JOURNAL OF BACTERIOLOGY, Nov. 1985, p. 925-928 Vol. 164, No. 20021-9193/85/110925-04$02.00/0Copyright © 1985, American Society for Microbiology

Export Defect Adjacent to the Processing Site of StaphylococcalNuclease Is Suppressed by a prlA MutationLAURA R. LISS, BARBARA L. JOHNSON, AND DONALD B. OLIVER*

Department of Microbiology, State University ofNew York at Stony Brook, Stony Brook, New York 11794

Received 29 April 1985/Accepted 13 August 1985

Plasmids have been constructed in which the Escherichia coli alkaline phosphatase promoter and signalsequence have been fused to the staphylococcal nuclease gene to promote the high-level expression and secretionof this gene product in E. coli. We determined that the first amino acid residue after the signal sequence candetermine whether this protein was processed and exported to the periplasmic space. Fractionation andprotease accessibility studies were used to show that the export-defective, nuclease precursor is internal to thecytoplasmic membrane barrier of the cell. Furthernore, this export defect was suppressed in a straincontaining a prUi mutation. These findings are novel in that this region of the polypeptide chain has beenimplicated in processing but not export and that priA mutations have not been previously known to suppresssuch defects.

While engineering the high-level expression and secretionof the nuclease from Staphylococcus aureus Foggi in Esch-erichia coli, we noticed that the first amino acid after thesignal sequence can determine whether staphylococcal nu-clease was processed and exported to the periplasmic space.Furthermore, thiS export defect was suppressed in a straincontaining a prlA mutation (4). These findings are novelsince mutations adjacent to the processing site generallyaffect signal peptide cleavage but not export (8, 16), and prlAmutations have been shown to suppress defects in thehydrophobic core region of the signal sequence (3, 4) but notdefects adjacent to the processing site. Therefore, we char-acterized this phenomenon in greater detail, and the resultsare described below.To promote the high-level expression and secretion of

staphylococcal nuclease in E. coli, plasmids have beenconstructed in which this gene lacking its own signal se-quence (18) was fused to the E. coli alkaline phosphatasepromoter and signal sequence (7). pFOG403 is a pBR322derivative plasmid containing a precise fusion of theseelements (Fig. 1). To maintain an amino acid sequenceadjacent to the processing site similar to that found in thealkaline phosphatase precursor, oligonucleotide-directedmutagenesis was performed on pFOG403 to insert the tripletCGG encoding an arginine residue between the alkalinephosphatase signal sequence and the staphylococcal nucle-ase gene (Fig. 1). This procedure resulted in the creation ofpFOG406. The details of these plasmid constructions will bedescribed elsewhere (D. Shortle, manuscript in preparation).

Strain MC4100 (F- AlacUJ69 araD136 relA rpsL thi) (10)or its isogenic derivative SE6004 containing the prlA4 muta-tion t3) was used as a host for pFOG403, pFOG406, orpBR322. Overnight cultures grown in L broth were diluted50-fold into MOPS (morpholinopropanesulfonic acid) me-dium (14) containing 1.0% glucose, 0.15% vitamin-freecasein hydrolysate, 0.0001% thiamine, and 0.1 mM KH2PO4and grown at 37°C with shaking for 7 h to allow induction andaccumulation of the plasmid-encoded staphylococcal nucle-ase. For radiolabeling protein, strains were grown for 4 h inMOPS medium, as described above except lacking 0.15%casein hydrolysate and containing 0.05% yeast extract and

* Corresponding author.

