Proc. Natl. Acad. Sci. USA Vol. 89, pp. 1548-1552, March 1992 Immunology
Transmission-blocking antibodies recognize microfilarial chitinase in brugian lymphatic filariasis (Brugia malyi/parasitic nematodes/PCR)
JULIET A. FUHRMAN*t, WILLIAM S. LANEt, RANDALL F. SMITH§, WILLY F. PIESSENS*, AND FRANCINE B. PERLERt *Department of Tropical Public Health, Harvard School of Public Health, 665 Huntington Avenue, Boston, MA 02115; tNew England Biolabs, Inc., 32 Tozer Road, Beverly, MA 01915; *Harvard University Microchemistry Facility, 16 Divinity Avenue, Cambridge, MA 02138; and §Molecular Biology Computer Research Resource, Dana-Farber Cancer Institute, 44 Binney Street, Boston, MA 02115
Communicated by Phillips W. Robbins, November 12, 1991
ABSTRACT Brugia malayi is a parasitic nematode that causes lymphatic filariasis in humans. The monoclonal antibody MF1, which mediates clearance of peripheral microfilaremia in a gerbil infection model, recognizes two stage-specific proteins, p70 and p75, in B. malayi microrflariae. cDNA coding for the MF1 antigen was sequenced, and the predicted protein sequence shows significant similarities to chitinases from bacteria and yeast. When microfilarial extracts and purified preparations of the MF1 antigen were tested for chitinase activity, strong bands of chitin-degrading activity comigrated in SDS/PAGE with p70 and p75 and showed a reductiondependent mobility shift characteristic of the MFl antigen. Thus, the MF1 antigen is microfrarial chitinase, which may function to degrade chitin-containing structures in the microfilaria or in its mosquito vector during parasite development and transmission.
MATERIALS AND METHODS Parasite Material. Gerbils (Meriones unguiculatus) were infected intraperitoneally with B. malayi either in our own facilities or at the Filariasis Repository Research Service,
Athens, GA. mf were purified and proteins were extracted as described (3). Poly(A) mRNA was isolated from purified mf according to Cox and Smulian (4). Genomic DNA was made from adult B. malayi as described (5). Protein Purification and Amino Acid Sequencing. The MF1 antigen was partially purified from extracts of mf by DE-52 chromatography as described (6). N-terminal sequence analysis was performed on the combined p70 and p75 following elution with 0.5 M NaCI from DE-52, using an ABI model 477A protein sequencer. The sample was subjected to automated Edman degradation using the program NORMAL-i, which was modified using the manufacturer's recommendations for faster cycle time (37 min) by decreasing dry-down times and increasing reaction cartridge temperature to 530C during coupling. The resultant phenylthiohydantoin amino acid fractions were manually identified using an on-line ABI model 120A HPLC apparatus and Shimadzu CR4A integrator. Trypsin, Asp-N, and Glu-C (sequencing grade, Boehringer Mannheim) were used to digest the separated p70 and p75. Proteolytic digests were performed directly on nitrocellulose according to Aebersold et al. (7), using DE-52-purified material that was separated by SDS/PAGE (8, 9) and transblotted (10). Following proteolytic digestion, peptides were separated by narrow-bore reverse-phase HPLC on a Vydac 2.1 mm x 150 mm C18 column. The gradient employed was a modification of that described by Stone et al. (11). Briefly, where buffer A was 0.06% trifluoroacetic acid/H20 and buffer B was 0.055% trifluoroacetic acid/acetonitrile, a gradient of 5% buffer B at 0 min, 33% buffer B at 63 min, 60%o buffer B at 95 min, and 80% buffer B at 105 min with a flow rate of 0.15 ml/min was used. CNBr peptides were generated by cleavage of the separated p70 and p75 in gel slices following SDS/PAGE, using 7 mg of CNBr per ml in 70%6 formic acid. Cleavage products were then electrophoresed and transblotted to poly(vinylidene difluoride) (Immobilon, Millipore) according to Matsudaira (12), stained briefly with Coomassie blue, excised, and sequenced directly. Glycosidase treatment was performed on total mf extracts, using N-glycanase and 0-glycanase from Genzyme, according to the supplier's protocols. Library Screening, DNA Sequencing, and PCR Amplification. MF1-specific DNA probes were synthesized by PCR amplification of B. malayi genomic DNA using degenerate primers derived from the protein sequence of the N terminus (sense strand, primers P1 and P4) and a tryptic peptide (antisense strand, primers P5 and P6). PCR amplifications were performed using Taq polymerase (Cetus), and amplification products were purified from low-melting-point agarose
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Abbreviation: mf, microfilaria(e). $The sequence reported in this paper has been deposited in the GenBank data base (accession no. M73689).
Lymphatic filariasis afflicts nearly 100 million people worldwide (1). Of the three species of lymphatic filariids that infect humans, Brugia malayi can most readily be maintained in the laboratory by passage between gerbils (another natural host) and a mosquito vector. Thus, B. malayi has been useful for studying pathogenesis and transmission of filariasis at the molecular level. Passive transfer of the monoclonal antibody MF1 was previously shown to mediate the transient clearance of first stage larvae, or microfilariae (mf), from the peripheral blood of gerbils infected with B. malayi (2). The MF1 antibody recognizes two stage-specific proteins (p70 and p75) in extracts of mf that appear only after the mf have matured for several days in the vertebrate host (3). The appearance of the MF1 antigen corresponds with the onset of the parasite's ability to infect the mosquito (3). It was thus of great interest to define more precisely the role of the MF1 antigen in filarial development and transmission. With the expectation that the proteins' sequences might clarify their function, we obtained partial amino acid sequences for p70 and p75 and then used this information to isolate and to sequence the corresponding cDNA.1 We report here that the predicted protein sequence of the MF1 antigen has several regions of significant similarity to bacterial and yeast chitinases. Furthermore, the native MF1 antigen is demonstrated to degrade chitin in vitro.
1548
Immunology: Fuhrman et al. (SeaKem, FMC). The genomic probes were 32P-labeled by nick-translation (13) and used to screen a B. malayi mf cDNA library in Agtll (gift of Timothy Nilsen, Case Western Reserve University, Cleveland). MF1-specific single-stranded cDNA was synthesized from mf mRNA with the cDNA Synthesis Plus kit (Amersham), using an antisense primer (P21) derived from the DNA sequence of phage 4A2 (see Results). First strand cDNA was then PCR amplified using two primer combinations. The cDNA fragment 52.4 was generated using P48 and P21, and fragment 4'V was generated using PSL and P21.2 (see Fig. 3). Oligonucleotides for PCR amplifications and for sequencing primers were synthesized on a Biosearch model 8750 DNA synthesizer or were purchased from New England Biolabs. The cDNA clones and PCR-amplified cDNA products were sequenced in both directions using a nested set of oligonucleotide primers and single-stranded DNA templates (14). All nucleotide sequencing was performed by the dideoxy method using Sequenase Version 2.0 (United States Biochemical) according to the manufacturer's protocol. Computer Sequence Analyses. Protein sequence data base searches were conducted with the GENINFLO BLAST Network Service provided by the National Center of Biotechnology Information, National Library of Medicine, National Institutes of Health (15). Multisequence alignments were performed using the PATTERN-INDUCED MULTI-ALIGNMENT program (R. F. Smith and T. F. Smith, Molecular Biology Computer Research Resource, 1990) based on the pattern construction algorithm previously described (16). Chitinase Activity Assays. Assays for chitinolytic activity were performed in SDS gels cast with glycol chitin according to Trudel and Asselin (17). Extracts of native mf proteins were prepared as described (6), but in the absence of protease inhibitors. Under these conditions, p70 and p75 consistently migrate with a lower molecular mass than previously described for extracts prepared in the presence of protease inhibitors. Following electrophoresis, proteins were renatured in the gel by overnight incubation at 37°C in 1% Triton X-100 (Boehringer Mannheim, 789-704). Areas of digested chitin were visualized as dark bands on a fluorescent background after staining with Calcofluor M2R (17).
RESULTS Protein Sequence of the Native Antigen. The MF1 monoclonal antibody detects two bands in Western blots with mobilities corresponding to approximately 70 and 75 kDa under reducing conditions (6). The HPLC profiles of tryptic digests of p70 and p75 are highly homologous, with most major peaks shared between the two (Fig. 1). Three tryptic peptides, as well as two CNBr peptides, from p75 had sequences identical to peptides of the same mobility from p70 (Fig. 2). N-terminal sequence analysis of a mixture of p70 and p75 generated a single amino acid residue at each cycle (Fig. 2). CNBr peptides derived from the individual proteins generated the same N-terminal sequence (YVRG-YY), suggesting that p70 and p75 are identical at the N terminus. These comparisons indicate a high degree of primary structure similarity between p70 and p75. To examine differences between the primary structures of p70 and p75, we sequenced the double peak (retention time = 67 min) indicated by the arrow in the p75 profile (Fig. 1A) and the corresponding single peak (retention time = 67 min) from the p70 profile. The p70 peak gave a single sequence (GYGGAFIWALDFDDFTGK). The p75 double peak gave this sequence and a secondary sequence in addition (GPYPLLNAISSELEGESEN-). This suggested that the secondary sequence could be part of a peptide unique to the p75 molecule.
Proc. Natl. Acad. Sci. USA 89 (1992)
1549
00A
E 1
-10,
10
20
30
40
50
60
70
20
30 40 50 Time (min)
60
70
60 B 5040
:D30 E 20
10
0-
-10
10
FIG. 1. Vydac C18 HPLC profiles of tryptic digests of isolated p75 (A) and p70 (B). The arrow indicates the mobility of the peptide unique to the p75 map that was partially sequenced (amino acids 377-395, see Fig. 3). mAU, milliabsorbance units.
