From hundreds to thousands: Widening the normal human Urinome (1)
-
Upload
gian-marco -
Category
Documents
-
view
216 -
download
3
Transcript of From hundreds to thousands: Widening the normal human Urinome (1)
1
2
3Q1
4
5
6
7
8
1011
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47484950
J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 4 ) X X X – X X X
Ava i l ab l e on l i ne a t www.sc i enced i r ec t . com
ScienceDirectwww.e l sev i e r . com/ loca te / j p ro t
JPROT-01898; No of Pages 10
From hundreds to thousands: Widening the normalhuman Urinome (1)
OO
F
Laura Santuccia,1, Giovanni Candianoa,1, Andrea Petrettob, Maurizio Bruschia,Chiara Lavarellob, Elvira Ingleseb, Pier Giorgio Righettic, Gian Marco Ghiggeria,⁎aNephrology, Dialysis, Transplantation Unit and Laboratory on Pathophysiology of Uremia, Istituto Giannina Gaslini, 16148 Genova, ItalybLaboratory of Mass Spectrometry, Core Facility, Istituto Giannina Gaslini, 16148 Genova, ItalycPolitecnico di Milano, Department of Chemistry, Materials and Chemical Engineering “Giulio Natta”, Via Mancinelli 7, 20131 Milano, Italy
RA R T I C L E I N F O
UNCO
R
DOI of original article: http://dx.doi.org/10⁎ Corresponding author at: Division of Nephro
Gaslini, Largo G. Gaslini 5, 16148 Genova, ItaE-mail address: GMarcoGhiggeri@ospedal
1 These two authors have equally contribu
http://dx.doi.org/10.1016/j.jprot.2014.07.0211874-3919/© 2014 Elsevier B.V. All rights rese
Please cite this article as: Santucci L, et alhttp://dx.doi.org/10.1016/j.jprot.2014.07.02
A B S T R A C T
Q3
Article history:Received 20 March 2014Accepted 15 July 2014Q2
RECTED PIt is currently unknown howmany proteins can be detected in urine. Improving the analytical
approachwould increase their number and potentially strengthen their predictive potential indiseases.We developed a combination of analytical procedures for maximizing sensitivity andreproducibility of normal human urinary proteome analysis based on ultracentrifugation,vesicles separation, combinatorial peptide ligand libraries (CPLLs) and solvent removal ofpigments. Proteins were identified by an Orbitrap Velos Mass Spectrometry.Overall, 3429 proteinswere characterized:most components (1615) were contained in vesicleswhile the remaining 1794 were equally distributed among CPLL and butanol insolublefractions. Several proteins were detected exclusively in one of the phases of the procedure,suggesting that each step is crucial in the fractionation strategy. Many (1724) proteins aredescribed here whose presence in urine has never been reported and represents a potentialsource of information considering that urine is the unique site of excretion of products ofinteraction of metabolic processes. Improving the characterization of normal urinaryproteome would also represent the basis for the analysis of urine biomarkers in humandiseases.
Biological significanceSub-fractionating normal urine by successive steps (vesicle separation, CPLL and solventtreatments) allowed the identification of 3429 proteins, a relevant part (1724) being detectedfor the first time in urine. Several proteins of new description have been implicatedin physiology pathways and in pathologies thus representing a potential source of newinformation on both metabolic processes and diseases.
© 2014 Elsevier B.V. All rights reserved.
Keywords:Urinary proteomeVesiclesCombinatorial peptideligand librariesMass spectrometry
.1016/j.dib.2014.08.006.logy, Dialysis and Transplantation and Laboratory on Pathophysiology of Uremia, Istituto G.ly. Tel.: +39 010 380742; fax: +39 010 395214.e-gaslini.ge.it (G.M. Ghiggeri).ted to the work and share first authorship.
rved.
, From hundreds to thousands: Widening the normal human Urinome (1), J Prot (2014),1
5152
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
2 J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 4 ) X X X – X X X
1. Introduction
T
111
112
113
114
115
116
117
118
119Q4
120
121
122123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143Q5
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
UNCO
RREC
The composition of biological fluids represents a potentiallyrelevant source of information for research in human diseasesparticularly in the area of biomarker development. Urine is anideal fluid for such investigations, since it is the physiologic siteof excretion ofmostmolecules deriving from biological process.Products of cell lesions or cell transformation could be alsodetected provided that they are freely filtered through theglomerular sieves. Urine also drains fluids from the urinarytract and is for this reason particularly enriched in proteinsderiving from the kidney, bladder and prostate. Finally, urinecan be easily collected, does not require invasive proceduresand can be handily manipulated and stored [1–7]. Improvingour knowledge on urine protein composition is crucial to anyevolution in the field of mechanisms, early detection andmonitoring of diseases, in particular in those conditions affect-ing the urinary tract.
In recent years, technology advancements have markedlymodified the number of proteins that can be detected in urineandprotocols havebeen developed for the analysis of intra- andinter-individual variability, e.g. age, diet, and time of collection.[8–12]. In fact, proteomic research on urine has a long historicalrecord. In 1979, theAnderson's group published the first studiesby two-dimensional electrophoresis (2-DE) on normal urine[13,14] in which they identified only the major components,in general the same ones present in serum plus some lowmolecular mass peptides. Up to the end of the century nomoreimprovements were obtained [15] when it became clear thatmore than one technique should be utilized for improvingsensitivity. Thus several groups analyzed the “normal humanurinary proteome” by combining liquid chromatography withtandemmass spectrometry (LC–MS) [16,17] and/or utilizing twomethods of sample fractionation, e.g. acetone precipitation andultracentrifugation. [18–21]. These new approaches notablyimproved sensitivity, so that in 2011 the total normal urinaryproteome listed about 2000 proteins [22–30].
