Characterization of Vibrio cholerae protease activities with peptide digest analysis

0095-1137/81/010080-05$02.00/0

Characterization

of

Vibrio

cholerae Protease Activities with

Peptide

Digest

Analysis

DENNIS R.SCHNEIDER,t* SUZANNE P. SIGEL, AND CHARLOTTE D. PARKERt

Department ofMicrobiology, University ofTexasatAustin,Austin, Texas 78712

Asimple method for the analysis ofmicrobial proteases isdescribed thatwas usedtocharacterize theproteolyticactivities of various Vibriocholerae isolates. Thismethod utilized the unique peptides generated from the

degradation

ofa standard protein byproteasesof variousspecificities. Thesepeptideswere anaâ

lyzed by sodium dodecyl

sulfate-polyacrylamide gel

electrophoresis.

Theunique

patternsofpeptidesseeningelscanbe usedtotype proteases

according

totheir relative specificities. Culture supernatants of V. cholerae isolates fromavariety ofenvironmental and humansources wereanalyzedfor the presence ofaprotease previously isolated and characterized in thislaboratoryfrom V. cholerae strain CA401. Supernatants from most isolates showing

dimethyl

casein proteolytic activity exhibited the presence of enzymes similartotheCA401proteaseintheir peptidedigestpatternsagainstbovineserumalbumin and intheir immunological reactivities. The probable widespreadpresence of thisvirulence-associated pro-teasein V. cholerae isolates is discussed.

Microbial proteases are

increasingly

recog-nizedasimportantvirulence factors foravariety ofpathogens. Evidence exists that proteases may

beimportant virulencefactors in Pseudomonas

aeruginosa, Vibrio cholerae, and

possibly

Ser-ratia marcescens infections(6-9).Proteasesare oftendifficult enzymestocharacterize

biochem-ically. A single protease may display

heteroge-neity in molecular weight or

charge.

This may

be dueto post-translational processingor

auto-or heterocatalytic destruction during

purifica-tion. With growing evidence of the importance of proteases ininfections, it isimportantto de-velop simple methods which can be used to characterize proteases from clinical or environ-mental isolates.Afrequentlyusedtechniquefor protease typing has been the electrophoretic

separation of proteases in agarose, starch, or

polyacrylamide gelsfollowedbyanactivitystain

(zymogramtechnique).Thistechniqueproduces

unique spots or zones ofenzymatic activity on

thegel,and haschieflybeenused fortaxonomic

purposes (2, 5, 11). The zymogram technique is limited by a lack of specificity in the staining

procedures andthe electrophoretic

heterogene-ity which can be induced ina single precursor

enzymeby various degreesofpost-translational cleavage. Thus, a single enzyme may produce multiple activity spots which represent catalyt-ically identical but electrophoretcatalyt-ically

heterog-enousenzymes.

t Present address: Department of Microbiology, School of

Medicine, UniversityofMissouri-Columbia,Columbia, MO 65212.

Thetechnique described hereutilizes the cat-alyticspecificity of proteases as a basis for typing strains of V. cholerae.Astandard protein is used as a substrate for proteases found in culture supernatants. The unique peptidesgenerated by cleavage are separated on the basis of relative molecular weight by using sodium dodecyl

sul-fate-polyacrylamide gel electrophoresis

(SDS-PAGE). The unique banding patterns generated in the gels are used as the basis for catalytic typing.

This technique was inspired by the work of Clevelandetal. (1), whoutilized SDS-PAGEto perform peptide digest analysis of phage pro-teins. We have in essence reversed this technique and utilized the various cleavage patterns of a reference protein by different protease prepara-tions to establish the catalytic relaprepara-tions of the proteases to one another.

Theprotease of V.cholerae strainCA401 has recently been isolated and characterized (D. R.

