85
Molecular and Biochemical Parasitology, 55 (1991) 85-94 © 1992 Elsevier Science Publishers B.V. All rights reserved. / 0166-6851/92/$05.00 MOLBIO 01806
Neutralization-sensitive merozoite surface antigens of Babesia bovis encoded by members of a polymorphic gene family Stephen A. Hines a'b* Guy H. Palmer b, Douglas P. Jasmer b, Travis C. McGuire b and Terry F. McElwain a'b* aDepartment of Infectious Diseases, University of Florida, Gainesville, FL, USA; and bDepartment of Veterinary Microbiology and Pathology, Washington State University, Pullman, WA, USA (Received 6 February 1992; accepted 11 June 1992)
Monospecific antibodies against native and recombinant versions of the major merozoite surface antigen (MSA-1) of
Babesia bovis neutralize the infectivity of merozoites from Texas and Mexico strains in vitro. Sequence analysis shows that MSA-1 and a related, co-expressed 44 kDA merozoite surface protein (MSA-2) are encoded by members of a multigene family previously designated BabR. BabR genes, originally described in Australia strains of B. bovis, are notable because their marked polymorphism is apparently mediated by chromosomal rearrangements, but protein products of BabR genes have not previously been identified. The 3' terminal 173 nucleotides of the MSA-1 gene, including 60 nucleotides of untranslated sequence, are highly similar to the 3' terminal sequences of BabR 0.8 (84% identity) and MSA-2 (94% identity). Alignment of the predicted protein sequences demonstrates significant overall homology between MSA-I and MSA-2, and between both proteins and the amino terminal BabR sequence. MSA-1 nucleic acid probes also hybridize weakly to genomic DNA from the Australia 'L' strain, even though this strain does not express merozoite surface epitopes cross-reactive with MSA-1 or MSA-2. Hybridization of these same probes to genomic DNA from the cloned Mexico strain reveals a pattern of bands compatible with two copies each of MSA-1 and MSA-2. Proteins encoded by this B. bovis gene family have been designated variable merozoite surface antigens (VMSA). The extent and mechanism of VMSA polymorphism among strains will be important when evaluating the role these surface proteins have in the host-parasite interaction, including immunity to blood stages. Key words: Gene rearrangement; Antigenic diversity; BabR; MSA-1; MSA-2
Introduction Correspondence address. StephenA. Hines, Dept. of Veterinary Microbiology and Pathology, Washington State University. Pullman, WA 99164-7040, USA. Tel.: (509) 3356069; Fax (509) 335-6094.
*Present address: Dept. of Veterinary Microbiology and Pathology, Washington State University, Pullman, WA, USA. Abbreviations: MSA-1, merozoite surface antigen 1; MSA-2, merozoite surface antigen 2; VMSA, variable merozoite surface antigens; PBS, phosphate-buffered saline; SDSPAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Note: Nucleotide sequence data reported in this paper have been submitted to the GenBankT M data base with the accession number M77192.
Surface exposed epitopes of merozoite proteins that mediate attachment, invasion and early intraceUular survival are potential targets of the protective immune response to bovine babesiosis. We have previously shown that cattle protected against a cloned strain of Babesia boris produce high titered antibodies that are preferentially directed against an integral membrane glycoprotein designated merozoite surface antigen 1 (MSA-1, formerly Bv42) [1]. This observation is important since antibody is considered a primary mediator of the protective immune response [2]. Two critical considerations for development of MSA-1 as a component of a non-infectious
86 vaccine are the ability of monospecific antibody against MSA-1 to neutralize infectivity of merozoites and the extent of MSA-1 polymorphism among strains. Recently, we have demonstrated that surface exposed B-lymphocyte epitopes of MSA-1 and a co-expressed 44-kDa merozoite surface protein (MSA-2) are not conserved in strains from Australia and Israel [3]. In this paper, we report that monospecific antibody to recombinant and native MSA-1 neutralize the infectivity of homologous merozoites for bovine erythrocytes and analyze the genetic basis for antigenic diversity of MSA-1 among strains. Sequence analysis reveals that MSA-1 and MSA-2 are members of a highly polymorphic variable merozoite surface antigen (VMSA) gene family that includes genes earlier designated BabR [4,5]. BabR genes are notable because of their marked polymorphism, but the function of the encoded proteins was previously unknown. The extent of polymorphism among strains and the mechanism for generating antigenic variants will be important to define the function of VMSAs and determine their potential as vaccine antigens.
