary findings of Nlrp12 interacting with Hck, and 437
Nlrp12 co-occurs with Hck in AML patients. The binding data presented here will expand the 438
database regarding the interactors of Nlrp12, and additional bioinformatic and functional in vitro 439
and in vivo analyses may shed light on the functional consequence of the interaction of Nlrp12 440
with Hck. 441
442
Material and Methods 443
Reagents and cells 444
U937, THP-1, RAW 264.7, and 293T cells were all purchased from American Type Culture 445
Collection (ATCC) (Manassas, VA). pCDNA-Nlrp12 plasmid and pCDNA-Nlrp3 plasmid were 446
gifts from Dr. Jenny P. Y. Ting (University of North Carolina-Chapel Hill). pCDNA-Nlrp8 447
plasmid was purchased from Genescript (Piscataway, NJ). All other chemicals were reagent 448
grade. The mouse monoclonal anti-FLAG antibody and the rabbit monoclonal anti-FLAG 449
antibody were purchased from Sigma-Aldrich (St. Louis, MO). The mouse monoclonal anti-Hck 450
antibody, the mouse monoclonal anti-TRAF3IP3 antibody, and the mouse IgGκ binding protein-451
horseradish peroxidase (HRP) for detecting the anti-Hck antibody were purchased from Santa 452
Cruz Biotechnology (Dallas, TX). The rabbit monoclonal anti-V5 antibody and polyclonal 453
rabbit HRP-conjugated anti-mouse IgG secondary antibodies were purchased from Cell 454
Signaling (Beverly, MA). The goat anti-rabbit IgG HRP-linked secondary antibody was 455
Nlrp12 binds to Hck
21
purchased from Cell Signaling. The Clean-Blot™ IP Detection Reagent was purchased from 456
Thermo Fisher Scientific (Grant Island, NY). Goat anti-mouse IgG (H+L) cross-adsorbed 457
secondary antibody, conjugated to Alexa Fluor 488 and goat anti-rabbit IgG (H+L) cross-458
adsorbed secondary antibody, conjugated to Alexa Fluor 555 were purchased from Thermo 459
Fisher Scientific (Grant Island, NY). The 4',6-diamidino-2-phenylindole, dihydrochloride (DAPI) 460
was purchased from Thermo Fisher Scientific. 461
462
Cloning 463
Nlrp12’s PYD + NBD (a.a. 1-531) was amplified by polymerase chain reaction (PCR) (primers 464
see Table 3), from a pCDNA-Nlrp12 plasmid, was digested with enzymes NdeI and SalI, and 465
subcloned into the pGBKT7 vector (Takara). The sequence encoding the PYD + NBD of Nlrp3 466
(a.a. 1-536) was amplified by PCR (see table 3 for primers), digested with enzymes NdeI and 467
BamHI, and subcloned in the pGBKT7 vector. The sequence encoding the PYD + NBD of 468
Nlrp8 (a.a. 1- 527) was amplified by PCR (see Table 3 for primers), digested with enzymes 469
EcoRI and BamHI, and subcloned into the pGBKT7 vector. The entire sequences of the C-470
terminal 42 amino acids of Fgr, Src, Yes, and Fyn, and 40 amino acids of Blk, Lck, Lyn, and 471
Hck (The C terminal 40 (or 42) amino acids is called “C1”), and Hck constructs C2, C3, C4, C5, 472
C6, C7, N1, N2, N3, N4, N5, M1, M2, M3, M4, and Hck C1 mutations for alanine scanning with 473
F503A, E504A, Y505A, I506A, Q507A, S508A, V509A, L510A, and D511A, were all 474
synthesized from IDT DNA (Coralville, IA) and cloned into the pACT2 vector (Takara). Nlrp12 475
was amplified by PCR (see Table 3 for primers), digested with AflII and XbaI, and subcloned 476
into pEFIRES-P-3FLAG (Hobbs et al., 1998) and pEFIRES -P-V5 (Hobbs et al., 1998). 3FLAG 477
and V5 are homemade, i.e., added later into pEFIRES-P. The peptide sequence of the 3FLAG 478
Nlrp12 binds to Hck
22
epitope is DYKDDDDKDYKDDDDKDYKDDDDK. The peptide sequence of the V5 epitope is 479
GKPIPNPLLGLDST. The multiple cloning sites of pEFIRES-P were modified according to our 480
needs. Sequences encoding Hck p61 and p59 isoforms were amplified by PCR (see Table 3 for 481
primers), digested with AflII and XbaI, and subcloned into pEFIRES -3FLAG vector. Final 482
plasmids were sequenced to confirm the constructs. 483
484
Yeast two-hybrid assay 485
Yeast transformed with the plasmid PGBKT7- Nlrp12’s PYD + NBD 486
Yeast strain AH109 (Takara, Mountain View, CA) was stored frozen at -80°C. Upon use, it was 487
streaked onto an agar plate supplemented with yeast extract, peptone, dextrose, and adenine 488
hemisulfate (YPDA) (Takara), which was then placed in a 30°C incubator overnight. The 489
growing colonies were picked and placed into 10 ml of YPDA media, incubated overnight with 490
shaking at 30°C. Then they were transformed with Nlrp12’s PYD + NBD plasmid, following 491
recommendations in the Takara manual (MatchmakerTM Gal4 Two-hybrid System 3 and 492
Libraries User Manual). A 4ml of 1 M of 3-AT was added into 800 mL of the YPDA medium 493
that were used for making the plates due to a leakage problem. 3-AT is a competitive inhibitor 494
for the product of the HIS3 reporter gene. Yeast cells were lysed using a trichloroacetic acid 495
(TCA) method, and western blots were performed to check Nlrp12’s PYD + NBD protein 496
expression (data not shown). 497
498
Library scale transformation of cDNA library into yeast that have been transformed with the 499
PGBKT7- Nlrp12 PYD + NBD plasmid 500
Nlrp12 binds to Hck
23
A human leukocyte MatchmakerTM cDNA library was purchased from Takara. The library was 501
titered according to the MatchmakerTM Gal4 Two-hybrid System 3 and Libraries User Manual. 502
The colony-forming units (CFU) for the library were calculated to be about 3.1 x 109 CFU/ml. 503
Four large scale transformations (equal to one library scale) were done at the same time 504
following the MatchmakerTM Gal4 Two-hybrid System 3 and Libraries User’s Manual. To 505
calculate the efficiency of yeast two-hybrid (i.e., the number of colons obtained per µg of library 506
plasmid DNA) and to check the expressions of “bait” and “prey” plasmids, small amounts of the 507
solution, diluted 10X, 100X, 1000X, and 10,000X were plated in SD/-Trp-Leu 2DO (drop out) 508
plates. The yeast two-hybrid efficiency obtained was 3 x 106 CFU per µg of library plasmid 509
DNA. In addition, if the yeast grew on the SD/-Trp-Leu-His-Ade + 3-AT plates, the “bait” and 510