One of the general problems encountered with normal urineanalysis is that low expression components are practicallyundetectable since fifty spots represent 99% of totality thusmasking the low-abundance species. Actually, the same dispro-portion between high and low expression proteins also exists inserum and with respect to the composition in major protein,i.e. albumin, transferrin, immunoglobulins, etc., urine is similarto serum. It is for this reason that combinatorial peptide ligandlibraries (CPLLs) [31–33] were developed to increase sensitivitytowards low expression peptides (for reviews, see [34–37]).
Separation of exosomes has been the second evolution[38–44]. Exosomes are extracellular membrane-bound nano-vesicles secreted by diverse cell types along the urinary tract[45,46]. They are expected to play a role in immune systemmodulation and in the transport of proteins, mRNAs andmicroRNA, overall influencing angiogenesis, cell proliferationand tumor cell invasion [47,48]. Vesicles are ubiquitous andurine is very enriched of them [49–57]. However, their purifica-tion isnot an easy task, since, in urine, they interactwith solubleproteins such as of uromodulin and, in spite of minimizingaggregation phenomena by treatment with reducing agents,they appear to be frequently contaminated [41,58–61]. Overall,
Please cite this article as: Santucci L, et al, From hundreds to thouhttp://dx.doi.org/10.1016/j.jprot.2014.07.021
CPLL treatment and purified vesicle collection, coupled withmodern mass spectrometry, substantially increase our analyt-ical view on the global urinary proteome and should representthe basis to develop common laboratory strategies.
In this study, we merged different methods of proteinseparation to optimize the efficiency of low-abundance proteindiscovery. The objective was to bring the field close to a finalsurvey, and put the basis for planning studies in humandiseases. The result is the development of a sequence oftechniques that altogether allow the detection of 3500 singleproteins in normal urine and constitutes a good starting pointfor the objectives above.
ED P
RO
OF
2. Materials and methods
2.1. Materials
ProteoMiner™, urea, thiourea, 3-[3-cholamidopropyl di-methylammonio]-1-propansulfonate (CHAPS), dithioerythritol(DTE), Tris(hydroxymethyl)aminomethane (Tris), ethylendiami-no tetracetic acid (EDTA) and sodiumdodecyl sulfate (SDS) werepurchased by Sigma-Aldrich (St Louis, Mo, USA). Completeprotease inhibitor cocktail tablets were from Roche Diagnostics,(Basel, CH). Sequencing grade bovine trypsinwas from Promega(Madison,WI, USA). All other chemicals of analytical gradewerefrom J.T. Baker (Deventer, Holland).
2.2. Sample collection and urine preparation
Second morning urine samples (approximately 160 mL) fromhealthy volunteers (12 individuals, age 35–45 years, 6 malesand 6 females) were collected after informed consent andwere immediately added with tablets of protease inhibitor,chilled on ice and centrifuged at 4 °C for 10 min at 1000 ×g toeliminate cell debris in accord to Standard Protocols (seeEuroKUP-European Kidney and urine proteomics at http://www.eurokup.org). At this point urine was frozen and main-tained at −80 until further processing. Second step wascentrifugation at 17,000 ×g in a JA-20 rotor (Beckman Avanti™J-25) for 30 min at 16 °C to remove the urinary microscopicparticle sediments. Then the collected urine was pooled (twodifferent pools, each consisting of three males and threefemales) and analyzed in technical triplicates to confirm thereproducibility of the experiments. An aliquot of the total17,000 ×g urinary supernatant, after Bradford protein assay,was dialyzed three times against 25 mM sodium phosphatepH 7.4 in 3500 MWCO Spectra/Por® cellulose membranes at4 °C. After lyophilization this material was stored at −80 °Cuntil use.
2.3. Urinary vesicles isolation
The 17,000 ×g urinary supernatant (80 mL) was ultra-centrifuged in 8 tubes at 48,000 rpm in a Ti 90 rotor (BeckmanOptima™ l-90K) for 75 min at 18 °C. The ultracentrifugationstep was repeated by adding the same volume used beforeuntil vesicles were isolated; DTT (200 mg/mL) was added to48,000 rpm pellets and incubated for 2 min at 95 °C to avoidthe formation of high molecular weight complexes of vesicles
sands: Widening the normal human Urinome (1), J Prot (2014),
T
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
3J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 4 ) X X X – X X X
UNCO
RREC
with uromodulin. The pellets were pooled together andwashed three times with Tris–HCl 65 mM pH 8.8 and centri-fuged a 14,000 rpm for 10 min at 4 °C and stored at −80 °Cuntil mass spectrometry analysis.
2.4. Butanol precipitation
The 48,000 rpm supernatant was dialyzed versus water;aliquots of 50 mL were added with 100 μL of acetic acid(about pH 3–4) and after a few minutes with 10 mL ofn-butanol; this mixture was rotated in agitation for 90 minand centrifuged at 4,000 rpm for 10 min at 18 °C. Threedifferent phases were obtained: a first phase consistedin a protein pellet, a second was the supernatant (forProteoMiner™, see below) and a third phase consisted inpigments (that were discarted), yielding two fractions: CPLL-beads and unbound or flow-through.
2.5. Combinatorial solid-phase ligand library(ProteoMiner™)
The phase deriving from butanol extraction was lyophilizedand loaded onto a column of 150 μL peptide library beads(ProteoMiner™) equilibrated in 25 mM phosphate buffer,pH 7.4 as already described by Candiano et al. [33]. Sodiumazide and EDTA were added to samples. The eluate in Tris–SDS (CPLL-beads) and the unbound fraction were preserved at−80 °C until analysis by mass spectrometry.
2.6. Mass spectrometry
2.6.1. ApparatusA linear Trap Quadropole (LTQ) Orbitrap Velos Pro MassSpectrometry was used for the analysis of all urinary sub-fractions.