Isolation of a cDNA Clone Coding for the C Terminus of the MF1 Antigen. Degenerate oligonucleotide primers were synthesized corresponding to the N-terminal region (sense) and a tryptic peptide shared by p70 and p75 (antisense). These primers (P1, P4, P5, and P6, Fig. 3) were used to amplify genomic DNA from B. malayi by PCR. Combinations of P1 and P5 or P6 generated a 2.3-kilobase (kb) amplification product, whereas combinations of P4 and P5 or P6 generated a 2.0-kb product. This suggests the presence of an intron between P1 and P4. The 2.3-kb product was partially sequenced, and the predicted amino acid sequence of one region matched the sequence of a tryptic peptide from p75 (amino acids 207-221) distinct from that used to construct the oligonucleotide primers. This strongly implicated the 2.3-kb N-terminus p75/p70 YVRG-YYTNW AQYRDGEGKF
LPGNIPNGL-
Tryptic peptides, p75 GYGGAFIWAL DFDDFTGK* NNFDGFDLDW EYPVGVAEQH A** KNNFDGFDL IIIGIPMYAQ GWTLDNP' ETVHQEGVGA YMVK GPYPLLNAIS SELEGESEN GTIDGS
GDQWYG -DELGDSKPF E-NDEDTEWS KG GEVSGIGIFN TEFAADYWAS
T-ILYAFAKV D
Tryptic peptides, p7O GYGGAFIWAL DFDDFTGK* NNFDGFDLD- EY** YLK
-IIGIPMYAQ G-TLD' ETVHQEGVGA YMVK
Asp-N peptides, p70 DGSYNVESLG KNF DQWYGY DNPSETAI-A AASRPS-A
Glu-C peptides, p70
LVEAMK
AMKTAFVE
NFDLLFLMSY DLHGS-E TNPAGGTASY -EI S-GK
GVGAYMVKGD Q-YGYDN
CNBr peptides, p75
CNBr peptides, p70
-QY'
YVRG-YYTNW AQYRDGEGKF LPGNI'
YVRG-YYDN-
YSAVT5 YAQG
YSAVTKLRET DPGLKVLLSY G-YNF' YAQGWTLDNP SETAIG KTAFVEEAKT SGKQRLL
FIG. 2. Peptide sequences from p75 and p70. Peptides marked with identical superscripts have equivalent retention times (tryptic peptides) or molecular masses on SDS/PAGE (CNBr peptides).
1550
'C,A|PIe.;f:-40GW~LaFh.A' Immunology: Fuhrman et al.
Proc. Natl. Acad. Sci. USA 89 (1992)
amplified genomic DNA fragment as part ofthe coding region for the MF1 antigen. The 2.3- and 2.0-kb DNA fragments were used to probe a B. malayi microfilarial cDNA library in Agtll. Clone 4A2 contained the largest insert of all hybridizing clones. The insert was sequenced and found to encode seven peptides sequenced from native p75, including the peptide unique to the p75 tryptic profile (Fig. 3). No discrepancies could be found between the amino acid sequences of the native peptides and the deduced sequence of clone 4A2. The open reading frame of 718 nucleotides was followed by a stop codon, a 50-nucleotide untranslated region, and a poly(A) tail. The 4A2 insert encoded sequences from p70 and p75 (Figs. 2 and 3). It was therefore used as a probe for Southernblotted, EcoRI-restricted genomic DNA (from B. malayi) to test the possibility that two separate genes code for p70 and p75. Two bands (1.4 and 3.7 kb) hybridized strongly, and two bands (0.8 and 3.1 kb) hybridized very weakly with the 4A2-derived probe (data not shown). PCR amplification of genomic DNA spanning the region coding for 4A2 generated a single 2.5-kb product, which yielded a 1.4-kb product following digestion with EcoRI (data not shown). Although this suggests that the two strongly hybridizing bands (1.4 and 3.7 kb) are part of a single gene coding for these proteins, we cannot rule out the possibility that the weakly hybridizing bands represent a second gene coding for an MF1-reactive molecule. Sequencing the 5' End of the mRNA by PCR Amplification. None of the isolated cDNA clones contained sequence encoding the N terminus of the MF1 protein. Therefore, the
A
200
1
nt
I
800
600
400
1400
1200
1000
i
I
I
I
I
I
I
a
remainder of the sequence was obtained from uncloned cDNA amplified in two steps by PCR. First strand cDNA was synthesized using mf mRNA and an antisense primer from the 4A2 phage insert (P21, Fig. 3). The cDNA was then amplified by PCR using P48 (corresponding to sequence obtained from the 2.3-kb genomic amplification product) and P21. The sequence of this amplified cDNA (fragment 52.4, Fig. 3) yielded an open reading frame of 814 nucleotides. The final 132 nucleotides overlapped with the 5' end of the 4A2 insert, indicating that the EcoRI site at the beginning of the 4A2 insert was derived from the MF1 cDNA and not from linker addition. The deduced amino acid sequence of 52.4 corresponded to six peptides sequenced from native p75, including the last three residues from the N-terminal sequence. One discrepancy was noted at amino acid residue 156, which was predicted to be a glutamate from the nucleotide sequence but appeared as glutamine in the tryptic peptide. To obtain the 5' terminus of the message, we amplified the first strand cDNA using a forward primer (PSL, Fig. 3) corresponding to the sequence of the trans-spliced leader RNA from B. malayi (18) and a reverse primer from the antisense strand of 52.4 (P21.2). This yielded fragment 4'V (Fig. 3), which contained an open reading frame of 516 nucleotides beginning 28 nucleotides from the 3' end of the spliced leader, with a potential 22-residue signal peptide preceding the N terminus of the mature protein. Sequence Similarity to Chitinase and Enzymatic Activity of Native MF1. A search of the Protein Identification Resource (release 27.0), Swiss-Prot (European Molecular Biology Laboratory, release 17.0), and Translated GenBank (release 64.3)
f
EcoRI
4'V
B
i
A-A'-AG :':
AAE
t
A
4A2
52-4
cDNA
G
T
Y
C
N
W
A
5
P
r
E
G
G
K
S.
F
P
G
-;A -'.--AA"' ,A-A, --;ATA A:,--:A 4; A,e, A ; 4 N A A YA AF F F. '1 i' N , GA
-
--
: : -' N FI . hA;
A
:
r F
N
-
A A
-
V
C'
N
-
K T
T
,
T
"
A
*
*
-A-AAAAA---AAA ALGIA -_L AA K
E
T
N
p
A
X
NF'rGF
A.'-A''"k-GWFT:. AA -.AF AA ;X h. .LA.....AA..A. AA-,A A MAY F L A A
-.'' A
A -AA- 'A-i'7A'-':AAAAAA A:ACA. A' '-A G--AAA AAA'A V PAA D w EA Y V A E H A A mS 1. AY E A M-N L
*f'
A A Y T
I
A ': TAhGA -'A-- :.
;,,T AGA AA';: TGAATTT'f AX;ATC-AATTGGAAA';A-A;:;AAA .-'AXA:,A'A"A'..GAALePPAArATA'AATrAAI GE V S G F A A D Y N A At FL S A F
A"ATA"-'AAALAZ~~~~~j
Y
',A':.:T:'*r,':A AIAAA; 'Ar AM, .......
o-
*'A
''
M
Y
A' ni7' -AA.'g
t
-: AT AS TA A A- 4-AA. G S Y N AG A rA M 4H IA
*AATA
'.AL -GAA '
G ~
S
A
A
~
-a
AA
AAAAA' ':: A::
( A' AA
P
A:
:
'A
A'
A 'A.-.
A
r
A
.'
A .A'A:.sA-.:.I;AA'. A:. :AAhA-
A:r.FAA s
A'.. -AA s
A
A' R
.
AAh . .... 7: A . 1: .A ...
AA:' GAG' 'CA'' -GA' A A: - -A
A A.' -A
V F " :0 &Y G '3 A F I W A L D F DD r F' ~~~~~~~~~~~wr
.';A .'.' A ,A;;.| A.: '--:,AAA- AA..A.:A-`-'a *LA./ AA ': A
A
AA
A AA. .AA.;"A.'' T: A '' A .'-A
sA
T( JG TuT TCATAT..................... T T AC ;('. AA:
:
hA'.A 1 -AA hAA-Ah .A
.
A *A -ACAA GA _ATA ALA
r A 'T;:;1T........... A'CsvA sTfm TAS 'GG
FA
;
A.
's-:
AAAAAT
A:-AA ','A,CAGA'I
C x A AAG:A' A A
'A
G
K
C S
G
G pP
.,AA......................... \A AM' A A...-A 7':A!% -
A
?ALA ;.;- e hA lAAf e ';A
.AA-''AAA'-:-.'-: T Ah'.
::'AA.'.?s*.- ".A
A A-A.
-- -A -
-GA' ?
L
N
A
AAA A
-
-A
-A
r
,A A-
FIG. 3. (A) Alignment of cDNA fragments used to sequence the MF1 message. 4A2 is a Agtll clone. 52-4 and 4'V are cDNA fragments amplified from mRNA using primers P48 and P21 or PSL and P21.2, respectively. Nucleotide numbering begins with the first nucleotide of the initiation codon. (B) Nucleotide sequence and deduced amino acid sequence for the MF1 antigen. Nucleotide numbering corresponds to that in A. Amino acid numbering begins at the initiating methionine, with the mature N terminus beginning at amino acid 23. Lowercase letters on the nucleotide line repre-
sent untranslated regions. Primers used for PCR amplifications are marked as arrows above the nucleotide line, with open arrows denoting degenerate primers derived from protein sequence and closed arrows denoting nondegenerate primers corresponding to known nucleotide sequence. The EcoRI site is overscored on the nucleotide line. The consensus poly(A) signal is underlined. Lowercase letters on the amino acid line represent the predicted signal peptide. The N-terminal sequence derived from the combined p70 and p75 is underlined with dashes ----). Peptides sequenced from native p75 are shaded, and peptides sequenced from native p70 are boxed. A triangle marks the start of the peptide unique to p75
tryptic profile.