“prey” plasmids were considered to be interacting with each other. 511
512
Small scale transformation 513
The yeast two-hybrid co-transformation on a small scale for all of the direct screenings was 514
performed following the small-scale transformation protocol in the MatchmakerTM Gal4 Two-515
hybrid System 3 & Libraries User’s Manual. The direct screens were to test whether Nlrp12 516
PYD + NBD and Hck specifically interact; identification of Hck F503, Q507, L510, and D511 517
being critical for binding with Nlrp12 PYD + NBD; and the characterization of the Hck C-518
terminal 30 amino acid fragment binding to Nlrp12 PYD + NBD. 519
520
Cell transfection 521
293T cells, THP-1 cells, U937 cells, and RAW 264.7 cells were seeded at a density of 5 x 105 522
per well in a 6-well plate. Transfection followed the Lipofectamine® 3000 reagent (Thermo 523
Nlrp12 binds to Hck
24
Fisher Scientific) protocol with the following modifications. Briefly, 3.75 µl of Lipofectamine® 524
3000 was added into 125 µl of Opti-MEM® medium (Thermo Fisher Scientific). 2.5 µg of 525
plasmid DNA (prepared by isolation through cesium chloride (CsCl) density gradient 526
purification (Brakke, 1951) and 5 µl of P3000TM reagent were added into another aliquot of 125 527
µl of Opti-MEM® medium. Then these two tubes were mixed together, and the mixed tube was 528
incubated at room temperature for 15 min. 250 µl of this mixture was added to the cells. The 529
cells were incubated at 37 °C with 5% CO2 for two days until analysis. 530
531
Nlrp12 stable cell lines 532
THP-1 cells, U937 cells, and RAW 264.7 cells were transfected with pEFIRES-P-Nlrp12-533
3FLAG plasmid, and 2.5 μg/ml puromycin (InvitrogenTM by Thermo Fisher Scientifc) was added. 534
After two weeks, the cells were considered to be stably expressing Nlrp12 after screening for 535
expression. The stable cells were “pooled” clones, and they are maintained by adding 2.5 μg/ml 536
puromycin. 537
538
Cell lysis and immunoprecipitation 539
After trypsinization, cells were lysed in 500 μl of lysis buffer (1% NP-40, 4mM Tris-HCl, pH 8.0, 540
150 mM NaCl). Cells were then treated with freeze-thaw cycles, i.e., frozen in the -80C freezer 541
for 5 min and thawed at 37C in a water bath for 3 min, for a total of five times. Cells were then 542
centrifuged at 12,000 x g for 10 min at 4C. After centrifugation, the supernatant was saved at -543
20C until further use. 544
545
Nlrp12 binds to Hck
25
For immunoprecipitations, cell lysates were pre-cleared by adding 40 μl of protein A/G agarose 546
beads (Thermo Fisher Scientific) from a 50% slurry, and the sample was rotated for 1 hr at 4C. 547
The sample was spun at 1000 x g for 1min. Then, 30 μl of pre-cleared cell lysate were removed 548
and run on immunoblotting as a control. The remaining cell lysate was incubated with 1 μg anti-549
Hck mouse monoclonal primary antibody overnight. On the second day, 30 μl of protein A/G 550
agarose, from a 50% slurry and previously washed with lysis buffer three times, were added to 551
the tube, and the tube was incubated for 3 hrs at 4C. The beads were then washed with lysis 552
buffer, but with the NaCl concentration increased to 500 mM. The tube was briefly centrifuged 553
at 1000 x g for 1min., and the supernatant was removed. This step was performed for a total of 5 554
times. The proteins were eluted from the beads by addition of SDS-PAGE sample buffer. 555
556
Immunoblotting 557
Protein concentrations were determined by the PierceTM BCA protein assay (Thermo Fisher 558
Scientific). About 20µg of cell lysate samples and immunoprecipitated samples were loaded 559
onto the 8% acrylamide gel. After overnight transfer of the proteins from an SDS-gel to a 0.45 560
μm pore size polyvinylidene difluoride (PVDF) membrane, the membrane was blocked in for 1 561
hr at room temperature 5% non-fat dry milk dissolved in (Tris-buffered saline with Tween 20, 562
pH=8.0) (TBST) (Sigma-Aldrich). Mouse monoclonal anti-FLAG antibody (diluted 1:1000) and 563
mouse monoclonal anti-Hck antibody (diluted 1:1000) were added and the membrane incubated 564
overnight. For detection of bands in the cell lysate, rabbit anti-mouse IgG HRP-linked secondary 565
antibody (diluted 1:2000) (for anti-FLAG antibody) and mouse IgGκ binding protein-HRP 566
secondary antibody (for anti-Hck antibody) (diluted 1:1000) were added, and the membrane was 567
incubated for 1 hr at room temperature. For immunoprecipitations, Clean-Blot™ IP detection 568
Nlrp12 binds to Hck
26
reagent (HRP) secondary antibody (1:300) (Thermo Fisher Scientific) was used. The detection 569
signal was visualized by using enhanced chemiluminescence (ECL) reagent (Thomas Scientific, 570
Swedesboro, NJ). Immoblots were analyzed by chemiluminescence on a Bio-Rad ChemiDocTM 571
Touch Gel imaging system (Bio-Rad, Hercules, CA). 572
573
Immunofluorescence 574
A protocol for immunofluorescence was followed 575
(https://www.cellsignal.com/contents/resources-protocols/immunofluorescence-general-576
protocol/if), with the following modifications. Coverslips were autoclaved and coated with 0.1 577
mg/ml of poly-D-lysine hydrobromide (mol wt 70,000-150,000) (Sigma-Aldrich). Cells were 578
plated onto the coverslips. Fixation was stopped by adding 50 mM NH4Cl in PBS for 15 min at 579
room temperature. Fixed cells were permeabilized by adding 0.5% Triton®-X-100 in PBS for 580
10 min. The cells were then washed with PBS three times and blocked with 0.1% of BSA in 581
PBS at room temperature for 1 hr. The rabbit monoclonal anti-FLAG antibody (diluted 1:100) 582
and mouse monoclonal anti-Hck antibody (diluted 1:100) were added to each cover slip and 583
incubated overnight. Goat anti-mouse IgG (H+L) cross-adsorbed secondary antibody, 584
conjugated to Alexa Fluor 488 (diluted 1:100) and goat anti-rabbit IgG (H+L) cross-adsorbed 585
secondary antibody, conjugated to Alexa Fluor 555 (1: 100) were added. After incubation with 586
