Cloning and characterization of a second complementary DNA for human tryptase

Cloning and characterization of a second

complementary DNA for human tryptase.

J S Miller, … , G Moxley, L B Schwartz

J Clin Invest.

1990;

86(3)

:864-870.

https://doi.org/10.1172/JCI114786

.

A second cDNA for human tryptase, called beta-tryptase, was cloned from a mast cell cDNA

library in lambda ZAP. Its nucleotide sequence and corresponding amino acid sequence

were determined and compared with those of a previously cloned tryptase cDNA, now

called alpha-tryptase. The 1,142-base sequence of beta-tryptase encodes a 30-amino acid

leader sequence of 3,089 D and a 245-amino acid catalytic region of 27,458 D. The amino

acid sequence of beta-tryptase is 90% identical with that of alpha-tryptase, the first 20 amino

acids of the catalytic portions being 100% identical. This identity, together with recognition

of each recombinant protein by monoclonal antibodies directed against purified tryptase

validate the tryptase identity of both alpha-tryptase and beta-tryptase cDNA molecules.

Modest differences between the nucleic acid sequences of alpha- and beta-tryptase

occurred throughout the cDNA molecules except in the 3' noncoding regions, which were

identical. Although most highly conserved regions of amino acid sequence in the trypsin

superfamily are conserved in both tryptase molecules, beta-tryptase has one carbohydrate

binding site compared to two in alpha-tryptase, and one additional amino acid in the

catalytic sequence. Regions of the substrate binding pocket in beta-tryptase (DSCQ,

residues 218-221; SWG, residues 243-245) differ slightly from those in alpha-tryptase

(DSCK, residues 217-220; SWD, residues 242-244). The presence of both alpha- and

beta-tryptase sequences in each haploid genome was […]

Research Article

Cloning and Characterization

DNA

for

Human

Tryptase

of

a

Second Complementary

Jeffrey S. Miller, George Moxley, and Lawrence B. Schwartz

Department ofMedicine, Medical College of Virginia, Virginia Commonwealth University, Richmond, Virginia23298

Abstract

A second cDNA for human tryptase, called

fl-tryptase,

was

cloned from a mast cell cDNA library in lambda ZAP. Its nucleotide sequence and corresponding amino acid sequence weredetermined and compared with those ofa previously

clonedtryptasecDNA,nowcalleda-tryptase.The1,142-base

sequence of ,B-tryptase encodes a 30-amino acid leader

se-quenceof3,089 D anda 245-amino acidcatalytic region of

27,458 D. The amino acidsequenceof t-tryptaseis 90%

iden-tical with that of a-tryptase, the first 20 aminoacids of the

catalytic portions being100%identical. Thisidentity, together with recognition of each recombinant protein by monoclonal antibodiesdirectedagainst purifiedtryptasevalidatethe tryp-tase identity ofboth a-tryptase and ,B-tryptase cDNA mole-cules. Modest differences between the nucleicacidsequences

ofa-and

,8-tryptase

occurred throughoutthe cDNA molecules except in the 3' noncodingregions, whichwereidentical.

Al-thoughmosthighlyconservedregionsofaminoacidsequence

inthe trypsin superfamilyareconservedinboth tryptase

mole-cules,

fl-tryptase

hasonecarbohydrate bindingsitecompared

to two ina-tryptase, andoneadditionalamino acid in the

cata-lytic sequence. Regionsofthe substrate bindingpocketin

,-tryptase(DSCQ, residues 218-221;SWG,residues243-245)

differ slightly from those in a-tryptase (DSCK, residues

217-220; SWD, residues 242-244). Thepresenceofboth a-and,-tryptase sequences ineach haploid genome was indi-cated by finding a- and j-tryptase specific fragments after

amplification byPCRofgenomicDNA in10unrelated individ-uals. Localization of both a- and 8-tryptase sequences to

human chromosome 16 was then performed by analysis of DNA preparations from 25 human/hamster somatic hybrids byPCR. It is now possibleto assessthe expressionofeach tryptase cDNAbymast cells andtherelationshipofeachgene

product totheactiveenzyme.(J. Clin. Invest. 1990.

86:864-870.) Keywords: mastcell*protease*polymerasechain reac-tion*chromosome localization-trypsin superfamily

Introduction

Human tryptase isatetramericserine protease that is

concen-trated and stored selectively in the secretory granules ofall

types of human mastcells, from where it is secretedduring

AddressreprintrequeststoDr.Jeffrey S.Miller, Medical Collegeof

Virginia,Box263, Richmond,VA23298.

Receivedforpublication12March1990andinrevisedformIMay

1990.

mastcell degranulation (1-4). When detected in the

circula-tionorin otherbiologic fluids,the tryptase level serves as an

indication ofmastcellactivation(5, 6). Subunitsofpurified

tryptase from lung (1, 7), skin (8), and pituitary (9) exhibit

diffuse bandingpatternsin the31,000-35,000-Drange,often with two predominantbands - 2,000-D equivalents apart

fromoneanother.Theactive form ofhuman tryptase has four

subunits,

each

having

certain

epitopes

incommon(10), iden-tical

NH2-terminal

20-amino acidsequences (9, 11), and one

active enzymatic site (1). On this basis,thesize heterogeneity ofthesubunitswaspostulatedtobe due toposttranslational

eventsratherthan todifferentgeneproducts, but the presence

ofmorethanonegenefortryptase was not excluded.

ComplementaryDNAmoleculesfortryptase were cloned andsequenced from human (1 1) and dog (12)mast cell cDNA

libraries.Thecorresponding amino acidsequences show 71%

identity and include a 30-amino acid leader and 245- and

244-amino acid catalytic portions for thedog and human se-quences,

respectively.

Inthestudy of dogtryptase, cDNAfora

related trypsin-like enzyme, called dog protease, also was

cloned andsequenced and foundtohave 53%identity withthe

dogtryptaseamino acidsequence. Thecurrent work reports

the finding and characterization ofa secondcomplementary DNA forhuman tryptase, called

fl-tryptase,

whichis highly homologous tothe previously described cDNA sequence for human tryptase, now called a-tryptase. Furthermore, se-quencesfor botha-and

_(l-tryptase

are

typically

presentinthe

haploidgenomeof normal subjects and havebeenlocalizedto chromosome 16.