2.6.2. Sample preparation and procedureSamples for mass spectrometry were solubilized in 0.1 mL of 4%SDS, 50 mM DTT, and 0.1 M Tris/HCl, pH 7.6, at 90 °C for 5 minand briefly sonicated and were processed by the FASP procedureusing 30 kVivacom filtrationdevices (Sartorius) [62]. Briefly, 50 μLaliquots were mixed with 0.2 mL of 8 M urea in 0.1 M Tris/HCl,pH 8.5 (UA solution), loaded into the filtration devices andcentrifuged at 14,000 g for 15 min. The concentrates werediluted in the deviceswith 0.2 mLofUA solution and centrifugedagain. After centrifugation, the concentrates were mixed with0.1 mL of 50 mM iodoacetamide in UA solution and incubatedin darkness at room temperature (RT) for 30 min followed bycentrifugation for 15 min. Then, the concentrate was dilutedwith 0.2 mL of 8 M urea in 0.1 M Tris/HCl, pH 8.5 (UB solution),and concentrated again. This step was repeated twice. Thesamples were then diluted with 120 μL of 40 mM NaHCO3
containing 1 μg of trypsin. Following anovernight digestion, thepeptides were collected by centrifugation of the filter units for20 min and the filter device was rinsed with 50 mL 0.5 M NaCland centrifuged in order to suppress hydrophobic interactions.
2.6.3. NanoLC setupThe sample was first loaded from the sample loop onto atrapping column (Acclaim PepMap C18; 2 cm × 100 μm × 5 μm,
Please cite this article as: Santucci L, et al, From hundreds to thouhttp://dx.doi.org/10.1016/j.jprot.2014.07.021
ED P
RO
OF
100 Å) using the loading solvent at a flow rate of 5 μL/min for5 min. The trapping column was then switched in-line with theseparation column and the peptides eluted with increasingorganic solvent at a flow rate of 300 nL/min. The peptideseparations were carried out at 30 °C in a 75-μm ID × 15 cm2 μm, 100 Å C18 column mounted in a thermostated columncompartment. The peptides were separated and eluted with alinear gradient of 5–35% solution B (99.9% ACN, 0.1% formic acid)in 150 min.
2.6.4. Mass spectrometer setupThe mass spectrometer LTQ-Orbitrap Velos Pro was operatedin the positive ionization mode. Single MS survey scans wereperformed in the Orbitrap, recording a mass window between350 and 1650 m/z using a maximal ion injection time of250 ms. The resolution was set to 60,000 and the automaticgain control was set to 1,000,000 ions. The lock mass optionwas enabled allowing the internal recalibration of spectrarecorded in the Orbitrap by polydimethylcyclosiloxane back-ground ions (protonated (Si(CH3)2O)6; m/z 445.120025). Theexperiments were done in data-dependent acquisition modewith alternating MS and MS/MS experiments. The minimumMS signal for triggering MS/MS was set to 500 ions, with themost prominent ion signal selected for MS/MS using anisolation window of 2 Da. The m/z values of signals alreadyselected for MS/MS were put on an exclusion list for 60 s usingan exclusion window size of ±10 ppm. In all cases, onemicro-scan was recorded. CID was done with a target valueof 5000 ions in the linear ion trap, a maximal ion injectiontime of 150 ms, normalized collision energy of 35%, aQ-value of 0.25 and an activation time of 10 ms. A maximumof 20 MS/MS experiments were triggered per MS scan.
The standard filter of 10 ppm was chosen on the basis ofcurrent accepted criteria for deep and unbiased proteomecoverage (see for example [63]). As reported below and inaccord with the literature [24], we did not restrict our searchto the “two-peptide” rule and adopted as well the singlepeptide identification approach, as suggested by Gupta andPevzner [64]. These authors confirmed that the “two-peptide”rule reduces the number of protein identifications in thetarget databasemore significantly than in the decoy databaseand results in increased false discovery rates, compared tothe case when single-hit proteins are not discarded.
2.6.5. Data analysisRaw MS files were processed with the Thermo ScientificProteome Discoverer software version 1.3. Peak list files weresearched by the MASCOT and SEQUEST search engine againstUniprot human database (release 2012_07) containing bothforward and reversed protein sequences. The resulting peptidehits were filtered for a maximum 1% FDR using Percolator, thepeptide mass deviation was set to 10 ppm and a minimum ofsix 6 amino acids per identified peptide were required. TheDatabase search parameters were mass tolerance precursor20 ppm [65], mass tolerance fragment CID 0.8 Da with dynamicmodification of deamidation (N, Q), oxidation (M) and staticmodification of alkylation with IAM (C). For all searches theoption trypsinwith twomissed cleavageswas selected. Proteinswere grouped by applying the maximum parsimony rule. Thecomputational methods used to determinate the absolute
sands: Widening the normal human Urinome (1), J Prot (2014),
T
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332Q6
333
334
4 J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 4 ) X X X – X X X
C
protein quantification was average of the three most abundantpeptides.
2.6.6. Bioinformatic analysisLabel free experiments were analyzed with the freely availablePerseus software, which includes all necessary functionalities(http://www.perseusframework.org/Perseus_1.4.1.3.zip). Proteingroups were filtered to require in at least one experi-mental group three valid values. Label free intensities werelogarithmized and missing values were imputed with randomnumbers from a normal distribution,whosemean and standarddeviation were chosen to best simulate low abundance valuesclose to noise level. Single peptide identificationwas accepted incase the same peptide could be identified 4 times on twobiological replicates performed in triplicate. A modified ANOVAandT-testwith permutation based FDR statistics (6)was appliedto filter the first results. We performed 250 permutationsand required an FDR of 0.05 and 0.01 for Anova and T-testrespectively. The parameter S0 was empirically optimizedto separate outliers from the background distribution. Hierar-chical clustering of resulting proteins was performed onlogarithmized intensities after z-score normalization of thedata, using Euclidean distances. The protein categorical anno-tation was supplied by Gene Ontology biological process,molecular function, and cellular component. All annotationwas extracted from the UniProt database. The “annotationmatrix algorithm”, implemented in Perseus software, and theHeatmap were used to describe the differential enrichment ofspecific Gene Ontology biological process terms in the samples.The software filters protein annotation terms by testing thedifference among means for any protein annotation from theoverall ratio distribution for all cells of the expression matrix.The statistical test is a two-dimensional non-parametric
UNCO
RRE
Fig. 1 – A: Flowchart of the procedure used for analysis of the urisupernatant was treated with a butanol precipitation and with PCandiano et al. [33]. The 5 fractions thus obtained (untreated, veprocessed by mass spectrometry analysis. B: Heat map of Pearsonthe label-free replicates (identified with “1 or 2 or 3” and “a or b” rthe five different fractions.