Materials and Methods
Parasites and cDNA Library. The origin, in vitro propagation and metabolic radiolabeling of the biologically cloned Mexico (Mo7) and the uncloned Texas, Australia 'L', Australia'S' and Israel strains of B. bovis have been described previously [1,3]. Production of the blood stage cDNA expression library from asynchronous cultures of Mo7 was also described previously [6] Immunosereening. Plaque lifts onto isopropyl thiogalactopyranoside-soaked nitrocellulose were screened using monospecific rabbit anti MSA-1 antiserum (R-914) followed by ~2~IProtein A and autoradiography [7,8]. Rabbit R-914 was immunized with native MSA-1 protein (Mexico strain) immunoaffinity purified using Babb 35A4, a previously described
monoclonal antibody [9]. Positive plaques were tested for reactivity with normal rabbit serum, two monoclonal antibodies (Babb35A4 and 23.10.36) that recognize surface-exposed epitopes of MSA-1 [8,9], and an isotype control monoclonal antibody. The cloned insert in plaque purified 2 phage was subcloned into Bliaescript SK phagemid using the in vivo excision capabilities of 2Zap II [10]. Recombinant phagemid (pBv42) was tested for expression using colony lifts from transformed, ampicillin resistant Escherichia coli (XL1-Blue strain) [7]. Immunoblotting. E. coli host strain XL1-Blue containing pBv42 phagemid, a phagemid encoding an unrelated Anaplasma marginale antigen, or no phagemid were grown in liquid culture to A6o0=0.9-1.2. Bacteria washed 3 times in phosphate-buffered saline (PBS), pH 7.4, were lysed in PBS by sonication and several cycles of freeze/thaw. The resulting bacterial lysate was cleared by ultracentrifugation, filtered, and sonicated again before freezing in aliquots at - 7 0 ° C [11]. Lysates from equivalent numbers of bacteria and from 2.5× 10 7 B. boris infected erythrocytes were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transblotted to nitrocellulose filters [1,12,13]. Nitrocellulose filters were probed as described for immunoscreening. Antisera against recombinant MSA-1. A lysate of E. coli containing pBv42 phagemid was prepared as described above. Two rabbits (R-S21 and R-$22) were immunized 4 times with lysate that had been cleared by ultracentrifugation as above and emulsified in Freund's adjuvant. Two additional rabbits (R-S19, R$20) were immunized with a crude lysate that was not cleared by centrifugation before emulsification. Antibodies to control E. coli were produced by immunizing rabbits with a lysate of XL1-Blue without Bluescript phagemid (R-126) or XL1-Blue with Bluescript phagemid encoding an unrelated A. marginale antigen (R-108), prepared without ultracentrifugation as described above.
87 Radioimmunoprecipitation / SDS-PAGE / Indirect immunofluorescence. Radiolabeled B. bovis antigen was processed and immunoprecipitated using Protein A-bearing Staphylococcus aureus as previously described [11,14]. The immunoprecipitates were electrophoresed in 7.5 to 17.5% gradient SDS-polyacrylamide gels and detected by fluorography [8]. Antibodies were tested for reactivity with the surface of live merozoites by epifluorescent microscopy using a described method [8,14]. In vitro neutral&ation of merozoites. In vitro parasitemia was maximized by a previously described method [15]. Free merozoites were collected by differential centrifugation [16], washed twice in Puck's saline G, and shown to be 95-99% viable by retention of 6-carboxy fluorescein diacetate stain (6-CFDA) [14]. 5 × 105 merozoites were incubated in microtiter wells at 4°C for 30 rain with heat inactivated (56°C for 30 rain) test sera diluted 1:5 in complete medium. Complete medium consists of M-199 (Gibco) supplemented with 40% normal bovine serum [17]. An equal volume of complete media (100/A) containing 5% packed normal bovine erythrocytes were then added to each well and the plates incubated at 37°C, 5% O2 and 5% CO2 for 96 h. Sera were tested in triplicate for the ability to neutralize merozoites from the Texas strain (Expt. 1) and the Mo7 Mexico strain (Expt. 2). Giemsa-stained slides were made from each well and read by two independent observers with no knowledge of the experimental design. Each observer counted infected cells and total erythrocytes in 20 representative fields (a minimum of 4000 erythrocytes). Merozoites were also collected following incubation with antiserum, washed, incubated with 6-CFDA and examined by epifluorescent microscopy to determine if antiserum caused agglutination or lysis. DNA sequencing / sequence analysis. The cDNA insert was sequenced as double-stranded DNA using sequentially derived primers initiating dideoxy chain reactions [6,18] (Sequenase Version 2.0 kit, USB, Cleveland,
