Protective immunity induced by an intranasal multivalent vaccine comprising 10 Lactococcus lactis strains expressing highly prevalent M-protein antigens derived from Group A Streptococcus

Abstract

Streptococcus pyogenes (group A Streptococcus) causes diseases ranging from mild pharyngitis to severe invasive infections. The N-terminal fragment of Streptococcal M protein elicits protective antibodies and is an attractive vaccine target. However, this N- terminal fragment is hypervariable and there are more than 200 different M types. We are developing an intranasal live bacterial vaccine comprised of 10 strains of Lactococcus lactis, each expressing one N-terminal fagment of M protein. Live bacterial-vectored vaccines have lower associated costs because of its less complex manufacturing processes compared to protein subunit vaccines. Moreover, intranasal administration does not require syringe or specilized personnel. The evaluation of individual vaccine types (M1, M2, M3, M4, M6, M9, M12, M22, M28 and M77) showed that most of them protected mice against challenge with virulent S. pyogenes. All of the 10 strains combined in a 10-valent vaccine (Mx10) induced serum and bronchoalveolar lavages IgG titers that ranged from 3 to 10-fold those of unimmunized mice. Survival of Mx10-immunized mice after intranasal challenge with M28 streptococci is significantly higher than unimmunized mice. In contrast, when mice were challenged with M75 streptococci, survival of Mx10-immunized mice was not significantly different from unimmunized mice. Mx-10 immunized mice were significantly less colonized with S. pyogenes in oropharyngeal washes and developed less severe disease symptoms after challenge compared to unimmunized mice. Our L. lactis-based vaccine may provide an alternative solution to the development of broadly protective group A streptococcal vaccines.

INTRODUCTION

Streptococcus pyogenes (S. pyogenes) is responsible for a wide range of illnesses ranging from uncomplicated pharyngitis and impetigo to toxic shock syndrome and necrotizing fasciitis (1). Moreover, autoimmune sequelae like acute rheumatic fever (ARF) and rheumatic heart disease (RHD) may arise after an inappropriately treated infection (2, 3). Streptococcal pharyngitis is not a severe illness but is very common and is associated with high costs (4). In contrast, toxic shock syndrome and necrotizing fasciitis occur much less frequently but are very severe with mortality rates of 10–32% (5). ARF and RHD following a S. pyogenes infection are effectively controled in most developed countries. However, they are still frequent in developing countries and indigenous populations worldwide (6, 7). Despite ongoing efforts, there is no vaccine currently available to prevent these infections and their complications.

One of the most important virulence factors of this pathogen is M protein, which is involved in phagocytosis resistance through impairment of complement deposition on the bacterial surface (8). This protein is hypervariable in its N-terminal region, resulting in more than 200 different M protein types (9). Antibodies produced against this hypervariable region have the greatest protective efficacy against infection (10, 11). For this reason, M protein has been considered an attractive vaccine target (12). However, anti-S. pyogenes vaccine development faces safety concerns related to the possibility that vaccines may elicit autoinmune responses like ARF and RHD. Regions of the M protein, other than the hypervariable one, mimic human epitopes and are responsible for autoinmune disorders (13). The identification of these regions made possible the design of safe vaccines that exclude these “harmful” regions and are least likely to cross-react with human tissues (14). Multivalent M protein subunit vaccines delivered parenterally have been evaluated in clinical trials and elicited serum antibody responses (15, 16). Protein subunit vaccines have high production costs because they require complex manufacturing processes. An alternative lower cost approach is lactic acid bacteria expressing heterologous antigens (17, 18). We are developing an intranasal live bacterial vaccine comprised of 10 strains of Lactococcus lactis, each expressing a different M protein fragment. We have previously reported a live bacterial vaccine based in L. lactis expresing the hypervariable region of M9 protein (19). We have now evaluated nine additional strains of L. lactis expressing different M protein fragments individually (M1, M2, M3, M4, M6, M12, M22, M28 and M77) and all of them combined in a 10-valent vaccine, Mx10. We report here that Mx10 vaccine elicits specific serum and mucosal antibodies against most M types included in the vaccine and it protects against lethal challenge with virulent S. pyogenes M28.