0.2 mM KH2PO4, and then labeled with [35S]methionine(1,050 Ci/mmol) at 20 ,uCi/ml for 10 min. Staphylococcalnuclease was immunoprecipitated from radiolabeled culturesas described previously (15) with an antiserum directedagainst this protein. Total cell protein and immunoprecipi-tates were examined on a 15% polyacrylamide gel containingsodium dodecyl sulfate (Fig. 2). Both strains MC4100 andSE6004 containing pFOG403 had high nuclease activity uponphosphate starvation (360 and 450 U/mg of protein, respec-tively [6]) and produced large quantities of a protein thatcomigrates with purified staphylococcal nuclease (Worth-ington Diagnostics) and is immunoprecipitable with staphylo-coccal nuclease antiserum (Fig. 2, lanes 1, 2, 6, 8, and 9). Incontrast, strain MC4100 containing pFOG406 had little nu-clease activity upon phosphate starvation (20 U/mg of pro-tein) and instead produced large quantities of a somewhathigher-molecular-weight protein that is the presumed staph-ylococcal nuclease precursor since it is immunoprecipitablewith staphylococcal nuclease antiserum (Fig. 2, lanes 3 and10). Thus, the additional arginine codon present after thealkaline phosphatase signal sequence encoded on pFOG406results in the production of an enzymatically inactive, un-processed precursor. This result is somewhat unexpectedsince an arginine residue is normally found at this position inalkaline phosphatase (Fig. 1). Furthermore, when pFOG406was introduced into strain SE6004 containing a priA muta-tion, this export or processing defect was suppressed, result-ing in the processing and restored enzymatic activity of amajor portion of staphylococcal nuclease produced by thisstrain (425 U/mg of protein) (Fig. 2, lanes 4 and 11). Thisresult is also somewhat unexpected since prlA suppressormutations have been shown previously to suppress defectsin the hydrophobic core region of the signal sequence (3, 4)but not defects adjacent to the processing site.We determined the cellular location of the staphylococcal

nuclease precursor and mature form that are produced inthese strains. Periplasmic fractions were prepared fromstrains grown without labeling as described above by thelysozyme-EDTA method described previously (15)..We alsoused an alternative method that specifically releases staph-ylococcal nuclease that has been secreted into the periplas-mic space. In this method cells are sedimented, resuspendedin 1/40 volume of cold 1 M Tris (pH 10.5)-2.5 mM EDTA,

925

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

16

Oct

ober

202

1 by

120

.51.

50.1

9.

926 NOTES

alkaline phosphatass

ACA AAA GCC CGG ACA CCA

Thr - Lys - Ala - Arg - Thr - Pro

-3 -2 -1 +1 +2 +3

pFOG 403

ACA AAA GCC GCA ACT TCA

Thr- Lys -Ala -Ala -Thr - Ser

-3 2 -1 +1 +2 +3

above, and lysozyme-EDTA spheroplasts were preparedfrom each sample. Spheroplasts were incubated on ice withvarious amounts of trypsin in the absence or presence of0.05% Triton X-100 for 30 min, followed by the addition ofsoybean trypsin inhibitor to a final concentration of 200,ug/ml. Samples were heated at 1009C for 2 min in satiplebuffer (15) and analyzed on 15% polyacrylamide gels con-taining sodium dodecyl sulfate (Fig. 4). It can be seen thatstaphylococcal nuclease present in spheroplasts of strainMC4100 containing pFOG403 is de'graded by the addition oftrypsin either when the cytoplasmic membrane is intact orwhen it is disrupted by the addition of Triton X-100, indicat-ing 'that this protein is secreted through the cytoplasmicmembrane (Fig. 4, cf. lanes 2 through 5 with lanes 6 through9). In contrast, the staphylococcal huclease precursor pres-ent in spheroplasts of stra'in MC4190\ containing pFOG406 isdegr,aded only when the cytoplasmic membrane is disruptedby the addition of Triton X-100, indicating that this protein isinternal to the cytoplasmic membrane (Fig. 4, cf. lanes 10

pFOG 406

ACA AAA GCC CGG GCA ACT

O o o 0 m B

I0 C.)e u zz 2

Cm Cl) ID (r

r II

cn

Thr -Lys-Ala-A rg -Ala -Thr

-3 -2 1 + 1 +2 +3FIG. 1. Sequences adjacent to the processing site of alkaline

phosphatase and staphylococcal nuclease containing the alkalinephosphatase signal sequence. The DNA and amino acid sequencesadjacent to the processing site of alkaline phosphatase (7) andalkaline phosphatase signal sequence fus$ed to staphylococcal nucle-ase are shown. pFOG406 is derived from pFOG403 by the insertionof CGG between the alkaline phosphatase signal sequence andstaphylococcal nuclease gene by oligonucleotide-directed mutagen-esis. This alteration creates a cleavage site for SmaI. The sequenceofthe entire signal sequence through the SmaI site for pFOG406wasconfirmed by DNA sequence analysis. Negative numbers indicateamino acid residues in the alkaline phosphatase signal sequence,starting from the normal processing site. Positive numbers indicateamino acid residues in mature alkaline phosphatase or maturestaphylococcal nuclease.