OH, USA). Sequence analysis was done using the Genetics Computer Group (GCG) package of the University of Wisconsin, version 6.2 on a VAX 11/785 computer [19] The FASTP program was used to compare the MSA-1 sequence to known sequences in the NBRF library and to MSA-2, the gene encoding a 44kDa surface protein co-expressed on the biologically cloned Mexico isolate of B. bovis [3,8,20]. FASTP scores regions of homology by comparing paired amino acids using the replaceability matrix PAM250 [21,22]. The sequences are ranked by their initial score, and an optimized score allowing for insertions and deletions is computed. The significance of the FASTP relationship was evaluated with the related program RDF, by comparing FASTP scores to the mean and standard deviation of 50 randomly permuted versions of the potentially related sequence [21]. Southern blots / RNA probes. A pBv42 [32p]RNA probe prepared using T7 RNA polymerase (Riboprobe System, Promega, Madison, WI, USA) was hybridized to restriction endonuclease digested genomic DNA as described previously [6,23]. Region specific RNA probes were made by linearizing the phagemid with an enzyme (AMID that cuts near the junction of the highly conserved 3' sequence and the 5' region which varies among MSA-1, MSA-2 and BabR 0.8 (see Fig. 3). A 3' specific 32p-labeled RNA probe was transcribed using T7 RNA polymerase and the T7 promoter located just 3' to the insert. A 5' probe was generated using T3 RNA polymerase and the T3 promoter at the other end of the insert. Results
cDNA clone encoding MSA-1. A plaque purified clone, designated 2rBv42, was identified by specific reactivity with monospecific rabbit antiserum to MSA-1. E. coli transformed with in vivo excised recombinant phagemid expressed a protein that was recognized by two surface reactive monoclo-
88
hal antibodies against native MSA-1 (Babb 35A4 and 23.10.36) [8,9] and which migrated in SDS-polyacrylamide gels at an apparent size of 45 kDa, slightly larger than the native glycoprotein (data not shown). None of the MSA-1 reactive antibodies bound to control recombinant bacteria, and none of the control antibodies recognized the recombinant clone. To determine if the recombinant protein reproduced B-lymphocyte epitopes of native MSA-1, 4 rabbits were immunized with bacterial lysates containing recombinant MSA-1 (rMSA-1). Antibodies from all 4 rabbits immunoprecipitated a 42-kDa native protein that was metabolically labeled in infected erythrocyte cultures with [35S]methionine (Fig. 1) or [3H]glucosamine, or was radioiodinated on live merozoites using a surface specific method [1,11]. Rabbit antiserum against rMSA-1 also reacted specifically with the surface of live merozoites by
indirect immunofluorescence (titer 1000). To confirm that the 42-kDa surface glycoprotein immunoprecipitated by anti-recombinant serum was native MSA-1, rabbit antibodies to rMSA-1 were shown to completely inhibit binding of the monoclonal antibody Babb 35A4 to native MSA-1 in immunoprecipitation (data not shown) [8]. In vitro neutralization. Since parasite surface proteins are postulated to mediate host cell recognition and invasion, antibody to MSA-1 was tested for the ability to inhibit merozoite infectivity for bovine erythrocytes. Monospecific antiserum against both native and recombinant MSA-1 significantly inhibited infectivity of the Texas and Mo7 strains in vitro (Table I). Examination of merozoites after incubation with antiserum confirmed that the mechanism of inhibition was not agglutination or lysis of merozoites. sequence. The phagemid encoding recombinant MSA-1 contained a 0.95-kb cDNA insert that was excised as a single fragment with EcoRI. The DNA sequence of D NA
1
2
3
4
5
6
7
8
9
10
< 200
< 92.5 69
TABLE 1 In vitro neutralization of B. bovis merozoites using monospecific antisera to native and recombinant MSA-I Experiment No. a
-< 4 6
I. Bovine sera Group A (n = 5)
< 30
Group B (n = 5)
-~ 1 4 . 3
Fig. 1. Antibodies to recombinant MSA-1 immunoprecipitare a 42-kDa native B. bovis antigen. A [3SS]methioninelabeled 42-kDa protein was immunoprecipitated using serum from rabbits immunized with recombinant MSA-I (lanes 4,6,8,10). No parasite antigen was precipitated using preimmunization sera from the same rabbits (lanes 3,5,7,9) or sera from rabbits immunized with lysates of E. coli host strain XLI-Blue (R-126, lane 2) or XLI-Blue containing a recombinant phagemid encoding an unrelated A. marginale protein (R-108, lane 1). Rabbit serum defined in this experiment were also used for in vitro neutralization (Table I).
lmmunogen Mean PPE b Level of (S.D.) significance nMSA-1
1.04 (.90)
Ovalbumin
3.69 (.47)
rMSAI
0,33 (.21)
G r o u p C pre (n - 4)
None
1,83 (.65)
Group D (n = 4)
E. coli lysate 1.73 (1.3)
2. Rabbit sera Group C (n = 4)
P < 0.001 ~
P = 0,017 d P = 0.04 c
Expt. 1, Texas strain merozoites; Expt. 2, = Mo7 strain. b Percentage of parasitized erythrocytes (96 h after initiation of in vitro infection). c Two-sample t-test (one tailed). o Paired t-test (one tailed)
89
position 72 but is incomplete at the 5' end and does not include a predicted N-terminal signal peptide [27]. Five adenosine residues at the 3' end of the insert suggests a poly(A) tail.