Schneider, Ph.D. thesis, University of Texas at

Austin, 1978). Strain CA401 syncase

superna-tants appearto containa protease of one

cata-lytic specificity and antigenicity which exists in twoformsofdifferingmolecular weight. A mod-ification of the catalytic typing technique de-scribed here was used in the aforementioned publicationto help establish the identical cata-lytic specificities of these two forms of the en-zyme. Dueto the extensivecharacterization of the protease of strain CA401, it was chosen as

the reference strain forthis study. This paper

80

on February 7, 2020 by guest

http://jcm.asm.org/

A

B

C

VIBRIO CHOLERAE PROTEASE 81

D

E

F

TABLE 1. Summary ofprotease analysis ofVibrio

cholerae strains

FIG. 1. SDS-PAGE gelsofBSAdigestionproducts

of various protease preparations. Lanes: A, BSA

alone(no enzymeadded); B, bovine trypsin treated

(10pg/100,ulof digestion mix, 30-min incubation); C,

S.griseus protease (10pg/UX0,il, 15-minincubation);

D, V. cholerae CA401 culture supernatant (15-min

incubation); E, P. aeruginosa strain 2141 culture

supernatant(2-h incubation); F, A. hydrophila

cul-turesupernatant(2-h incubation). Gels run with

en-zymepreparationsalone showed nosignificant levels ofstaining.

extends the

applicability

of this technique for

proteasetyping.

MATERIALS AND METHODS

Bacterial strains. All strainsareV.cholerae0-1,

exceptfor V40 which is V. choleraenon-O-i.Strains

1to 34have beendescribed previously (12, 13).

Strains1 to9wereisolated fromcholerapatientsin

Calcutta, India. Strains 10 to 17were isolated from

cholera patients in Bahrain. Strains 18 to 21 were

isolated from environmental sources in the western

hemisphere. Strain22 wasisolated fromapatientin

Alabama with chronic gall bladder disease (a

non-diarrhealillness).Strain23wasisolatedfromacholera

patient in Texas. Strains 24, 25, and 27to 33 were

isolated frompatientsandsewersduringanoutbreak

of cholera in Louisiana. Strains26and34wereisolated

fromcrustaceansintheareaduringthesameoutbreak.

Crabs,whichyielded strain33,wereeatenbysomeof

the cholera patients. Strain 35 was isolated from. a

cholerapatientin Vietnamby Finkelsteinet al. (4).

Strains36to38wereisolatedfromcholerapatientsin

Dacca, Bangladesh, byI.Huq.Strains39to 44were

isolated from cholerapatientsinDacca,Bangladesh,

byJ. P.Craig.Strain45wasobtainedfromJ. P.Craig. P. aeruginosastrains wereclinical isolatesfrom

the-TexasDepartment of Health. The Aeromonas

hydro-phila strainwasobtained from, E. Matthew,

Brack-Strain 1. CA401 2. CA321 3. CA325 4. CA381 5. CA383 6. CA412 7. CA414 8. CA415 9. CA416 10. 7967 11. 7969 12. 7972 13. 7974 14. 7975 15. 7979 16. 7986 17. 7990 18. 1074-78 19. 2634-78 20. V40 21. V69 22. 1166-77 23. 479935

24. Lou4

25. Lou7 26. Lou18 27. Lou 15

28. Lou22 29. Lou33 30. Lou 17

31. Lou11 32. Lou20 33. Lou 13 34. Lou14

35. 3083

36. RV 47 37. RV53

38. RV 59

39. Q3987 40. N15,503 41. N15,870 42. 17,486 43. 16,117 44. 27,459 45. Suny I

Dimethyl casein protease activity U/mg" U/mlb 0.22 0.81 0.18 0.65 0.32 1.15 0.27 1.09 0.31 1.15 0.21 0.72 0.12 0.42 0.13 0.46 0.13 0.44 0.35 1.15 0.04 0.07 0.25 0.79 0.02 0.07 0.28 0.92 0.22 0.62 0.32 1.04 0.32 1.06 0.16 0.51 0.13 0.40 0.71 1.46 0.01 0.28 0.34 0.97 0.28 0.97 0.29 1.04 0.31 1.09 0.30 1.08 0.25 0.83 0.14 0.48 0.14 0.46 0.24 0.85 0.24 0.83 0.26 0.79 0.32 1.02 0.23 0.77 0.15 0.46 0.12 0.37 0.22 0.72 0.22 0.72 0.12 0.39 0.23 0.67 0.21 0.67 0.12 0.37 0.19 0.60 0.24 0.76 0.04 0.14 Ouchter-lony reac-tion' Reference Notdone + + + + BSA diges-tionpattern Reference Typical Typical Typical Typical Typical Typical Typical Typical Typical Nottypable' Typical Not typable' Nottypablef Typical Typical Typical Typical Typical Typical Not typable' Notdone Typical Typical Typical Typical Typical Typical Typical Typical Typical Typical Typical Typical Typical Typical Typical Typical Typical Typical Typical Typical Typical Typical

Typical'

aMilligrams(dryweight)ofcell.

bMilliliters of supernatant.