Methods

Materials. The cDNA human mastcell library wasconstructed in lambda ZAP withpurifiedpolyadenylatedRNAobtained frommast

cells purifiedfrom human lungaspreviously described (11). DNA

restrictionendonucleaseswereobtained from Bethesda Research Lab-oratories(Gaithersburg, MD).[32P]gammadATP used for 5' end label-ing ofsyntheticoligonucleotidesand[35S]alphadATP used for dideoxy nucleotide sequencingwere obtained from New England Nuclear (Boston, MA). Sequenase polymerasewaspurchased fromU.S. Bio-chemicalCorp.(Cleveland, OH)and Taq polymerase with the

poly-merasechain reaction

(PCR)'

reagents werepurchased from Perkin-ElmerCetusCorp. (Norwalk, CT). PCRexperimentswereperformed usingathermal cycler (Perkin-ElmerCetus Corp.)providedby the Medical College ofVirginia/Virginia Commonwealth University

(MCV/VCU)NucleicAcid CoreLaboratory.Nitroplus2000 nitrocel-lulose filterswereobtained from MicronSeparations,Inc.(Westboro,

MA). NuSieve GTG and SeaKem GTGagarosewerepurchasedfrom FMCBioproducts (Rockland, ME), and PCRableDNAfrom human/

hamsterhybrid cell lines waspurchased from Bios Corp. (New

Ha-ven,CT).

Oligonucleotides usedasprimersforDNAsequencing,PCR

ex-periments,andashybridization probesweresynthesizedonDNA

syn-1.Abbreviations usedinthispaper:PCR,polymerasechain reaction.

J.Clin.Invest.

©The AmericanSociety for Clinical Investigation,Inc.

0021-9738/90/09/0864/07 $2.00

thesizer (Model 380A; Applied Biosystems, Inc., Foster City, CA) and purified by reverse phase HPLC on a Dynamax-300-A

_12-,um

(10 mm i.d. X 25 cm 1.) C18 reverse phase column (Rainin Instrument Co., Woburn, MA). Two Rabbit HP high-pressure pumps were used to generate anelution gradient from 7.25 to 50% acetonitrile in 0.1 M triethylammonium acetate. Oligonucleotide synthesis and HPLC puri-fication wereperformed by the MCV/VCU Nucleic Acids Core Labo-ratory. Alternatively, deblocked oligonucleotides for PCR experiments were obtained from Oligos, Etc. (Ridgefield, CT) and used without furtherpurification.

Screeningand sequence analysis oftryptase cDNA molecules. Indi-vidual tryptase cDNA clones identified previously were detected using animmunoscreening technique with three antitryptase mAbs (1 1). In this study, additional clones were detected by screening with a radiola-beled 19-mer oligonucleotide probe corresponding to the

fl-tryptase

sequence (TGCAACTGCGGGAGCAGCA). Duplicate plaque lifts weremade from BB4 Escherichiacolihostcells infected with an am-plified portion of the lambda ZAP human mast cell cDNA library. Filters were prehybridized with 5X Denhardt's solution (0.1 g Ficoll 400, 0.1 g polyvinylpyrrolidone, 0.1 g BSA [fraction V] in 100 ml H20), 0.5% SDS. The 19-mer oligonucleotide was radiolabeled by the

T4polynucleotidekinasetechniquetoaspecific radioactivityof 108 cpm/ug. Thehybridization mixture contained 6X NET (0.9 M NaCl, 6

mMEDTA, 0.3MTrisHCl, pH 8.0), 5X Denhardt's solution, 0.1% SDS, 250 Mg/ml of heat-denatured yeast RNA, and oligonucleotide (6-8 ng/ml).Anovernight hybridizationat520Cwasfollowed by two washes at370Cand one wash at 52°C with 6X NET/0.5% SDS. Auto-radiographywasthenperformed, and positive plaques were selected.

Plaque-derived phage of interest were purified by selecting single plaques containing material hybridizing to tryptase nucleic acid probes and usedtoinfectE.coli (XL1 Blue) in the presence of R408 helper phage, which caused the autoexcision of the pBluescript plasmid

con-tainingthecDNAinserts of interest. Plasmidpreparations ofDNA

weresubjected to digestion with EcoRItoexcise the cDNA insert and

Eco RV to selectinserts likely to contain the fl-tryptasesequence. a-TryptasecDNAhas a known EcoRVrestriction site at the position

correspondingtonucleotides 368/369 of

fl-tryptase

that isnotpresent infl-tryptasecDNA.EachplasmidDNAcontaininganEco

RV-resis-tantinsertwaspurifiedbyatechniqueinvolvingalkalinelysis andused

directly for double stranded DNAsequencingwith Sequenase (13). PlasmidDNA(10

_,d)

wasdilutedto 24

;LI

with dH20 and denatured by

adding6

,d

of 1MNaOH andheatingat85°Cfor 5 min. Ice-cold5 M

ammoniumacetate(10 ul)and80 Mlof 100% ethanolweresequentially

added and the mixturewasthen chilled at-70°C for 20 min. The pellet obtainedby centrifugationat 12,000rpm in acentrifuge rotor

(Model HB4; Beckman Instruments, Fullerton, CA)waswashed in 70%ethanol three times and dried invacuo.Annealingwasperformed by resuspendingthe pelletin 10

Al

ofsequencingbuffer containing

primer(10ng),heatingthe mixtureto65°C,andthenslowlycoolingit

to roomtemperature.Astandarddideoxy nucleotide-sequencing pro-tocol withSequenasewasthenperformed usingbothdGTP and dITP nucleotidestoresolve anycompression artefactsaccordingtothe in-structions of the manufacturer(United States BiochemicalCorp.).

Each cDNAinsertcorrespondingtotryptasewasdesignated GRA-n,

wherenisanintegeruniquetoeach cloned insert.

PCRfragmentevaluationof normalhumangenomicDNA.