secondary antibodies, the cells were washed five times with PBS. During the first step, DAPI 587
(InvitrogenTM by Thermo Fisher Scientific) was also added at 1:1000 in PBS for 2 min. Finally, 588
the coverslips were mounted in ProLong™ Gold Antifade mounting medium (Life Technologies 589
Corporation by Thermo Fisher Scientific, Eugene, OR). The stained slides were viewed on a 590
Nlrp12 binds to Hck
27
ZEISS LSM 880 (Pleasanton, CA) with Airyscan. The images were viewed and processed using 591
Fiji software (Schindelin et al., 2012). 592
593
IPA diagram 594
Yeast two-hybrid data were analyzed by IPA software (QIAGEN Inc., 595
https://www.qiagenbioinformatics.com/products/ingenuity- pathway-analysis). 596
597
Statistical analysis 598
The Spearman coefficient and Pearson coefficient used for co-expression was generated by 599
cBioPortal website (Cerami et al., 2012b; 2013; Gao et al., 2013). Clinical samples from the 600
Oncomine website (https://www.oncomine.org) were downloaded and checked for Nlrp12 601
expression levels and Hck levels. The Fisher exact test and the odds ratio test used for co-602
occurrence were generated by the cBioPortal website. Analysis of variance (ANOVA) was 603
performed to check whether a significant difference of the treatment exists among all the groups. 604
A p-value < 0.05 was considered statistically significant. If P < 0.05, then the Tukey test was 605
used to determine whether a pair-wise difference exists. Quantification of the western blot bands 606
was performed by FIJI software (Schindelin et al., 2012) and made into graphs by Prism 6.0 607
(GraphPad Software, Inc., San Diego, CA).608
609
Acknowledgement 610
Nlrp12 binds to Hck
28
Thank you to Dr. Houda Alachkar for providing valuable suggestions to this paper. We also 611
thank Ms. Meng Li and Dr. Yibu Chen at the University of Southern California’s Norris Medical 612
Library for their suggestions for bioinformatic analysis. Appreciation is given to Mr. Noam 613
Morningstar-Kywi and Dr. Ian Haworth for their helps on structural analysis. 614
615
Abbreviations 616
3-AT, 3-Amino-1,2,4-triazole 617
AML, acute myeloid leukemia 618
ANOVA, analysis of variance 619
ASC, apoptosis-associated speck-like protein containing a caspase activation and 620
recruitment domain 621
ATCC, American Type Culture Collection 622
C1, self-named truncations of C-terminal 40 (or 41) amino acids fragment of Src family non-623
receptor tyrosine kinases 624
C2-C7 and N1-N5, self-named truncations Hck C1 fragment 625
CD40, cluster of differentiation 40 626
CDK6, cyclin D-dependent kinase 627
CFU, colony-forming units 628
COSMIC, catalogue of somatic mutations in cancer 629
CsCl, cesium chloride 630
DAMP, damage-associated molecular patterns 631
DAPI, 4',6-diamidino-2-phenylindole, dihydrochloride 632
DC, dendritic cells 633
Nlrp12 binds to Hck
29
DO, drop out (2DO: SD/-Trp-Leu; 4DO: SD/-Trp-Leu-His-Ade); 634
ECL, enhanced chemiluminescence 635
FAf1, Fas associated factor 1 636
FLT3-ITD, FLT3-internal tandem duplications 637
FLT3, feline McDonough sarcoma (fms)-like tyrosine kinase 3 638
GSK3b, glycogen synthase kinase 3 bet 639
Hck C1 fragment, Hck C-terminal 40 amino acids 640
Hck, Hematopoiesis cell kinase 641
HRP, horseradish peroxidase 642
HSP90, heat shock protein 90 643
IL, interleukin 644
IPA, ingenuity pathway analysis 645
IRAK1, IL-1 receptor-associated kinase 1 646
KO, knock out 647
KRAS, Ki-ras2 Kirsten rat sarcoma viral oncogene homolog 648
LRR, leucine rich repeat domain 649
M1, the amino acids in the R1 region were mutated to the amino acids identical to the R1 region 650
in Lyn; 651
M2, the amino acids in the R2 region were all mutated to Alanine 652
M3, the amino acids in the R3 region were all mutated to Alanine 653
M4, the amino acids in the R4 region were all mutated to Alanine 654
NBD, nucleotide binding domain 655
NF-κB, nuclear factor kappa-light-chain-enhancer of activated B cells 656
Nlrp12 binds to Hck
30
NIK, NF-κB-inducing kinase 657
NLR, nucleotide binding domain and leucine-rich repeat 658
Nlrp, NLR pyrin domain-containing protein 659
PAMP, pathogen-associated molecular patterns 660
PCR, polymerase chain reaction 661
PDB, protein data bank 662
PMA, phorbol 12-myristate 13-acetate 663
PRR, pattern recognition receptors 664
PVDF, polyvinylidene difluoride 665
PYD, pyrin domain 666
R1, R2, R3, and R4, self-named four regions in the Hck C terminal 40 amino acids fragment 667
SD, synthetic dropout 668
SD/-Trp-Leu, SD base plus amino acids supplements minus tryptophan and leucine 669
SD/-Trp-Leu-His-Ade, SD base plus amino acids supplements minus tryptophan, leucine, and 670
histidine, and adenine 671
STRING: search tool for recurring instances of neighboring genes 672
TBST: tris-buffered saline with Tween 20, pH=8.0 673
TCA, trichloroacetic acid 674
TCGA, the cancer genome atlas 675
TLR, toll-like receptor 676
TRAF3, tumor necrosis factor receptor associated factor 3 677
TRAF3IP3, TRAF3 interacting protein 3 678
UBA, ubiquitin-associated domains 679
Nlrp12 binds to Hck
31
WT, wild type 680
YPDA, yeast extract, peptone, dextrose, and adenine hemisulfate. 681
682
Conflict of Interest 683
The authors declare no conflict of interest regarding publication of this article. 684
685
Author Contributions Statement 686
YZ design the experiment and wrote and revised the manuscript. CO helped to design the 687
experiments and to revise the manuscript. 688
689
References 690
Allen, I. C., Wilson, J. E., Schneider, M., Lich, J. D., Roberts, R. A., Arthur, J. C., et al. (2012). 691
NLRP12 suppresses colon inflammation and tumorigenesis through the negative 692
regulation of noncanonical NF-kappaB signaling. Immunity. 36(5), 742-754. 693
doi:10.1016/j.immuni.2012.03.012 694
Arthur, J. C., Lich, J. D., Aziz, R. K., Kotb, M., and Ting, J. P. (2007). Heat shock protein 90 695
associates with monarch-1 and regulates its ability to promote degradation of NF-696
kappaB-inducing kinase. J. Immunol. 179(9), 6291-6296. 697