Immunology: Fuhrman et al.
1551
Proc. Natl. Acad. Sci. USA 89 (1992)
15 16 30 31 90 75 76 45 46 60 61 ---YVRGCYYTNWAQ YRDGEGKFLPGNIPN GLCTHILYAFAKVDE LGDSKPFEWNDEDTE WSKGMYSAVTKLRET *L* 2 CHI1$BACCI 62/VADIDPTKVTHINYA FADICWNGIHGNPDP SGPNPVTWTCQNEKS QTINVPNGTIVLGDP WIDTGKTFAGDTWDQ PIAGNINQLNKLKQT 3 CHIB$SERMA ------MSTRKAVIG YYFIPTNQINNYTET DTSVVPFPVSNITPA KAKQLTHINFSFLDI NSNLECAWDPATNDA KARDVVNRLTALKAH 4 CHIA$SERMA 173/NFTVDKIPAQNLTHL LYGFIPICGGNGIND SLKEIEGSFQALQRS CQGREDFKISIHDPF AALQKAQKGVTAWDD PYKGNFGQLMALKQA 5 KTXA$KLULA 349/DYCDKKSSTTGAPGT DGCFSNCGYGSTSNV KSSTFKKIAYWLDAK DKLAMDPKNIPNGPY DILHYAFVNINSDFS IDDSAFSKSAFLKVT 1
1 MF1
---------------
1 MF1
NPGLXVLLSYGGYNF G IFTGIA------ KSAQK TERFIKSAl * * S*GG* * l** F * ** NPNLKTIISVGGWTW SN FSDVAA------ -TAAT EVFANSAV NPSLRIMFSIGGWYY ISN LGVSHANYVNAV KTPAA TKFAQSCV HPDLKILPSIGGWTL SD PFFFMG.------- -DK DRFVGSVK SS--XKIPSFGGWDF S PSTYT-IFRNAV KTDQN NTFANNLI
90 2 CHI1$BACCI 3 CHIBSSERMA 4 CHWIASERMA 5 KTXA$KLULA
181
IMFI 2 CHI1$BACCI 3 CHIMBSERMA 4 CHIASSERMA
5 KTXASKLULA
1 MF1 2 CHII$BACCI 3 CHIBSSERMA 4 CHIASSERMA
5 KTXA$KLULA
105 106
195 196
120 121
135 136 150 AFLRKN-NFDGFDLDW **** *DG*D*DW DFLRKY-NFDGVDLDW RIMKDY-GFDGVDIDW
180 151 165 166 EYPV4AEEHAKLVE AMKT--AFV-----E*P * *
EYPVYSGLDGNSKRP EDKQNYTLLLSKIRE
EYPQA .EVDGFIAAL QEIR--TLL-----N
EFLQTWKFFDGVDIDW EFPGC KGANPNLGSP QDGETYVLLMKELRA NFMNKY-NLDGIDLDW EYPGPDIPDIPADD SSSG-SNYL-----T
210 211
225 226 240 DL LFLMSYDLHGS -KN D * *M*YD* G ** KLDAAGAVDGKKY 14L TIASGA TYAAN-- TELAKIAA-W INIMTYDFNGA -KI QQTIADGRQALPYqL TIAGAGC 4FFLSRYY SKLAQIVA-- DY INLMTYDLAGP -KI MLDQLSAETGRKY EL TSAISAC KDKI---- DKVAYNVAQN MDH IFLMSYDFYGAF LKN FLKLLKGKMPSGK' LSIAIPS WYL---- KNFPISDIQNI YMWMTYDIHGI-EY
270 241 255 256 VDLHGKLHPTXGE-- --------------* SAHNAPLNYDPAAS - --------------TNHQAALFGDAAGPT FYNALREANLGWSWE
271 285 286 300 301 315 316 330 -------VSGIGIFN TEFAADYWAS MPK EKIIIGIPMYAQGW LDNPSETAIGAAASR ~ * * K** G * Y***** ** ----AAGVPDANTFN VAAGAQGHLDA GVPA AKLVLGVPFYGRGW GCAQAGNGQYQTCTG ELTRAFPSPFSLTVD -AAVQQH LMMGVPS AKIVMGVPFYGRAF GVSGGNGGQYSSHST ----------RHRLH HGERRECAAG VKP GKIVVGTAMYGRGWI GVNGYQNNIPFTGTH
360 345 346 331 PSSASKTNPAGG--- -------------TA
--EEAKTSGKQRL]L TAAVSA GTI--DG SYNVESLGK* *L * A ***
-----------RKEIE DA--IKMLDK GVKF NKVFGGVANYGRSY MVNTNCYNYGCGFQR
LGHQTALNARPGS-GKANSYINCHTP---
-----------------------------
*
*
GSSVGTWEAGSF --- -------F------DF PGEDPYPNADYWLVG CDECVRDKDPRIASY RAVKGTWENGIV--- --------------D EGGNSRDMTNTP--- --GVLSDSEIIDIDS
5 KTXA$KLULA
361 375 376 390 391 405 406 435 420 421 SYWEICKYLKEGGKE TVHQEGV YMVKGD Q--WYGYDNEETI I KMKWLKEKGYGGAFI WALDFDDF KSCGK * D **IW D** ** * * *** YDLEANYINKNGYTR YWNDTAK YLYNAS NKRFISYDDAES Y KTAYIKSKGLGGAMF WELSGDRNK rL-QNK RQLEQMLQGNYGYQR LWNDKTK YLYHAQ NGLFVTYDDAESF Y KAKYIKQQOLGGVMF WHLGODNRN ----D YRQIASQFMSGEWQY TYDATAE YVFKPS TGDLITFDDARS A KGKYVLDKQLGGLFS WEIDADNGD --LNS SDKKNDRWVDTNTDC IFMKYDGN SSWPK SR-YDLEDMKN F AGTSLWAANYFKHDE WKNDEDD ----D
IMF1
451 465 466 480 481 540 495 496 510 511 525 526 ESENPEITTEEPSIT ETEAYETDETEETSE TEAYDTDETEETSET EATTYDTDETEGQEC PERDGLFPHPTDCHL FIQCANNIAYVMQCP
2 CHI1$BACCI 3 CHIB$SERMA 4 CHIASSERMA
TTAPSVPGNARSTGV TANSVTLAWNASTDN VGVTGYNVYNGANLA TSVTGTTATISGLTA GTSYTFTIKAKDAAG NLSAASNAVTVSTTA DSQLDMGTGLRYTGV G--PGNLPIMTAPAY VPGTTYAQGALVSYQ GYVWQTKWGYITSAP GSDSAWLKVGRLA-- ---------------
S KTXASKLULA
DCKNKAGYDLDNPVY GCRLETAINI I IWNG TESVNTVLNILNDYD NYIKYYEALTRAHYD SVMEKYEKWLFEEDG YYTYYTDVDGDDI II
1
541 555 556 580 ATTFFNDAIKVCDHM TNAPDTCI-------
1MF1 2 CHI1$BACCI 3 CHIBSSERMA 4 CHIASSERMA
MF1
2 CHI1SBACCI 3 CHIB$SERMA 4 CHIASSERMA
5 KTXA$KLULA
450 436 GPYPLLNAISSELEG
*
LKADLPTGGTVPPVD LLAALDRYFNAADYD MNASLGNSAGVQ--TEDPFDEENVYFDVY
QPGGDTQAPTAPTNL ASTAQTTSS ITLSWT\ TPPDKKKRDYIQEKY SFEKEFMMSQNMTEL\
FIG. 4. Amino acid sequence alignment ofthe MF1 protein with three bacterial chitinases and one yeast chitinase: chitinase Al of B. circulans (Swiss-Prot locus CHI1$BACCI), chitinases B and A of S. marcescens (CHIB$SERMA and CHIA$SERMA), and killer toxin RF2 a subunit of K. lactis (KTXA$KLULA). Major regions of primary sequence similarity shared among this set are shown boxed. Identical residues conserved in the set are shown as uppercase letters beneath sequence 1; conservative amino acid groups are indicated by asterisks.
data bases revealed significant sequence similarities between the MF1 protein sequence and several chitinases (E.C. 3.2.1.14). An alignment of the MF1 protein sequence with three bacterial chitinases (chitinase Al of Bacillus circulans, chitinases A and B of Serratia marcescens) and one yeast chitinase [killer toxin RF2 a subunit of Kluyveromyces lactis (19)] shows seven major regions of primary sequence conservation (shown as boxed regions in Fig. 4). Following the last conserved region, the MF1 sequence diverges and contains an acid-rich triplication of 14 amino acids (starting at reference position 466, Fig. 4). To test the MF1 antigen for chitinase activity, extracts of mf were electrophoresed in SDS gels containing glycol chitin (17). Chitinase activity was localized to two prominent bands that copurified with the MF1 antigen in DE-52 chromatography and comigrated with the MF1 antigen in onedimensional SDS/PAGE, demonstrating the reductiondependent shift in mobility characteristic of the MF1 antigen (Fig. 5; ref. 6). Since the predicted molecular mass (=55 kDa) for the translation product of the cDNA sequence is lower than that observed for the enzymatically active bands in SDS/PAGE, we tested the possibility that glycosylation could account for the discrepancy. Following treatment with 0-glycanase, but not N-glycanase, a sizeable shift in molecular mass was observed for p70 and p75 by Western blot (Fig. 5).