Tem-plateDNAused for PCRexperimentsincluded either a-tryptase or

B3-tryptase

cDNA fromplasmid minipreparations preparedasabove,

human genomic DNA prepared from peripheral leukocytes as

de-scribed before(14)orbuffer alone. GenomicDNA(500-1,000 ng)and

a 107-fold dilution ofplasmidDNAper reaction mixture with 500 ng of eachprimerwerefoundtobeoptimal. Denaturation occurredat

95°Cfor2min,annealingat62°Cfor2min, and extensionat720C for 3 min,typicallyover30cycles.Theoptimal MgCl2concentrationwas 1.5 mM. Otherconditions and commercial reagents were used

ac-cordingtothePerkin-Elmer Cetus instructions. PCR reactionproducts

were concentrated - 10-fold onmicroconcentrators (Centricon 10;

AmiconCorp. [Danvers, MA]),diluted backtotheoriginalvolume

withdH20, and then reconcentrated 10-fold. REACT 2 buffer(iOX;

Bethesda Research Laboratories) was added to the concentrated mate-rial (1:10). A mixture consisting of 20 Ml of this buffered concentrate and 20 U of HaeIIIwas then incubated for 2 h at 370C. Agarose gel electrophoresis of digested material was performed with 5% NuSieve GTG/0.5% SeaKem GTG agarose with Tris borate buffer at 100 V for 1-1.5 h. Southern transfers to Zeta Probe-cationized nylon mem-branes (Bio-Rad Laboratories, Richmond, CA) were performed with 0.4 N NaOH. Hybridizations with 13-mer radiolabeled oligonucleo-tides were performed at420Cusing the buffer described above except that Denhardt's solution was omitted because of interference with hybridization of small oligonucleotides. DNA sequence data were stored and analyzed using the Genetics Computer Group Sequence Analysis Software Package from the University of Wisconsin (version 6.1; August, 1989) (15).

Chromosome localization. Purified template DNA preparations for PCRs were obtained from 25 human/hamster somatic cell hybrid lines (100ng/reaction), from normal human cells (50 ng/reaction) and from normal hamster cells (50 ng/reaction) as described by Bios Corp. Buffer controls without template DNA were performed in parallel to assesscontamination. Primers were used at 0.2MAM,MgCl2 at 1.5 mM, and other reagents as above ina I00-Mreaction volume under mineral oil. Denaturation was performed at950Cfor 1 min, annealing at620C

forImin, and extensionat720Cfor 2 min over 30 cycles. Agarose gel electrophoresiswasperformed on a medium-size apparatus with 5% NuSieve GTG/0.5% SeaKem GTG agarose with Tris borate buffer at 60 V for 5 h. Amplified material from human chromosome 16 was thensubjected to HaeIIIendonuclease digestion, agarose gel electro-phoresis and Southern blot analysis and compared with results ob-tained with human genomicDNAabovefor the presence of a-tryptase and,B-tryptase sequences in the humangenome.

Results

Isolation from a human lung mast cell library of a nucleotide sequence

encoding

_fl-tryptase.

Of the

100,000

plaques pre-viously analyzed (produced by phage obtained from a mast

cell-derivedcDNAlambda ZAP library)10plaques expressed

protein recognized by each of three different murine MAbs

directedagainst humantryptase (11). Seven ofthesecontained inserts of various sizes, with nucleotidesequences identical to

the

previously

reportedtryptase cDNA sequence, nowcalled

a-tryptase cDNA.Three otherclonescontained insertsthat are

highly similartoa-tryptase cDNA, butwithaslightly different nucleotide sequence, even though each coded for a protein withtryptaseepitopes.Theinserts in each ofthese three clones were identical over the regions sequenced and extended to

different degrees toward the 5' and 3' termini. Only one

f3-tryptase cDNAclone, Gra- 1, containedtheentire3'noncoding regionand none reachedfurtherthannucleotide residue233 toward the 5'terminus.Toobtainthe 5'end ofthe sequence,a

19-meroligonucleotide complementaryto aregiontoward the 5'end of GRA-1

(nucleotide

residues 268-286) and

having

three nucleotide substitutionsaswell as threeadditional nu-cleotides compared with the

corresponding region

of a-tryp-tase was

synthesized

and used to screen

-100,000

new

plaques produced with the lambda ZAP

library.

10

plaques

appearedtobe

positive

and

plasmids

were

prepared

from these recombinant phageasdescribedin Methods. The cDNA insert

from each

corresponding

plasmid preparation

was

subjected

to

digestion

withEcoRV, for whicha

single cleavage

siteexistsin a-tryptase, butnotin the

corresponding region

of

j-tryptase.

These digests were

analyzed

by agarose

gel

electrophoresis.

Seven ofthe insertswere not

digested by

Eco RV andwere

cDNAinserts were not further analyzed. Three of the seven Eco RV-resistant inserts contained the apparent methionine initiation codon for

_fl-tryptase,

but none appeared to contain the entire 3'noncoding region.Complete sequence data were obtained with GRA- 15 from the new cDNA isolates and GRA-1 from the previous cDNA isolates according to the strategy shown inFig. 1.Thenucleotide sequence

correspond-ingtotheentirecodingand 3'noncoding regions and a portion of the 5'noncodingregion wasobtainedand is called 3-tryp-tase cDNA. Fig. 2 shows the entire

f3-tryptase

cDNA and

amino acid sequences and also the nucleotide residues of a-tryptase cDNA that differ. Limited sequence analyses per-formed on theeightother

f3-tryptase

isolates were identical to the sequence illustrated.

The CCAGG sequence preceding the presumed initiator Metisidentical inaand f3 tryptase and corresponds to optimal

translation initiation sequencesdescribed previously(16, 17). The 1,142-base cDNA sequence is predicted to encode a

30-amino acidleader sequenceof3,089 D and a 245-amino acid

catalytic portionof27,458D.The20-aminoacid NH2-termi-nal sequenceofthecatalyticportion is identical to the amino acid sequences previously determined from purified tryptase and that predicted from a-tryptase cDNA. Thus, the shared

amino acidsequenceandtherecognitionbymurine antitryp-taseMAbsreportedpreviously (11)suggesttheidentity of this

fl-tryptase

cDNA as anauthentic tryptase cDNA.