incubated on ice for 20 min, and sedimented, and thesupernatant fraction is recovered (D. Shortle, personal com-munication) (Fig. 3). Strains MC4100 and SE6004 containingpFQG403 as well as strain SE6004 containing pFOG406produced mature periplasmic staphylococcal nuclease thatcould be specifi,cally released by either of these methods(Fig. 3, lanes 1, 2, and 4, and 7, 8, and 10). In'contrast,staphylococcal nuclease precursor made by strain MC4100containing pFOG406 was not released by either method (Fig.3, lanes 3 and 9). We conclude either that the staphylococcalnuclease precursor made by this strain exhibits a secretiondefect or that it is secreted but not released fhto theperiplasmic fraction because it is insoluble or associatedwith the exterior face of the cytoplasmic membrane, as hasbeen found with several other exported proteins with abnor-mal structures (11, 12).To determine whether the unprocessed staphylococcal

nuclease precursor produced from strqjn 44C4100 containingpFOG406 is exterior to the cytoplasmic membrane bwrier,we performed trypsin accessibility studies on spheroplasts.Cultures were labeled with [3'S]methionine as described

_:

_.

RF_ F~~prSNAm 4 M_ _ _ _ _ ~ -S

1 2 3 4 5 6Total cell protein

7 8 9 10 11 12Immunoprecipitations

FIG. 2. Different forms of staphylococcal nuclease produced bystrains containing pFOG403 or pFOG406. Strains were grown inMOPS medium containing 1.0% glucose, 0.15% vitamin-free caseinhydrolysate, 0.0001% thiamine, and 0.1 mM KH2PO4 at 37°C withshaking for 7 h to examine total cell protein. Cells were sedimented,suspended in 1/10 volume of 'sample' buffer, and heated to 100°C for2 min, and 10 ,ul of each sample was analyzed on the polyacrylamidegel shown. For radiolabeling, cells were'grown for 4 h in MOPSmedium as described above except lacking 0.15% casein hydrolys-ate and containing 0.05% yeast extract and 0.2 mM KH2PO4, andthen 1 ml of culture was labeled with 2.0 pCi of '[35S]methionine(1,050 Ci/mmol) per ml for 10 min. Immunoprecipitation of staphy-lococcal nuclease was performed as described previously (15). Eachimmunoprecipitate (20 ,ul) corresponding to 0.08 ml of the respectiveculture was analyzed on the polyacrylamide gel shown. Lanes: 1through 5, gel stained with Coomassie brilliant blue R of total cellprotein from MC4100(pFOG403), SE6004(pFOG403),MC4100(pFOG406), SE6004(pFOG406), and MC4100(pBR322); 6,staphylococcal nuclease standard; 7' radiolabeled precursor andmature form of staphylococcal nuclease; 8 through 12, autoradio-gram of immunoprecipitates of the cultures given in lanes 1 through5. preSN and SN indicate the positions of staphylococcal nucleaseprecursor and mature form, respectiv'ely.

J. BACTERIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

16

Oct

ober

202

1 by

120

.51.

50.1

9.

NOTES 927

through 13 with lanes 14 through ,17). Thus the unprocessedstaphylococcal nuclease precursor produced by strainMC4100 containing pFOG406 has not crossed the cytoplas-mic membrane barrier and therefore exhibits a secretiondefect.