DISCUSSION Transmission of lymphatic filariasis in nature depends on the persistence of mf in the peripheral blood of infected verte-
brates and on the capacity of those mf to infect and develop in the insect vector. Female adults of Brugia spp. shed live mf, which retain a remnant of their eggshells as the extracuticular sheath (20). This sheath is maintained while the mf circulate in the vertebrate bloodstream but is lost during the initial phases of development in the arthropod vector. 2-ME ---+ + + ..v|-
+-
+
-
95 -
5.R' E
95 5
-95.5
5-55 -
2 3
4 5 6
7
8
9
10
43
1112 13
FIG. 5. Copurification of microfilarial chitinase with the MF1 antigen. Molecular masses are indicated in kDa. Lanes 1-6, chitinase activity is demonstrated following electrophoresis in glycol chitincontaining SDS/polyacrylamide gels for total microfilarial extract (lanes 1 and 6) and DE-52-purified fractions of extracts eluted with 0.3 M (lanes 2 and 5) and 0.5 M NaCl (lanes 3 and 4), electrophoresed in the absence (lanes 1-3) or presence (lanes 4-6) of 2-mercaptoethanol (2-ME, - or +). Lanes 7-10, silver-stained SDS/PAGE (lanes 7 and 8) and Western blot (lanes 9 and 10) of 0.5 M NaCl fraction electrophoresed in the absence (lanes 8 and 9) and presence (lanes 7 and 10) of 2-mercaptoethanol. Lanes 11-13, total microfilarial extract following sham (lane 11), N-glycanase (lane 12), or 0-glycanase (lane 13) treatment was Western blotted and developed with the MF1 monoclonal antibody.
1552
Proc. Natl. Acad. Sci. USA 89 (1992)
Immunology: Fuhrman et al.
Exsheathment is necessary for subsequent molting and development of the filarial larvae in the vector. We previously demonstrated that the microfilarial sheath contains oligomers of f,1-4-linked N-acetylglucosamine and that sheath morphogenesis can be altered by inhibitors of chitin synthesis (21). It is thus of great interest to find a larval chitinase whose temporal appearance coincides with the onset of microfilarial infectivity for the mosquito. The chitinase could be involved in either the degradation of microfilarial structures, such as the sheath, or the interaction with chitinous mosquito structures that the mf must disrupt or penetrate during infection. The peritrophic membrane is a chitinous structure elaborated in the mosquito midgut following a bloodmeal, but mf appear to escape from the midgut long before this membrane is well formed (22). The possibility that mosquitoes secrete a midgut lectin that can inhibit infection was suggested by Phiri and Ham (23), who demonstrated enhanced microfilarial penetration of the midgut in the presence of N-acetylglucosamine. A microfilarial chitinase could generate inhibitory oligosaccharides and promote infection by a similar mechanism. Chitinase activity was previously described in Onchocerca gibsoni (24). In this filarial species, the activity is present in the gravid female adult and presumably acts to remodel chitin fibrils in the developing eggshell or to free the developing mf from their eggshells prior to birth. The mf of Onchocerca spp. are born without sheaths and therefore do not need a mechanism for exsheathment in the vector. Brugian mf do not hatch from their eggshells in utero but seem to remodel the shell and elongate it ultimately to form the sheath (20). This remodeling may also involve chitinase, but this enzyme should be antigenically distinct from the chitinase sequenced here, as the MF1 antibody does not react with extracts of adult female B. malayi (6). The sequence reported here was compiled from three overlapping sequences: the cDNA clone 4A2 (nucleotides 794-1606), the cDNA PCR product 52.4 (nucleotides 112925), and the cDNA PCR product 4'V (nucleotides -48-516). The 5' and 3' untranslated regions are short, and a poly(A) addition signal (ATAAAA) occurs at nucleotides 1547-1552. In sequencing the 5' terminus of the message, we utilized a primer specific for the spliced leader present on some, but not all, mRNAs in B. malayi. Since this technique selectively amplified SL-containing messages, any versions of the MF1 message not containing this leader would not have been detected. Nevertheless, at least a subpopulation of the messages coding for this antigen does contain this spliced leader and a signal peptide preceding the mature N terminus. Although the complete cDNA sequence was compiled from three separate fragments, the extensive overlap at the nucleotide and peptide levels supports the contiguity of the sequence. Furthermore, a unique 1.5-kb fragment can be amplified from microfilarial cDNA using primers corresponding to the sequence coding for p70/p75 N terminus (forward) and the 3' end of the 4A2 clone upstream of the poly(A) tail (reverse) (M. Southworth, personal communication). Chitinase activity is demonstrated here for both proteins, p70 and p75, recognized by the MF1 antibody. However, the relationship between p70 and p75 is still unclear. Two separate messages could exist for p75 and p70, or posttranslational modification (such as C-terminal truncation) could account for the difference in their apparent molecular masses. Since O-glycanase treatment caused equivalent mobility shifts for both proteins, we cannot at this point ascribe the difference between the two to this particular class of posttranslational modification. The region of the sequence that spans amino acids 377-453, which includes the acid-rich triplication, is characterized by a high PEST score [where PEST indicates proline (P),
glutamic acid (E), serine (S), and threonine (T); ref. 25], suggesting a calcium-binding site. p70 and p75 were previously shown to bind calcium (6), and this region of the protein could be responsible for that property. The translation of the full cDNA predicts a molecular mass of -55 kDa for the encoded protein, which is smaller than the molecular mass for either protein recognized by the MF1 antibody. As demonstrated in Fig. 5, 0-glycosylation accounts for part of this discrepancy. In addition, the very acidic composition of this PEST region could result in anomolous migration in SDS gels. Expression of the recombinant protein from full-length cDNA should resolve this question. As chitin is absent from the vertebrate host, chitin synthesis and processing offer potential targets for chemotherapy, and the enzymes involved in chitin metabolism could be suitable candidates for vaccine development. The demonstration of chitinase activity for the purified MF1 antigen and the availability of the nucleotide sequence coding for the enzyme favor its analysis as a therapeutic target in this and other parasitic helminths. We thank Dr. Sen Dissanayake for expert advice, Sherry Roemer for sequencing the genomic PCR product, and Dr. Maurice W. Southworth for helpful discussions. Peptide preparation and sequencing were performed by Renee Robinson and Ruth Davenport at the Harvard Microchemistry Facility. Materials used in this study were provided by a National Institute of Allergy and Infectious Diseases supply contract (Al 02642), U.S.-Japan Cooperative Medical Science Program. This work was supported in part by a National Institute of Allergy and Infectious Diseases grant (Al 24858) to J.A.F. 1. World Health Organization Expert Committee on Filariasis (1984) Tech. Rep. Ser. No. 702 (WHO, New York). 2. Canlas, M., Wadee, A., Lamontagne, L. & Piessens, W. F. (1984) Am. J. Trop. Med. Hyg. 33, 420-424. 3. Fuhrman, J. A., Urioste, S. S., Hamill, B., Spielman, A. & Piessens, W. F. (1987) Am. J. Trop. Med. Hyg. 36, 70-74. 4. Cox, R. A. & Smulian, N. (1983) FEBS Lett. 155, 73-80. 5. Arasu, P., Philipp, M. & Perler, F. (1987) Exp. Parasitol. 64,
281-291. 6. Fuhrman, J. A. & Piessens, W. F. (1989) Mol. Biochem. Parasitol.
35, 249-258. 7. Aebersold, R. H., Leavitt, J., Saavedra, R. A., Hood, L. E. & Kent, S. B. H. (1987) Proc. Natl. Acad. Sci. USA 84, 6970-6974. 8. Laemmli, U. K. (1970) Nature (London) 227, 680-685. 9. Hunkapillar, M. W., Lujan, E., Ostrander, F. & Hood, L. E. (1983) Methods Enzymol. 91, 227-236. 10. Towbin, M., Staehelin, T. & Gordon, J. (1979) Proc. Natl. Acad. Sci. USA 76, 4350-4354. 11. Stone, K. L., Lopresti, M. B., Williams, N. D., Crawford, J. M., DeAngelis, R. & Williams, K. R. (1989) in Techniques in Protein Chemistry, ed. Hugh, T. (Academic, New York), pp. 377-391. 12. Matsudaira, P. (1987) J. Biol. Chem. 262, 10035-10038. 13. Rigby, P. W. J., Dieckmann, M., Rhodes, C. & Berg, P. (1977) J. Mol. Biol. 113, 237-251. 14. Higuchi, R. G. & Ochman, H. (1989) Nucleic Acids Res. 17, 5865. 15. Altschul, S. F., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J. (1990) J. Mol. Biol. 215, 403-410. 16. Smith, R. F. & Smith, T. F. (1990) Proc. Natl. Acad. Sci. USA 87, 118-122. 17. Trudel, J. & Asselin, A. (1989) Anal. Biochem. 178, 362-366. 18. Takacs, A. M., Denker, J. A., Perrine, K. G., Maroney, P. A. & Nilsen, T. W. (1988) Proc. Natl. Acad. Sci. USA 85, 7932-7936. 19. Bradshaw, H. D. (1990) Nature (London) 345, 299. 20. Rogers, R., Ellis, D. S. & Denham, D. A. (1976) J. Helminthol. 50, 251-257. 21. Fuhrman, J. A. & Piessens, W. F. (1985) Mol. Biochem. Parasitol. 17, 93-104. 22. Sutherland, D. R., Christensen, B. M. & Lasee, B. A. (1986) J. Invertebr. Pathol. 47, 1-7. 23. Phiri, J. & Ham, P. J. (1990) Trans. R. Soc. Trop. Med. Hyg. 84, 462. 24. Gooday, G. W., Brydon, L. J. & Chappell, L. H. (1988) Mol. Biochem. Parasitol. 29, 223-225. 25. Rogers, S., Wells, R. & Rechsteiner, M. (1986) Science 234, 364-368.