'+,Reaction ofidentity withCA401;-,noreaction. dTypical,

Sinilar

to

reference

'Low

degradation

of BSA.

fHighdegradationof BSA.

'Gaveonlymoderatedegradation,butwastypable.

enridgeHospital, Austin,Tex.

Enzymepreparationandassay.Preparation of

culture supernatants for the enzyme assay and the

dimethyl casein assay for proteolytic activity were

done as describedpreviously (12). One enzyme unit

equals 1

pmol

of-NH2 groups generated per min.

Thepseudomonas strains weregrown in Trypticase

soybroth(BBL Microbiology Systems)rather thanin

syncase.

Peptide digest preparation. The protein

on February 7, 2020 by guest

http://jcm.asm.org/

82 SCHNEIDER, SIGEL, AND PARKER

B~~~~~~~~~~~~

ri

E S

FIG. 2. GelscansofBSAdigestiontimecourses.Strains and incubation times:A, CA401, 15min; B, 7990,

15min;C,CA401,30min; D, 7990,30min; E,CA401,1h; F. 7990,1h;G,CA401,2h;H,7990,2h.Migrations

of molecularweight (Mr)standardsaremarked in A:LY, lysozyme, 14,300; STI, soybean trypsininhibitor,

21,000;CA,carbonicanhydrase, 30,000;OA,ovalbumin,43,000;BSA, 68,000.

tionstock consists of4.0mgof bovineserumalbumin

(BSA;Boehringer-Mannheim) per ml dissolved in a

solution of 20% (vol/vol) glycerol (Difco

Laborato-ries)-0.5% SDS (Sigma Chemical Co.)-0.1%

bromo-phenol blue (Sigma)in 0.05 Mborate buffer,pH9.0.

Thisdigestion stockwasheated for 5min at 100°C

and storedat-20°C.The stockwaswarmedto37°C

before the addition of theenzymepreparation.A 1:10

dilution of the bacterial culturesupernatants in0.05

M borate buffer, pH 9.0, was used as the enzyme

preparation.This enzymepreparationwasmixedwith

anequal volumeof thedigestion stock (usually40 or

80,ul)andincubatedat37°C. Incubationtimes varied

asnoted in text, but for routinescreeningexperiments,

30-minand2-hincubationswereused and gave

satis-factoryresults. Afterincubation, the enzymewas

in-activatedby boiling for5min after the additionof

2-mercaptoethanol and SDSto afinalconcentration of

10and1%, respectively. (Forexample,an80-,lsample

[40 tlofenzymeplus40

pl

ofdigestion stock]had10

pl

of 2-mercaptoethanol and 10 ,l ofa 10% SDS

solutionadded beforeboiling.) Enzyme-treated stocks

werethen storedat-20°Cuntiluse.Theywerestable

as judged by SDS-PAGE patterns for at least 3

months.

Trypsin (Sigma) andStreptomyces griseus type VI

protease (Sigma) digestions were done in a similar

digestion stock prepared with 0.05 M

tris(hy-droxymethyl)aminomethane, pH 7.5, with 0.001 M

CaCl2 instead of borate buffer. This bufferwasalso

usedas adiluent for theseenzymes.

SDS-PAGEwascarried out in 15% gelsessentially

asdescribed by Cleveland et al. (1). Electrophoresis wascarried out in a verticalslab gel system by using

platesmeasuring 1.5 mmby 10 cm by 15 cm. BSA (32

,ug perslot,basedoninitialconcentrationbyweight)

was added. Power settings were 75 mA/gelusing a

"constantcurrent" modeonanISCOmodel 490 power

source.Electrophoresiswasperformeduntil the

track-ing dye reached the bottomof the gel. Staining and

destainingwasperformedwithCoomassieblue G-250

by the method of Fairbanksetal. (3). Gel scans were

doneon aJoyce-Loebl model200 densitometer with

a620-nmfilter. Molecular weightstandardswere from

Bio-Rad Laboratories.

Immunological analysis of supernatants.

on February 7, 2020 by guest

http://jcm.asm.org/

A

B

C D

E

F

G

2A,

bracket

1).

Forroutinescreening of strains,

30-minand 2-h incubation periods were used.

,'

',.-:AThe

widespread

presenceofthe

peptide

cleav-age

pattern

of strain

CA401

in V. cholerae iso-" X =;lates

X

is shown in

Fig.

