Richard right,^ John C. ~ e l r , ~ .... v) plus glycerol (5%, v/v) and were stored at - 70°C. For ...... Burton 2, Burgess RR, Lin J, Moore S, Holder S, Gross CA.
BIOLOGY OF REPRODUCTION 53, 462-471 (1995)
Oral Immunization with Attenuated Salmonella Expressing Human Sperm Antigen Induces Antibodies in Serum and the Reproductive Tract' Jay ~rinivasan,'.~ Steven ~ i n ~Richard e , ~ right,^ ~ ~ John C. ~ e l rand , ~ Roy Curtiss llf Department of ~ i o l o g yWashington ,~ University, St. Louis, Missouri 63130 Department of Cell ~iology,' University of VifginiaHealth Sciences Center, Charlottesuille, Virginia 22908 ABSTRACT Induction of immune responses in the reproductive tract will be crucial for a functional gamete antigen-based antifertilityvaccine. Here we describe the construction and development of an avirulent Salmonellaas an oral vaccine delivery vectorto elicit sperm-specific immune responses in reproductive tract secretions. A cDNA sequence encoding the human sperm antigen SPlO was cloned on an asd t vector and expressed to a high level in an avirulent Acya, Acrp, and Aasdvaccine strain of Salmonella typhimurium. Oral immunization of female BALBlc mice with this recombinant Salmonella elicited high-titer anti-SP10 IgG antibodies in serum and IgA antibodies in vaginal secretions. Anti-SP10 antibody titers could be increased by secondary and tertiary oral administrations of the recombinant Salmonella. Induction of sperm-specific antibodies in the reproductive tract following oral administration of a recombinant Salmonella could lead to the development of a simple, safe, efficient, and easy-to-use antifertility vaccine.
INTRODUCTION
Inducing immune responses against gamete-specific antigens represents one approach to developing antifertility vaccines. If the gamete-specific antigen is one involved in the processes of fertilization, antibody induced by vaccination should bind to the active site of the sperm or egg and prevent pregnancy prior to conception. Thus, for a gamete-specific antifertility vaccine to be effective, it is essential for the anti-gamete antibodies to be present and available in the female reproductive tract at the sites of sperm deposit (vagina), transport (cervical mucushterine and oviductal mucosa), or gamete interaction (oviduct). Although gamete-specific antigens are integral components of the body, they are not usually recognized by the immune system as "self" antigens. Under normal conditions, these antigens are shielded from the immune system, and once they are exposed, the immune system elicits an immune response against them. Such situations occur clinically, where infertile patients have anti-sperm antibodies in the serum and reproductive tract [I, 21. Gamete antigenbased antifertility vaccines are designed to mimic this situation. A number of gamete-specdic antigens that are involved in sperm-egg interaction have been identfied 13-61, One such antigen that has been extensively studied is the sperm-specific antigen SPlO [31. SPlO is a human sperm acrosomal antigen that has been designated a "primary vaccine candidate" by a Accepted March 27, 1995. Received November 14, 1994. h his project was supported by research grants ROlDEO6669 and U54 HD29099 from the U.S. Public Health Service, National Institutes of Health. a grant from Bnstol-Myers Squibb, and CSA-94-129from the contraceptive Research and Development Program, Eastern Virginia Medical School, under a Cooperative Agreement with the United States Agency' for International Development (USAID) (DPE-3044-19-00-2015-00). Z~orrespondence. FAX. (314) 935-4432. 3 ~ u r r e naddress: t MEGAN Animal Health, 3655 Vista Ave., St. Louis, MO 63110.
World Health Organization (WHO) task force on contraceptive vaccines [41. Human SPlO exists in the acrosome and can be detected as a series of polymorphic peptides ranging in size from 18 to 45 kDa in sperm and 55 kDa in testis extracts [71, with a majority of them having an isoelectric point of 4.9 [51. SPlO is conserved between the different species and is presumed to play a role in the fertilization process. A monoclonal antibody (MHSl0) [61directed against human SPlO has been shown to inhibit human sperm penetration in a hamster egg penetration assay [41. This and other monoclonal antibodies also cross-react with bovine SPlO and inhibit bovine in vitro fertilization (S. Coonrod, manuscript in preparation). Analogs of SPlO with similar immunoreactive and electrophoretic properties have been identified in macaque, baboon, porcine, and fox sperm [3, 81. SPlO has been shown to be expressed only in the testis, making it a potentially useful tissue-specfic vaccinogen [91. Conventional perenteral methods of immunization, while adequate to induce a serum immune response, are very poor in inducing an antigen-specific response in mucosal secretions. In order to study the feasibility of inducing a secretory immune response in the female reproductive tract, the cDNA sequence coding for SPlO was expressed in an attentuated vaccine strain of Salmonella t@himurium. S. typhirnuriurn strains with deletions of the adenylate cyclase (cya) and cyclic AMP receptor protein (clp)genes are avirulent and immunogenic while retaining their ability to colonize gut-associated lymphoid tissue (GALT) and internal organs [lo].The advent of recombinant DNA technology has enabled us to express foreign genes in Salmonella and use it as an efficient delivery system for the induction of immune responses to the expressed antigens. Since S. typhimurium naturally invades and persists in GALT [Ill, oral immunization with attenuated Salmonella expressing foreign antigens stimulates antigen-specificsecretory, humoral
INDUCTION OF ANTI-SPERM ANTIBODIES BY RECOMBINANT SALMONELLA and cellular immune responses (for a review see [12]). Such a system may be particularly advantageous in eliciting an immune response against a gamete-specific antigen in the female reproductive tract for the purpose of developing a contraceptive vaccine. The cDNA sequences coding for human, macaque, and baboon SP10 have been cloned and sequenced [13, 14]. Analysis of the human SP10 nucleotide sequence shows that the codon usage is conducive to its expression in a prokaryotic expression system like Salmonella. Various expression vectors have been used to express foreign genes in Salmonella. The stability of these vectors is necessary for prolonged expression of foreign antigens in vivo. For this purpose, the 3-aspartate semialdehyde dehydrogenase (asd)-based vector system has been shown to be ideal for use in live vaccine strains of Salmonella [15]. In this report, we describe a modification of the asd-based vectors for an even higher level of expression of the gene for human sperm antigen SP10 and discuss the feasibility of using the recombinant Salmonella to induce anti-sperm antibodies. Expression of SP10 in Salmonella will make it possible to study the possibility of using this system to induce an antigen-specific immune response in the female reproductive tract and to develop a potential antifertility vaccine.