The complete amino acid sequence of ,B-tryptase is 90%

identical with that of a-tryptase. Highly conserved regions in the trypsin superfamily and in a-tryptase include theIVGG

activation site region (residues 31-34), the WVLTAAHC

ac-tive site histidine region (residues 68-75), theDIALL

active-site aspartic acid region (residues 121-125),the M-CAG on the

NH2-terminal side oftheactivesite serine (residues209-213) and theGDSGGPactive site serine region (residues 222-227).

These regions also are conserved in

p3-tryptase.

The aspartic

acid at position 218 in

_fl-tryptase,

like that in a-tryptase,

confers the acknowledged trypsin-like substrate specificity of human mast cell tryptase (18). Putative cystine linkages in j3-tryptase between Cys residues atpositions 59 and 75, 155 and230, 188 and 211, and 220 and 248 also correspondtothe

analogousresiduesin a-tryptase.

Although a-tryptase cDNA encodes two potential N-linked carbohydrate binding regions and a274-amino acid protein,

fl-tryptase

cDNA encodes only one potential N-linked

carbo-hydrate binding region (NGT. at positions 233-235) and a 275-amino acid protein. The additional amino acid in

fl-tryp-tase occurs between the second and third conserved regions (see below) of the serine protease family (QLREQ, residues 89-93 in ,B-tryptase compared with NSGT, residues 89-92, in a-tryptase). There are also differences in the predicted

sub-strate binding pockets ofa- and

f,-tryptase.

The location of these residues is based on extensive homology with other ser-ine proteases, but may need to be revised after x-ray crystallo-graphic data on tryptase are obtained. On one side, DSCQ (residues 218-221) occurs in

_fl-tryptase,

DSCK (residues 217-220) in a-tryptase. Another side of the putative binding pocket contains SWG (residues 243-245) in

_fl-tryptase,

a typi-cal sequence for serine proteases,asopposed to SWD (residues 242-244) in a-tryptase (11, 19). The Gln in

_fl-tryptase

is analo-gous totrypsin (20), the Lys in a-tryptasetoplasma kallikrein (21), and Factor XI (22), each locatedneartheentrance tothe substrate binding pocket.

The predictedamino acid sequences for thecatalytic por-tions of a- and (3-tryptase molecules, respectively, permitted calculation of net charge values of -4 and -3 and pI values of 6.65 and 6.92 (15). The amino acidcompositionof,B-tryptase is notsubstantially different from that of a-tryptase (11), both being compatible with previously determined amino acid

compositions(1, 7).

Evaluation of thepresenceofnucleotide sequences coding fora-and f3-tryptaseingenomicDNA ofnormalsubjects. To determine whether the putative genes for a-tryptase and

#-GRA- 1

GRA-15

loom No.M Dram ApaU But= goo571 NuMiApaI Dam

N N N * K * * 0 0

600 750

o, o

100i 1143

NucleoUdes(bp)

Figure1.Sequencingstrategyandrestrictionmapof,B-tryptase cDNA.Hexanucleotide restrictionenzymeswithone_recognitionsite in

f3-tryp-tasecDNAareshownattheirsites ofcleavage.Inaddition,the tetranucleotide restrictionenzyme Hae IIIisshownatthecleavagesite of

inter-estintheanalysis of amplifiedPCRfragments(seeFig. 3).Restriction sitespresent in ,B-tryptase andnota-tryptaseareshown withan_asterisk,

MetLeuAsnLeuLeuLeuLeuAlaLeuProValLeuAlaSerArgAlaTyrAlaA

CCAGGATGCTGAATCTGCTGCTGCTGGCGCTGCCCGTCCTGGCGACCGCCCCTACGCCG GC

laProAlaProGlyGlnAlaLeuGlnArjValGlyIleValGlyGlyGlnGluAlaProA

CCCCTGCCCCAGGCCAGGCCCTGCAGCGAGTGGGCATCGTTGGGGGTCAGGAGGCCCCCA

T A C T C

rgSerLysTrpProTrpGlnValSerLeuArgValHisGlyProTyrTrpMetHisPheC GGAGCAAGTGGCCCTGGCAGGTGAGCCTGAGAGTCCACGGCCCATACTGGATGCACTTCT

G A G

ysGlyGlySerLeuIleHisProGlnTrpValLeuThrAlaAlaHisCysValGlyProA GCGGGGGCTCCCTCATCCACCCCCAGTGGGTGCTGACCGCAGCGCACTGCGTGGGACCGG

G C

spValLysAspLeuAlaAlaLeuArgValGlnLeuArgGluGlnHisLeuTyrTyrGlnA ACGTCAAGGATCTGGCCGCCCTCAGGGTGCAACTGCGGGAGCAGCACCTCTACTACCAGG

A C ..CA

spGlnLeuLeuProValSerArgIleIleValHisProGlnPheTyrThrAlaGlnIleG ACCAGCTGCTGCCGGTCAGCAGGATCATCGTGCACCCACAGTTCTACACCGCCCAGATCG

A G T AT CT

lyAlaAspIleAlaLeuLeuGluLeuGluGluProValLysValSerSerHisValHisT GAGCGGACATCGCCCTGCTGGAGCTGGAGGAGCCGGTGAAGGTCTCCAGCCACGTCCACA

T C CA G

hrValThrLeuProProAlaSerGluThrPheProProGlyMetProCysTrpValThrG

CGGTCACCCTGCCCCCTGCCTCAGAGACCTTCCCCCCGGGGATGCCGTGCTGGGTCACTG

TG G

lyTrpGlyAspValAspAsnAspGluArgLeuProProProPheProLeuLysGlnValL

GCTGGGGCGATGTGGACAATGATGAGCGCCTCCCACCGCCATTTCCTCTGAAGCAGGTGA

C C

t~o

1..

120

240

300

360

4'20

4..9

S...Il

ysValProIleMetGluAsnHisIleCysAspAlaLysTyrHisLeuGlyAlaTyrThrG

:10.9.