Please cite this article as: Santucci L, et al, From hundreds to thouhttp://dx.doi.org/10.1016/j.jprot.2014.07.021
Mann–Whitney test, with a Benjamini–Hochberg multiplehypothesis testing correction that was controlled by using afalse discovery rate threshold of 0.05. All categories that survivedthe test for at least one of the samples were converted to onerow of the annotation matrix. The values for each row arethe differences between groups of the Mann–Whitney testmentioned above. Fisher's exact test was performed witha false discovery rate value of 0.02. See data available atChorusProject.org under project name “From hundreds tothousands: widening the normal human Urinome”.
ED P
RO
OF3. Results
3.1. Protein composition of un-fractionated normal urine
Without fractionation, urine is, on average, found to contain1176 unique proteins. This number has been obtained byanalyzing with mass spectrometry the second morning urinesamples from 12 young normal volunteers (6 females, 6males,age 35–45 years.) pooled in two samples that were evaluatedthree times each. We calculated the coefficient of variation(CV = standard deviation/mean) across the three replicatesof the same sample for each protein, identified to providea direct quantitative measure of technical variability, andbetween the several samples for biological variability [11]. Themedian CV value of the six runs (technical variability) rangedfrom 0.012 to 0.047 while that of the biological variabilityranged between 0.11 and 0.15.
The list of single proteins detected in un-fractioned urine isgiven in Table 1 in Data in Brief article. Technical variabilityamong six determinations was negligible in terms of presenceor absence, relevant if concentration was considered.
nary proteome. After obtaining vesicles from the pellet, theroteoMiner followed by Tris & SDS–DTE elution as persicles, precipitate, CPLL-beads eluate and unbound) werecorrelation coefficients showing good reproducibility amongespectively for technical triplicates of biological duplicates) of
sands: Widening the normal human Urinome (1), J Prot (2014),
T335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
Fig. 2 – Venn Diagram of the 3429 non-redundant proteinsdetected in a normal urinary proteome with a peptide ID of1% FDR, at a peptide mass deviation of 10 ppm and aminimum of six 6 amino acids per identified peptide. Thetotal proteins identified in each fraction was: 1176(untreated), 1615 (vesicles), 993 (precipitate), 1488(CPLL-beads) and 1345 (unbound).
5J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 4 ) X X X – X X X
3.2. Normal urine composition step by step
Fractionating urine by a multistep procedure increased thenumber of detectable proteins to an overall of 3429, differentlydistributed among the successive steps of the sequence. Thismeans that each fractionation step increases the concentra-tion of single proteins that could not otherwise be detected in
UNCO
RREC
Fig. 3 – Unsupervised hierarchical clustering analysis of protein expand filtering for three valid values for each technical replicate ofshows the segregation of the each urinary fraction in five clustersanalysis, of the same fractions, represents in each row the clusteurine fractions with six values of technical replicates are displayproteins in the different fractions.
Please cite this article as: Santucci L, et al, From hundreds to thouhttp://dx.doi.org/10.1016/j.jprot.2014.07.021
ED P
RO
OF
unfractionated samples. When the results were matchedwith the Arabidopsis database (most frequently utilizing onepeptide and in a few three peptides as matching system) thenumber of identified proteins corresponded to 77 differentaccessions, represented in most cases by interspecies pre-served molecules with basic fundamental functions for celllife (e.g. heat shock proteins, histone chaperone proteins,peroxidase, RNA polymerase). Comparing peptides identifiedby using the human database (n 14,023) with Arabidopsisdatabase (n 92) supported the accuracy of the results reportedin our work.
This improvement was obtained by using about 150 mL ofurine and utilizing the multi-step flow chart illustrated inFig. 1A, that includes: 1) ultracentrifugation for obtainingvesicles, 2) extraction with butanol of supernatant derivingfrom ultracentrifugation to eliminate pigments and 3) CPLL toenhance the signal of low-expression components. Variabilityin purification efficiency was evaluated by repeating eachsingle passage three times on the same pool of urine and wasfound very low. The heatmap (Fig. 1B) analysis of compositionof each purification step (untreated, vesicles, precipitateafter butanol extraction, bound and unbound material toCPLL-beads) shows a good reproducibility among the repli-cates of five sample preparation methods. In this case, thecolor map highlights the groups of samples that have similarcorrelations on a scale of colors from black (maximun) towhite (minimum) where the strict replication is confirmed bydrawing a theoretical diagonal (in gray in Fig. 1B).