MATERIALS AND METHODS BacterialStrains, Media, and Plasmid Vectors Attenuated S. typhimurium triple mutant X4 55 0 (Acya, Acrp, Aasd), an asd-derivative of X4 0 6 4 [10], was used in these studies. The bacterial strains were grown in Luria broth (LB: Tryptone 10 g/l, NaCl 10 g/l, and yeast extract 5 g/l; Difco, Detroit, MI; Sigma, St. Louis, MO), supplemented with diaminopimelic acid (DAP) (50 jig/ml) when Aasdmutant strains without an asd complementing plasmid were grown [16]. Bacterial stocks were made in peptone (1%, w/ v) plus glycerol (5%, v/v) and were stored at - 70°C. For Western blotting (immunoblotting) or immunizing animals, the bacteria were grown from the frozen stock as a static overnight culture, diluted 1:20 in prewarmed LB, and grown with aeration at 37°C to an optical density at 600 nm (OD60 0 ) of 0.8-0.9 (-109 colony-forming units, CFU). CFU were determined by serial dilutions on LB agar plates or MacConkey agar plates supplemented with 1% maltose. S. typhimurium X4 550 was transformed by electroporation as previously described [15]. Part of the SP10 cDNA was amplified by polymerase chain reaction (PCR) from the plasmid pGEX2T-SP10 and purified after separation of an agarose gel by use of a commercial kit (Geneclean; BIO101, LaJolla, CA). pGEX2T-SP10 was obtained by PCR amplification of the entire SP10-specific cDNA [13] and cloned as a 0.75-kb BamHI fragment into the plasmid pGEX2T (Pharmacia, Piscataway, NJ). The Salmonella expression vector,
463
pYA3098, contains the trc promoter, multiple cloning sites, and the asd gene, and is an Ncol I derivative of the low copy number plasmid (p15A origin of replication) pYA292 [16]. pYA810 is identical to pYA292 except for the deletion of the lac Z a gene. pYA3148 was generated by substituting the p15A origin of replication of pYA3098 with the pUC origin of replication as described in the construction of pYA3146. All initial cloning and characterizations of the recombinant plasmids were done in the Eschericiacoli strain X4212 an asd-derivative of DH5a, containing the lac Iq repressor gene on the plasmid pYA232 [151. Immunoblotting andImmunoscreening Assays for PlasmidStability and Expression of SP1O in S. typhimurium Growth characteristics of S. typhimurium X4 550 containing the appropriate recombinant plasmids were determined in LB. Stability of the plasmids in X4550 was checked by repeated subcultures and continuous growth in LB for approximately 50 generations as described [16]. After 50 generations of growth, the cells were plated on LB agar plates, and random colonies were picked, grown to late log phase, and assayed for the expression of SP10 by Western blots (described below). To determine the in vivo stability of the SP10 constructs, spleens, Peyer's patches, and mesenteric lymph nodes were removed from immunized mice 18 days after immunization and homogenized in buffered saline with gelatin (BSG) [17], and samples were plated on Mac Conkey agar plates containing 1% maltose. Expression of SP10 by bacteria in these colonies was determined by colony immunoblot using SP10-specific monoclonal antibodies and a chemiluminescent detection kit (ECL; Amersham, Arlington Heights, IL). The levels of SP10 expressed by S. typhimurium as a percentage of the total protein was determined by SDSPAGE with use of equal amounts of protein from lysed X4550 containing the SP10 clones. Gels were then either stained with Coomassie brilliant blue dye and scanned with a laser densitometer (Molecular Dynamics, Sunnyville, CA) or transferred to a nitrocellulose membrane and processed for immunoblot assays essentially as described [18]. Blocking buffer containing BSA (2%) in Tris-buffered saline (pH 7.4) with 0.1% Tween 20 (Sigma) was used to reduce nonspecific binding of the antibodies. All antibody dilutions were made in blocking buffer, and incubations were carried out for 2-3 h at room temperature. SP10-specific monoclonal antibodies (MHS10 [6]) or anti-SP10 polyclonal antibodies raised in rabbits were used as the primary antibodies. An appropriate second antibody conjugated to avidin and biotin conjugated to horse radish peroxidase was used sequentially in these experiments. Color was developed by means of 5-bromo-4-chloro-3-indolylphosphate toluidinium (BCIP) (Sigma) and nitroblue tetrazolium (Sigma). A
464
SRINIVASAN ET AL.
FIG. 1 Plasmid constructions. SP10 cDNA was cloned into pYA3098, an Nco I derivative of pYA292 at the Nco I and EcoRI sites, to give rise to plasmid pYA3130. Translational termination signals were inserted at end of SP10 sequence by filling in EcoRI site and then by addition of a universal translational terminator inserted at Sma I (pYA3144). Origin of replication in pYA3144 was changed to that of high copy number pUC origin to give rise to the plasmid pYA3146.
INDUCTION OF ANTI-SPERM ANTIBODIES BY RECOMBINANT SALMONELLA
465
comparison of the amount of SP10 expressed by the low and high copy number clones was done by a direct densitometric scanning of the Western blot. Integrated areas of the SP10 bands in X4550(pYA3144) lanes and X4550(pYA3146) lanes were used to calculate the total increase in SP10 expression.
cial S. typhimurium lipopolysaccharide (LPS) preparation (Sigma), was used in the ELISAs to determine anti-Salmonella LPS antibody titers.
Animal Model and Immunizations
Immunized mice were anesthetized with Metofane (Pitman-Moore, Mundelein, IL), and retro-orbital bleeds were collected at 2-wk intervals. Serum was separated by standard procedures and stored individually at - 700C until the time of assay. Mice were bled prior to immunization (preimmune), and the serum was assayed for the presence of antiSalmonella and anti-SP10 specific antibodies. Vaginal secretions were collected by repeatedly pipetting (5-6 times) 50 gll of sterile PBS (pH 7.4) into the vaginal opening. The fluid and mucus was mixed vigorously by a vortex mixer and spun down briefly in a microfuge, and the supernatant fluid was stored at - 20C until the time of assay. Preimmune mucosal secretions were also collected from the mice in a similar manner for use as negative controls.