AGGTCCCCATAATGGAAAACCACATTTGTGACGCAAAATACCACCTTGGCGCCTACACGG 600

lyAspAdpValArgIleValArgAspAspMetLeuCysAlaGlyAsnThrArgArgAspS

ZIA

GAGACGACGTCCGCATCGTCCGTGACGACATGCTGTGTGCCGGGAACACCCGGAGGGACT 660

A G A

erCysGlnGlyAspSerGlyGlyProLeuValCysLysValAsnGlyThrTrpLeuGlnA CATGCCAGGGCGACTCCGGAGGGCCCCTGGTGTGCAAGGTGAATGGCACCTGGCTGCAGG

C A T A

laGlyValValSerTrpGlyGluGlyCysAlaGlnProAsnArgProGlyIleTyrThrA

CGGGCGTGGTCAGCTGGGGCGAGGGCTGTGCCCAGCCCAACCGGCCTGGCATCTACACCC A

rgValThrTyrTyrLeuAspTrpIleHisHisTyrValProLysLysPro*

GTGTCACCTACTACTTGGACTGGATCCACCACTATGTCCCCAAAAAGCCGTGAGTCAGGC CTGGGTGTGCCACCTGGGTCACTGGAGGACCAACCCCTGCTGTCCAAAACACCACTGCTT CCTACCCAGGTGGCGACTGCCCCCCACACCTTCCCTGCCCCGTCCTGAGTGCCCCTTCCT GTCCTAAGCCCCCTGCTCTCTTCTGAGCCCCTTCCCCTGTCCTGAGGACCCTTCCCCATC

CTGAGCCCCCTTCCCTGTCCTAAGCCTGACGCCTGCACTGCTCCGGCCCTCCCCT( GGCAGCTGGTGGTGGGCGCTAATCCTCCTGAGTGCTGGACCTCATTAAAGTGCAT(

TCAn

tryptaseareallelic or

duplicated

ineach

haploid

genome,

ge-nomicDNAwasobtained from 1Osubjectsandanalyzed.An 86-bpnucleotidefragmentfrom either a-tryptaseor

_fl-tryptase

templateswasamplified byPCRusing20-mer(,B-tryptase

nu-cleotides 135-154) and 19-mer

(f,-tryptase

nucleotides

202-220)primers.Theprimerswerecomplementaryto

iden-tical regionsin both a-and f3-tryptasecDNAmolecules, but

the intervening region contained restriction site differences. The

j-tryptase

product containedaHae III

site;

the a-tryptase productdidnothavethis restriction site.

Fig.

3Ashows that the PCRproducedanucleotide

fragment

of 86 apparent

base-pairs with each ofthe three

templates,

a-tryptase

cDNA,

i-tryptase cDNA and

genomic

DNA. Treatmentof the

ampli-720

..4.4...

900

960 1020

Figure 2. Nucleotide and amino acid sequence of fi-tryptase cDNA. Theaminoacid sequence ofB-tryp-taseisshownfirstwith the number of the last residue shadowed. Theentire nucleotidesequence of the

compositefl-tryptasecDNAisshown on the next line

withthe unshadowednumber representing the last

nucleotideonthatline.Nucleotide sequence differ-ences ina-tryptase cDNA are shown below the ,B-tryptasenucleotidesequence. The threenucleotides

notpresent in a-tryptase areindicatedwith periods.

Catalytic triad residues (***),leader sequence(-), putativesubstratebindingpocket regions(- ),and theputative asparagine-linked carbohydrate binding site ( )areshownfortheaminoacidsequence as

indicated,and theregion fromwhich the

oligonucle-otidesequence usedforselectingLambdaZAP

carry-ing5' ,B-tryptasecDNAinserts is underlined. These

sequencedataareavailable from EMBL/GenBank/

DDBJ underaccessionnumbersM37488and

M30038.

fiedDNAfragments withHaeIII(GG'CC)didnotchangethe

electrophoretic

mobility ofthe 86-bpa-tryptase

fragment.

In

contrast, complete cleavageof the 86-bp f,-tryptase fragment

into20- and66-bpportionsoccurred(Fig. 3B).Treatment of the86-bpfragments generatedfrom thegenomicDNA of each

subjectyielded

fragments

of86, 66,and 20bp,consistentwith

the presenceofbotha-and

fl-tryptase genetic

material in the PCR products from each person. Identical results were ob-tained with DNA from anadditional five unrelated

subjects

(notshown).

Thepossibilitythat the observed pattern of

fragments (86,

66, and 20 bp)was due to

partial cleavage

of the

,3-tryptase

PCRproduct from

genomic

DNA wasconsidered. Two

bp S a

_p

1 2 3 4 5

.i.

C

S.

...~

E

Ea

F

i

2

3 4

s

...s

..:

F

Figure 3. Southern blot analysis for thepresenceofsequencesspecific fora-andf3-tryptaseinthehumangenome.PCR reactionswere per-formed witha-tryptasecDNA(lane a),f3-tryptasecDNA(lanefl)andgenomic DNA obtained from five unrelated individuals(lanes 1-5)as templates.Productsobtainedfromthe PCRreactionsweresubjectedtoelectrophoresis inagarosegelseither directly (A)oraftertreatmentwith HaeIII(B), stained with ethidium bromide and photographed under UV light. Material in each of these gelswasblotted and bakedonto nitro-cellulosepaper.The a-tryptasespecific 13-merprobewaslabeledand hybridizedtoHae Ill-untreated (C) and -treated (D) material;the blot

wasthen subjectedtoautoradiography. Radiolabeled hybridizeda-tryptaseprobewasremovedbyboiling;thefilterwasthensubjectedto

auto-radiographytoinsurethatradioactivity had been removed. Rehybridization with the13-merfl-tryptaseprobewasperformed withHae III-un-treated(E) and -treated (F) material. Molecular weight standardsareshown inthe S lane. Arrowsshow fragment positions withapparent mo-lecularweights of 86, 60, and 26 bp.