Proteins that were purified uniquely in a step of theprocedure were numerous (overall 1678), as shown by the
ression profiles of the five urine fractions after normalizationbiological duplicates. A) Principal component analysis (PCA), according to the different sample preparations. B) Heat mapring of protein expression profiles, while in each column theed. The strongest colors indicate the relative abundances of
sands: Widening the normal human Urinome (1), J Prot (2014),
371
372
373Q7
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
6 J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 4 ) X X X – X X X
number in bold in the Venn diagram of Fig. 2 summarizingthese aspects; details on the identity of single proteins arelisted in Table 1 in the Data in Brief article. This figure alsoreports that most proteins were contained in vesicles (1615),followed byCPLL (1488) and by other precipitated andunboundfractions (993 and 1345 proteins respectively). Fig. 3B displaysa heat map analysis relative to the concentration of eachsingle protein in different purification phases: this image (inwhich black means high levels and white minimum) clearlyshows clusters of proteins in vesicles, CPLL, etc., withminimalnumbers of proteins shared by the different procedures. Thisfinding suggests a good purification efficacy overall. The sameconcept is illustrated in Fig. 3A that shows, with the principalcomponent analysis (PCA), clusters of proteins in differentfractions of the method. Also a good reproducibility was
UNCO
RRECT
Fig. 4 – Volcano plot used graphically to represent the results of tonly species that exceeded the threshold for significance in the Tthe − log 10 of the P-value. In all panels the proteins of the untrea(panel b), unbound (panel c) and vesicles (panel d) are shown.
Please cite this article as: Santucci L, et al, From hundreds to thouhttp://dx.doi.org/10.1016/j.jprot.2014.07.021
F
confirmed by the near overlap of the six samples (tripledtechnical, duplicate biological). In other words, the multistepprocedure described here augments the possibility of isolatingsingle proteins from urine with a good degree of purity (seesection below).
How much every single step improves the purificationefficacy in respect to un-fractioned urine and which proteinstake major benefit from this approach is shown in Fig. 4 and inSupplementary Table 1. In the former case, by exploiting aVolcanoplot, the level of proteinswhich exceeded the thresholdfor significance in the T-test was compared to un-fractionedurine. The statistically significant under-expressed (on the left)and over-expressed (on the right) proteins are plotted in blackvs. gray. The proteins considered in all these plots were 1526and were differently distributed among the different phases of
ED P
RO
O
he T-tests for differential expression of proteins. It includes-test. The log 2 fold change for each protein is plotted againstted fraction (control) versus CPLL-beads (panel a), precipitate
sands: Widening the normal human Urinome (1), J Prot (2014),
OO
F
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416417
418
419
420
421
422
423
424
425
426
427
428
Fig. 6 – Venn diagram comparing our study with four otherrelevant reports on urinary proteome. A total of 2883proteins of new description have been identified(see Supplementary Table 2).
7J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 4 ) X X X – X X X
the purification procedure. For example, when consideringCPLL (shown in a), at least 323 proteins were concentrated by afactor of 4 or more and the same was for other conditions ofpurification (Fig. 4b, c, d; the identification of over-expressedproteins in the Volcano plot is reported in SupplementaryTable 1).
3.3. Functional clusters and single protein, where and how
Fig. 5 shows that, when considering the function of any singleprotein purified in different steps, there was a sort of clusterseparation mainly related to the presence of nuclear proteinsin vesicles. In fact, vesicles contain proteins deriving from thenucleus andmitochondria that are, in general, nucleotide andATP binding families (in red in Fig. 5) while, vice versa,receptor, cell adhesion, signal, trans-membrane and secretedproteins are maximally contained in the fraction derivingfrom CPLL capture (Fig. 5).
Remarkably, in the data presented in SupplementaryTable 2 there are 1724 proteins that have never been detectedin urine before, which, on this basis, can be defined as‘proteins of new description’ in this biological fluid. Detectionof new proteins in urine means that other recent studies onthe normal urine proteome failed to detect their presence.However, as shown by the Venn diagram presented in Fig. 6,
UNCO
RRECT
429
430
431
432
433
Fig. 5 – Gene Ontology analysis of the five urine fractions,obtained via the scheme of Fig. 1A, and displaying theclassification of biological processes of proteins exhibitingdifferent physic-chemical proprieties.
Please cite this article as: Santucci L, et al, From hundreds to thouhttp://dx.doi.org/10.1016/j.jprot.2014.07.021
ED P
R
other relevant studies on urine proteome [23,24,30,66] coulddetect urine proteins of new description on their own. Overall,considering our and other studies, 2883 new proteins havebeen described (Supplementary Table 2 identifies all singlenew urinary proteins). A detailed discussion consideringevery component, its significance and possible use as abiomarker is far beyond the scope of the present workand will be presented separately for different organs andapparatuses.
434435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
4. Discussion
The main objective of our study was to widen the knowledgeon protein composition of normal urine while maintaining ahigh reproducibility. The strategy was to utilize moderntechniques of analysis (mass spectrometry of last generation,CPLLs) and traditional approaches (ultracentrifugation, sol-vent extraction, etc.) while combining them in a virtuoussuccession of steps altogether allowing for maximal efficien-cy. We thus identified 3429 non-redundant proteins in normalurine with a net increment of 2253 proteins not detectable inun-fractioned urine. One basic methodological aspect relatedto the sensitivity of the method needs to be discussed here. Infact, as reported above and in accord with the literature [64],we adopted as standard a cut off 10 ppm. The same limit iswidely utilized in the literature and represents, in our opinion,the gold standard of the technique (see for example [63]).Studies on urine proteome almost uniquely utilized the filterof 10 ppm and the most up-dated databases [23,24,30,66]utilized this condition (Fig. 6). Actually, when adopting morestringent conditions of analysis (i.e. 3 ppm) on the same urine,only 3071 proteins and 13,123 peptides could be detected witha decrease of 10% and 6% respectively (Supplementary Table3). However, we chose to present the results generatedfollowing similar parameters as described in current litera-ture; comparison to results obtained with less or higherstringent conditions will also be feasible when other data-bases will appear. Many of the newly described proteins that
sands: Widening the normal human Urinome (1), J Prot (2014),
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495496
497
498499
500
501
502503
504
505
506
507
508
509
510
511Q8
51 2
513514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
8 J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 4 ) X X X – X X X
could be visualized were much more concentrated in one ormore steps of the procedure. It is of interest that functionalclusters of proteins were purified by selected procedures (e.g.nuclear proteins in vesicles) suggesting that analysis and/orpurification strategies are simplified.