3andTable 1. Most strains

produced

proteolytic activity

as measured

by

the dimethyl casein

assay.

Isolates from both

Eastern and Western

hemispheres

of both

Ogawa and Inaba serotypes show very similar

peptide cleavage

patterns.

Of

the 44 strains tested, 39 gave patterns similar to that of strain

CA401.

Of

the five

remaining strains,

three

(7969,7974,and V69) did not producesufficient

levels

of

degradation

to

give

a

digest pattern,

and one

(7975)

degraded

the BSA too

exten-sively for the peptidedigest pattern to be

deter-mined (timecourseanalysiswas not

performed).

One strain

(1166-77)

was nottested.

Of

the 39 strains

giving

the

typical

digest

pattern,28reacted with

immunological

identity

~~~~

~~~with

strain

CA401

when

antiserum. prepared,

against

purified

strain

CA401

enzymewas

used..

Nine strains did

not react (one strain was not FIG. 3. SDS-PAGE digestionpatterns of various tested). Of these nine, only onestrain, Suny I, V. cholerae strains. Strains, serotypes, and incuba- showed low (<0.15U/ml)levelsof enzyme. tion times: A, CA401, Inaba, 30 min;B, Q3987, Inaba, Ofthe three low degradation strains fromthe 2 h; C, RV53, Inaba, 2h; D, 3083, Ogaba, 2h;E, five in the group not typable by BSA digest 479935, Inaba, 2 h;F,Lou14, Inaba, 2h; G, 1074-78, analysis, two (7974 and V69) were immunologi-Ogawa, 2 h. The photograph of G was reduced

cally

reactive,

and one (7969) was

not.

Of

the slightly more than the other gels during photo- two

stra.is

producing

of'dgradation,

high

levels

processing.

one (V40) reacted immunologically, and one ('hirhteprlnnvLyelimmunodiffusion (10)wasDerformed

(7975)

did not.

by using antiserumprepared againstthepurified

ma-jor higher-molecular-weight form of the protease

found in CA401 (EFIIA). This antiserum was

de-scribed inearlier work(Schneider,Ph.D.thesis).

RESULTS

Figure

1illustratesthe

specificity

of the

cata-lytic

typing technique.

Trypsin, S.

grieseus,

V.

cholerae,P.

aeruginosa,

and A.

hydrophila

pro-teasesall showed different

peptide banding

pat-ternsinSDS-PAGE. Some similarityin

banding

patterns isevidentin thelast three

organisms.

The importance of incubation time in this

analysis

isshownin the

gel

scansin

Fig.

2.Some

V.cholerae strains

produced

such

high

levels of

degradation

that

only

asmearinthe

10,000-

to

20,000-dalton

rangewasevidentat2h. No

good

comparison between strains was

possible. By

following the

degradation

with

time, however,

the

complete

identity

of the

degradation

pat-ternsoftwostrains could beseen. CA401at 15

min (Fig.2A) and7990 at 30 min

(Fig.

2D)

are

practically

superimposable.

Particularly

striking

in V.choleraesupernatant

digests

werethe four

peaksinthe29,000-to

35,000-dalton

range

(Fig.

DISCUSSION

Thisstudydescribesa simplemethod for de-termining the relative specificity of microbial proteases based upon the unique peptides

gen-erated bycleavage ofagiven protein. It issimple

toperformand isreadily reproducible.We have

successfully

used thistechnique with other pro-teins (rabbit aldolaseand human immunoglob-ulinA) andalsoundernondenaturing conditions (absence of SDS in the digestion stock) (unpub-lished results).

The apparentwidespreadpresencein V.

chol-eraeisolates ofproteasesimilartothatfoundin

strainCA401isinteresting. Previous workfrom

thislaboratoryhas suggestedthat the protease

in strain CA401 is required for virulence and

may be involved in either maintenance or growth of the organisminthegut. Thefrequent occurenceof theenzyme in both environmental

andclinicalisolates may suggest that itplays a

role in maintenance of V. cholerae in other environmentsaswellasin thegut.

Thefailure ofsomestrainstoreact

immuno-logically

despite

positive

peptide

digest

patterns

identical to that of CA401 may be

ex-13,

on February 7, 2020 by guest

http://jcm.asm.org/

84 SCHNEIDER, SIGEL, AND PARKER

plained in several ways. It is known that the

enzyme is very unstable in culture supernatants (Schneider and Parker, unpublished data) and may lose virtually all activity in a matter of hours at room temperature. The prolonged in-cubationrequiredfor thegeldiffusiontechnique may mean that in these strains the proteasemay

loseantigenicity (perhaps bydegradation) along

with activity.