Four groups of five female mice each (BALB/c, 4 wk old; Sasco, NE) were used in the study. The mice were deprived of food and water for 4 h and then fed 30 ll of 10% (w/v) sodium bicarbonate to neutralize stomach acidity. Thirty minutes later, 25 pl sterile BSG or BSG containing --109 CFU of X4550 (pYA3144), X4 550 (pYA3146), or X4 55 0 (pYA810) was administered orally (pipetted into the back of the mouth with a micropipette). Food and water were restored 30 min later. Mice were boosted at 4, 9, and 28 wk with 109 CFU of X4 550 containing the respective recombinant plasmid or sterile BSG, as earlier. Colonization of spleen, Peyer's patches, and mesenteric lymph nodes of the immunized mice was determined according to our previously published method [10]. Antigens and Antibodiesfor Immunoassays A recombinant human SP10 protein (rSP10) consisting of amino acids 17-181 [13] was produced in liquid culture as a glutathione S-transferase (GST)-SP10 fusion protein by use of the clone pGEX2T-SP10. Recombinant SP10 expression was induced by isopropylthio-13-D-galactoside (IPTG), and the cells were collected by centrifugation at 3200 X g for 10 min and disrupted by sonication. The cell debris was separated by centrifugation, and the supernatant was incubated with glutathione sepharose 4B beads (New England Biolabs, Boston, MA). Recombinant SP10 was subsequently cleaved from the GST by means of thrombin, and the supernatant was dialyzed against 0.01 M ammonium bicarbonate and 0.02% SDS, lyophilized on a Virtis Freeze Mobile 12SL lyophilizer (The Virtis Company, Gradiner, NY), and separated by preparative SDS-PAGE. SP10 protein bands were visualized by staining with 4 M sodium acetate and were cut out of the gel. The SP10 protein was electroeluted in TBE buffer (90 mM Tris, 90 mM borate, 2 mM EDTA, pH 8.0), dialyzed against TBE buffer, and lyophilized. Three rabbits received i.m. injections of 1 mg purified recombinant human SP10 in Freund's complete adjuvant and were subsequently boosted at 2, 4, and 6 wk with 1 mg in Freund's incomplete adjuvant. Serum was collected at 2, 5, and 7 weeks and 2-wk intervals thereafter. Sera from these rabbits were pooled and assayed for anti-SP10 antibodies. The lack of cross-reactivity of the rabbit anti-SP10 antiserum with E. coli was verified by immunoblots with use of whole E. coli lysate (unpublished data). A commer-
Collection of Sera and FemaleReproductive Tract Secretions
Assays for the Detection ofAntibodies to SP10 and Salmonella LPS in Serum and Reproductive Tract Secretions Anti- SP10 antibody titers in the serum and vaginal washings were determined by an ELISA. The assay was standardized by using 400 ng/ml -' of purified E. coli-expressed recombinant SP10 in a carbonate buffer (pH 9.6) as the coating antigen. ELISA plates (Immunolon 1; Dynatect, Alexandria, VA) were coated at 370 C for 4 h or 4°C overnight. Nonspecific binding was blocked by preincubation of the antigen-coated ELISA plates with the blocking buffer (10 mM Tris, pH 8, 0.1% Tween 80, and 0.5% BSA), and all antibody dilutions were made in the blocking buffer. Assays were done in duplicate with appropriately diluted test and control samples, and the results were averaged. Primary incubations were carried out overnight at room temperature in a moist chamber. Goat anti-mouse IgG (or IgA) conjugated to biotin (Southern Biotechnology, Birmingham, AL) and extravidin conjugated to alkaline phosphatase (Sigma), each diluted 1:5000, were sequentially used to reveal the presence of anti-SP10 antibodies. ELISA plates were washed five times with Tris-buffered saline, pH 7.4 (TBS) containing 0.1% Tween 20 (Sigma) in an ELISA plate washer (BioTek, Winooski, VT). p-Nitrophenylphospate (Sigma) (1 mg/ml) in 0.1 M diethanolamine buffer (pH 9.8) was used as the substrate, and the color resulting from the substrate-enzyme reaction was read at 405 nm with an automated ELISA reader (BioTek). A similar ELISA using 1 .ig/ml- of purified LPS to coat the ELISA plates was standardized to determine the presence of anti-S. typhimurium antibodies. ELISA values that
SRINIVASAN ET AL.
FIG. 2 Expression of human SPtO by recombinant Salmonella. Four different ~4550(pYA3144)colonies and four ~4550(pYA3146)colonies were picked from LB agar plates containing fresh transformants and grown at 37°C with aeration. One milliliter of each culture (equal to 6 X lo8cells) was spun down and lysed. One tenth of total protein from the four ~4550(pYA3144)lysates (lanes 1-41and a similar amount of ~4550(pYA3146)(lanes 7-10) were electrophoresed along with negative control ~4550(pYA810)(lane 5) on 12.5% acrylamide gel under denaturing conditions and electro-blotted onto nitrocellulose paper. Blot was probed with SP10specific monoclonal antibody MHS10.
-
were three times greater than that of the negative control were taken as positive, and the dilution at which the sera were no longer positive was taken as the end-point of dilution. Antigen-specific and total IgA antibody present in the vaginal washings were quantified by an ELISA from a standard curve obtained with known quantities of purified mouse myeloma IgA (Cappel, Durham, NC).