Arthur, J. C., Lich, J. D., Ye, Z., Allen, I. C., Gris, D., Wilson, J. E., at al. (2010). Cutting edge: 698
NLRP12 controls dendritic and myeloid cell migration to affect contact hypersensitivity. 699
J. Immunol. 185(8), 4515-4519. doi:10.4049/jimmunol.1002227 700
Brakke, M. K. (1951). Density Gradient Centrifugation: A New Separation Technique1. Journal 701
of the American Chemical Society. 73(4), 1847-1848. doi:10.1021/ja01148a508 702
Carreno, S., Gouze, M. E., Schaak, S., Emorine, L. J., and Maridonneau-Parini, I. (2000). Lack 703
of palmitoylation redirects p59Hck from the plasma membrane to p61Hck-positive 704
lysosomes. J. Biol. Chem. 275(46), 36223-36229. doi:10.1074/jbc.M003901200 705
Cerami, E., Gao, J., Dogrusoz, U., Gross, B. E., Sumer, S. O., Aksoy, B. A., et al. (2012). The 706
cBio Cancer Genomics Portal: An Open Platform for Exploring Multidimensional Cancer 707
Genomics Data. Cancer Discovery. 2(5), 401. doi:10.1158/2159-8290.CD-12-0095 708
Cerami, E., Gao, J., Dogrusoz, U., Gross, B. E., Sumer, S. O., Aksoy, B. A., et al. (2012). The 709
cBio cancer genomics portal: an open platform for exploring multidimensional cancer 710
genomics data. Cancer Discov. 2(5), 401-404. doi:10.1158/2159-8290.cd-12-0095 711
Chen, L., Wilson, J. E., Koenigsknecht, M. J., Chou, W. C., Montgomery, S. A., Truax, A. D., et 712
al (2017). NLRP12 attenuates colon inflammation by maintaining colonic microbial 713
Nlrp12 binds to Hck
32
diversity and promoting protective commensal bacterial growth. Nat. Immunol. 18(5), 714
541-551. doi:10.1038/ni.3690 715
Chen, S. T., Chen, L., Lin, D. S., Chen, S. Y., Tsao, Y. P., Guo, H., et al (2019). NLRP12 716
Regulates Anti-viral RIG-I Activation via Interaction with TRIM25. Cell Host Microbe. 717
25(4), 602-616.e607. doi:10.1016/j.chom.2019.02.013 718
Damiano, J. S., Oliveira, V., Welsh, K., and Reed, J. C. (2004). Heterotypic interactions among 719
NACHT domains: implications for regulation of innate immune responses. The 720
Biochemical journal. 381(Pt 1), 213-219. doi:10.1042/BJ20031506 721
Dos Santos, C., Demur, C., Bardet, V., Prade-Houdellier, N., Payrastre, B., and Récher, C. 722
(2008). A critical role for Lyn in acute myeloid leukemia. Blood. 111(4), 2269. 723
doi:10.1182/blood-2007-04-082099 724
Faustin, B., Lartigue, L., Bruey, J. M., Luciano, F., Sergienko, E., Bailly-Maitre, B., et al. (2007). 725
Reconstituted NALP1 inflammasome reveals two-step mechanism of caspase-1 726
activation. Mol. Cell. 25(5), 713-724. doi:10.1016/j.molcel.2007.01.032 727
Fu, G., Xu, Q., Qiu, Y., Jin, X., Xu, T., Dong, S., et al. (2017). Suppression of Th17 cell 728
differentiation by misshapen/NIK-related kinase MINK1. J. Exp. Med. 214(5), 1453-729
1469. doi:10.1084/jem.20161120 730
Gao, J., Aksoy, B. A., Dogrusoz, U., Dresdner, G., Gross, B., Sumer, S. O., et al. (2013). 731
Integrative analysis of complex cancer genomics and clinical profiles using the 732
cBioPortal. Sci. Signal. 6(269), pl1. doi:10.1126/scisignal.2004088 733
Genomic and Epigenomic Landscapes of Adult De Novo Acute Myeloid Leukemia. (2013). New 734
England Journal of Medicine. 368(22), 2059-2074. doi:10.1056/NEJMoa1301689 735
Greenbaum, D., Colangelo, C., Williams, K., and Gerstein, M. (2003). Comparing protein 736
abundance and mRNA expression levels on a genomic scale. Genome Biol. 4, 117. 737
Guiet, R., Poincloux, R., Castandet, J., Marois, L., Labrousse, A., Le Cabec, V., and 738
Maridonneau-Parini, I. (2008). Hematopoietic cell kinase (Hck) isoforms and phagocyte 739
duties - from signaling and actin reorganization to migration and phagocytosis. Eur. J. 740
Cell Biol. 87(8-9), 527-542. doi:10.1016/j.ejcb.2008.03.008 741
Hafner-Bratkovič, I., Sušjan, P., Lainšček, D., Tapia-Abellán, A., Cerović, K., Kadunc, L., el al. 742
(2018). NLRP3 lacking the leucine-rich repeat domain can be fully activated via the 743
canonical inflammasome pathway. Nature Communications. 9(1), 5182. 744
doi:10.1038/s41467-018-07573-4 745
Hobbs, S., Jitrapakdee, S., and Wallace, J. C. (1998). Development of a Bicistronic Vector 746
Driven by the Human Polypeptide Chain Elongation Factor 1α Promoter for Creation of 747
Stable Mammalian Cell Lines That Express Very High Levels of Recombinant Proteins. 748
Biochemical and Biophysical Research Communications. 252(2), 368-372. 749
doi:https://doi.org/10.1006/bbrc.1998.9646 750
Hornick, E. E., Banoth, B., Miller, A. M., Zacharias, Z. R., Jain, N., Wilson, M. E., et al. (2018). 751
Nlrp12 Mediates Adverse Neutrophil Recruitment during Influenza Virus Infection. J. 752
Immunol. 200(3), 1188-1197. doi:10.4049/jimmunol.1700999 753
Hu, Y., Liu, Y., Pelletier, S., Buchdunger, E., Warmuth, M., Fabbro, D., et al. (2004). 754
Requirement of Src kinases Lyn, Hck and Fgr for BCR-ABL1-induced B-lymphoblastic 755
leukemia but not chronic myeloid leukemia. Nat. Genet. 36(5), 453-461. 756
doi:10.1038/ng1343 757
Nlrp12 binds to Hck
33
Janowski, A. M., Kolb, R., Zhang, W., and Sutterwala, F. S. (2013). Beneficial and Detrimental 758
Roles of NLRs in Carcinogenesis. Frontiers in immunology. 4, 370-370. 759
doi:10.3389/fimmu.2013.00370 760
Jeru, I., Duquesnoy, P., Fernandes-Alnemri, T., Cochet, E., Yu, J. W., Lackmy-Port-Lis, M., et al. 761
(2008). Mutations in NALP12 cause hereditary periodic fever syndromes. Proc. Natl. 762
Acad. Sci. U. S. A. 105(5), 1614-1619. doi:10.1073/pnas.0708616105 763
Jeru, I., Le Borgne, G., Cochet, E., Hayrapetyan, H., Duquesnoy, P., Grateau, G., et al. (2011). 764
Identification and functional consequences of a recurrent NLRP12 missense mutation in 765
periodic fever syndromes. Arthritis Rheum. 63(5), 1459-1464. doi:10.1002/art.30241 766
Kanneganti, T. D., Lamkanfi, M., and Nunez, G. (2007). Intracellular NOD-like receptors in host 767
defense and disease. Immunity. 27(4), 549-559. doi:10.1016/j.immuni.2007.10.002 768
Kinnaird, J. H., Singh, M., Gillan, V., Weir, W., Calder, E. D., Hostettler, I., et al. (2017). 769