meroligonucleotide probes corresponding toresidues 156 to

168 in a-and 13-tryptase cDNA,butdistinguished from each

otherby three nucleotide mismatches, weresynthesized and labeled with 32P. These probes werethen used under condi-tionsofhigh stringencyto identify a-andf3-tryptase-specific

sequences.The a-tryptaseproberecognizedthe86-bp

amplifi-cationproductfroma,butnotfrom tryptase cDNA(Fig. 3 C); the f3-tryptase probe recognized onlythe86-bp fragment

derived from 3-tryptasecDNA(Fig. 3E).Bothprobes

recog-nizedthe86-bp fragmentfromgenomicDNAbefore Hae III

treatment(Figs. 3,C andE).After Hae IIItreatment(Figs. 3,

DandF), apositive signal appeared only with thea-tryptase

probe,whichhybridizedtothe 86-bp fragment. HaeIIIcutthe

fl-tryptase

PCR fragment within the sequence corresponding

tothe 13-merprobe, leaving complementaryendsof 4 and 9

bp. Hybridization of these fragmentswiththe 1 3-mer f3-tryp-tase probe does not occurunder the conditions used. Thus, PCRproducts fromboth a-and

fl-tryptase

DNAwerepresent

in the DNA of 10 of 10 persons. The putative genes for a-tryptaseandf,-tryptase, therefore,arenotallelic, butareeach presentinthenormalhaploidgenome.Although quantitation

ofgenomic sequences based on analysis of PCRproducts is somewhat uncertain, both sequences appear to be present

there inequalamountsbasedonthe relativequantities visual-izedoftheamplified 86-bp fragments.

Chromosome localizationofthenucleotidesequencesfora

and,B tryptase. DNApreparations fromthe 25

human/ham-stersomatichybridsinadditiontocontrol human and hamster

DNA wereused as PCRtemplatesasdescribed in Methods.

Only the humanDNAcontroland two of the somatic

hybrid-derived DNApreparations yieldedareactionproductafterthe initial 30-cycle amplification reaction, all comigratingat an apparent size of 86bp. Onehybridhad beencharacterizedas having chromosomes5, 8,and16,theotherhad chromosomes

3, 5, 10, 15,and 16.Onlychromosome 16wasuniquetothis

pair ofheterohybrids and not found in others, indicating

that theputative gene(s) fortryptaseare on human chromo-some 16.

Toassesswhether the nucleotidesequencesfor bothaand

(3tryptasewerepresenton human chromosome 16, the PCR

product(s) from each of the two positive hybrid lineswas

treated withHae III and subjectedasabovetoSouthern blot

analysis under stringent conditions with the 13-mer probes specificfora-and(3-tryptasecDNA.Bothprobes hybridizedto the 86-bp DNA fragments amplified from each of the two

hybrid lines. Hae III only cleaved about half of the 86-bp 118

'8614

72

bp

118

'86'4 72

'60' 4

-:.X.

Af

A.. I .._!.::::.: .:.': _-3

fragment, indicating that the putative genes for both a-tryptase and

f3-tryptase

arelocated on human chromosome 16.

Discussion

Twodistinct cDNA molecules for human tryptase have now been cloned from a mast cell cDNA library; the one reported previously, now called a-tryptase cDNA, and the one reported currently called

_fl-tryptase

cDNA. Verification of each as rep-resenting tryptase was based upon complete identity between theNH2-terminal20-amino acid sequence obtained with pu-rified authentic tryptase and those obtained from the

corre-sponding region ofthe cloned cDNA molecules. Recognition

oftheexpressedrecombinant protein derived from a- as well

as

f3-tryptase

cDNAby three different monoclonal antibodies

directed against purified tryptase (11), further validates the tryptaseidentities ofthe cloned cDNAs.

Comparison ofthecorresponding nucleotide sequences for

a-and

#-tryptase

cDNA(Fig. 2) showed 96% identity.

Differ-ences occurred throughout the coding region. The 3' non-translated region of the single f-tryptase cDNA cloned with

thisregion intactand thefivenucleotides 5'to the start methio-nine codon forf3-tryptasewere identical to the 3' nontranslated region offive different a-tryptase cDNA molecules and the 5' nontranslated region ofthe single a-tryptase cDNA cloned withthisregion intact. Conservation of the noncoding regions

is unusual among different serine proteases. Perhaps these identical 3' noncoding regions reflect a conserved regulatory

functioninvolved with theexpressionof each tryptase geneor

thestability ofthecorrespondingmRNA molecules. Another

possibility isthat only onecopy of thisregion resides in the genome, sharedbythetwocoding regions, perhaps byan

al-ternative splicing mechanism. Whether the genes for a- and ,-tryptaseareallelicorboth present in thehaploidgenomecan

be addressed with the present data. The nucleotide differences observedthroughout thecoding regionsof the a-and 3-tryp-tase cDNAs and the observation that no

homozygosity

for eithera-or

fl-tryptase

PCRproductswasobserved in 10 of 10 unrelated individualsindicate the presence ofadistinct gene

foreachform oftryptase.

ThefindingthatmRNAfor both a-tryptase and

f3-tryptase

musthave been presentforbothtobeclonedfromthe cDNA

librarysuggests thatmastcells alsosynthesizethe

correspond-ingproteinproducts. However, thecapacityof eachproductto

be incorporated into the functional tetrameric enzyme, the

relativemolarratiosofa-and

f-tryptase proteins possible

ina

functionalenzyme and the relative

expression

ofeach

putative

tryptase genebydifferent types ofmastcellsare notknown. In

addition, this studydoesnotruleoutthe

possibilities

of

poly-morphic variations withinthe tryptase genesorthe presence of

additional genes for tryptase. At present, the terms a- and

3-tryptase

should berestrictedtorefertothe

protein products

ofthecorrespondingcDNA

molecules,

whereas tryptase refers

tothe authentic enzyme.

Whether the expression in vivo ofa- and

3-tryptase

ac-count for the size

heterogeneity

observed among subunits of

authentictryptasebySDS PAGE is unknown. The calculated sizesofthe

corresponding

mature

proteins, 27,423

D for

a-tryptase, and27,458 Dfor

3-tryptase,

do not accountforthe difference of -

2,000

D

actually

observed between the two

dominant

protein

bands.

Carbohydrate

differences

_(two

puta-tiveglycosylation sites on a-, one on f-tryptase) may account for muchof the size heterogeneity. However, endoglycosidase treatmentof lung-derived human tryptase had no effect on the electrophoretic mobility of the subunits by SDS PAGE, whereas treatment of pituitary-derived human tryptase con-verted subunits of 36,300 and 34,600 D to subunits of 33,400 and 32,400 D (9). Thus, the relative contributions of post-translational events and primary sequence differences to the sizeheterogeneity of authentic tryptase remain problematic.