the cDNA insert is shown in Fig. 2. A single open reading frame starting at nucleotide position 3 is in frame with fl-galactosidase, encoding a predicted fusion protein of 347 amino acids. A 17 amino acid hydrophobic region at the carboxy-terminus [25] of the encoded protein is compatible with signal sequences for attachment of a glycosyl-phosphaditylinositol anchor, consistent with the post-translational modification of native MSA-1 [1,26]. The sequence has a potential N-linked glycosylation site at amino acid
Homology of MSA-1 with BabR and MSA2. The MSA-1 sequence was analyzed for similarities to known sequences. The 3' terminal 173 nucleotides of the MSA-1 gene, including 60 nucleotides of untranslated sequence, were highly similar to the 3' sequences of two previously described B. boris genes,
MSA-I
CTTCAATCGTCCTTCCCGAAGGATCATTCTACGATGACATGTCTAAGTTCTACGGTGCTGTTGGAAGTTTCGACCAGACCAAATTGTATAG•GTTCTTTC
100
MSA-I
TG•TAACTTCAAAGCCG•TAAAATGGATGATCAGAAGGTAAAAGACACATTCAAAAATTTATACAAAGTCAACGCATTGATAAAGAACAATC•TATGATT
200
MSA-I
•G••CTGATCTATTTAATGCAA•TATTGTTAGCGGTTTTT•AACTAAGAATGA•GAGGAAAAATTCAATGCTATATTTGATTCCATTAAGGGAATGTACT
300
MSA-I
ATAGAGCT•AACACATGGACAAATATTTGAAGTCACTAAGGTGGAATACTGATATTGTTGAGGAAGAT•GTGAGAAGGCAGTTGAATATTTCAAGAAG•A
400
MSA-I
TGTTTATACGGGGGAACACGTTGTTGACGTCAACGGTATGGCTGGTGTTTGCAAGGAGTTTTTAAGCCCGGCCT•TGATTTCTACAAA•TTGTTGAGTCT
500
MSA-I
TTTGATGCGTTTGCACATGCTAAGGTGCACGCTCAAGTAGGAAATTTTGTTAAA••TGGAACTGACATCGCTCCTCCTAAGGATGTTACTGATG•ATTAG
600
MSA-I
AAAAGGAATTGCAAGAG•AAAAACCTGCA•GAAGTGAGAG•AC•GAAGTACCCGCTCCAGGTGATG•ATCTGGCGTC•AA•AACCGCCTGCATCAGGAAC
700
vAvell
/Begin
conserved
sequence
3'
MSA-I MSA'2 BabR 0.8
ATCCCCGCAAGGACCTGCT•CGACTACACC^CAG•CCATCTCCAGAGT•CTCAGGAAACCTCCAAGGACAACAGGGTACAACCAAGCCAGCCGGATCTT• TGAAAGC-C-TCC-AA---GACCACC--A-^--AA--TA---AGACAC-TGA---T ............................................ - A - - T G T - - - A C - - - G - A G G A - G T - G A - - - A - - A .... G-G-T . . . . . . . . T---A-T ..... G ...... T--G .......... T ........
799 (895) (636)
MSA-I MSA-2 BabR 0.8
TTTCACCTATGGCGGATTGACTGTGGCTACTCTCTGCTACTTCG^TT~TCT~TGCATTTTA.-~AAAACTAATGGTAGTGACACAATAGTTTTGTAAACTCAT ........................... C ............................... TAATAA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
898 (995)
A. . . . . . .
(736)
MSA-I MSA-2
BabR 0.8
MSA-2
T-CCG . . . . . .
T. . . . . . . . . .
(:::1.I..
:
I
I1.., . : l ' " , ., - - . J .
.. . . . . .
.
.
:::
T--
I . . . , . - - : :.
.
.
.
.
::
I . . . . .
:
-,-.
II II
:
.
.
.
.
.
.
.
.
.
.
.
I I
.
. . . . . . . . . . . .
I
125t
. . . . . . . . . . . . . .
....
I
.lll.
I..
I ....
II
100
I.II
MEAQSATETQKNLKTLMELIKTKRPFKSSDFDTLNLDYLSGQSNEELFKLLIDAINGMKE
BabR 0 . 8
ML•EFNAFLNDNP.••HMLTNGKEKMTEYYKKN•SKEDGEVK•YKTMVKFCNDFL•SKSPFMRLYKHLNEY•ELVKKKPAQE•SPA•SSPQRPAETQQTQ :