Besides the general considerations above, it is noteworthythat many of the proteins here described have never beenreported in urine before and have been defined here as ‘ofnew description’ on this basis. Several of them are of interestto clinical chemistry as biomarkers of relevant human dis-eases as the list is quite long and involves pathologies ofdifferent organs and systems. A detailed description herewould appear too long and out of the scope of the presentinvestigation. Reports focused on specialized journals wouldbetter suit the objective to attract clinician interest and givethem the opportunity to look at urine for diagnosis of selecteddiseases.
T531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
EC
5. Conclusions
By coupling Orbitrap Mass spectrometry with a few simplefractionation steps almost 3500 proteins could be detected innormal urine, 50% of which are of new description. Assem-bling databases dedicated to different medical specialitieswould constitute the basis for looking at human pathologies.Given the number of proteins detected for the first timein urine, the wealth of information gathered here could formthe basis to consider studies in human pathologies. Diseasesaffecting the kidney and the urinary tract are of particularinterest considering that tissue proteins (for malformation),mediators of inflammation and renal autoantibodies (forglomerulonephritis) are preferentially excreted into urine.
Supplementary data to this article can be found online athttp://dx.doi.org/10.1016/j.jprot.2014.07.021.
R 552
553
554
555
556
557
RConflict of interest
All authors declare no conflicts of interest.
O 558559
560
561
562
563
564
NCTransparency document
The Transparency document associated with this article canbe found, in the online version.
565
566
567
568
569
570
571
572
573
574
575
576
577
578
U
Acknowledgfments
The Giannina Gaslini Institute provided financial and logisticsupport to the study. This work was also supported bythe Italian Ministry of Health ‘Ricerca Corrente’ and fromcontributions derived from ‘Cinque per mille dell'IRPEF’. Wealso acknowledge contributions from the Renal Child Foun-dation, Fondazione La Nuova Speranza (‘Progetto integrato perla definizione dei meccanismi implicati nella glomerulo sclerosifocale’) and Italian Society of Nephrology (Progetto Ricercando).
579
Please cite this article as: Santucci L, et al, From hundreds to thouhttp://dx.doi.org/10.1016/j.jprot.2014.07.021
ED P
RO
OF
R E F E R E N C E S
[1] Decramer S, Gonzalez de Peredo A, Breuil B, Mischak H,Monsarrat B, Bascands JL, et al. Urine in clinical proteomics.Mol Cell Proteomics 2008;7:1850–62.
[2] Lee RS, Monigatti F, Briscoe AC,Waldon Z, FreemanMR, SteenH. Optimizing sample handling for urinary proteomics.J Proteome Res 2008;7:4022–30.
[3] Thongboonkerd V. Practical points in urinary proteomics.J Proteome Res 2007;6:3881–90.
[4] Thongboonkerd V, Chutipongtanate S, Kanlaya R. Systematicevaluation of sample preparation methods for gel-basedhuman urinary proteomics: quantity, quality, and variability.J Proteome Res 2006;5:183–91.
[5] Thongboonkerd V, Saetun P. Bacterial overgrowth affectsurinaryproteomeanalysis: recommendation for centrifugation,temperature, duration, and the use of preservatives duringsample collection. J Proteome Res 2007;6:4173–81.
[6] Vlahou A, Schanstra J, Frokiaer J, El Nahas M, Spasovski G,Mischak H, et al. Establishment of a European Network forurine and kidney proteomics. J Proteomics 2008;71:490–2.
[7] Zurbig P, Renfrow MB, Schiffer E, Novak J, Walden M, WittkeS, et al. Biomarker discovery by CE-MS enables sequenceanalysis via MS/MS with platform-independent separation.Electrophoresis 2006;27:2111–25.
[8] Candiano G, Santucci L, Petretto A, Bruschi M, Dimuccio V,Urbani A, et al. 2D-electrophoresis and the urine proteomemap: where do we stand? J Proteomics 2010;73:829–44.
[9] Court M, Selevsek N, Matondo M, Allory Y, Garin J, MasselonCD, et al. Toward a standardized urine proteome analysismethodology. Proteomics 2011;11:1160–71.
[10] Molina L, Salvetat N, Ameur RB, Peres S, Sommerer N, JarrayaF, et al. Analysis of the variability of human normal urine by2D-GE reveals a “public” and a “private” proteome. J Proteomics2011;75:70–80.
[11] Nagaraj N, Mann M. Quantitative analysis of the intra- andinter-individual variability of the normal urinary proteome.J Proteome Res 2011;10:637–45.
[12] Mischak H, KolchW, Aivaliotis M, Bouyssie D, Court M, DihaziH, et al. Comprehensive human urine standards forcomparability and standardization in clinical proteomeanalysis. Proteomics Clin Appl 2010;4:464–78.
[13] Anderson NG, Anderson NL, Tollaksen SL. Proteins of humanurine. I. Concentration and analysis by two-dimensionalelectrophoresis. Clin Chem 1979;25:1199–210.
[14] Anderson NG, Anderson NL, Tollaksen SL, Hahn H, Giere F,Edwards J. Analytical techniques for cell fractions. XXV.Concentration and two-dimensional electrophoretic analysisof human urinary proteins. Anal Biochem 1979;95:48–61.
[15] Edwards JJ, Tollaksen SL, Anderson NG. Proteins of humanurine. III. Identification and two-dimensional electrophoreticmap positions of some major urinary proteins. Clin Chem1982;28:941–8.
[16] Spahr CS, Davis MT, McGinley MD, Robinson JH, Bures EJ,Beierle J, et al. Towards defining the urinary proteome usingliquid chromatography–tandemmass spectrometry. I. Profilingan unfractionated tryptic digest. Proteomics 2001;1:93–107.
[17] Bueler MR, Wiederkehr F, Vonderschmitt DJ. Electrophoretic,chromatographic and immunological studies of humanurinary proteins. Electrophoresis 1995;16:124–34.