Transmission-blocking antibodies recognize microfilarial chitinase in brugian lymphatic filariasis (Brugia malyi/parasitic nematodes/PCR)
JULIET A. FUHRMAN*t, WILLIAM S. LANEt, RANDALL F. SMITH§, WILLY F. PIESSENS*, AND FRANCINE B. PERLERt *Department of Tropical Public Health, Harvard School of Public Health, 665 Huntington Avenue, Boston, MA 02115; tNew England Biolabs, Inc., 32 Tozer Road, Beverly, MA 01915; *Harvard University Microchemistry Facility, 16 Divinity Avenue, Cambridge, MA 02138; and §Molecular Biology Computer Research Resource, Dana-Farber Cancer Institute, 44 Binney Street, Boston, MA 02115
Communicated by Phillips W. Robbins, November 12, 1991
ABSTRACT Brugia malayi is a parasitic nematode that causes lymphatic filariasis in humans. The monoclonal antibody MF1, which mediates clearance of peripheral microfilaremia in a gerbil infection model, recognizes two stage-specific proteins, p70 and p75, in B. malayi microrflariae. cDNA coding for the MF1 antigen was sequenced, and the predicted protein sequence shows significant similarities to chitinases from bacteria and yeast. When microfilarial extracts and purified preparations of the MF1 antigen were tested for chitinase activity, strong bands of chitin-degrading activity comigrated in SDS/PAGE with p70 and p75 and showed a reductiondependent mobility shift characteristic of the MFl antigen. Thus, the MF1 antigen is microfrarial chitinase, which may function to degrade chitin-containing structures in the microfilaria or in its mosquito vector during parasite development and transmission.
MATERIALS AND METHODS Parasite Material. Gerbils (Meriones unguiculatus) were infected intraperitoneally with B. malayi either in our own facilities or at the Filariasis Repository Research Service,
Athens, GA. mf were purified and proteins were extracted as described (3). Poly(A) mRNA was isolated from purified mf according to Cox and Smulian (4). Genomic DNA was made from adult B. malayi as described (5). Protein Purification and Amino Acid Sequencing. The MF1 antigen was partially purified from extracts of mf by DE-52 chromatography as described (6). N-terminal sequence analysis was performed on the combined p70 and p75 following elution with 0.5 M NaCI from DE-52, using an ABI model 477A protein sequencer. The sample was subjected to automated Edman degradation using the program NORMAL-i, which was modified using the manufacturer's recommendations for faster cycle time (37 min) by decreasing dry-down times and increasing reaction cartridge temperature to 530C during coupling. The resultant phenylthiohydantoin amino acid fractions were manually identified using an on-line ABI model 120A HPLC apparatus and Shimadzu CR4A integrator. Trypsin, Asp-N, and Glu-C (sequencing grade, Boehringer Mannheim) were used to digest the separated p70 and p75. Proteolytic digests were performed directly on nitrocellulose according to Aebersold et al. (7), using DE-52-purified material that was separated by SDS/PAGE (8, 9) and transblotted (10). Following proteolytic digestion, peptides were separated by narrow-bore reverse-phase HPLC on a Vydac 2.1 mm x 150 mm C18 column. The gradient employed was a modification of that described by Stone et al. (11). Briefly, where buffer A was 0.06% trifluoroacetic acid/H20 and buffer B was 0.055% trifluoroacetic acid/acetonitrile, a gradient of 5% buffer B at 0 min, 33% buffer B at 63 min, 60%o buffer B at 95 min, and 80% buffer B at 105 min with a flow rate of 0.15 ml/min was used. CNBr peptides were generated by cleavage of the separated p70 and p75 in gel slices following SDS/PAGE, using 7 mg of CNBr per ml in 70%6 formic acid. Cleavage products were then electrophoresed and transblotted to poly(vinylidene difluoride) (Immobilon, Millipore) according to Matsudaira (12), stained briefly with Coomassie blue, excised, and sequenced directly. Glycosidase treatment was performed on total mf extracts, using N-glycanase and 0-glycanase from Genzyme, according to the supplier's protocols. Library Screening, DNA Sequencing, and PCR Amplification. MF1-specific DNA probes were synthesized by PCR amplification of B. malayi genomic DNA using degenerate primers derived from the protein sequence of the N terminus (sense strand, primers P1 and P4) and a tryptic peptide (antisense strand, primers P5 and P6). PCR amplifications were performed using Taq polymerase (Cetus), and amplification products were purified from low-melting-point agarose
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Abbreviation: mf, microfilaria(e). $The sequence reported in this paper has been deposited in the GenBank data base (accession no. M73689).
Lymphatic filariasis afflicts nearly 100 million people worldwide (1). Of the three species of lymphatic filariids that infect humans, Brugia malayi can most readily be maintained in the laboratory by passage between gerbils (another natural host) and a mosquito vector. Thus, B. malayi has been useful for studying pathogenesis and transmission of filariasis at the molecular level. Passive transfer of the monoclonal antibody MF1 was previously shown to mediate the transient clearance of first stage larvae, or microfilariae (mf), from the peripheral blood of gerbils infected with B. malayi (2). The MF1 antibody recognizes two stage-specific proteins (p70 and p75) in extracts of mf that appear only after the mf have matured for several days in the vertebrate host (3). The appearance of the MF1 antigen corresponds with the onset of the parasite's ability to infect the mosquito (3). It was thus of great interest to define more precisely the role of the MF1 antigen in filarial development and transmission. With the expectation that the proteins' sequences might clarify their function, we obtained partial amino acid sequences for p70 and p75 and then used this information to isolate and to sequence the corresponding cDNA.1 We report here that the predicted protein sequence of the MF1 antigen has several regions of significant similarity to bacterial and yeast chitinases. Furthermore, the native MF1 antigen is demonstrated to degrade chitin in vitro.
1548
Immunology: Fuhrman et al. (SeaKem, FMC). The genomic probes were 32P-labeled by nick-translation (13) and used to screen a B. malayi mf cDNA library in Agtll (gift of Timothy Nilsen, Case Western Reserve University, Cleveland). MF1-specific single-stranded cDNA was synthesized from mf mRNA with the cDNA Synthesis Plus kit (Amersham), using an antisense primer (P21) derived from the DNA sequence of phage 4A2 (see Results). First strand cDNA was then PCR amplified using two primer combinations. The cDNA fragment 52.4 was generated using P48 and P21, and fragment 4'V was generated using PSL and P21.2 (see Fig. 3). Oligonucleotides for PCR amplifications and for sequencing primers were synthesized on a Biosearch model 8750 DNA synthesizer or were purchased from New England Biolabs. The cDNA clones and PCR-amplified cDNA products were sequenced in both directions using a nested set of oligonucleotide primers and single-stranded DNA templates (14). All nucleotide sequencing was performed by the dideoxy method using Sequenase Version 2.0 (United States Biochemical) according to the manufacturer's protocol. Computer Sequence Analyses. Protein sequence data base searches were conducted with the GENINFLO BLAST Network Service provided by the National Center of Biotechnology Information, National Library of Medicine, National Institutes of Health (15). Multisequence alignments were performed using the PATTERN-INDUCED MULTI-ALIGNMENT program (R. F. Smith and T. F. Smith, Molecular Biology Computer Research Resource, 1990) based on the pattern construction algorithm previously described (16). Chitinase Activity Assays. Assays for chitinolytic activity were performed in SDS gels cast with glycol chitin according to Trudel and Asselin (17). Extracts of native mf proteins were prepared as described (6), but in the absence of protease inhibitors. Under these conditions, p70 and p75 consistently migrate with a lower molecular mass than previously described for extracts prepared in the presence of protease inhibitors. Following electrophoresis, proteins were renatured in the gel by overnight incubation at 37°C in 1% Triton X-100 (Boehringer Mannheim, 789-704). Areas of digested chitin were visualized as dark bands on a fluorescent background after staining with Calcofluor M2R (17).
RESULTS Protein Sequence of the Native Antigen. The MF1 monoclonal antibody detects two bands in Western blots with mobilities corresponding to approximately 70 and 75 kDa under reducing conditions (6). The HPLC profiles of tryptic digests of p70 and p75 are highly homologous, with most major peaks shared between the two (Fig. 1). Three tryptic peptides, as well as two CNBr peptides, from p75 had sequences identical to peptides of the same mobility from p70 (Fig. 2). N-terminal sequence analysis of a mixture of p70 and p75 generated a single amino acid residue at each cycle (Fig. 2). CNBr peptides derived from the individual proteins generated the same N-terminal sequence (YVRG-YY), suggesting that p70 and p75 are identical at the N terminus. These comparisons indicate a high degree of primary structure similarity between p70 and p75. To examine differences between the primary structures of p70 and p75, we sequenced the double peak (retention time = 67 min) indicated by the arrow in the p75 profile (Fig. 1A) and the corresponding single peak (retention time = 67 min) from the p70 profile. The p70 peak gave a single sequence (GYGGAFIWALDFDDFTGK). The p75 double peak gave this sequence and a secondary sequence in addition (GPYPLLNAISSELEGESEN-). This suggested that the secondary sequence could be part of a peptide unique to the p75 molecule.
Proc. Natl. Acad. Sci. USA 89 (1992)
1549
00A
E 1
-10,
10
20
30
40
50
60
70
20
30 40 50 Time (min)
60
70
60 B 5040
:D30 E 20
10
0-
-10
10
FIG. 1. Vydac C18 HPLC profiles of tryptic digests of isolated p75 (A) and p70 (B). The arrow indicates the mobility of the peptide unique to the p75 map that was partially sequenced (amino acids 377-395, see Fig. 3). mAU, milliabsorbance units.