Characterization of HSP90 isoforms in transformed bovine leukocytes infected with 770
Theileria annulata. Cell Microbiol. 19(3). doi:10.1111/cmi.12669 771
Kinoshita, T., Kondoh, C., Hasegawa, M., Imamura, R., and Suda, T. (2006). Fas-associated 772
factor 1 is a negative regulator of PYRIN-containing Apaf-1-like protein 1. Int. Immunol. 773
18(12), 1701-1706. doi:10.1093/intimm/dxl104 774
Lich, J. D., Williams, K. L., Moore, C. B., Arthur, J. C., Davis, B. K., Taxman, D. J., and Ting, J. 775
P. (2007). Monarch-1 suppresses non-canonical NF-kappaB activation and p52-776
dependent chemokine expression in monocytes. J. Immunol. 178(3), 1256-1260. 777
Linz, B. M., Neely, C. J., Kartchner, L. B., Mendoza, A. E., Khoury, A. L., Truax, A., et al. 778
(2017). Innate Immune Cell Recovery Is Positively Regulated by NLRP12 during 779
Emergency Hematopoiesis. J. Immunol. 198(6), 2426-2433. 780
doi:10.4049/jimmunol.1601048 781
Liu, Y., Beyer, A., and Aebersold, R. (2016). On the Dependency of Cellular Protein Levels on 782
mRNA Abundance. Cell. 165, 535-50. 783
Lopez, S., Voisset, E., Tisserand, J. C., Mosca, C., Prebet, T., Santamaria, D., et al. (2016). An 784
essential pathway links FLT3-ITD, HCK and CDK6 in acute myeloid leukemia. 785
Oncotarget. 7(32), 51163-51173. doi:10.18632/oncotarget.9965 786
Martin, B. N., Wang, C., Zhang, C. J., Kang, Z., Gulen, M. F., Zepp, J. A., et al. (2016). T cell-787
intrinsic ASC critically promotes T(H)17-mediated experimental autoimmune 788
encephalomyelitis. Nat. Immunol. 17(5), 583-592. doi:10.1038/ni.3389 789
Martinon, F., Agostini, L., Meylan, E., and Tschopp, J. (2004). Identification of bacterial 790
muramyl dipeptide as activator of the NALP3/cryopyrin inflammasome. Curr. Biol. 791
14(21), 1929-1934. doi:10.1016/j.cub.2004.10.027 792
Naganna, N., Opoku-Temeng, C., Choi, E. Y., Larocque, E., Chang, E. T., Carter-Cooper, B. A., 793
et al. (2019). Amino alkynylisoquinoline and alkynylnaphthyridine compounds potently 794
inhibit acute myeloid leukemia proliferation in mice. EBioMedicine. 40, 231-239. 795
doi:10.1016/j.ebiom.2019.01.012 796
Normand, S., Waldschmitt, N., Neerincx, A., Martinez-Torres, R. J., Chauvin, C., Couturier-797
Maillard, A., et al. (2018). Proteasomal degradation of NOD2 by NLRP12 in monocytes 798
promotes bacterial tolerance and colonization by enteropathogens. Nature 799
Communications. 9(1), 5338. doi:10.1038/s41467-018-07750-5 800
Patel, R. K., Weir, M., Hellwig, S., Dorman, H., and Smithgall, T. E. (2017). Abstract 2363: 801
Targeting Src-family kinases to combat acquired inhibitor resistance in FLT3-802
Nlrp12 binds to Hck
34
ITD<sup>+</sup> AML. Cancer Research. 77(13 Supplement), 2363-2363. 803
doi:10.1158/1538-7445.am2017-2363 804
Pene-Dumitrescu, T., Peterson, L. F., Donato, N. J., and Smithgall, T. E. (2008). An inhibitor-805
resistant mutant of Hck protects CML cells against the antiproliferative and apoptotic 806
effects of the broad-spectrum Src family kinase inhibitor A-419259. Oncogene. 27(56), 807
7055-7069. doi:10.1038/onc.2008.330 808
Pinheiro, A. S., Eibl, C., Ekman-Vural, Z., Schwarzenbacher, R., and Peti, W. (2011). The 809
NLRP12 pyrin domain: structure, dynamics, and functional insights. Journal of molecular 810
biology. 413(4), 790-803. doi:10.1016/j.jmb.2011.09.024 811
Poh, A. R., O'Donoghue, R. J., and Ernst, M. (2015). Hematopoietic cell kinase (HCK) as a 812
therapeutic target in immune and cancer cells. Oncotarget. 6(18), 15752-15771. 813
doi:10.18632/oncotarget.4199 814
Prada-Arismendy, J., Arroyave, J. C., and Rothlisberger, S. (2017). Molecular biomarkers in 815
acute myeloid leukemia. Blood Rev. 31(1), 63-76. doi:10.1016/j.blre.2016.08.005 816
Prochnicki, T., and Latz, E. (2017). Inflammasomes on the Crossroads of Innate Immune 817
Recognition and Metabolic Control. Cell Metab. 26(1), 71-93. 818
doi:10.1016/j.cmet.2017.06.018 819
Robbins, S. M., Quintrell, N. A., and Bishop, J. M. (1995). Myristoylation and differential 820
palmitoylation of the HCK protein-tyrosine kinases govern their attachment to 821
membranes and association with caveolae. Mol. Cell Biol. 15(7), 3507-3515. 822
doi:10.1128/mcb.15.7.3507 823
Roversi, F. M., Pericole, F. V., Duarte, A. d. S. S., Ferro, K. P., Corrocher, F. A., Palodetto, B., 824
et al. (2015). Knockdown of HCK Reduces Cell Death and Erythroid Differentiation in 825
Human CD34<sup>+</sup> Hematopoietic Progenitor Cells. Blood. 126(23), 2860. 826
Roversi, F. M., Pericole, F. V., Machado-Neto, J. A., da Silva Santos Duarte, A., Longhini, A. L., 827
Corrocher, F. A., et al. (2017). Hematopoietic cell kinase (HCK) is a potential therapeutic 828
target for dysplastic and leukemic cells due to integration of erythropoietin/PI3K pathway 829
and regulation of erythropoiesis: HCK in erythropoietin/PI3K pathway. Biochim. 830
Biophys. Acta Mol. Basis Dis. 1863(2), 450-461. doi:10.1016/j.bbadis.2016.11.013 831
Ryu, S. W., Lee, S. J., Park, M. Y., Jun, J. I., Jung, Y. K., and Kim, E. (2003). Fas-associated 832
factor 1, FAF1, is a member of Fas death-inducing signaling complex. J. Biol. Chem. 833
278(26), 24003-24010. doi:10.1074/jbc.M302200200 834
Saito, Y., Yuki, H., Kuratani, M., Hashizume, Y., Takagi, S., Honma, T., et al. (2013). A 835
pyrrolo-pyrimidine derivative targets human primary AML stem cells in vivo. Sci. Transl. 836
Med. 5(181), 181ra152. doi:10.1126/scitranslmed.3004387 837
Schindelin, J., Arganda-Carreras, I., Frise, E., Kaynig, V., Longair, M., Pietzsch, T., et al. (2012). 838
Fiji: an open-source platform for biological-image analysis. Nat. Methods. 9(7), 676-682. 839
doi:10.1038/nmeth.2019 840
Sharif, H., Wang, L., Wang, W. L., Magupalli, V. G., Andreeva, L., Qiao, Q., et al. (2019). 841
Structural mechanism for NEK7-licensed activation of NLRP3 inflammasome. Nature. 842
570(7761), 338-343. doi:10.1038/s41586-019-1295-z 843
Shi, C., Miley, J., Nottingham, A., Morooka, T., Prosdocimo, D. A., and Simon, D. I. (2019). 844
Leukocyte integrin signaling regulates FOXP1 gene expression via FOXP1-IT1 long non-845