RESULTS
Cloning of Human SPlO cDNA Sequence in S. typhimurium The SPlO protein has a hydrophobic I6aa N-terminal secretory signal sequence that is cleaved during protein processing to give rise to the mature protein. The hydrophilic middle portion of the SPlO molecule from amino acid position 17-213 contains SP10-specific antigenic epitopes, including the one recognized by the monoclonal antibody MHSlO [131. The C-terminal 52aa contains 6 of the 10 cysteine residues found in the human SPlO molecule. It is largely hydrophobic and unlikely to possess important imnlunogenic epitopes. In order to clone the cDNA sequence coding for the middle portion of human SP10, the entire SPlO cDNA sequence was cut out as a 759-bp BamHI fragment from the plasmid pGEX2T-SP10. This fragment was modified by the addition of a 10mer Nco I linker (CAG CC ATGG CTG, New England Biolabs) and cut to completion with Nco I and EcoRI enzymes to release a 596-bp Nco I-
EcoRI fragment coding for the middle antigenic sequences of SPlO (aa 17-213). This fragment was cloned into the Nco I-EcoRI sites of the vector pYA3098, an Nco I derivative of pYA292 I161 to generate the plasmid pYA3130. pYA3130 had the human SPlO cDNA sequence in-frame with the vector-derived translational initiation codon ATG and read into the lac z-a sequence at the 3' end. A translational termination signal TAA was introduced at the 3' end of the gene by digesting pYA3130 with EcoRI followed by an end-filling reaction and self-ligation (GAATTC to GAATTAATTC; pYA3142). A second set of translational termination sequences in all three reading frames (GCTTAATTAATTAAGC, New England Biolabs) was also introduced into the Sma I site of pYA3142 to generate the plasmid pYA3144. pYA3144 contains the p15A (10-1 5 copies/cell) origin of replication. The p15A origin of replication was removed from pYA3144 as a 830-bp fragment by Acc I partial digestion followed by an Xba I digestion and was replaced by a similarly cut 884-bp pUC origin of replication that was PCRamplified from the plasmid pUC18 (191. This resulted in the plasmid pYA3146, with a pUC origin of replication (Fig. 1). High-Level Expression of Human Sperm Antigen by S. typhimurium Initial experiments on the expression of SPlO by E. coli, ~ 6 2 1 2containing , either pYA3144 or pYA3146, revealed the presence of a 37-kDa protein that reacted with the SP10specific monoclonal antibody MHS10. A similar 37-kDa band was detected when the total cell lysate from recombinant Salmonella ~ 4 5 5 0(pYA3144 or pYA3146) was run under denaturing conditions and probed with either the monoclonal antibody (MHS10) or SP10-specific rabbit polyclonal antibodies. All detectable SPlO was found in the cytoplasm, and none was present in the periplasmic space or in the culture supernatant fluid (data not shown). The total amount of SPlO expressed by ~ 4 5 5 0(pYA3146) was 56-fold greater than that expressed by ~ 4 5 5 0(pYA31441, whereas the negative control ~4550(pYA810)showed no iminunoreactivity for the SPlO protein (Fig. 2). S. typhimu?$urn,~ 4 5 5 0(~YA3146),expresses SPlO u p to 5% of the total cellular protein content as determined on the basis of a Coomassie brilliant blue dye-stained gel (data not shown). Growth, Stability, and Colonization of Mice by Recombinant S. typhimurium Ekpressing SPlO Recombinant S. t@himu~-ium 14550 (pYA3144 or pYA3146) was grown in LB medium at 37°C with aeration. The growth of ~4550-SP10recombinants and the nonrecombinant vectors ~ 4 5 5 0(pYA810 or pYA3148) were compared (Fig. 3). All the four S. typhimurium constructs grew equally well and reached a maximum cell density by 8 h. However, ~ 4 5 5 0containing the high copy number clone,
INDUCTION OF ANTI-SPERM ANTIBODIES BY RECOMBINANT SALMONELLA -In
E
467
10s
1 04
1 0-
Ca s
10
-1
03
C C
E C 0 O
. 02' 10 10
1
I
I
5
10
1 05 0
100
200
300
400
500
0
Time in Minutes FIG. 3 Growth curve. Growth of recombinant cultures was determined inLB with aeration. Standing overnight cultures of 4550 containing pYA3144 (--), pYA3146 (O-0), pYA3148 (1-[), or pYAB10(0-) were diluted 1:20 inLBand allowed to grow at 37°C. Number of CFU was determined at periodic intervals.
pYA3146, had a prolonged lag phase. After 50 generations of growth in LB media, all the analyzed colonies (13/13) containing the low copy number plasmid X4 550 (pYA3144) continued to express SP10, while 12/13 of the high copy number X4 550 (pYA3146) were positive for SP10. Three groups of five mice each were orally immunized with 109 CFU of S. typhimurium, X4550, containing pYA3144, pYA3146, or pYA810. After immunization, none of the immunized mice exhibited any clinical symptoms of infection. Internally, the Peyer's patches, spleen, and mesenteric lymph nodes had a moderate infection, with the bacterial content ranging from 102 on Day 6 to 104 by Day 18 (Table 1). S. typhimurium ( 4 550) strains containing either pYA3144 (low copy number SP10 clone, p15A origin of replication) or pYA3146 (high copy number SP10 clone, pUC origin of replication) colonized mice as efficiently as X4 550 (pYA810; vector control). In vivo stability of S. typhimurium X4 550 (pYA3146) was determined by colony immunoblot of the bacterial cells obtained from spleen of mice 18 days postimmunization. At this stage, 94% (79/84) of the bacteria retrieved from the infected animals continued to express SP10.
4.4.
II/I 15
/ 20
25
Weeks
4.
30
SP PP MLN
FIG. 4 Anti-SP10 immune response inserum. Groups of 5 BALB/c mice each were immunized orally with 109 CFU of recombinant S. typhimurium, X4550 , containing either pYA3144 (0-0), pYA3146 (--), or negative controls 4550(pYA810) or BSG ([-O). Boosters were given at 4,9, and 28 wk. Anti-human SP10 IgG antibody titers in pooled serum from each group were measured at periodic intervals by ELISA. All assays were done induplicate, and values greater than three times that of negative control were taken as positive. Reciprocal of end-point dilution of antisera isexpressed as antibody titer.