Isolation of a cDNA Clone Coding for the C Terminus of the MF1 Antigen. Degenerate oligonucleotide primers were synthesized corresponding to the N-terminal region (sense) and a tryptic peptide shared by p70 and p75 (antisense). These primers (P1, P4, P5, and P6, Fig. 3) were used to amplify genomic DNA from B. malayi by PCR. Combinations of P1 and P5 or P6 generated a 2.3-kilobase (kb) amplification product, whereas combinations of P4 and P5 or P6 generated a 2.0-kb product. This suggests the presence of an intron between P1 and P4. The 2.3-kb product was partially sequenced, and the predicted amino acid sequence of one region matched the sequence of a tryptic peptide from p75 (amino acids 207-221) distinct from that used to construct the oligonucleotide primers. This strongly implicated the 2.3-kb N-terminus p75/p70 YVRG-YYTNW AQYRDGEGKF
LPGNIPNGL-
Tryptic peptides, p75 GYGGAFIWAL DFDDFTGK* NNFDGFDLDW EYPVGVAEQH A** KNNFDGFDL IIIGIPMYAQ GWTLDNP' ETVHQEGVGA YMVK GPYPLLNAIS SELEGESEN GTIDGS
GDQWYG -DELGDSKPF E-NDEDTEWS KG GEVSGIGIFN TEFAADYWAS
T-ILYAFAKV D
Tryptic peptides, p7O GYGGAFIWAL DFDDFTGK* NNFDGFDLD- EY** YLK
-IIGIPMYAQ G-TLD' ETVHQEGVGA YMVK
Asp-N peptides, p70 DGSYNVESLG KNF DQWYGY DNPSETAI-A AASRPS-A
Glu-C peptides, p70
LVEAMK
AMKTAFVE
NFDLLFLMSY DLHGS-E TNPAGGTASY -EI S-GK
GVGAYMVKGD Q-YGYDN
CNBr peptides, p75
CNBr peptides, p70
-QY'
YVRG-YYTNW AQYRDGEGKF LPGNI'
YVRG-YYDN-
YSAVT5 YAQG
YSAVTKLRET DPGLKVLLSY G-YNF' YAQGWTLDNP SETAIG KTAFVEEAKT SGKQRLL
FIG. 2. Peptide sequences from p75 and p70. Peptides marked with identical superscripts have equivalent retention times (tryptic peptides) or molecular masses on SDS/PAGE (CNBr peptides).
1550
'C,A|PIe.;f:-40GW~LaFh.A' Immunology: Fuhrman et al.
Proc. Natl. Acad. Sci. USA 89 (1992)
amplified genomic DNA fragment as part ofthe coding region for the MF1 antigen. The 2.3- and 2.0-kb DNA fragments were used to probe a B. malayi microfilarial cDNA library in Agtll. Clone 4A2 contained the largest insert of all hybridizing clones. The insert was sequenced and found to encode seven peptides sequenced from native p75, including the peptide unique to the p75 tryptic profile (Fig. 3). No discrepancies could be found between the amino acid sequences of the native peptides and the deduced sequence of clone 4A2. The open reading frame of 718 nucleotides was followed by a stop codon, a 50-nucleotide untranslated region, and a poly(A) tail. The 4A2 insert encoded sequences from p70 and p75 (Figs. 2 and 3). It was therefore used as a probe for Southernblotted, EcoRI-restricted genomic DNA (from B. malayi) to test the possibility that two separate genes code for p70 and p75. Two bands (1.4 and 3.7 kb) hybridized strongly, and two bands (0.8 and 3.1 kb) hybridized very weakly with the 4A2-derived probe (data not shown). PCR amplification of genomic DNA spanning the region coding for 4A2 generated a single 2.5-kb product, which yielded a 1.4-kb product following digestion with EcoRI (data not shown). Although this suggests that the two strongly hybridizing bands (1.4 and 3.7 kb) are part of a single gene coding for these proteins, we cannot rule out the possibility that the weakly hybridizing bands represent a second gene coding for an MF1-reactive molecule. Sequencing the 5' End of the mRNA by PCR Amplification. None of the isolated cDNA clones contained sequence encoding the N terminus of the MF1 protein. Therefore, the
A
200
1
nt
I
800
600
400
1400
1200
1000
i
I
I
I
I
I
I
a
remainder of the sequence was obtained from uncloned cDNA amplified in two steps by PCR. First strand cDNA was synthesized using mf mRNA and an antisense primer from the 4A2 phage insert (P21, Fig. 3). The cDNA was then amplified by PCR using P48 (corresponding to sequence obtained from the 2.3-kb genomic amplification product) and P21. The sequence of this amplified cDNA (fragment 52.4, Fig. 3) yielded an open reading frame of 814 nucleotides. The final 132 nucleotides overlapped with the 5' end of the 4A2 insert, indicating that the EcoRI site at the beginning of the 4A2 insert was derived from the MF1 cDNA and not from linker addition. The deduced amino acid sequence of 52.4 corresponded to six peptides sequenced from native p75, including the last three residues from the N-terminal sequence. One discrepancy was noted at amino acid residue 156, which was predicted to be a glutamate from the nucleotide sequence but appeared as glutamine in the tryptic peptide. To obtain the 5' terminus of the message, we amplified the first strand cDNA using a forward primer (PSL, Fig. 3) corresponding to the sequence of the trans-spliced leader RNA from B. malayi (18) and a reverse primer from the antisense strand of 52.4 (P21.2). This yielded fragment 4'V (Fig. 3), which contained an open reading frame of 516 nucleotides beginning 28 nucleotides from the 3' end of the spliced leader, with a potential 22-residue signal peptide preceding the N terminus of the mature protein. Sequence Similarity to Chitinase and Enzymatic Activity of Native MF1. A search of the Protein Identification Resource (release 27.0), Swiss-Prot (European Molecular Biology Laboratory, release 17.0), and Translated GenBank (release 64.3)
f
EcoRI
4'V
B
i
A-A'-AG :':
AAE
t
A
4A2
52-4
cDNA
G
T
Y
C
N
W
A
5
P
r
E
G
G
K
S.
F
P
G
-;A -'.--AA"' ,A-A, --;ATA A:,--:A 4; A,e, A ; 4 N A A YA AF F F. '1 i' N , GA
-
--
: : -' N FI . hA;
A
:
r F
N
-
A A
-
V
C'
N
-
K T
T
,
T
"
A
*
*
-A-AAAAA---AAA ALGIA -_L AA K
E
T
N
p
A
X
NF'rGF
A.'-A''"k-GWFT:. AA -.AF AA ;X h. .LA.....AA..A. AA-,A A MAY F L A A
-.'' A
A -AA- 'A-i'7A'-':AAAAAA A:ACA. A' '-A G--AAA AAA'A V PAA D w EA Y V A E H A A mS 1. AY E A M-N L
*f'
A A Y T
I
A ': TAhGA -'A-- :.
;,,T AGA AA';: TGAATTT'f AX;ATC-AATTGGAAA';A-A;:;AAA .-'AXA:,A'A"A'..GAALePPAArATA'AATrAAI GE V S G F A A D Y N A At FL S A F
A"ATA"-'AAALAZ~~~~~j
Y
',A':.:T:'*r,':A AIAAA; 'Ar AM, .......
o-
*'A
''
M
Y
A' ni7' -AA.'g
t
-: AT AS TA A A- 4-AA. G S Y N AG A rA M 4H IA
*AATA
'.AL -GAA '
G ~
S
A
A
~
-a
AA
AAAAA' ':: A::
( A' AA
P
A:
:
'A
A'
A 'A.-.
A
r
A
.'
A .A'A:.sA-.:.I;AA'. A:. :AAhA-
A:r.FAA s
A'.. -AA s
A
A' R
.
AAh . .... 7: A . 1: .A ...
AA:' GAG' 'CA'' -GA' A A: - -A
A A.' -A
V F " :0 &Y G '3 A F I W A L D F DD r F' ~~~~~~~~~~~wr
.';A .'.' A ,A;;.| A.: '--:,AAA- AA..A.:A-`-'a *LA./ AA ': A
A
AA
A AA. .AA.;"A.'' T: A '' A .'-A
sA
T( JG TuT TCATAT..................... T T AC ;('. AA:
:
hA'.A 1 -AA hAA-Ah .A
.
A *A -ACAA GA _ATA ALA
r A 'T;:;1T........... A'CsvA sTfm TAS 'GG
FA
;
A.
's-:
AAAAAT
A:-AA ','A,CAGA'I
C x A AAG:A' A A
'A
G
K
C S
G
G pP
.,AA......................... \A AM' A A...-A 7':A!% -
A
?ALA ;.;- e hA lAAf e ';A
.AA-''AAA'-:-.'-: T Ah'.
::'AA.'.?s*.- ".A
A A-A.
-- -A -
-GA' ?
L
N
A
AAA A
-
-A
-A
r
,A A-
FIG. 3. (A) Alignment of cDNA fragments used to sequence the MF1 message. 4A2 is a Agtll clone. 52-4 and 4'V are cDNA fragments amplified from mRNA using primers P48 and P21 or PSL and P21.2, respectively. Nucleotide numbering begins with the first nucleotide of the initiation codon. (B) Nucleotide sequence and deduced amino acid sequence for the MF1 antigen. Nucleotide numbering corresponds to that in A. Amino acid numbering begins at the initiating methionine, with the mature N terminus beginning at amino acid 23. Lowercase letters on the nucleotide line repre-
sent untranslated regions. Primers used for PCR amplifications are marked as arrows above the nucleotide line, with open arrows denoting degenerate primers derived from protein sequence and closed arrows denoting nondegenerate primers corresponding to known nucleotide sequence. The EcoRI site is overscored on the nucleotide line. The consensus poly(A) signal is underlined. Lowercase letters on the amino acid line represent the predicted signal peptide. The N-terminal sequence derived from the combined p70 and p75 is underlined with dashes ----). Peptides sequenced from native p75 are shaded, and peptides sequenced from native p70 are boxed. A triangle marks the start of the peptide unique to p75
tryptic profile.