MSA-I
T. . . . . . . . . . . . . .
•IVLPEG•FYDDMSKFYGAVG•FDQTKLY•VL•AHFKAAKM•DQKVKDTFKNLYKVNALIKNNPMIRPDLFNATIVsGFsTKNDEEKFNA•FD•IKGMYY
Molecular and Biochemical Parasitology, 55 (1991) 85-94 © 1992 Elsevier Science Publishers B.V. All rights reserved. / 0166-6851/92/$05.00 MOLBIO 01806
Neutralization-sensitive merozoite surface antigens of Babesia bovis encoded by members of a polymorphic gene family Stephen A. Hines a'b* Guy H. Palmer b, Douglas P. Jasmer b, Travis C. McGuire b and Terry F. McElwain a'b* aDepartment of Infectious Diseases, University of Florida, Gainesville, FL, USA; and bDepartment of Veterinary Microbiology and Pathology, Washington State University, Pullman, WA, USA (Received 6 February 1992; accepted 11 June 1992)
Monospecific antibodies against native and recombinant versions of the major merozoite surface antigen (MSA-1) of
Babesia bovis neutralize the infectivity of merozoites from Texas and Mexico strains in vitro. Sequence analysis shows that MSA-1 and a related, co-expressed 44 kDA merozoite surface protein (MSA-2) are encoded by members of a multigene family previously designated BabR. BabR genes, originally described in Australia strains of B. bovis, are notable because their marked polymorphism is apparently mediated by chromosomal rearrangements, but protein products of BabR genes have not previously been identified. The 3' terminal 173 nucleotides of the MSA-1 gene, including 60 nucleotides of untranslated sequence, are highly similar to the 3' terminal sequences of BabR 0.8 (84% identity) and MSA-2 (94% identity). Alignment of the predicted protein sequences demonstrates significant overall homology between MSA-I and MSA-2, and between both proteins and the amino terminal BabR sequence. MSA-1 nucleic acid probes also hybridize weakly to genomic DNA from the Australia 'L' strain, even though this strain does not express merozoite surface epitopes cross-reactive with MSA-1 or MSA-2. Hybridization of these same probes to genomic DNA from the cloned Mexico strain reveals a pattern of bands compatible with two copies each of MSA-1 and MSA-2. Proteins encoded by this B. bovis gene family have been designated variable merozoite surface antigens (VMSA). The extent and mechanism of VMSA polymorphism among strains will be important when evaluating the role these surface proteins have in the host-parasite interaction, including immunity to blood stages. Key words: Gene rearrangement; Antigenic diversity; BabR; MSA-1; MSA-2
Introduction Correspondence address. StephenA. Hines, Dept. of Veterinary Microbiology and Pathology, Washington State University. Pullman, WA 99164-7040, USA. Tel.: (509) 3356069; Fax (509) 335-6094.
*Present address: Dept. of Veterinary Microbiology and Pathology, Washington State University, Pullman, WA, USA. Abbreviations: MSA-1, merozoite surface antigen 1; MSA-2, merozoite surface antigen 2; VMSA, variable merozoite surface antigens; PBS, phosphate-buffered saline; SDSPAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Note: Nucleotide sequence data reported in this paper have been submitted to the GenBankT M data base with the accession number M77192.
Surface exposed epitopes of merozoite proteins that mediate attachment, invasion and early intraceUular survival are potential targets of the protective immune response to bovine babesiosis. We have previously shown that cattle protected against a cloned strain of Babesia boris produce high titered antibodies that are preferentially directed against an integral membrane glycoprotein designated merozoite surface antigen 1 (MSA-1, formerly Bv42) [1]. This observation is important since antibody is considered a primary mediator of the protective immune response [2]. Two critical considerations for development of MSA-1 as a component of a non-infectious
86 vaccine are the ability of monospecific antibody against MSA-1 to neutralize infectivity of merozoites and the extent of MSA-1 polymorphism among strains. Recently, we have demonstrated that surface exposed B-lymphocyte epitopes of MSA-1 and a co-expressed 44-kDa merozoite surface protein (MSA-2) are not conserved in strains from Australia and Israel [3]. In this paper, we report that monospecific antibody to recombinant and native MSA-1 neutralize the infectivity of homologous merozoites for bovine erythrocytes and analyze the genetic basis for antigenic diversity of MSA-1 among strains. Sequence analysis reveals that MSA-1 and MSA-2 are members of a highly polymorphic variable merozoite surface antigen (VMSA) gene family that includes genes earlier designated BabR [4,5]. BabR genes are notable because of their marked polymorphism, but the function of the encoded proteins was previously unknown. The extent of polymorphism among strains and the mechanism for generating antigenic variants will be important to define the function of VMSAs and determine their potential as vaccine antigens.