Immunogenicity of Acya A crp Aasd Mutant Vaccine Strain of S. typhimurium Expressing SPlO After Oral Administration Two groups of five mice each were orally immunized with S. typhimurium 4 550 containing pYA3144 or pYA3146. Two control groups were immunized with either X4550 (pYA810) or 25 (il of BSG. An oral booster was given to all the animals after 4 wk followed by a second oral booster after 9 wk. Mice were bled every 2 wk, and a part of the serum from each group was pooled together prior to antibody assay. Anti-SP10 and anti-S. typhimurium LPS antibody titers in the pooled serum samples were assayed. Vaginal secretions from each group of mice were also pooled and checked for the presence of anti-SP10 and antiS. typhimurium LPS antibodies. In the X4 550 (pYA3146)immunized group, the serum anti-SP10 IgG antibody titers in the pooled sample were at 1:700 by 4 wk. After the pri-
pYA810
pYA3144
pYA3146
Day 6
Day 18
Day 6
Day 18
Day 6
Day 18
3.7 ± 0.30 4.1 + 0.15 3.8 0.50
3.7 ±+0.24 4.0 0.20 4.0 ±+0.28
2.7 ± 0.34 2.9 0.58 2.7 0.50
4.2 ± 0.12 4.0 0.06 4.2 0.12
1.6 ± 0.21 4.0 0.09 2.3 0.04
4.2 ± 0.12 3.9 0.38 4.2 0.06
aExpressed as logo0 ± SE (where n = 3). bNo S. typhimurium was detectable in unimmunized mice. CSP, spleen; PP, Peyer's patches; MLN, mesentric lymph nodes.
40
A
TABLE 1. Bacterial counts logot) in mice infected with recombinant S. typhimurium (X4550).a 'b Tissue colony
35
468
SRINIVASAN ET AL. 4-
250 200 -
0
150
-
100
c
3-
'
50-
2-
0-
1 -*" -uV
I
o
I
10
5
I
15
I
20
I
25
Weeks
t
I
30
3 35
A 0*
to ng)
pYA810
(0 ng)
pYA3144
pYA3146
(O ng)
-Cont.
Construct in X4550 Used for Immunization FIG. 6 Quantitative anti-SP10 and anti-S. typhimurium LPS immune responses in vaginal secretions. Vaginal secretions from four groups of five mice each, immunized with X4550(pYA3144), X4550(pYA3146), X4550(pYA810), or BSG were collected and independently pooled. Anti-SP10-specific (solid bar) and anti-S. typhimurium LPS-specific (hatched bar) IgA antibody titer in each group were measured by ELISA and compared with total IgA found in vaginal secretions of respective groups. Numbers in parentheses show actual amount of IgA antibodies recovered in ng/mouse.
390
-
280
=
170
60
._n
I
-
0
5
.+
A
I
10
I
15
Weeks
I
I
20
25
I
30
35
4
FIG. 5 Anti-SP10 and anti-S. Typhimurium (LPS) IgA antibody titers in vaginal secretions. Antibody titers obtained from pooled vaginal washings in four groups of 0 X4 55 (pYA3144) (0-), mice immunized with X4550(pYA3146) (--), are expressed as endstudy X4550(pYA810) (I-I), or BSG (A-A) for duration of point dilution values. a) Anti-SP10 IgA antibody titers and (b) anti-LPS IgA antibody titers.
mary booster, the serum anti-SP10 IgG antibody titers in this group went up to 1:5000 by wk 8 and after the second booster reached a maximum of 1:26000 by 11 wk (Fig. 4). Periodically, individual animals in each group were checked for anti-SP10 antibody titers to determine the percentage of immunized animals responding to immunization. In the X4 550 (pYA3146)-immunized group all the animals sero-converted (100% response) after the primary booster. In the S. typhimurium X4 55 0 (pYA3144)-immunized group, no significant antibody titers were detectable in any of the immunized animals following primary immunization. After two boosters, an anti-SP10 antibody titer of 1:800 was detectable in the pooled serum at 11 weeks. However, anti-SP10 antibodies were detectable in only 3 of the 5 animals even after the second booster. Control groups 4 0 of mice given only BSG or immunized with X 55 (pYA810)
did not have any anti-SP10 antibody titers. Serum anti-S. typhimurium LPS IgG response was consistently high in all the X4 550 immunized groups of mice by 5 wk after immunization and reached a maximum of 1:50000 by 11 wk in the X4550 (pYA810)-immunized mice and 1:45000 in the X4 550 (pYA3146)- and X4 550 (pYA3144)-immunized mice. Mice given BSG did not have any detectable anti-S. typhimurium LPS antibody titers. A third booster, given orally at 28 wk with the respective recombinant S. typhimurium constructs, restimulated the serum anti-SP10 IgG, anti-LPS IgG, and mucosal anti-LPS IgA in all the groups and also antiSP10 IgA in the X4550 (pYA3146)-immunized group. Vaginal anti-SP10 IgA response in the X4550 (pYA3146)immunized group remained low (1:10) prior to booster immunization and went up to a maximum of 1:200 by wk 11. In the x4550(pYA3144)-immunized group, anti-SP10 IgA antibody titers reached a maximum of 1:50 by 15 wk (Fig. 5a). The S. typhimurium LPS LgA antibody titers in the three 44 groups of mice immunized with X4550(pYA31 ), X4550(pYA3146), or X4550(pYA810) ranged from 1:100 to 1:400 by 11 wk (Fig. 5b). Anti-S. typhimurium LPS-specific IgA antibodies were detectable in vaginal secretions 2 wk after the primary booster and persisted for about 14 wk. Anti-SP10 IgA antibodies were detectable in the X4550(pYA3146)-immunized group by wk 2 after the primary booster and persisted for 11 wk. 0 4 At 11 wk, the anti-SP10 IgA response in the X 55 (pYA3146)-immunized group was 1% of the total IgA, and
INDUCTION OF ANTI-SPERM ANTIBODIES BY RECOMBINANT SALMONELLA