Immunology: Fuhrman et al.
1551
Proc. Natl. Acad. Sci. USA 89 (1992)
15 16 30 31 90 75 76 45 46 60 61 ---YVRGCYYTNWAQ YRDGEGKFLPGNIPN GLCTHILYAFAKVDE LGDSKPFEWNDEDTE WSKGMYSAVTKLRET *L* 2 CHI1$BACCI 62/VADIDPTKVTHINYA FADICWNGIHGNPDP SGPNPVTWTCQNEKS QTINVPNGTIVLGDP WIDTGKTFAGDTWDQ PIAGNINQLNKLKQT 3 CHIB$SERMA ------MSTRKAVIG YYFIPTNQINNYTET DTSVVPFPVSNITPA KAKQLTHINFSFLDI NSNLECAWDPATNDA KARDVVNRLTALKAH 4 CHIA$SERMA 173/NFTVDKIPAQNLTHL LYGFIPICGGNGIND SLKEIEGSFQALQRS CQGREDFKISIHDPF AALQKAQKGVTAWDD PYKGNFGQLMALKQA 5 KTXA$KLULA 349/DYCDKKSSTTGAPGT DGCFSNCGYGSTSNV KSSTFKKIAYWLDAK DKLAMDPKNIPNGPY DILHYAFVNINSDFS IDDSAFSKSAFLKVT 1
1 MF1
---------------
1 MF1
NPGLXVLLSYGGYNF G IFTGIA------ KSAQK TERFIKSAl * * S*GG* * l** F * ** NPNLKTIISVGGWTW SN FSDVAA------ -TAAT EVFANSAV NPSLRIMFSIGGWYY ISN LGVSHANYVNAV KTPAA TKFAQSCV HPDLKILPSIGGWTL SD PFFFMG.------- -DK DRFVGSVK SS--XKIPSFGGWDF S PSTYT-IFRNAV KTDQN NTFANNLI
90 2 CHI1$BACCI 3 CHIBSSERMA 4 CHWIASERMA 5 KTXA$KLULA
181
IMFI 2 CHI1$BACCI 3 CHIMBSERMA 4 CHIASSERMA
5 KTXASKLULA
1 MF1 2 CHII$BACCI 3 CHIBSSERMA 4 CHIASSERMA
5 KTXA$KLULA
105 106
195 196
120 121
135 136 150 AFLRKN-NFDGFDLDW **** *DG*D*DW DFLRKY-NFDGVDLDW RIMKDY-GFDGVDIDW
180 151 165 166 EYPV4AEEHAKLVE AMKT--AFV-----E*P * *
EYPVYSGLDGNSKRP EDKQNYTLLLSKIRE
EYPQA .EVDGFIAAL QEIR--TLL-----N
EFLQTWKFFDGVDIDW EFPGC KGANPNLGSP QDGETYVLLMKELRA NFMNKY-NLDGIDLDW EYPGPDIPDIPADD SSSG-SNYL-----T
210 211
225 226 240 DL LFLMSYDLHGS -KN D * *M*YD* G ** KLDAAGAVDGKKY 14L TIASGA TYAAN-- TELAKIAA-W INIMTYDFNGA -KI QQTIADGRQALPYqL TIAGAGC 4FFLSRYY SKLAQIVA-- DY INLMTYDLAGP -KI MLDQLSAETGRKY EL TSAISAC KDKI---- DKVAYNVAQN MDH IFLMSYDFYGAF LKN FLKLLKGKMPSGK' LSIAIPS WYL---- KNFPISDIQNI YMWMTYDIHGI-EY
270 241 255 256 VDLHGKLHPTXGE-- --------------* SAHNAPLNYDPAAS - --------------TNHQAALFGDAAGPT FYNALREANLGWSWE
271 285 286 300 301 315 316 330 -------VSGIGIFN TEFAADYWAS MPK EKIIIGIPMYAQGW LDNPSETAIGAAASR ~ * * K** G * Y***** ** ----AAGVPDANTFN VAAGAQGHLDA GVPA AKLVLGVPFYGRGW GCAQAGNGQYQTCTG ELTRAFPSPFSLTVD -AAVQQH LMMGVPS AKIVMGVPFYGRAF GVSGGNGGQYSSHST ----------RHRLH HGERRECAAG VKP GKIVVGTAMYGRGWI GVNGYQNNIPFTGTH
360 345 346 331 PSSASKTNPAGG--- -------------TA
--EEAKTSGKQRL]L TAAVSA GTI--DG SYNVESLGK* *L * A ***
-----------RKEIE DA--IKMLDK GVKF NKVFGGVANYGRSY MVNTNCYNYGCGFQR
LGHQTALNARPGS-GKANSYINCHTP---
-----------------------------
*
*
GSSVGTWEAGSF --- -------F------DF PGEDPYPNADYWLVG CDECVRDKDPRIASY RAVKGTWENGIV--- --------------D EGGNSRDMTNTP--- --GVLSDSEIIDIDS
5 KTXA$KLULA
361 375 376 390 391 405 406 435 420 421 SYWEICKYLKEGGKE TVHQEGV YMVKGD Q--WYGYDNEETI I KMKWLKEKGYGGAFI WALDFDDF KSCGK * D **IW D** ** * * *** YDLEANYINKNGYTR YWNDTAK YLYNAS NKRFISYDDAES Y KTAYIKSKGLGGAMF WELSGDRNK rL-QNK RQLEQMLQGNYGYQR LWNDKTK YLYHAQ NGLFVTYDDAESF Y KAKYIKQQOLGGVMF WHLGODNRN ----D YRQIASQFMSGEWQY TYDATAE YVFKPS TGDLITFDDARS A KGKYVLDKQLGGLFS WEIDADNGD --LNS SDKKNDRWVDTNTDC IFMKYDGN SSWPK SR-YDLEDMKN F AGTSLWAANYFKHDE WKNDEDD ----D
IMF1
451 465 466 480 481 540 495 496 510 511 525 526 ESENPEITTEEPSIT ETEAYETDETEETSE TEAYDTDETEETSET EATTYDTDETEGQEC PERDGLFPHPTDCHL FIQCANNIAYVMQCP
2 CHI1$BACCI 3 CHIB$SERMA 4 CHIASSERMA
TTAPSVPGNARSTGV TANSVTLAWNASTDN VGVTGYNVYNGANLA TSVTGTTATISGLTA GTSYTFTIKAKDAAG NLSAASNAVTVSTTA DSQLDMGTGLRYTGV G--PGNLPIMTAPAY VPGTTYAQGALVSYQ GYVWQTKWGYITSAP GSDSAWLKVGRLA-- ---------------
S KTXASKLULA
DCKNKAGYDLDNPVY GCRLETAINI I IWNG TESVNTVLNILNDYD NYIKYYEALTRAHYD SVMEKYEKWLFEEDG YYTYYTDVDGDDI II
1
541 555 556 580 ATTFFNDAIKVCDHM TNAPDTCI-------
1MF1 2 CHI1$BACCI 3 CHIBSSERMA 4 CHIASSERMA
MF1
2 CHI1SBACCI 3 CHIB$SERMA 4 CHIASSERMA
5 KTXA$KLULA
450 436 GPYPLLNAISSELEG
*
LKADLPTGGTVPPVD LLAALDRYFNAADYD MNASLGNSAGVQ--TEDPFDEENVYFDVY
QPGGDTQAPTAPTNL ASTAQTTSS ITLSWT\ TPPDKKKRDYIQEKY SFEKEFMMSQNMTEL\
FIG. 4. Amino acid sequence alignment ofthe MF1 protein with three bacterial chitinases and one yeast chitinase: chitinase Al of B. circulans (Swiss-Prot locus CHI1$BACCI), chitinases B and A of S. marcescens (CHIB$SERMA and CHIA$SERMA), and killer toxin RF2 a subunit of K. lactis (KTXA$KLULA). Major regions of primary sequence similarity shared among this set are shown boxed. Identical residues conserved in the set are shown as uppercase letters beneath sequence 1; conservative amino acid groups are indicated by asterisks.
data bases revealed significant sequence similarities between the MF1 protein sequence and several chitinases (E.C. 3.2.1.14). An alignment of the MF1 protein sequence with three bacterial chitinases (chitinase Al of Bacillus circulans, chitinases A and B of Serratia marcescens) and one yeast chitinase [killer toxin RF2 a subunit of Kluyveromyces lactis (19)] shows seven major regions of primary sequence conservation (shown as boxed regions in Fig. 4). Following the last conserved region, the MF1 sequence diverges and contains an acid-rich triplication of 14 amino acids (starting at reference position 466, Fig. 4). To test the MF1 antigen for chitinase activity, extracts of mf were electrophoresed in SDS gels containing glycol chitin (17). Chitinase activity was localized to two prominent bands that copurified with the MF1 antigen in DE-52 chromatography and comigrated with the MF1 antigen in onedimensional SDS/PAGE, demonstrating the reductiondependent shift in mobility characteristic of the MF1 antigen (Fig. 5; ref. 6). Since the predicted molecular mass (=55 kDa) for the translation product of the cDNA sequence is lower than that observed for the enzymatically active bands in SDS/PAGE, we tested the possibility that glycosylation could account for the discrepancy. Following treatment with 0-glycanase, but not N-glycanase, a sizeable shift in molecular mass was observed for p70 and p75 by Western blot (Fig. 5).