Materials and Methods
Parasites and cDNA Library. The origin, in vitro propagation and metabolic radiolabeling of the biologically cloned Mexico (Mo7) and the uncloned Texas, Australia 'L', Australia'S' and Israel strains of B. bovis have been described previously [1,3]. Production of the blood stage cDNA expression library from asynchronous cultures of Mo7 was also described previously [6] Immunosereening. Plaque lifts onto isopropyl thiogalactopyranoside-soaked nitrocellulose were screened using monospecific rabbit anti MSA-1 antiserum (R-914) followed by ~2~IProtein A and autoradiography [7,8]. Rabbit R-914 was immunized with native MSA-1 protein (Mexico strain) immunoaffinity purified using Babb 35A4, a previously described
monoclonal antibody [9]. Positive plaques were tested for reactivity with normal rabbit serum, two monoclonal antibodies (Babb35A4 and 23.10.36) that recognize surface-exposed epitopes of MSA-1 [8,9], and an isotype control monoclonal antibody. The cloned insert in plaque purified 2 phage was subcloned into Bliaescript SK phagemid using the in vivo excision capabilities of 2Zap II [10]. Recombinant phagemid (pBv42) was tested for expression using colony lifts from transformed, ampicillin resistant Escherichia coli (XL1-Blue strain) [7]. Immunoblotting. E. coli host strain XL1-Blue containing pBv42 phagemid, a phagemid encoding an unrelated Anaplasma marginale antigen, or no phagemid were grown in liquid culture to A6o0=0.9-1.2. Bacteria washed 3 times in phosphate-buffered saline (PBS), pH 7.4, were lysed in PBS by sonication and several cycles of freeze/thaw. The resulting bacterial lysate was cleared by ultracentrifugation, filtered, and sonicated again before freezing in aliquots at - 7 0 ° C [11]. Lysates from equivalent numbers of bacteria and from 2.5× 10 7 B. boris infected erythrocytes were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transblotted to nitrocellulose filters [1,12,13]. Nitrocellulose filters were probed as described for immunoscreening. Antisera against recombinant MSA-1. A lysate of E. coli containing pBv42 phagemid was prepared as described above. Two rabbits (R-S21 and R-$22) were immunized 4 times with lysate that had been cleared by ultracentrifugation as above and emulsified in Freund's adjuvant. Two additional rabbits (R-S19, R$20) were immunized with a crude lysate that was not cleared by centrifugation before emulsification. Antibodies to control E. coli were produced by immunizing rabbits with a lysate of XL1-Blue without Bluescript phagemid (R-126) or XL1-Blue with Bluescript phagemid encoding an unrelated A. marginale antigen (R-108), prepared without ultracentrifugation as described above.
87 Radioimmunoprecipitation / SDS-PAGE / Indirect immunofluorescence. Radiolabeled B. bovis antigen was processed and immunoprecipitated using Protein A-bearing Staphylococcus aureus as previously described [11,14]. The immunoprecipitates were electrophoresed in 7.5 to 17.5% gradient SDS-polyacrylamide gels and detected by fluorography [8]. Antibodies were tested for reactivity with the surface of live merozoites by epifluorescent microscopy using a described method [8,14]. In vitro neutral&ation of merozoites. In vitro parasitemia was maximized by a previously described method [15]. Free merozoites were collected by differential centrifugation [16], washed twice in Puck's saline G, and shown to be 95-99% viable by retention of 6-carboxy fluorescein diacetate stain (6-CFDA) [14]. 5 × 105 merozoites were incubated in microtiter wells at 4°C for 30 rain with heat inactivated (56°C for 30 rain) test sera diluted 1:5 in complete medium. Complete medium consists of M-199 (Gibco) supplemented with 40% normal bovine serum [17]. An equal volume of complete media (100/A) containing 5% packed normal bovine erythrocytes were then added to each well and the plates incubated at 37°C, 5% O2 and 5% CO2 for 96 h. Sera were tested in triplicate for the ability to neutralize merozoites from the Texas strain (Expt. 1) and the Mo7 Mexico strain (Expt. 2). Giemsa-stained slides were made from each well and read by two independent observers with no knowledge of the experimental design. Each observer counted infected cells and total erythrocytes in 20 representative fields (a minimum of 4000 erythrocytes). Merozoites were also collected following incubation with antiserum, washed, incubated with 6-CFDA and examined by epifluorescent microscopy to determine if antiserum caused agglutination or lysis. DNA sequencing / sequence analysis. The cDNA insert was sequenced as double-stranded DNA using sequentially derived primers initiating dideoxy chain reactions [6,18] (Sequenase Version 2.0 kit, USB, Cleveland,