the anti-S. typhimurium LPS response was 1.6% of the total IgA detectable in vaginal secretions (Fig. 6). In the x4 550 (pYA810)- and 4550(pYA3144)-immunized groups, the anti-S. typhimurium LPS response ranged from 0.9-3.1% of the total IgA. No anti-SP10 IgG antibody titers were detectable in the vaginal secretions of any of these groups whereas anti-S. typhimurium LPS IgG antibodies were detectable at 1:50 dilution in the vaginal secretions by 11 wk. DISCUSSION World human population currently is 5.6 billion, and approximately 90 million persons are added each year. The rapid population growth, disproportionate to the available natural resources, argues for the need of a user-friendly, cost-effective means of birth control. Immunocontraception would be an efficient method of contraception, devoid of the risk of user failure. In spite of the fact that antifertility vaccines are designed for use by healthy, immunocompetent adults of reproductive age, immunocontraception remains an achievable but still challenging field of study [20]. Since anti-sperm antibodies were identified as a cause of infertility in some women [2, 21], a number of sperm-specific antigens have been identified as potentital antifertility vaccinogens [22-27]. Of these, the sperm acrosomal antigen, SP10, has been more extensively studied [3, 5-8, 14]. AntiSP10 monoclonal antibodies are capable of preventing sperm-egg interaction [4], and SP10 fulfills all the criteria for a sperm-specific vaccinogen [5, 9]. Induction of an anti-sperm immune response in the female reproductive tract at the sites of sperm entry, transportation, and fertilization will be of significance in developing an antifertility vaccine. Although sperm bear antigens that are foreign in the female body and are capable of being recognized as such by the female immune system, an antisperm immune response does not usually occur in the course of their natural introduction into the female reproductive tract. Vaginal application of washed, sonicated sperm could lead to a local immunity against sperm [28]; however, repeated topical applications of large doses of antigen are required to induce an immune response in the reproductive tract [29]. In this report, we show that recombinant Salmonella can be efficiently used as a vaccine delivery system to elicit a sperm-specific immune response in the female reproductive tract. Avirulent recombinant Salmonella as a vaccine delivery vehicle has been shown to stimulate humoral and cell-mediated immunity [10, 30-33]. In addition to the benefits of a recombinant vaccine, oral administration of an attenuated Salmonella can induce serum and secretory immune responses to the Salmonella-expressed antigens [32, 34, 35]. The presence of a common mucosal immune system [36, 37], with the antibody secreting plasma cells originating from the Payer's patch [38] and the peritoneal cavity [39],
469
predicts the possibility of inducing a secretory immune response in all muscosal secretions including the reproductive tract secretions, following oral administration of attenuated recombinant Salmonella. In the present study, we have cloned and expressed a cDNA sequence encoding the human sperm-specific antigen SP10 in an attenuated vaccine strain of S. typhimurium. This recombinant S. typhimurium induces antigen-specific serum and mucosal immune responses in BALB/c mice. The human SP10 cDNA sequence was modified to remove nucleotide sequences that specify the hydrophobic N-terminal 16-aa signal and the C-terminal 52 aa of SP10. The middle portion coding for the hydrophilic 192-aa sequence was cloned into the Salmonellavector pYA3098 under the control of the bacterial trc promoter and the vectorderived initiation codon. A termination codon, TAA, was introduced after the gene as a part of the cloning strategy. Salmonella-expressed SP10 had a higher apparent molecular size on SDS-PAGE when compared to that deduced from the cDNA sequence. Introduction of a translational termination sequence in all three reading frames downstream of the actual termination signal did not alter the apparent size difference of the protein. A similar shift in mobility in SDS-PAGE was observed when the SP10 cDNA sequence was expressed in E. coli (Wright and Herr, unpublished data). This differential mobility in an acrylamide gel might be characteristic of the SP10 molecule, which possesses an overall pI of 4.9. Similar shifts in protein mobility have been observed for a number of other proteins [40-42]. Expression of SP10 by the low copy number (p15A replicon) plasmid pYA3144 was < 1% of the total cellular protein. Since the expressed protein did not in any way affect the growth of the bacteria, the plasmid copy number was increased by substituting the p15A origin of replication for that of the pUC replicon. This high copy number clone pYA3146 expressed SP10 at 5% of the total cellular protein. Increased expression of SP10 in the balanced lethal asd selection system did not significantly reduce the stability of the plasmid. Higher expression levels did not significantly alter the in vitro growth characteristics of the recombinant S. typhimurium after the extended lag phase, nor was the ability of the recombinant S. typhimurium to colonize tissues of the Peyer's patches, spleen, and mesenteric lymph nodes adversely affected. However, higher expression levels made a difference in the amount of antigen-specific immune responses induced. Female BALB/c mice orally immunized with X4550(pYA3146) had approximately 30-fold higher serum anti-SP10 IgG antibody titers than the X4550(pYA3144)-immunized group of mice. All the mice in the former group responded to immunization while only 3 of 5 mice in the latter group had anti-SP10 antibody titers, even after two boosters. Anti-SP10 antibody titers came down gradually by 28 wk
470
SRINIVASAN ET AL.