DISCUSSION Transmission of lymphatic filariasis in nature depends on the persistence of mf in the peripheral blood of infected verte-
brates and on the capacity of those mf to infect and develop in the insect vector. Female adults of Brugia spp. shed live mf, which retain a remnant of their eggshells as the extracuticular sheath (20). This sheath is maintained while the mf circulate in the vertebrate bloodstream but is lost during the initial phases of development in the arthropod vector. 2-ME ---+ + + ..v|-
+-
+
-
95 -
5.R' E
95 5
-95.5
5-55 -
2 3
4 5 6
7
8
9
10
43
1112 13
FIG. 5. Copurification of microfilarial chitinase with the MF1 antigen. Molecular masses are indicated in kDa. Lanes 1-6, chitinase activity is demonstrated following electrophoresis in glycol chitincontaining SDS/polyacrylamide gels for total microfilarial extract (lanes 1 and 6) and DE-52-purified fractions of extracts eluted with 0.3 M (lanes 2 and 5) and 0.5 M NaCl (lanes 3 and 4), electrophoresed in the absence (lanes 1-3) or presence (lanes 4-6) of 2-mercaptoethanol (2-ME, - or +). Lanes 7-10, silver-stained SDS/PAGE (lanes 7 and 8) and Western blot (lanes 9 and 10) of 0.5 M NaCl fraction electrophoresed in the absence (lanes 8 and 9) and presence (lanes 7 and 10) of 2-mercaptoethanol. Lanes 11-13, total microfilarial extract following sham (lane 11), N-glycanase (lane 12), or 0-glycanase (lane 13) treatment was Western blotted and developed with the MF1 monoclonal antibody.
1552
Proc. Natl. Acad. Sci. USA 89 (1992)
Immunology: Fuhrman et al.
Exsheathment is necessary for subsequent molting and development of the filarial larvae in the vector. We previously demonstrated that the microfilarial sheath contains oligomers of f,1-4-linked N-acetylglucosamine and that sheath morphogenesis can be altered by inhibitors of chitin synthesis (21). It is thus of great interest to find a larval chitinase whose temporal appearance coincides with the onset of microfilarial infectivity for the mosquito. The chitinase could be involved in either the degradation of microfilarial structures, such as the sheath, or the interaction with chitinous mosquito structures that the mf must disrupt or penetrate during infection. The peritrophic membrane is a chitinous structure elaborated in the mosquito midgut following a bloodmeal, but mf appear to escape from the midgut long before this membrane is well formed (22). The possibility that mosquitoes secrete a midgut lectin that can inhibit infection was suggested by Phiri and Ham (23), who demonstrated enhanced microfilarial penetration of the midgut in the presence of N-acetylglucosamine. A microfilarial chitinase could generate inhibitory oligosaccharides and promote infection by a similar mechanism. Chitinase activity was previously described in Onchocerca gibsoni (24). In this filarial species, the activity is present in the gravid female adult and presumably acts to remodel chitin fibrils in the developing eggshell or to free the developing mf from their eggshells prior to birth. The mf of Onchocerca spp. are born without sheaths and therefore do not need a mechanism for exsheathment in the vector. Brugian mf do not hatch from their eggshells in utero but seem to remodel the shell and elongate it ultimately to form the sheath (20). This remodeling may also involve chitinase, but this enzyme should be antigenically distinct from the chitinase sequenced here, as the MF1 antibody does not react with extracts of adult female B. malayi (6). The sequence reported here was compiled from three overlapping sequences: the cDNA clone 4A2 (nucleotides 794-1606), the cDNA PCR product 52.4 (nucleotides 112925), and the cDNA PCR product 4'V (nucleotides -48-516). The 5' and 3' untranslated regions are short, and a poly(A) addition signal (ATAAAA) occurs at nucleotides 1547-1552. In sequencing the 5' terminus of the message, we utilized a primer specific for the spliced leader present on some, but not all, mRNAs in B. malayi. Since this technique selectively amplified SL-containing messages, any versions of the MF1 message not containing this leader would not have been detected. Nevertheless, at least a subpopulation of the messages coding for this antigen does contain this spliced leader and a signal peptide preceding the mature N terminus. Although the complete cDNA sequence was compiled from three separate fragments, the extensive overlap at the nucleotide and peptide levels supports the contiguity of the sequence. Furthermore, a unique 1.5-kb fragment can be amplified from microfilarial cDNA using primers corresponding to the sequence coding for p70/p75 N terminus (forward) and the 3' end of the 4A2 clone upstream of the poly(A) tail (reverse) (M. Southworth, personal communication). Chitinase activity is demonstrated here for both proteins, p70 and p75, recognized by the MF1 antibody. However, the relationship between p70 and p75 is still unclear. Two separate messages could exist for p75 and p70, or posttranslational modification (such as C-terminal truncation) could account for the difference in their apparent molecular masses. Since O-glycanase treatment caused equivalent mobility shifts for both proteins, we cannot at this point ascribe the difference between the two to this particular class of posttranslational modification. The region of the sequence that spans amino acids 377-453, which includes the acid-rich triplication, is characterized by a high PEST score [where PEST indicates proline (P),
glutamic acid (E), serine (S), and threonine (T); ref. 25], suggesting a calcium-binding site. p70 and p75 were previously shown to bind calcium (6), and this region of the protein could be responsible for that property. The translation of the full cDNA predicts a molecular mass of -55 kDa for the encoded protein, which is smaller than the molecular mass for either protein recognized by the MF1 antibody. As demonstrated in Fig. 5, 0-glycosylation accounts for part of this discrepancy. In addition, the very acidic composition of this PEST region could result in anomolous migration in SDS gels. Expression of the recombinant protein from full-length cDNA should resolve this question. As chitin is absent from the vertebrate host, chitin synthesis and processing offer potential targets for chemotherapy, and the enzymes involved in chitin metabolism could be suitable candidates for vaccine development. The demonstration of chitinase activity for the purified MF1 antigen and the availability of the nucleotide sequence coding for the enzyme favor its analysis as a therapeutic target in this and other parasitic helminths. We thank Dr. Sen Dissanayake for expert advice, Sherry Roemer for sequencing the genomic PCR product, and Dr. Maurice W. Southworth for helpful discussions. Peptide preparation and sequencing were performed by Renee Robinson and Ruth Davenport at the Harvard Microchemistry Facility. Materials used in this study were provided by a National Institute of Allergy and Infectious Diseases supply contract (Al 02642), U.S.-Japan Cooperative Medical Science Program. This work was supported in part by a National Institute of Allergy and Infectious Diseases grant (Al 24858) to J.A.F. 1. World Health Organization Expert Committee on Filariasis (1984) Tech. Rep. Ser. No. 702 (WHO, New York). 2. Canlas, M., Wadee, A., Lamontagne, L. & Piessens, W. F. (1984) Am. J. Trop. Med. Hyg. 33, 420-424. 3. Fuhrman, J. A., Urioste, S. S., Hamill, B., Spielman, A. & Piessens, W. F. (1987) Am. J. Trop. Med. Hyg. 36, 70-74. 4. Cox, R. A. & Smulian, N. (1983) FEBS Lett. 155, 73-80. 5. Arasu, P., Philipp, M. & Perler, F. (1987) Exp. Parasitol. 64,
281-291. 6. Fuhrman, J. A. & Piessens, W. F. (1989) Mol. Biochem. Parasitol.
35, 249-258. 7. Aebersold, R. H., Leavitt, J., Saavedra, R. A., Hood, L. E. & Kent, S. B. H. (1987) Proc. Natl. Acad. Sci. USA 84, 6970-6974. 8. Laemmli, U. K. (1970) Nature (London) 227, 680-685. 9. Hunkapillar, M. W., Lujan, E., Ostrander, F. & Hood, L. E. (1983) Methods Enzymol. 91, 227-236. 10. Towbin, M., Staehelin, T. & Gordon, J. (1979) Proc. Natl. Acad. Sci. USA 76, 4350-4354. 11. Stone, K. L., Lopresti, M. B., Williams, N. D., Crawford, J. M., DeAngelis, R. & Williams, K. R. (1989) in Techniques in Protein Chemistry, ed. Hugh, T. (Academic, New York), pp. 377-391. 12. Matsudaira, P. (1987) J. Biol. Chem. 262, 10035-10038. 13. Rigby, P. W. J., Dieckmann, M., Rhodes, C. & Berg, P. (1977) J. Mol. Biol. 113, 237-251. 14. Higuchi, R. G. & Ochman, H. (1989) Nucleic Acids Res. 17, 5865. 15. Altschul, S. F., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J. (1990) J. Mol. Biol. 215, 403-410. 16. Smith, R. F. & Smith, T. F. (1990) Proc. Natl. Acad. Sci. USA 87, 118-122. 17. Trudel, J. & Asselin, A. (1989) Anal. Biochem. 178, 362-366. 18. Takacs, A. M., Denker, J. A., Perrine, K. G., Maroney, P. A. & Nilsen, T. W. (1988) Proc. Natl. Acad. Sci. USA 85, 7932-7936. 19. Bradshaw, H. D. (1990) Nature (London) 345, 299. 20. Rogers, R., Ellis, D. S. & Denham, D. A. (1976) J. Helminthol. 50, 251-257. 21. Fuhrman, J. A. & Piessens, W. F. (1985) Mol. Biochem. Parasitol. 17, 93-104. 22. Sutherland, D. R., Christensen, B. M. & Lasee, B. A. (1986) J. Invertebr. Pathol. 47, 1-7. 23. Phiri, J. & Ham, P. J. (1990) Trans. R. Soc. Trop. Med. Hyg. 84, 462. 24. Gooday, G. W., Brydon, L. J. & Chappell, L. H. (1988) Mol. Biochem. Parasitol. 29, 223-225. 25. Rogers, S., Wells, R. & Rechsteiner, M. (1986) Science 234, 364-368.