coding RNA-mediated IRAK1 pathway. Biochim. Biophys. Acta Gene Regul. Mech. 846
1862(4), 493-508. doi:10.1016/j.bbagrm.2019.02.008 847
Nlrp12 binds to Hck
35
Sicheri, F., Moarefi, I., and Kuriyan, J. (1997). Crystal structure of the Src family tyrosine kinase 848
Hck. Nature. 385(6617), 602-609. doi:10.1038/385602a0 849
Takeuchi, O., and Akira, S. (2010). Pattern recognition receptors and inflammation. Cell. 140(6), 850
805-820. doi:10.1016/j.cell.2010.01.022 851
Tian, Q. Y., Zhao, Y. P., and Liu, C. J. (2012). Modified yeast-two-hybrid system to identify 852
proteins interacting with the growth factor progranulin. J. Vis. Exp. (59). 853
doi:10.3791/3562 854
Truax, A. D., Chen, L., Tam, J. W., Cheng, N., Guo, H., Koblansky, A. A., et al. (2018). The 855
Inhibitory Innate Immune Sensor NLRP12 Maintains a Threshold against Obesity by 856
Regulating Gut Microbiota Homeostasis. Cell Host Microbe. 24(3), 364-378.e366. 857
doi:10.1016/j.chom.2018.08.009 858
Tuncer, S., Fiorillo, M. T., and Sorrentino, R. (2014). The multifaceted nature of NLRP12. J. 859
Leukoc. Biol. 96(6), 991-1000. doi:10.1189/jlb.3RU0514-265RR 860
Vladimer, G. I., Weng, D., Paquette, S. W. M., Vanaja, S. K., Rathinam, V. A., Aune, M. H., et 861
al. (2012). The NLRP12 inflammasome recognizes Yersinia pestis. Immunity. 37(1), 96-862
107. 863
Vogel, C., and Marcotte, E.M. (2012). Insights into the regulation of protein abundance from 864
proteomic and transcriptomic analyses. Nat. Rev. Genet. 13, 227-32. 865
866
Wang, Q., Liu, F., Zhang, M., Zhou, P., Xu, C., Li, Y., et al. (2018). NLRP12 Promotes Mouse 867
Neutrophil Differentiation through Regulation of Non-canonical NF-kappaB and 868
MAPK(ERK1/2) Signaling. Int. J. Biol. Sci. 14(2), 147-155. doi:10.7150/ijbs.23231 869
Williams, K. L., Lich, J. D., Duncan, J. A., Reed, W., Rallabhandi, P., Moore, C., et al. (2005). 870
The CATERPILLER protein monarch-1 is an antagonist of toll-like receptor-, tumor 871
necrosis factor alpha-, and Mycobacterium tuberculosis-induced pro-inflammatory 872
signals. J. Biol. Chem. 280(48), 39914-39924. doi:10.1074/jbc.M502820200 873
Ye, Z., Lich, J. D., Moore, C. B., Duncan, J. A., Williams, K. L., and Ting, J. P. (2008). ATP 874
binding by monarch-1/NLRP12 is critical for its inhibitory function. Mol. Cell Biol. 875
28(5), 1841-1850. doi:10.1128/mcb.01468-07 876
Zaki, M. H., Vogel, P., Malireddi, R. K., Body-Malapel, M., Anand, P. K., Bertin, J., et al. 877
(2011). The NOD-like receptor NLRP12 attenuates colon inflammation and 878
tumorigenesis. Cancer Cell. 20(5), 649-660. doi:10.1016/j.ccr.2011.10.022 879
880
Table 1: yeast two hybrid results (hit frequency >= 3) 881
Name Frequency
Deleted in azoospermia (DAZ) associated protein 2 (DAZAP2) 10
Filamin A, alpha (FLNA) 9
Hematopoietic cell kinase (HCK) proto-oncogene, Src family tyrosine kinase (HCK) 9
Chaperone DnaJ (a.k.a., heat shock protein 40 kD(Hsp40)) homolog, subfamily B, 9
Nlrp12 binds to Hck
36
member 1 (DNAJB1)
Metallothionein 2A (MT2A) 7
Cluster of differentiation 14 molecule (CD14) 6
DEAD (Asp-Glu-Ala-Asp) box helicase 5 (DDX5) 6
Talin 1 (TLN1) 5
Calcium modulating ligand (CAMLG) 4
Squamous cell carcinoma antigen recognized by T cells (SART1) 4
Ras-related nuclear protein (RAN) binding protein 9 (RANBP9) 3
Tumor necrosis factor (TNF) receptor associated factor 3 (TRAF3) interacting protein 3
(TRAF3IP3) 3
Thyroid hormone receptor interactor 4 (TRIP4) 3
882
Table 2: Co-occurrence of mRNA expression level of Hck and all of the Nlrp family 883
proteins (from the provisional data, assessed on July 11 2019) 884
Hck + Nlrp
family proteins
Co-occurrence (C) or
mutual exclusivity (M)
P-value
Nlrp1 C 0.048
Nlrp2 M 0.546
Nlrp3 C 0.415
Nlrp4 M 0.936
Nlrp5 M 0.877
Nlrp6 M 0.670
Nlrp12 binds to Hck
37
Nlrp7 M 0.475
Nlrp8 M 0.877
Nlrp9 M 0.491
Nlrp10 M 0.626
Nlrp11 M 0.233
Nlrp12 C <0.001
Nlrp13 C 1.000
Nlrp14 M 0.509
885
886
887
Table 3: Primers used in cloning (The underline shows the restriction enzymes) 888
Name Primers
Nlrp12 5’ primer (cloned
into PGBKT7)
GATCTACATATGCTACGAACCGCAGGCAG
Nlrp12 3’ primer (cloned
into PGBKT7)
GATGTAGTCGACTCACCCCTCGTCCAGGATATAGTACATAGCT
Nlrp3 5’ primer (cloned
into pGBKT7)
GATCTACATATGAAGATGGCAAGCACCCGCTGCA
Nlrp3 3’ primer (cloned
into PGBKT7)
GATGTAGGATCCTCACAGCAGGTAGTACATGGCGGCAAAGA
Nlrp8 5’ primer (cloned GATCTAGAATTCATGAGTGACGTGAATCCACCCTCTG
Nlrp12 binds to Hck
38
into pGBKT7)
Nlrp8 3’ primer (cloned
into PGBKT7)
GATGTAGGATCCTCAGAGTCTTTGTGGGAAACAGAGAACA
Nlrp12 5’ (cloned into
pEFIRES-P-3FLAG and
pIRES-V5)
GTCCAGCTTAAGATGCTACGAACCGCAGGCAG
Nlrp12 3’ (cloned into
PIRES-3FLAG and
pEFIRES-P-V5)
CTGAATTCTAGAGCAGCCAATGTCCAAATAAGG
Hck p61 5’ (cloned into
pEFIRES-P-3FLAG)
GATCTACTTAAGGCCACCATGGGGGGGCGCTCAAGCTGCGAGG
Hck p59 5’ (cloned into
pEFIRES-P-3FLAG)
GATCTACTCGAGGCCACCATGGGGTGCATGAAGTCCAAGTTCC
Hck p61/p59 3’ (cloned
into pEFIRES-P-3FLAG)
GTCGATTCTAGATGGCTGCTGTTGGTACTGGCTCTCT
889
Figure legends: 890
Figure 1: (A) Schematic drawing of the domain structure of full-length Nlrp12 (top) and the 891
domains used as the construct for the yeast two-hybrid screen (bottom). PYD: Pyrin domain; NBD: 892
Nucleotide binding domain; and LRR: Leucine rich repeat domain. (B) The domain structures of the 893
two alternatively spliced isoforms of Hck, Hck p59 and Hck p61. (C) Identification of the clones 894
of Hck interacting with Nlrp12 (“hits”) obtained from the yeast two-hybrid screen. Among the 895
nine of the clones positive for Hck, one clone encoded for the smallest interacting fragment of 896
Nlrp12 binds to Hck
39
Hck, the C-terminal 40 amino acids of Hck. The numbers in parentheses indicate the total 897
number of hits obtained in the screen. 898
899
Figure 2: Specificity of binding of Nlrp12 with Hck. (A) Schematic drawing of Hck (p61 900
isoform, top) and, for comparison, the region of Hck (“Hck C1 fragment,” lower) used for testing 901
the specificity of binding of Hck to other members of the Nlrp family of proteins. The numbers 902