The marked similarity in predicted amino acid sequence among the catalytic regions of human a-tryptase (73%) and human

fl-tryptase

(78%) with that of dog tryptase, though not as high as between the two human products (91%), clearly showextensive sequence conservation within this family (Fig. 4). Dog tryptase and human

_f3-tryptase

have the identical QLREQsequence (residues 89-93) and one putative carbohy-drate binding site. By comparison, human a-tryptase has NSGT (residues 89-92) and two carbohydrate binding sites. Dog tryptase and human a-tryptase share a carbohydrate re-gion that is lacking in human fl-tryptase. In addition, dog tryptase and human

f3-tryptase

each have SWG (residues 243-245)rather than SWD (residues 242-244) residing in the predictedsubstrate binding pocket ( 19). They also have a Gln (residue 221) rather than a Lys residue covering the entrance into the substrate binding pocket. Comparisons of the 30-amino acid leader sequence for each of these three tryptase molecules also shows extensive homology, but somewhat less than for the catalytic portions. Leader sequences for a- and

f-tryptaseshow 87% identity, whereas these sequences,

respec-ALPHA 1 MLSLLLLALPVLASRAYAAPAPVQAL0QAGIVGGQEAPRSKWPWQVSLRV 50

1111111111111IItIIIIIIIIIIIlii

BETA 1 NLNLLLLALPVLASRAYAAPAPGQALQRVGIVGGQEAPRSKWPWQVSLRV 50

li1 II I1111111111111111 1111111111

DOG 1 MPSPLVLALALLGSLVPVSPAPGQALQRVGIVGGREAPGSKWPWQVSLRL 50

LEADER CRI

ALPHA 51 RDRYWMHFCGGSLIHPQWVLTAAHCLGPDVKDLATLRVN.SGTHLYYQDQ 99

1111111111111111111II1111111111 1111111

BETA 51 HGPYWI4HFCGGSLIHPQWVLTAAHCVGPDVKDLMLRVQLREQHLYYWQ 100

lI 11 IIIIIIIIIIIIIIIIIIII 111111111111

DOG 51 KGQYWRHICGGSLIHPQWVLTAAHCVGPNVVCPEEIRVQLREQHLYYQDH 100

CRII

ALPHA 100 LLPVSRIMVHPQFYIIQTGADIALLELEEPVNISSRVHTVMLPPASETFP 149

1111111

111111

1111111111111 11 11 111111111

BETA 101 LLPVSRIIVHPQFYTAQIGADIALLELEEPVKVSSHVHTVTLPPASETFP 150 111 11 11 11 IIIIIIIIII 111111 111111 III

DOG 101 LLPVNRIVMHPNYYTPENGADIALLELEDPVNVSAHVQPVTLPPALOTFP 150

~~

CRI1II

ALPHA 150 PGMPCWVTGWGDVDNDEPLPPPFPLKQVKVPIMENHICDAKYHLGAYTGD 199

BETA 151 PGNPCWTSGWVDNDERLPPPFPLKQVIVPIENHJICDAYHLAYT1D 200

1111111111 11111111111111 11 11 1111 III

DOG 151 TGTPCWVTGWGDVHSGTPLPPPFPLKOVKVPIVENSMCDVQYHLGLSTGD 200

-CRIV- CRV

ALPHA200 DVRIIPDDMLCAGNS0RDSCKGDSGGPLVCKVNGTWLQAGWSWDEGCAQ 249

1111 11111111 1111 IIIIIIIIIIITIIIIIIii 11111

BETA 201 DVRIVRDDNLCAGNTRRDSCWDSGGPLVCKVNGTWLQAGVVSUGECAQ 250

11111 1111111 1111111 1111111

DOG 201 GVRIVREDMLCAGNSKSDSCOGDSGGPLVCRVRGVWLQAGVVSWGEGCAQ 250

- CRVI- -

CRVII-ALPHA 250PNRPGIYTRVTYYLDWIHHYVPKKP 274

BETA 251 PNRPGIYTRVTYYLDWIHHYVPKKP 275

11111111II 1111111 111111

DOG 251 PNRPGIYTRVAYYLDWIHQYVPKEP 275

Figure 4.Comparison of the amino acid sequences of human a- and

f3-tryptase

and dog tryptase. The extentofeachofthe seven con-servedregions(CRI-CR VII)ofthetrypsinsuperfamily (27) are des-ignated with double lines, the leader with a single line, active site

tively, show53and 60%identity withthedogtryptase leader. Each leader has aterminal Gly residue, consistent with the novel modeof activation suggested for the dogprotease( 12). Generally,

f3-tryptase

is more similar than a-tryptase to the correspondingdog enzyme.Whetherasecond cDNA fordog

tryptase, moreclosely analogoustohumana-tryptase,exists is

not known at present. The enzyme called dog protease (12)

appears to have substantially less homology to dog tryptase

than the human enzymes do to one anothersuggesting that

dogprotease may not beof thetryptasefamily.

Tryptase (23) is the name originally used to describe a

trypsin-likeactivity in human and caninemastcells (24)and was later chosen as the namefor the corresponding purified enzyme inhuman(1), canine (25),and rat(26)mastcells. In

each species, only mast cells appear to contain substantial

amountsof thisenzyme.Functionalcriteriacommonto these enzymesistheir tetramericstructureandtheirbindingtoand

stabilization by heparin. Conserved sequence homologies

amongthe human anddogenzymes

provide

afurther criterion

todefine thisenzyme

family

both withinand betweenspecies.

We recommend the term tryptase be reserved for enzymes

havingthesecharacteristics.

Acknowledaments

This workwassupported inpartby National Institutes of Healthgrants

AI-20487 andAI-27517 and Pharmacia Award 87-0697.

References

1.Schwartz, L. B.,R. A.Lewis,and K.F. Austen. 1981.Tryptase

fromhumanpulmonarymastcells. Purification and characterization. J. Biol. Chem. 256:11939-11943.

2.Schwartz,L.B.,R. A.Lewis,D.Seldin,andK. F. Austen. 1981. Acid hydrolases and tryptase from secretory granules ofdispersed humanlungmastcells.J. Immunol. 126:1290-1294.