OH, USA). Sequence analysis was done using the Genetics Computer Group (GCG) package of the University of Wisconsin, version 6.2 on a VAX 11/785 computer [19] The FASTP program was used to compare the MSA-1 sequence to known sequences in the NBRF library and to MSA-2, the gene encoding a 44kDa surface protein co-expressed on the biologically cloned Mexico isolate of B. bovis [3,8,20]. FASTP scores regions of homology by comparing paired amino acids using the replaceability matrix PAM250 [21,22]. The sequences are ranked by their initial score, and an optimized score allowing for insertions and deletions is computed. The significance of the FASTP relationship was evaluated with the related program RDF, by comparing FASTP scores to the mean and standard deviation of 50 randomly permuted versions of the potentially related sequence [21]. Southern blots / RNA probes. A pBv42 [32p]RNA probe prepared using T7 RNA polymerase (Riboprobe System, Promega, Madison, WI, USA) was hybridized to restriction endonuclease digested genomic DNA as described previously [6,23]. Region specific RNA probes were made by linearizing the phagemid with an enzyme (AMID that cuts near the junction of the highly conserved 3' sequence and the 5' region which varies among MSA-1, MSA-2 and BabR 0.8 (see Fig. 3). A 3' specific 32p-labeled RNA probe was transcribed using T7 RNA polymerase and the T7 promoter located just 3' to the insert. A 5' probe was generated using T3 RNA polymerase and the T3 promoter at the other end of the insert. Results
cDNA clone encoding MSA-1. A plaque purified clone, designated 2rBv42, was identified by specific reactivity with monospecific rabbit antiserum to MSA-1. E. coli transformed with in vivo excised recombinant phagemid expressed a protein that was recognized by two surface reactive monoclo-
88
hal antibodies against native MSA-1 (Babb 35A4 and 23.10.36) [8,9] and which migrated in SDS-polyacrylamide gels at an apparent size of 45 kDa, slightly larger than the native glycoprotein (data not shown). None of the MSA-1 reactive antibodies bound to control recombinant bacteria, and none of the control antibodies recognized the recombinant clone. To determine if the recombinant protein reproduced B-lymphocyte epitopes of native MSA-1, 4 rabbits were immunized with bacterial lysates containing recombinant MSA-1 (rMSA-1). Antibodies from all 4 rabbits immunoprecipitated a 42-kDa native protein that was metabolically labeled in infected erythrocyte cultures with [35S]methionine (Fig. 1) or [3H]glucosamine, or was radioiodinated on live merozoites using a surface specific method [1,11]. Rabbit antiserum against rMSA-1 also reacted specifically with the surface of live merozoites by
indirect immunofluorescence (titer 1000). To confirm that the 42-kDa surface glycoprotein immunoprecipitated by anti-recombinant serum was native MSA-1, rabbit antibodies to rMSA-1 were shown to completely inhibit binding of the monoclonal antibody Babb 35A4 to native MSA-1 in immunoprecipitation (data not shown) [8]. In vitro neutralization. Since parasite surface proteins are postulated to mediate host cell recognition and invasion, antibody to MSA-1 was tested for the ability to inhibit merozoite infectivity for bovine erythrocytes. Monospecific antiserum against both native and recombinant MSA-1 significantly inhibited infectivity of the Texas and Mo7 strains in vitro (Table I). Examination of merozoites after incubation with antiserum confirmed that the mechanism of inhibition was not agglutination or lysis of merozoites. sequence. The phagemid encoding recombinant MSA-1 contained a 0.95-kb cDNA insert that was excised as a single fragment with EcoRI. The DNA sequence of D NA
1
2
3
4
5
6
7
8
9
10
< 200
< 92.5 69
TABLE 1 In vitro neutralization of B. bovis merozoites using monospecific antisera to native and recombinant MSA-I Experiment No. a
-< 4 6
I. Bovine sera Group A (n = 5)
< 30
Group B (n = 5)
-~ 1 4 . 3
Fig. 1. Antibodies to recombinant MSA-1 immunoprecipitare a 42-kDa native B. bovis antigen. A [3SS]methioninelabeled 42-kDa protein was immunoprecipitated using serum from rabbits immunized with recombinant MSA-I (lanes 4,6,8,10). No parasite antigen was precipitated using preimmunization sera from the same rabbits (lanes 3,5,7,9) or sera from rabbits immunized with lysates of E. coli host strain XLI-Blue (R-126, lane 2) or XLI-Blue containing a recombinant phagemid encoding an unrelated A. marginale protein (R-108, lane 1). Rabbit serum defined in this experiment were also used for in vitro neutralization (Table I).
lmmunogen Mean PPE b Level of (S.D.) significance nMSA-1
1.04 (.90)
Ovalbumin
3.69 (.47)
rMSAI
0,33 (.21)
G r o u p C pre (n - 4)
None
1,83 (.65)
Group D (n = 4)
E. coli lysate 1.73 (1.3)
2. Rabbit sera Group C (n = 4)
P < 0.001 ~
P = 0,017 d P = 0.04 c
Expt. 1, Texas strain merozoites; Expt. 2, = Mo7 strain. b Percentage of parasitized erythrocytes (96 h after initiation of in vitro infection). c Two-sample t-test (one tailed). o Paired t-test (one tailed)
89
position 72 but is incomplete at the 5' end and does not include a predicted N-terminal signal peptide [27]. Five adenosine residues at the 3' end of the insert suggests a poly(A) tail.