Alternatively,

the

sensitivitiy

of

the geldiffusion technique is not

high.

Ifsuch strainsproduceprotease ofsignificantly higher

specific activity than that of CA401 protease,

theymay possessequivalent enzymatic activity butinsufficientantigenicityto reactin the

Ouch-terlonytest.It is alsopossiblethatsomestrains

of V. cholerae may possess an enzyme that is

catalyticallysimilarbutserologically dissimilar.

The few strainsthat reactedimmunologically but had lowdimethylcaseinproteaseand diges-tion activityareparticularly interesting. These may bestrains whichproduceinactiveprecursor enzyme (zymogen).Alternatively, theymay

pro-duce enzymatically inactive, butstili antigenic,

degraded enzyme.

ACKNOWLEDGMENTS

Wethank Susan Crossland for thetypingof themanuscript. This workwassupported byPublic Health Service grant AI12819fromthe National Institutes of HealthtoC.P.

LITERATURE CITED

1. Cleveland, D.W., S. G. Fischer, M.W. Kirschner, andV.K.Laemmli.1977.Peptide mapping bylimited proteolysis in sodiumdodecyl sulfate and analysis by gelelectrophoresis. J. Biol. Chem.252:1102-1106. 2. Dahle, H. K., and O. Sandvik.1971.Comparative

elec-trophoretic andserologicalanalysis of Vibrio comma

and Aeromonasliquefaciens proteinases.ActaPathol.

Microbiol. Scand. Sect. B 79:686-690.

3. Fairbanks, G., T. L. Steek, and D. F. H. Wallach. 1971. Electrophoretic analysis of the major polypeptides of thehuman erythrocyte membrane. Biochemistry 10: 2606-2617.

4. Finkelstein, R. A., P. Z. Sobcinski, P. Atthasam-punna, andP.Charunmethee.1966.Pathogenesisof experimentalcholera: identificationofcholeragen (pro-choleragen A) by disc immunoelectrophoresis and its differentiation from cholera mucinase. J. Immunol. 97: 25-33.

5. Grimont,P.A. D.,F.Grimont, and H. L. C. Dulong deRosnay.1977.CharacterizationofSerratia

marce-sens,S.liquefaciens,S. plymuthiea, S. marinorubra by electrophoresis of their proteinases. J. Gen. Microbiol. 99:301-310.

6. Kawaharajo, K., C. Abe, J. Y. Homma, M. Kawano, E.Gotoh, N. Tanaka, and K. Morihara. 1979. Cor-neal ulcerscausedbyprotease andelastase from Pseu-domonasaeruginosa.Jpn. J. Exp. Med. 44:435-442. 7. Kawaharajo, K., Y.Homma, Y. Yuzuru, Y. Aoyama,

L.Okada, and K. Morihara. 1975. Effects of protease and elastase fromPseudomonas aeruginosa on skin. Jpn. J.Exp. Med. 45:79-88.

8. Kreger, A. S., and L. D. Gray. 1978. Purification of Pseudomonas aeruginosa protease and microscopic characterization of pseudomonal protease-induced rab-bitcornealdamage. Infect. Immun. 19:630-648.

9. Lyerly, D. and A. S. Kreger. 1979. Purification and characterization of a Serratia marcesens

metallopro-tease.Infect. Immun. 24:411-421.

10. Nowotny, A. 1969. Basic exercises in immunochemistry.

Alaboratory manual, p. 151-152. Springer-Verlag New York, Inc., New York.

11. Sandvik, O., and H. K.Dahle. 1971. Enzymoserological relationships between Vibrio comma and other gram-negativeorganisms. Acta Pathol Microbiol. Scand. Sect.

B79:285-290.

12. Schneider, D. R., and C. D. Parker. 1978. Isolation and characterization ofprotease-deficient mutants of Vibrio cholerae. J. Infect. Dis. 138:143-151.

13. Sigel, S. P., S. Lanier, V. S. Baselski, and C. D. Parker. 1980. Invivoevaluationofpathogenicity of clinical andenvironmental isolates of Vibrio cholerae. Infect.Immun.28:681-687.

J. CLIN. MICROBIOL.

on February 7, 2020 by guest

http://jcm.asm.org/