Previous studies on the export of proteins to the periplas-mic space of E. coli have implicated the signal sequence asserving a primary role in this process. Proteins with defectsin the hydrophobic core region of the signal sequence areexport defective and accumulate as unsecreted cytoplasmicprecursors (2, 5, 13). In contrast, proteins with defectsadjacent to the processing site either within the signalsequience (9) or within the early mature region (8, 16) showprocessing defects but are exported to the correct cellularlocation. The staphylococcal nuclease precursor encoded bypFOG406 represents a novel example in which a change inthe first amino acid residue after the signal sequence canprevent export of the protein through the cytoplasmic mem-brane. This result may indicate that the signal peptide andearly mature region of an envelope protein. assume a partic-ular structure that is necessary to promote export of theprotein. Such a structure would be required to correctlyinteract with the export machinery of the cell. In accord withthis proposal, a priA mutation was, found to suppress theexport defect caused by an amino acid alteration adjacent tothe normal processing site present in staphylococcal nucle-ase encoded by pFOG406.The priA gene product is thought to be a component of the

E. coli export machinery since certain priA mutations cansuppress defects in the hydrophobic core regiqn of signalsequences in an allelic specific manner (3), whereas other

co n ° ei0

X X o X

o Lu Z U U'i 2 C 2Co2 CO 2CO

xw 4

- -___e,

SN-

1 2 3 4 S B 7

Lysozyme- EDTA

8 9 10 11

High pH Tris-EDTA

FIG. 3. Periplasmic location of the processed form of staphylo-coccal nuclease. Strains were grown without labeling as described inthe legend to Fig. 2, and the periplasmic fractioh was isolated by thelysozyme-EDTA method descrnbed previously (15). Alternatively, aportion of each culture was subjected to the high-pH, Tris-EDTAprocedure described in the text. A 20-pJ volume of 10-fold concen-trated periplasmic fractions and 5 IL of 40-fold high-pH, Tris-EDTAsupernatants were heated at 100°C for 2 min in sample buffer andanalyzed on the polyacrylamide gel shown. Lanes: '1 through 5,periplasmic fractions from MC41bO(pFOG403), SE6004(pFOG403),MC4100(pFOG406), SE6004(pFOG406), and MC4100/pBR322, re-spectively; 6, staphylococcal nuclease standard; 7 through 11,high-pH, Tris-EDTA supernatants from the cultures given in lanes 1through 5. SN indicates the position of staphylococcal nuclease.

ugimi

MC-406o - Triton O+.Tr i ton

o w i ° oo 8o

I| hi lsI

Itt~~t: v

X *4p@ . .- *9* a 0

preSN- _ ***S

SN- '*l* * ** S *0 W 4

*S t? *eif ?w

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

FIG. 4. Trypsin sensitivity of exported staphylococcal nuclease.Strains were grown and labeled with [35S]-methionine as describedin the legend to Fig. 2. Lysozyme-EDTA spheroplasts were pre-pared and incubated on ice with 0, 50, 100, or 200 ,ug of trypsin perml in the absence or presence of 0.05% Triton X-100 for 30 min,followed by the addition of soybean trypsin inhibitor to 200 ,ug/ml.Samples were heated at 100°C for 2 min in sample buffer, and 20 ,ulof each sample was analyzed on the polyacryiamide gel shown.Lanes: 1, staphylococcal nuclease precursor and mature formstandard; 2 through 9, MC4100(pFOG403); 10 through 17,MC4100(pFOG406). Each set of eight lanes corresponds to 0, 50,100, or 200 ,ug of trypsin per ml in the absence or presence of 0.05%Triton X-100, respectively. preSN and SN indicates staphylococcalnuclease precursor and mature form, respectively.

conditional lethal prlA alleles show general export defectswith normal envelope proteins (17). It has been proposedthat the pr/A gene product interacts with the signal peptide ofan envelope protein to pro'mote its export (3, 4). Since analtered prlA gene product not only suppresses defects in thehydrophobic core region of the signal peptide but also adefect adjacent to the normal processing site, then it may bethe overall structure of the amino-terminal region of theenvelope protein that the prlA gene product is recognizingrather than a specific site within the signal peptide. This lineof reasoning is supported by the finding that primary defectswithin the hydrophobic core region of the signal peptide formaltose-binding protein can be corrected somewhat by sec-ondary amino acid alterations in the 19th mature amino acidresidue of this protein (1). Clarification of the role of theearly mature region of an envelope protein in promoting itsexport and the involvement of the priA gene product in thisprocess will require further genetic and biochemical analysisof these components.

We thank Hiroshi Inouye for the plasmid containing the alkalinephosphatase signal sequence and David Shortle for pFOG403 andpFOG406 and advice during the course of the project.