the cDNA insert is shown in Fig. 2. A single open reading frame starting at nucleotide position 3 is in frame with fl-galactosidase, encoding a predicted fusion protein of 347 amino acids. A 17 amino acid hydrophobic region at the carboxy-terminus [25] of the encoded protein is compatible with signal sequences for attachment of a glycosyl-phosphaditylinositol anchor, consistent with the post-translational modification of native MSA-1 [1,26]. The sequence has a potential N-linked glycosylation site at amino acid
Homology of MSA-1 with BabR and MSA2. The MSA-1 sequence was analyzed for similarities to known sequences. The 3' terminal 173 nucleotides of the MSA-1 gene, including 60 nucleotides of untranslated sequence, were highly similar to the 3' sequences of two previously described B. boris genes,
MSA-I
CTTCAATCGTCCTTCCCGAAGGATCATTCTACGATGACATGTCTAAGTTCTACGGTGCTGTTGGAAGTTTCGACCAGACCAAATTGTATAG•GTTCTTTC
100
MSA-I
TG•TAACTTCAAAGCCG•TAAAATGGATGATCAGAAGGTAAAAGACACATTCAAAAATTTATACAAAGTCAACGCATTGATAAAGAACAATC•TATGATT
200
MSA-I
•G••CTGATCTATTTAATGCAA•TATTGTTAGCGGTTTTT•AACTAAGAATGA•GAGGAAAAATTCAATGCTATATTTGATTCCATTAAGGGAATGTACT
300
MSA-I
ATAGAGCT•AACACATGGACAAATATTTGAAGTCACTAAGGTGGAATACTGATATTGTTGAGGAAGAT•GTGAGAAGGCAGTTGAATATTTCAAGAAG•A
400
MSA-I
TGTTTATACGGGGGAACACGTTGTTGACGTCAACGGTATGGCTGGTGTTTGCAAGGAGTTTTTAAGCCCGGCCT•TGATTTCTACAAA•TTGTTGAGTCT
500
MSA-I
TTTGATGCGTTTGCACATGCTAAGGTGCACGCTCAAGTAGGAAATTTTGTTAAA••TGGAACTGACATCGCTCCTCCTAAGGATGTTACTGATG•ATTAG
600
MSA-I
AAAAGGAATTGCAAGAG•AAAAACCTGCA•GAAGTGAGAG•AC•GAAGTACCCGCTCCAGGTGATG•ATCTGGCGTC•AA•AACCGCCTGCATCAGGAAC
700
vAvell
/Begin
conserved
sequence
3'
MSA-I MSA'2 BabR 0.8
ATCCCCGCAAGGACCTGCT•CGACTACACC^CAG•CCATCTCCAGAGT•CTCAGGAAACCTCCAAGGACAACAGGGTACAACCAAGCCAGCCGGATCTT• TGAAAGC-C-TCC-AA---GACCACC--A-^--AA--TA---AGACAC-TGA---T ............................................ - A - - T G T - - - A C - - - G - A G G A - G T - G A - - - A - - A .... G-G-T . . . . . . . . T---A-T ..... G ...... T--G .......... T ........
799 (895) (636)
MSA-I MSA-2 BabR 0.8
TTTCACCTATGGCGGATTGACTGTGGCTACTCTCTGCTACTTCG^TT~TCT~TGCATTTTA.-~AAAACTAATGGTAGTGACACAATAGTTTTGTAAACTCAT ........................... C ............................... TAATAA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
898 (995)
A. . . . . . .
(736)
MSA-I MSA-2
BabR 0.8
MSA-2
T-CCG . . . . . .
T. . . . . . . . . .
(:::1.I..
:
I
I1.., . : l ' " , ., - - . J .
.. . . . . .
.
.
:::
T--
I . . . , . - - : :.
.
.
.
.
::
I . . . . .
:
-,-.
II II
:
.
.
.
.
.
.
.
.
.
.
.
I I
.
. . . . . . . . . . . .
I
125t
. . . . . . . . . . . . . .
....
I
.lll.
I..
I ....
II
100
I.II
MEAQSATETQKNLKTLMELIKTKRPFKSSDFDTLNLDYLSGQSNEELFKLLIDAINGMKE
BabR 0 . 8
ML•EFNAFLNDNP.••HMLTNGKEKMTEYYKKN•SKEDGEVK•YKTMVKFCNDFL•SKSPFMRLYKHLNEY•ELVKKKPAQE•SPA•SSPQRPAETQQTQ :
MSA-I
T. . . . . . . . . . . . . .
•IVLPEG•FYDDMSKFYGAVG•FDQTKLY•VL•AHFKAAKM•DQKVKDTFKNLYKVNALIKNNPMIRPDLFNATIVsGFsTKNDEEKFNA•FD•IKGMYY