in the x4550(pYA3146)-immunized group of mice, while in the x4550(pYA3144) group, the serum anti-SP10 IgG antibody titers were below detectable limits by 22 wk. The antiSP10 IgA antibody titers in vaginal secretions of the X4550(pYA3146)-immunized mice, as determined by the ELISA end-point dilution method, reached a maximum of 1:200 at 11 wk and came down to less than 1:10 by wk 17. Although the mice immunized with 4550(pYA3146) seemed to have better vaginal anti-LPS antibody titers than the other immunized groups, the total amount of IgA recovered from vaginal fluids of the x4550(pYA3146)-immunized group of mice was also greater (1300 ng/mouse) than that recovered from the X4550(pYA3144)-immunized group (322.5 ng/mouse) or the negative control (4550[pYA810]; 562.5 ng/mouse). End-point dilution assay, however, is not quantitative because the vaginal washings become diluted at the time of the wash and are again diluted to obtain the antibody titers. Quantification of the antibodies in the vaginal washings showed that up to 1% and 0.15% of the total IgA is SP10-specific in x4550(pYA3146)- and X4550(pYA3144)immunized groups, respectively. The anti-LPS IgA titers in these two groups were 1.5% (4550[pYA3146]) and 3.1% (X4550[pYA31441), respectively, of the total recovered IgA antibodies, showing that X4550/pYA3144 had, in fact, a better anti-LPS titer than X4550/pYA3146. The levels of antibody detected in the vaginal secretions are also influenced by the time point in the estrous cycle at which they are collected (for a review see [431). Hormone-dependent changes in the vaginal antibody levels may also have contributed to the variations in the amounts of antibodies recovered from different groups of animals. In these experiments with the SP10 sperm antigen, a predominantly IgA response was detected in vaginal secretions despite high titers of anti-SP10 IgG antibodies in the serum. The serum anti-SP10 IgA titers were below detectable limits of the assay. The short half-life and rapid uptake of IgA antibodies by the liver in mice [44] might account for the low detectability of antigen-specific IgA in blood. We are unable to explain the absence of SP 10-specific IgG antibodies in the vaginal secretions-particularly since there is a large amount of IgG present in these secretions. The presence of SP10specific IgA and the absence of SP10-specific IgG in female reproductive tract secretions should have a definite advantage in developing an antifertility vaccine as they can bind to sperm and effectively "neutralize" them without causing an inflammatory reaction like that caused by the IgG- or IgMmediated complement activation (for review see [45]). The absence of inflammation in the reproductive tract would be useful in developing a reversible antifertility vaccine because the sperm will continue to be shielded from a direct exposure to the body's immune system while being actively neutralized by the anti-sperm antibodies. Anti-SP10-specific antibody titers in serum and reproductive tract secretions drop as a function of time. Serum
anti-SP10 LgG antibody levels in 4550(pYA3146)-immunized mice came down to 1:150 (end point dilution) by 7 mo, and the vaginal anti-SP10 LgA titers came down to below detectable levels by 4 mo. Such an immune response against the sperm-specific antigen SP10 strongly suggests the feasibility of developing a reversible antifertility vaccine using recombinant avirulent Salmonella. Immunological memory, however, was retained, and both the serum and mucosal immune responses could be restimulated after a booster given at 7 mo. The levels of antibody titers against the sperm-specific antigen SP10 induced in mice are very encouraging. However, the efficacy of the anti-human sperm antibodies elicited by recombinant Salmonella in mice could not be tested because of the species specificity of the SP10 molecule. In order to determine the effect of oral immunization with recombinant Salmonella expressing sperm antigens on the regulation of fertilty in murine and macaque models, species-specific SP10 genes have been cloned into Salmonella, and experiments are in progress to evaluate immunological responses and induction of infertility in mice and macaques. In an era of increasing awareness of the importance of mucosal immunity, the SP10-Salmonella model should be helpful in answering some of the basic questions pertaining to the induction of mucosal immune responses in the reproductive tract. Alternate routes of immunization with recombinant Salmonella are currently under investigation to study their effectiveness in stimulating immune responses to the expressed antigen in the female reproductive tract.
REFERENCES 1. Chen C, Jones WR. Application of a sperm micro-immobilization test to cervical mucusin the investigation of immunologic infertility. Fertil Steril 1981; 35:542-545. 2. Bronson RA, Cooper GW. Autoimmunity to sperm: effect on sperm penetration of cervical mucus as reflected by post-coital testing. Fertil Steril 1988; 41:609-614. 3. Herr JC, Wright RM, John E, Foster J, Kays T, Flickinger CJ. Identification of human acrosomal antigen SPIO in primates and pigs. Biol Reprod 1990; 42:377-382.
4. Anderson DJ, Johnson PM, Alexander NJ, Jones WR, Griffin PD. Monoclonal antibodies to human trophoblast and sperm antigens: report of two WHO sponsored workshops, June 30, 1986; Toronto, Canada. J Reprod Immunol 1987; 10:231-257. 5. Herr JC, Flickinger CJ, Homyk M,Klotz K, John E. Biochemical and morphological characterizatoin of the intra-acrosomal antigen SP10 from human sperm. Biol Reprod 1990; 42:181-193. 6. Herr JC, Wright RM, John E,Klotz K, Homyk M, Foster J, Flickinger CJ. Monoclonal antibody MHS-10. In: Alexander NJ, Griffin D, SpielerJM, Waites GMH (eds.), Gamete interaction: Prospects for Immunocontraception. New York: Wiley-Liss, Inc.; 1990: 13-36. 7. Foster JA, Klotz K, Flickinger CJ, Thomas T, Wright RM, Castelo JR, Herr JC. Human SPI0: acrosomal distribution, processing and fate after acrosome reaction. Biol Reprod 1994; 51:1222-1231. 8. Bradley MP. Experimental strategies for the development of an immunocontraceptive vaccine for the European red fox Vulpes vulpes. Reprod Fertil Dev 1995; (in press). 9. Freemerman AJ, Wright RM, Flickinger CJ, Herr JC. Tissue specificity of the acrosomal protein SP10: a contraceptive vaccine candidate molecule. Biol Reprod 1994; 50:615-621. 10. Curtiss III R, Kelly SM. Salmonella typhimurium deletion mutants lacking adenylate cyclase and cyclic amp receptor protein are avirulent and immunogenic. Infect Immun 1987; 53:3035-3043. 11. Carter PB, Collins FM. The route of enteric infection in normal mice. J Exp Med 1974; 139:1189-1203. 12. Curtiss III R. Attenuated Salmonella strains as live vectors for the expression of foreign