[18] Thongboonkerd V, McLeish KR, Arthur JM, Klein JB. Proteomicanalysis of normal human urinary proteins isolated byacetone precipitation or ultracentrifugation. Kidney Int 2002;62:1461–9.
[19] Weissinger EM, Wittke S, Kaiser T, Haller H, Bartel S, Krebs R,et al. Proteomic patterns established with capillaryelectrophoresis and mass spectrometry for diagnosticpurposes. Kidney Int 2004;65:2426–34.
sands: Widening the normal human Urinome (1), J Prot (2014),
T
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
9J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 4 ) X X X – X X X
UNCO
RREC
[20] Schaub S, Wilkins J, Weiler T, Sangster K, Rush D, NickersonP. Urine protein profiling with surface-enhancedlaser-desorption/ionization time-of-flight massspectrometry. Kidney Int 2004;65:323–32.
[21] Oh J, Pyo JH, Jo EH, Hwang SI, Kang SC, Jung JH, et al.Establishment of a near-standard two-dimensional humanurine proteomic map. Proteomics 2004;4:3485–97.
[22] Pieper R, Gatlin CL, McGrath AM, Makusky AJ, Mondal M,Seonarain M, et al. Characterization of the human urinaryproteome: a method for high-resolution display of urinaryproteins on two-dimensional electrophoresis gels with a yieldof nearly 1400 distinct protein spots. Proteomics 2004;4:1159–74.
[23] Adachi J, Kumar C, Zhang Y, Olsen JV, Mann M. The humanurinary proteome contains more than 1500 proteins, includinga large proportion of membrane proteins. Genome Biol 2006;7:R80.
[24] Marimuthu A, O'Meally RN, Chaerkady R, Subbannayya Y,Nanjappa V, Kumar P, et al. A comprehensive map of thehuman urinary proteome. J Proteome Res 2011;10:2734–43.
[25] Ru QC, Katenhusen RA, Zhu LA, Silberman J, Yang S, OrchardTJ, et al. Proteomic profiling of human urine usingmulti-dimensional protein identification technology.J Chromatogr A 2006;1111:166–74.
[26] Soldi M, Sarto C, Valsecchi C, Magni F, Proserpio V, Ticozzi D,et al. Proteome profile of human urine with two-dimensionalliquid phase fractionation. Proteomics 2005;5:2641–7.
[27] Sun W, Li F, Wu S, Wang X, Zheng D, Wang J, et al. Humanurine proteome analysis by three separation approaches.Proteomics 2005;5:4994–5001.
[28] Zerefos PG, Vougas K, Dimitraki P, Kossida S, Petrolekas A,Stravodimos K, et al. Characterization of the human urineproteome by preparative electrophoresis in combination with2-DE. Proteomics 2006;6:4346–55.
[29] Kentsis A, Monigatti F, Dorff K, Campagne F, Bachur R, SteenH. Urine proteomics for profiling of human disease usinghigh accuracy mass spectrometry. Proteomics Clin Appl 2009;3:1052–61.
[30] Zerefos PG, Aivaliotis M, Baumann M, Vlahou A. Analysis ofthe urine proteome via a combination of multi-dimensionalapproaches. Proteomics 2012;12:391–400.
[31] Castagna A, Cecconi D, Sennels L, Rappsilber J, Guerrier L,Fortis F, et al. Exploring the hidden human urinary proteomevia ligand library beads. J Proteome Res 2005;4:1917–30.
[32] Candiano G, Dimuccio V, Bruschi M, Santucci L, Gusmano R,Boschetti E, et al. Combinatorial peptide ligand libraries forurine proteome analysis: investigation of different elutionsystems. Electrophoresis 2009;30:1–7.
[33] Candiano G, Santucci L, Bruschi M, Petretto A, D'Ambrosio C,Scaloni A, et al. “Cheek-to-cheek” urinary proteome profilingvia combinatorial peptide ligand libraries: a novel,unexpected elution system. J Proteomics 2012;75:796–805.
[34] Boschetti E, Righetti PG. The art of observing rare proteinspecies in proteomes with peptide ligand libraries. Proteomics2009;9:1492–510.
[35] Righetti PG, Boschetti E. The ProteoMiner and theFortyNiners: searching for gold nuggets in the proteomicarena. Mass Spectrom Rev 2008;27:596–608.
[36] Righetti PG, Boschetti E, Zanella A, Fasoli E, Citterio A.Plucking, pillaging and plundering proteomes withcombinatorial peptide ligand libraries. J Chromatogr A 2010;1217:893–900.
[37] Santucci L, Candiano G, Bruschi M, D'Ambrosio C, Petretto A,Scaloni A, et al. Combinatorial peptide ligand libraries for theanalysis of low-expression proteins: validation for normalurine and definition of a first protein MAP. Proteomics 2012;12:509–15.
[38] Pisitkun T, Shen RF, Knepper MA. Identification and proteomicprofiling of exosomes in human urine. Proc Natl Acad Sci U S A2004;101:13368–73.
Please cite this article as: Santucci L, et al, From hundreds to thouhttp://dx.doi.org/10.1016/j.jprot.2014.07.021
ED P
RO
OF
[39] Zhou H, Yuen PS, Pisitkun T, Gonzales PA, Yasuda H, Dear JW,et al. Collection, storage, preservation, and normalization ofhuman urinary exosomes for biomarker discovery. Kidney Int2006;69:1471–6.
[40] Merchant ML, Powell DW,Wilkey DW, Cummins TD, DeegensJK, Rood IM, et al. Microfiltration isolation of human urinaryexosomes for characterization by MS. Proteomics Clin Appl2010;4:84–96.
[41] Rood IM, Deegens JK, Merchant ML, Tamboer WP, Wilkey DW,Wetzels JF, et al. Comparison of threemethods for isolation ofurinary microvesicles to identify biomarkers of nephroticsyndrome. Kidney Int 2010;78:810–6.