represent the amino acid residues of the protein or protein fragment (B) Phylogenetic tree 903
showing the relationships among the PYD + NBD region of all of the members of the human 904
Nlrp family of proteins. The phylogenetic tree was produced through Clustal Omega 905
(https://www.ebi.ac.uk/Tools/msa/clustalo/). For each Nlrp family member, the PYD + NBD of 906
the longest isoform was selected for generating the phylogenetic tree. Nlrp3 is phylogenetically 907
the closest Nlrp to Nlrp12, while Nlrp8 is phylogenetically among the most distant from Nlrp12. 908
Nlrp3 and Nlrp8 were highlighted in red color. (C) When Nlrp12’s PYD + NBD and Hck C1 909
fragment were co-transformed into yeast, they grew on the highest stringency plates (4DO + 910
3AT), consistent with an interaction between the two constructs. On the other hand, neither 911
Nlrp3’s PYD + NBD co-transformed with the C1 fragment nor Nlrp8’s PYD + NBD co-912
transformed the C1 fragment grew on the highest stringency plates. These results are consistent 913
with a specific interaction between Nlrp12 and the Hck C1 fragment, but not with the other Nlrps. 914
(D) Yeast clones grew on the high stringency plates 4DO + 3AT when Hck’s C1 fragment and 915
Nlrp12’s PYD + NBD were co-transformed, but they did not grow on the high stringency 4DO + 916
3AT plates when the C1 fragments of other members of the Src family of non-receptor tyrosine 917
kinases were co-transformed with Nlrp12’s PYD + NBD. These results are again consistent with 918
a specific interaction between Nlrp12 and Hck. 919
Nlrp12 binds to Hck
40
920
Figure 3: (A) Structure of the Hck C1 fragment generated by Chimera from the University of 921
California, San Francisco (version 1.13.1), extracted from its structure within the entire Hck 922
protein. Hck C1 fragment was thus divided into four helical domains: R1, R2, R3, and R4. The 923
side chains of amino acids F503, Q507, L510, and D511 are shown. However, it is not clear 924
whether the C1 fragment retains the structural features shown here when it expressed on its own. 925
(B) Primary amino acid sequence alignments showing of the C1 fragments of all members of the 926
Src non-receptor tyrosine kinase family (40 amino acids for Hck, Blk, Lck, and Lyn; 42 amino 927
acids for Fgr, Src, Yes, and Fyn). The black lines indicate the R1, R2, R3, or R4 regions. The 928
arrows denote the four critical amino acids for binding of Hck C1 fragment to Nlrp12’s PYD + 929
NBD. (C) Phylogenetic tree showing the relationship among all of the C1 fragments of the 930
members of the Src family of non-receptor tyrosine kinases. The phylogenetic tree was 931
produced through Clustal Omega (https://www.ebi.ac.uk/Tools/msa/ clustalo/). Hck was 932
highlighted in red color. (D) Four sets of distinct mutations were made within the Hck C1 933
fragment. M1: the amino acids in the R1 region were mutated to the amino acids identical to the 934
R1 region in Lyn (highlighted in red color); M2: the amino acids in the R2 region were all 935
mutated to alanine (Ala); M3: the amino acids in the R3 region were all mutated to Ala; and M4: 936
the amino acids in the R4 region were all mutated to Ala. (E) Yeast did not grow on high 937
stringency 4DO + 3AT plates when Hck C1 fragment with the M3 mutations was co-transformed 938
with Nlrp12’s PYD + NBD. But yeast can grow on the high stringency 4DO + 3AT plates when 939
Hck C1 fragment contained either M1, M2, or M4 mutations, suggesting that the R3 region 940
within the C1 fragment was necessary for binding to Nlrp12 NBD + PYD. (E) Yeast did not 941
grow on high stringency 4DO plates when Nlrp12’s PYD + NBD was co-transformed with Hck 942
Nlrp12 binds to Hck
41
C1 fragment with single Ala substitutions at either F503, Q507, L510, or D511. But yeast grew 943
on high stringency 4DO plates when Nlrp12’s PYD + NBD was co-transformed with Hck’s C1 944
fragment in which the following amino acids in the R3 region were individually mutated to Ala: 945
Y505, S508, or V509. Yeast did not grow on either 2DO nor 4DO plates when Nlrp12’s PYD + 946
NBD was co-transformed with Hck C1 fragment in which either E504 or I506 was individually 947
mutated to Ala, making the results with these particular mutants inconclusive. 948
949
Figure 4: (A) Schematic drawing of Hck truncations comprising the C1 to C7 fragments. (B) 950
Yeast grew on the high stringency 4DO plates and 4DO +3AT plates when Hck’s C-terminal 40 951
amino acid fragment (C1), Hck’s C-terminal 35 amino acid fragment (C2), and Hck’s C-terminal 952
30 amino acid fragment (C3), individually, were co-transformed with Nlrp12’s PYD + NBD. 953
But yeast did not grow on the high stringency 4DO plates nor 4DO+ 3AT plates when C4 to C7 954
fragments were co-transformed to with Nlrp12’s PYD + NBD. The lack of yeast growth upon 955
co-transformation of the C7 fragment and Nlrp12 are not shown. (C) Schematic drawing of Hck 956
truncations comprising the N1 to N5 fragments. (D) Yeast grew on the high stringency 4DO 957
plates and 4DO +3AT plates when Hck’s C-terminal 40 amino acid fragment (C1) was co-958
transformed with Nlrp12’s PYD + NBD (control). But yeast did not grow on the high stringency 959
4DO plates nor 4DO+ 3AT plates when N1 and N5 fragments were co-transformed to with 960
Nlrp12’s PYD + NBD, suggesting that the very C-terminal 5 amino acids are also necessary for 961
binding of Nlrp12. The colored boxes represent different fragments that were deleted each time. 962
963
Figure 5: Nlrp12 co-immunoprecipitates with Hck and co-localizes with Hck by 964
immunofluorescence. (A) Nlrp12 co-immunoprecipitated with Hck when both proteins were 965
Nlrp12 binds to Hck
42
transiently exogenously co-expressed in 293T cells, with Nlrp12 having to be expressed as 966
epitope-tagged forms. Cells were lysed using a non-denaturing cell lysis buffer (see Material 967
and Methods). Co-immunoprecipitations were done using a mouse monoclonal anti-Hck 968