INDUCTION OF ANTI-SPERM ANTIBODIES BY RECOMBINANT SALMONELLA antigens. In: Woodrow GC, Levine MM (eds.), New Generation Vaccines. New York: Marcel Dekker lnc; 1990: 161-188. 13. Wright RM, John E, Klotz K. Flickinger CJ, Herr JC. cDNAs for acrosomal protein SP-10. Biol Reprod 1990; 42:693701 14. Freemerman AJ, Wright RM, Flickinger CJ, Herr JC. Cloning and sequencing of baboon and cynomolgus monkey intraacrosomal protein SP10: homology with human SPlO and a mouse sperm antigen (MSA-63).Mol Reprod Dev 1993; 34:14&148. 15. Nakayama K, Kelly SM, Curtiss 111R. Constructionof an asdf expression-cloningvector: stable maintenance and high level expression of cloned genes ~na Salmonella vaccine strain. Biotechnology 1986; 6:693497. 16. Galan JE, Nakayama K, Curtiss 111 R. Cloning and characterization of the asd gene of Salmonella typhimurium: use in stable maintenance of recombinant plasmids in Salmonellavaccine strains. Gene 1990; 94:29-35. 17. Curtiss 111 R. Chromosomal alterations associated with mutations to bacteriophage resistance in Eschericbfacoli.J Bacterial 1965;89:2@0. 18. Sarnbrook J, Fritsch EF, Maniatis T, Molecular Cloning: A Laboratory Manual. 2nd ed. New York: Cold Spring Harbor Press; 1989. 19. Yanisch-Perron C, Vieria J, MessingJ. lmprovedM13phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene 1985; 33:105119. 20. Griffin PD, Johnson PM. The WHO special programme of research in human reproduction, its task force onvaccines for fertility regulation and the current status of birth control vaccine development:specific issues regarding local immunity in reproductive tract tissues. In: Griffin PD, Johnson PM (eds.), Local Immunity in Reproductive Tract Tissues. WHO publication: 1993: 1-8. 21. Clarke GN. Detection of anti spermatozoa1 antibod~esof IgG, IgA and IgM irnmunoglobulin classes in cervical mucus. An1J Reprod Immunol 1984; 5:179-181. 22. Goldberg E. Infetlillty in female rabbits immunized with lactate dehydrogenase X. Science 1973; 181:458-459. 23. O'Rand MG. Inhibition of fertility and sperm zona binding by antiserum to the rabbit sperm-membraneautoantigen RSA-1. Biol Reprod 1981; 25:621628. 24. Naz RK. The fertilization antigen (FA-]) causes a reduction of fertility in actively immunized female rabbits.J Reprod Immunol 1987; 11:117-133. 25. Prinlakoff P, Lathrop W, Woolman L, Cowan A, Myles D. Fully effective contraception in male and female guinea pigs immunized with the sperm protein, PH20. Nature 1988; 335543546, 26. Shaha C, Suri A, Talwar GP. Identification of sperm antigens that regulate fertility. Int J Androl 1988; 11:47!+491. 27. Henderson CJ, Hulme MJ, Atken RJ. Contraceptive potential of antibodies to zonapellucida. J Reprod Fertil 1989;83:32>343. 28. Yang SL, Schumacher GFB. Immune response after vaginal application of antigens in the rhesus moneky. Fertil Steril 1979; 32:5&598. 29. Pam EL, Pam MB. A comparison of antibody titers in mouse uterine fluid after immunization by several routs, and the effect of the uterus on antibody titers in vaginal fluid.J Reprod Fenil1990; 89:61G25.
471
30. Brown GA, Hormaeche CE, Demarco de Hormaeche R, Winther M, Dougan G, aske ell DJ, Stocker BAD. An attenuated aroA Salmonella typhimunum vaccine elicits humoral and cellular immunity to cloned b-galactosidase in mice. ] Infect Dis 1987; 15586-92. 31. ClemetsJD, Cardenas L. Vaccines against enterotoxigenicbacterial pathogens based on hybrid Salmonella that expresses heterologousantigens Res Microbiol 1990; 141:981-993. 32. Curtiss III R, Kelly SM, Gulig PA, Nakayama K. Selective delivery of antigens by recombinant bacteria. Curr Top Microbiol lmmunol 1989; 146:3519. 33. Maskell DJ, Sweeney KJ, O'Callaghan D, Hormaeche CE, Liew Pi,Dougan G. Salmonella typhimt6tiumaroAmutans as carriers of the EscherichiacoLiheat-labileexterotoxin B subunit to the murine secretory and systemic immune systems. Microb Pathog 1987; 2211-221. 34. Doggett TA, Jagusztyn-Krynicka EK, Curtiss I11 R. Immune responses to St~ptococcus sobn'nussurfaceprotein antigen A expressed by recombinant Salmonella typhimun'um Infect Irnmun 1993;61:1859-1866. 35. Katz J, Michalek SM, Curtiss I11 R, Harmon C, Richardson G, Mestecky J. Novel oral vaccines: the effectiveness of cloned gene products on inducing secretory immune responses. Adv Exp Med Biol1987; 216B:1741-1747. 36. Mestecky J. The common mucosal irnnlune system and current strategies for induction of immune response in external secretions.J Clin Immunol 1987; 7265-276. 37. Brandtzaeg P. Ovewiew of the mucosal immune system. Curr Top Microbiol Immunol 1989; 146:1325. 38. ScicchitanoR, Stansz A, Ernst P, Bienenstock S . A common mucosal system revisited. In. Husband AJ (ed.), Migration and Homing of LyrnphoidCells.Boca Raton, FL CRC Press; 1988: 12-24. 39. Kroese RGM, Butcher EC, Stall AM, Lalor PA, Adama S, Herzenberg LA. Many of the IgA producing plasma cells in murine gut are derived from self replenishing precursors in the peritoneal cavity. Int Immunol 198% 1:7584 40. Burton 2, Burgess RR, Lin J, Moore S, Holder S, Gross CA. The nucleotide sequence of the cloned tpoDgene for the RNA polymerase sigma subunit from E coli K12. Nucleic Acids Res 1981; 9:2883-2903. 41. Ishii S, Ihara M, Maekawa T, Nakamura Y, Uchida H, Imamoto F. The nucleotide sequence of the cloned nus A gene and its flanking region of Escherichia coli. Nucleic Acids Res 1984; 12:33333342. 42. Hollingshead SK, Fischetti VA, Scott JR. Complete nucleotide sequence of type 6M protein of the group A streptococcus.J Biol Chem 1986, 261:1677-1686. 43. Wira CR, Stem J. Endocrine regulation of the mucosal immune system in the female reproductive tract: control of IgA, IgG and secretory component during the reproductive cycle, at implantation and throughoutpregnancy.In: PasqualiniJR,SchollerR(eds.), Hormones and Foetal Pathophysiology Marcel Decker, Inc; 1991: 345367 44. MesteckyJ, MacGhee JR. The secretory IgA system. In: Griffin PD, Johnson PM (eds.), Local Immunity in Reproductive Tract Tissues. Oxford University Press; 1993: 5+72. 45. Mestecky J, MacGhee JR. llnmunoglobulin A (IgA): molecular and cellular interactions involved in IgA biosynthesis and immune response. Adv [nimunol 1987; 40:153-245.