This work was supported by grants 1RO1GM32958 (D.B.O.) and5T32GM07964 (B.L.J.) from the National Institute of General Med-ical Sciences and 5T32CA09176 (L.R.L.j from the National CancerInstitute.

Vol. 164, 1985

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

16

Oct

ober

202

1 by

120

.51.

50.1

9.

928 NOTES

LITERATURE CITED1. Bankgs, V. A., B. A. Rasmussen, and P. J. Bassford, Jr. 1984.

Intragenic suppressor mutations that restore export of maltosebinding protein with a truncated signal peptide. Cell 37:243-252.

2. Bedoueile, H., P. J. Bgssford, A. V. Fowler, I. Zabin, J.Beckwith, and M. Hofnung. 1980. Mutations which alter thefunction of the signal sequehce of the maltose binding protein ofEscherlchia coli. Nature (London) 285:78-81.

3. Emr, S. D., and P. J. Bassford. 1982. Localization and process-ing of outer membrane and periplasmic proteins in Escherichiacoli strains harboring export-specific suppressor mutations. J.Biol. Chem. 257:5852-5860.

4. Emr, S. I., S. Hanley-Way, and T. J. Silhavy. 1981. Suppressormutations that restore export of a protein with a defective signalsequence. Cell 23:79-88.

5. Emr, S. D., J. Hedgpeth, J.-M. Clement, T. J. Silhavy, and M.Hofnung. 1980. Sequence analysis of mutations that preventexport of X receptor, an Escherichia coli outer membraneprotein. Nature (London) 285:82-85.

6. Heins, J. N., J. R. Suriano, H. Taniuchi, and C. B. Anfilsen.1967. Characterization of a nuclease produced by Staphylococ-cus aureus. J. Biol. Chemn. 242:1016-1020.

7. Inouye, H., W. Barnes, and J. Beckwith. 1982. Signal sequenceof alkaline phosphatase of Escherichia coli. J. Bacteriol.149:434-439.

8. Inouye, S., T. Franceschiiii, M. Sato, K. Itakura, and M. Inouye.1983. Prolipoprotein signal peptidase of Escherichia coli re-quires a cysteine residue at the cleavage site. EMBO J. 2:87-91.

9. Inouye, S., C.-P. S. Hsu, K. Itakura, and M. Inouye. 1983.Requirement for signal peptide cleavage of Escherichia coli

prolipoprotein. Science 221:59-61.10. Ito, K., P. J. Bassford, and J. R. Beckwith. 1981. Protein

localization in Escherichia coli: is there a common step in thesecretion of periplasmic and outer membrane proteins? Cell24:707-714.

11. Ito, K., and J. R. Beckwith. 1981. Role of the mature prOteinsequence of maltose-binding protein in its secretion across theE. coli cytoplasmic membrane. Cell 25:143-150.

12. Koshland, D., and D. Botstqin. 1982. Evidence for posttransla-tional translocation of I3lactamase across the bacterial innermembrane. Cell 30:893-902.

13. Michaelis, S., H. Inouye, D. Oliver, and J. Beckwith. 1983.Mutations that alter the signal sequence of alkaline phosphataseof Escherichia coli. J. Bacteriol. 154:366-374.

14. Neidhardt, F. C., P. L. Bloch, and D. F. Smith. 1974. Culturemedium for enterobacteria. J. Bacteriol. 119:736-747.

15. Oliver, D. B., and J. Beckwith. 1982. Regulation of a membranecomponent required for protein secretion in Escherichia coli.Cell 30:311-319.

16. Russel, M., and P. Model. 1981. A mutation downstream fromthe signal peptidase cleavage site affects cleavage but notmembrane insertion of phage coat protein. Proc. Natl. Acad.Sci. USA 78:1717-1721.

17. Shiba, K., K. Ito, T. Yura, and D. P. Cerretti. 1984. A definedmutation in the protein export gene with the spc ribosomalprotein operon of Escherichia coli: isolation and characteriza-tion of a new temperature-sensitive secY mutant. EMBO J.3:631-635.

18. Shortle, D. 1983. A genetic system for analysis of staphylococ-cal nuclease. Gene 22:181-189.

J. BACTERIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

16

Oct

ober

202

1 by

120

.51.

50.1

9.