antibody followed by protein A/G agarose. The immunoprecipitated proteins were analyzed by 969
immunoblotting. The same method was used for experiments shown in (B), (C), and (D). (B) 970
Nlrp12 co-immunoprecipitated with endogenous Hck in macrophage-like RAW 264.7 cells, 971
when Nlrp12 is exogenously and stably expressed. Arrows show the two isoforms of Hck p59 972
and p61 that were immunoprecipitated by the anti-Hck antibody, although it is not clear whether 973
Nlrp12 co-immunopreciptitated with one form or the other, or both. (C) Exogenously and stably 974
expressed Nlrp12 co-immunoprecipitated with endogenously expressed Hck in monocyte-like 975
THP-1 cells. In these cells, phorbol 12-myristate 13-acetate (PMA) (1µM) was added to 976
enhance the expression of Nlrp12 under the PMA-sensitive cytomegalovirus promoter of the 977
Nlrp12 plasmid. (D) Left panel shows that Nlrp12 was exogenously and stably expressed in 978
lymphoblast-like U937 cells. Right panel shows that Nlrp12 co-immunoprecipitated with 979
endogenously expressed Hck in U937 cells after PMA was added to 1µM, as was done for THP-980
1 cells, to enhance expression of Nlrp12. (E) Immunofluorescent images showing the 981
colocalization of a cohort of Nlrp12 with either Hck p59 or p61. Hck was immunolabeled with a 982
mouse monoclonal anti-Hck antibody, followed by a secondary goat anti-mouse IgG (H+L) 983
cross-adsorbed, conjugated to Alexa Fluor 488. For Nlrp12, the primary antibody was a rabbit 984
monoclonal anti-FLAG antibody, and the secondary antibody was a goat anti-rabbit IgG (H+L) 985
cross-adsorbed secondary antibody, conjugated to Alexa Fluor 555. Nuclei were counterstained 986
by DAPI. 987
988
Nlrp12 binds to Hck
43
Figure 6: The co-expression of Hck protein significantly decreased the steady-state protein 989
expression levels of Nlrp12 (A) Epitope-tagged Nlrp12-V5 and Hck p61 were co-transfected into 990
293T cells. Cell lysates were immunoblotted with rabbit monoclonal anti-V5 antibody for 991
Nlrp12 and mouse monoclonal anti-Hck antibody. Comparing lane 2 (basal level of Nlrp12 992
expression) with lanes 5, 6, and 7, co-expression of Hck protein leads to a significant decrease of 993
protein expression levels of Nlrp12 over three time are they co-expressed. Increasing protein 994
expression of Hck, relative to the plasmid DNA added (ramp: lanes 5, 6, and 7; 2.5-20 μg of 995
DNA) was not observed possibly because the amount of plasmid for Hck transfection were 996
already too high, resulting in the maximum expression of Hck’s at all plasmid DNA amounts. 997
The figure is a representative of one of three results. (B) Quantification of the Nlrp12 protein 998
bands (mean + standard error) to show that Nlrp12 protein expression levels decreased in the 999
presence of co-expressed Hck, relative to equal amounts of cell lysate being assayed. The 1000
Nlrp12 protein expression levels are indicated as arbitrary units. (C) Quantification of the Hck 1001
protein bands (mean + standard error) confirming that Hck protein expression levels did not 1002
change although the amount of plasmid DNA for Hck increased. Figure (B) and (C) were 1003
generated by Prism 6.0 (GraphPad Software, Inc., San Diego, CA). Twenty μg of cell lysate 1004
proteins were loaded for each lane. ANOVA and Tukey tests were used to compare the Nlrp12 1005
and Hck protein bands. 1006
1007
Figure 7: Nlrp12 is co-expressed with Hck in acute myeloid leukemia (AML) patient samples. 1008
Both Nlrp12 expression and Hck expression data are shown as log2 values. The data are RNA 1009
seq data obtained from cBioportal (provisional) database assessed on July 11, 2019. The graph 1010
was taken from the cBioportal website. Spearman coefficient, indicating the association of 1011
Nlrp12 binds to Hck
44
Nlrp12 and Hck mRNA expression level (if Nlrp12 and Hck mRNA expression levels are non-1012
parametric) = 0.80. Pearson coefficient, indicating the degree of the linear relationship between 1013
the Nlrp12 mRNA expression level and Hck mRNA expression level (if Nlrp12 and Hck mRNA 1014
expression levels are parametric) = 0.73. The p-values for Spearman coefficient and Pearson 1015
coefficient are both < 0.001. 1016
1017
Figure 8: (A) Ingenuity Pathway Analysis (IPA) displaying the 13 distinctive hits from the total 1018
results of the yeast two-hybrid screen for proteins that interacted with Nlrp12. The figure only 1019
shows the Nlrp12-interacting proteins with hits that have a frequency larger than 2. The solid 1020
lines indicate interacting proteins found through our yeast two-hybrid screen. The dashed lines 1021
are the Nlrp12-interacting proteins previously reported in published data. The purple-colored 1022
molecules are those associated with AML. (B) To validate an interaction shown in the IPA 1023
diagram, tumor necrosis factor receptor-associated factor 3 interacting protein 3 (TRAF3IP3)-1024
3FLAG and Nlrp12-V5 were co-immunoprecipitated from co-transfected 293T cells. A non-1025
denaturing cell lysate buffer (see Material and Methods) was used to lyse the cell, and co-1026
immunoprecipitations were performed with a mouse monoclonal anti-TRAF3IP3 antibody 1027
followed by protein A/G agarose. The immunoprecipitates were analyzed for Nlrp12 by 1028
immunoblotting with rabbit monoclonal anti-V5 antibody. The results from the co-1029
immunoprecipitation are consistent with a confirmation of an interaction between TRAF3IP3 1030
and Nlrp12. 1031
1032
1033
1034
Nlrp12 binds to Hck
45
1035
Figure 1 1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
Nlrp12 binds to Hck
46
1046 Figure 2 1047
1048
1049
1050
1051
Nlrp12 binds to Hck
47
1052
Figure 3 1053
1054
Nlrp12 binds to Hck
48
1055
Figure 4 1056
1057
Nlrp12 binds to Hck
49
1058
Figure 5 1059
1060
Nlrp12 binds to Hck
50
1061
Figure 6 1062
Nlrp12 binds to Hck
51
1063
Figure 7 1064
1065
1066
1067
1068
1069
1070
1071
1072
Nlrp12 binds to Hck
52
1073
Figure 8 1074
1075