3.Craig, S.S.,N.M. Schechter,andL.B.Schwartz. 1988.

Ultra-structuralanalysisofhuman Tand TCmast cellsidentified by im-munoelectronmicroscopy. Lab.Invest. 58:682-691.

4.Irani,A.A.,N. M.Schechter, S. S.Craig,G.DeBlois,andL. B.

Schwartz. 1986. Two types of human mast cells that have distinct neutralproteasecompositions.Proc. Natl.Acad. Sci. USA. 83:4464-4468.

5. Schwartz, L. B., D. D. Metcalfe,J. S. Miller, H. Earl, and T.

Sullivan. 1987.Tryptaselevelsas anindicator of mast-cell activation in

systemicanaphylaxis and mastocytosis. N. Engl.J. Med. 316:1622-1626.

6.Schwartz,L.B.,J.W.Yunginger,J.S.Miller,R._Bokhari,and D.

Dull. 1989. The time course ofappearance and disappearance of humanmastcelltryptasein the circulation afteranaphylaxis.J. Clin.

Invest. 83:1551-1555.

7.Smith,T.J.,M. W.Hougland,andD. A.Johnson.1984.Human

lung tryptase, purification and characterization. J. Biol. Chem. 259:11046-11051.

8.Harvima,I.T.,N. M.Schechter,R. J.Harvima,andJ. E.Fraki.

1988.Humanskintryptase:purification,partialcharacterizationand

comparison with human lungtryptase. Biochim. Biophys. Acta.

957:71-80.

9.Cromlish, J.A.,N.G. Siedah,M. Marcinkiewicz,J.

Hamelin,

D. A. Johnson, and M.Chretein. 1987. Humanpituitary tryptase:

molecularforms,NH-terminalsequence,immunocytochemical local-ization,andspecificitywithprohormoneand

fluorogenic

substrates.J.

Biol. Chem. 262:1363-1373.

10.Schwartz, L. B. 1985. Monoclonal antibodiesagainsthuman

mast celltryptasedemonstrate sharedantigenic sitesonsubunits of

tryptaseand selective localization of theenzymetomastcells. J.

Im-munol. 134:526-531.

11. Miller,J.S.,E. H.Westin,and L.B.Schwartz. 1989.Cloning andcharacterization ofcomplementaryDNAfor human tryptase. J. Clin. Invest. 84:1188-1195.

12.Vanderslice, P.,C. S.Craik,J.A.Nadel,and G. H.Caughey.

1989. Molecularcloningofdogmastcelltryptaseandarelated

pro-tease:Structuralevidenceofauniquemodeofserineprotease activa-tion.Biochemistry. 28:4148-4155.

13.Toneguzzo, F.,S.Glyn,E.Levi,S.Mjolsness,andA.Hayday.

1988. Useofchemically modified T7 DNApolymerase formanual

and automated sequencing ofsupercoiled DNA. BioTechniques. 6:460-469.

14. Moxley, G. 1989. DNA polymorphism ofimmunoglobulin kappaconfers risk of rheumatoid arthritis. Arthritis Rheum. 32:634-637.

15. Devereux, J.,P.Haeberli,and0.Smithies. 1984. A

compre-hensivesetofsequenceanalysisprogramsfor theVAX.Nucleic Acids

Res. 12:387-395.

16. Kozak,M. 1986.Pointmutations defineasequenceflanking the AUG initiator codon that modulates translationby eukaryotic ribosomes.Cell. 44:283-292.

17.Avraham, S.,R. L.Stevens,C.F.

Nicodemus,

M.C.Gartner,

K. F.Austen, andJ. H.Weis. 1989. MolecularcloningofacDNAthat encodesthepeptidecoreofa mousemastcellsecretorygranule

pro-teoglycanandcomparisonwith the_analogousratandhumancDNA. Proc.Natl. Acad. Sci. USA. 86:3763-3767.

18.Steitz,T.A.,R.Henderson,andD. M.Blow.1969.Structureof crystallinealpha-chymotrypsin.J.Mol. Biol. 46:337-348.

19.Polgar,L. 1987. Structure and function ofserineproteases.In

HydrolyticEnzymes.A.Neubergerand K.Broddehurst,editors.

Else-vierSciencePublishers,Amsterdam. 159-200.

20.Krieger, M.,L. M.Kay,andR. M.Stroud. 1974. Structure and specific bindingof_trypsin:comparisonof inhibited derivatives anda modelfor substratebinding.J.Mol. Biol. 83:209-230.

21.Chung,D.W.,K.Fujikawa,B. A.McMullen,andE. W.Davie. 1986. Human plasmaprekallikrein, azymogen toaserineprotease thatcontains four tandem repeats.

Biochemistry.

25:2410-2417.

22. Fujikawa, K., D. W. Chung, L. E.

Hendrikson,

and E. W.

Davie. 1986. Amino acidsequenceof humanFactorXI,ablood

coag-ulation factor with four tandem repeatsthatare

highly

_homologous withplasmakallikrein.Biochemistry.25:2417-2424.

23.Lagunoff, D.,andE. P.Benditt. 1963.

_Proteolytic

enzymesof

mastcells.Ann.NYAcad. Sci. 103:185-198.

24.Glenner,G.C.,andL. A.Cohen. 1960.Histochemical demon-stration of

_{species-specific}

trypsin-like

enzymeinmast cells.Nature

(Lond.). 185:846-847.

25.Caughey, G.H.,N. F.

Viro,

J.Ramachondran,S. C.Lazarus,

D. B. Borson, and J. A. Nadel. 1987. Dog mastocytoma tryptase:

affinity

purification,

characterization and amino-terminal sequence.

Arch. Biochem.

_Biophys.

258:555-563.

26.Kido, H.,N.Fukusen,andN.Katunuma. 1985.

Chymotrypsin

and

trypsin-type

serine proteases in rat mast cells:

_properties

and

functions. Arch. Biochem.

Biophys.

239:436-443.

27. Furie, B., D. H.

Bing,

R. J. Feldman,D.J. Robinson,J. P.

Burnier,andB.C.Furie. 1982.

Computer-generated

modelsofblood

coagulation

factorXa,factorIXa,andthrombin baseduponstructural