[42] Zhang Y, Li Y, Qiu F, Qiu Z. Comprehensive analysis oflow-abundance proteins in human urinary exosomes usingpeptide ligand library technology, peptide OFFGELfractionation and nanoHPLC-chip-MS/MS. Electrophoresis2010;31:3797–807.
[43] Taylor DD, Zacharias W, Gercel-Taylor C. Exosome isolationfor proteomic analyses and RNA profiling. Methods Mol Biol2011;728:235–46.
[44] Wang Z, Hill S, Luther JM, Hachey DL, Schey KL. Proteomicanalysis of urine exosomes by multidimensional proteinidentification technology (MudPIT). Proteomics 2012;12:329–38.
[45] Gruenberg J, Stenmark H. The biogenesis of multivesicularendosomes. Nat Rev Mol Cell Biol 2004;5:317–23.
[46] Thery C, Zitvogel L, Amigorena S. Exosomes: composition,biogenesis and function. Nat Rev Immunol 2002;2:569–79.
[47] Li XB, Zhang ZR, Schluesener HJ, Xu SQ. Role of exosomes inimmune regulation. J Cell Mol Med 2006;10:364–75.
[48] Valadi H, Ekstrom K, Bossios A, Sjostrand M, Lee JJ, Lotvall JO.Exosome-mediated transfer of mRNAs and microRNAs is anovel mechanism of genetic exchange between cells. Nat CellBiol 2007;9:654–9.
[49] Moon PG, You S, Lee JE, Hwang D, Baek MC. Urinary exosomesand proteomics. Mass Spectrom Rev 2011;30:1185–202.
[50] van Balkom BW, Pisitkun T, Verhaar MC, Knepper MA.Exosomes and the kidney: prospects for diagnosis andtherapy of renal diseases. Kidney Int 2011;80:1138–45.
[51] Santucci L, Candiano G, Bruschi M, Bodria M, Murtas C,Petretto A, et al. Urinary proteome in a snapshot: normalurine and glomerulonephritis. J Nephrol 2013;26:610–6.
[52] Alvarez ML, Khosroheidari M, Kanchi Ravi R, Distefano JK.Comparison of protein, microRNA, and mRNA yields usingdifferentmethods of urinary exosome isolation for the discoveryof kidney disease biomarkers. Kidney Int 2012;82:1024–32.
[53] Cheruvanky A, Zhou H, Pisitkun T, Kopp JB, Knepper MA, YuenPS, et al. Rapid isolation of urinary exosomal biomarkers usinga nanomembrane ultrafiltration concentrator. Am J PhysiolRenal Physiol 2007;292:F1657–61.
[54] Moon PG, Lee JE, You S, Kim TK, Cho JH, Kim IS, et al.Proteomic analysis of urinary exosomes from patients ofearly IgA nephropathy and thin basement membranenephropathy. Proteomics 2011;11:2459–75.
[55] Pisitkun T, Johnstone R, Knepper MA. Discovery of urinarybiomarkers. Mol Cell Proteomics 2006;5:1760–71.
[56] Raj DA, Fiume I, Capasso G, Pocsfalvi G. A multiplexquantitative proteomics strategy for protein biomarkerstudies in urinary exosomes. Kidney Int 2012;81:1263–72.
[57] Simpson RJ, Lim JW, Moritz RL, Mathivanan S. Exosomes:proteomic insights and diagnostic potential. Expert RevProteomics 2009;6:267–83.
[58] Fernandez-Llama P, Khositseth S, Gonzales PA, Star RA,Pisitkun T, Knepper MA. Tamm–Horsfall protein and urinaryexosome isolation. Kidney Int 2010;77:736–42.
[59] Musante L, Saraswat M, Duriez E, Byrne B, Ravida A, Domon B,et al. Biochemical and physical characterisation of urinarynanovesicles following CHAPS treatment. PLoS One 2012;7:e37279.
sands: Widening the normal human Urinome (1), J Prot (2014),
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
10 J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 4 ) X X X – X X X
[60] Gonzales PA, Pisitkun T, Hoffert JD, Tchapyjnikov D, Star RA,Kleta R, et al. Large-scale proteomics and phosphoproteomicsof urinary exosomes. J Am Soc Nephrol 2009;20:363–79.
[61] Gonzales PA, Zhou H, Pisitkun T, Wang NS, Star RA, KnepperMA, et al. Isolation and purification of exosomes in urine.Methods Mol Biol 2010;641:89–99.
[62] Wisniewski JR, Zougman A, Nagaraj N, Mann M. Universalsample preparation method for proteome analysis. NatMethods 2009;6:359–62.
[63] Branca RM, Orre LM, Johansson HJ, Granholm V, Huss M,Perez-Bercoff A, et al. HiRIEF LC-MS enables deep proteomecoverage and unbiased proteogenomics. Nat Methods 2014;11:59–62.
UNCO
RRECT
Please cite this article as: Santucci L, et al, From hundreds to thouhttp://dx.doi.org/10.1016/j.jprot.2014.07.021
[64] Gupta N, Pevzner PA. False discovery rates of proteinidentifications: a strike against the two-peptide rule.J Proteome Res 2009;8:4173–81.
[65] Hsieh EJ, HoopmannMR, MacLean B, MacCoss MJ. Comparisonof database search strategies for high precursor mass accuracyMS/MS data. J Proteome Res 2010;9:1138–43.
[66] Li QR, Fan KX, Li RX, Dai J, Wu CC, Zhao SL, et al. Acomprehensive and non-prefractionation on the protein levelapproach for the human urinary proteome: touchingphosphorylation in urine. Rapid Commun Mass Spectrom2010;24:823–32.
ED P
RO
OF
sands: Widening the normal human Urinome (1), J Prot (2014),