MATERIALS AND METHODS

Strains, media and culture conditions

Each vaccine strain was constructed using the NICE genetic system (NICE System, MoBiTech, Germany). Portions of the A domain of M protein of 41 to 50 amino acids long were PCR amplified from S. pyogenes of M types 3, 6, 22 and 77 using primers with restriction enzyme sites. All of the strains used were clinical pharyngeal isolates. Epitopes responsible for autoimmune diseases are localized in the B domain and some regions of the A and C domains (20). The sequences were analyzed to verify the absence of sequences resembling human proteins and likely to elicit autoreactive antibodies. The portion of the A domain used to construct each vaccine correspond to the hypervariable region and is shown in Table 1. In order to achieve surface display of the M protein, the hypervariable peptide was fused to 148 amino acids of the C-terminal membrane anchoring domain of M6 protein. The resulting fusion fragment was ligated to the linearized vector pNZ8149 and introduced into L. lactis strain NZ3900 through electroporation. Transformed clones were selected in M17-lactose-bromocresol purple. The L. lactis wild-type (WT) control strain was obtained after transformation of L. lactis strain NZ3900 with the empty plasmid pNZ8149. The presence of the plasmid bearing the insert was verified through PCR. For protein expression, recombinant L. lactis strains were grown at 30°C in M17-lactose broth supplemented with nisin 20 ng/ml. Three hours after nisin addition, cultures were centrifuged and pellets sonicated. The resulting crude extracts were centrifuged and analyzed through 12.5% polyacrylamide gel electrophoresis with SDS (SDS-PAGE). All of the vaccine strains expressed a protein of the expected size (Figure 1A).

Table 1.

Sequence and size of the hypervariable M protein fragment included in each vaccine and isolates used to generate each vaccine strain

M protein	Sequence/final position number (in original S. pyogenes protein with signal peptide)	Hypervariable M protein fragment length (aa)	Expected Protein Molecular Weight (kDa)	Reference
M1	NGDGNPREVIEDLAANNPAIQNIRLRHENKDLKARLENAM/81	40	24.6	(21)
M2	NSKNPVPVKKEAKLSEAELHDKIK/65	24	22.7	(21)
M3	DARSVNGEFPRHVKLKNEIENLLDQVTQLYTKHNSNYQQYNAQAGRLDLR/91	50	25.14	(22)
M4	AEIKKPQADSAWNWPKEYNALLKENEELKVEREKYLSYADDKEKDPQYRA/91	50	26.1	(21)
M6	RVFPRGTVENPDKARELLNKYDVENSMLQANNDKLTTENKN/83	41	24.16	(22)
M9	EGVKKAEEAKLSVPKTEYDKLYDDYDKLQEKSAEYLERIGELEERQ/87	46	25.5	(21)
M12	DHSDLVAEKQRLEDLGQKFERLKQRSELYLQQYYDNKSNGYKGDWYVQQ/90	49	27.6	(21)
M22	ESSNNAESSNISQESKLINTLTDENEKLREELQQYYALSDAKEEEPRYKA/91	50	25.08	(22)
M28	AESPKSTETSANGADKLADAYNTLLTEHEKLRDEYYTLIDAKEEEPRYKA/83	50	26.1	(21)
M77	EGVSVGSDASLHNRITDLEEEREKLLNKLDKVEEEHKKDHEQ/83	42	24.24	(22)

Figure 1. SDS-PAGE analysis of heterologous M protein expression.

A: Cultures were induced for 3 h with nisin (20 ng/ml). Crude extracts of each vaccine strain were analyzed in a 12.5% polyacrylamide gel that was stained with Coomassie blue. Bands corresponding to each M protein fragment are indicated with a white arrow. B: The amount of each expressed M protein was estimated using a BSA (Bovine Serum Albumin) calibration curve. Ld, protein molecular mass standard ladder. V, crude extract of L. lactis bearing the empty plasmid (no M protein). The molecular mass (kDa) of each band of the standard is shown on the left side of the gel.

To calculate the amount of each M peptide, protein concentration in lysates of each vaccine strain was calculated using Bradford solution (Protein Assay Solution, Bio-Rad) in a microtiter plate read at 620 nm. BSA standard solutions of known concentrations were prepared (1, 1.5 and 2 mg/ml) and analyzed together with vaccine lysates in a 12.5% polyacrilamide gel with SDS. Bands of the expected molecular weight for BSA and samples were quantified through densitometry with the program Quantity One version 4.6.3 (Bio-Rad). A calibration curve was constructed with values obtained from BSA standards. The amount of recombinant M protein in each vaccine strain was extrapolated from the calibration curve and calculated with the program GraphPad Prism 5.03.

For the preparation of the infective inoculum, an erythromycin-resistant S. pyogenes clinical isolate of each M type that had been previously reported (21, 22) was grown in Todd-Hewitt broth at 37°C.

Immunization and challenge of mice

All animal experiments were performed according to Ethics and Biosafety Committee of the Pontificia Universidad Católica de Chile approved protocols. For each single M type and L. lactis-WT vaccine dose preparation, 1 ml of freshly induced culture (with 20 ng/ml nisin) was centrifuged and resuspended in 50 μl of PBS, which was used to immunize one mouse. For the 10-valent vaccine, each immunization dose was divided in two doses of five different M types to facilitate manipulation and ensure an appropriate immunization. Dose 1 was composed by the mixture of M1, M2, M3, M4, M6 and dose 2 by M9, M12, M22, M28 and M77 and it was prepared by mixing 1 ml of each M type, centrifuged and resuspended in 250 μl of PBS (Phosphate Buffered Saline), which was used to immunize five mice (1/5 of each M type per dose). Therefore mice receive 2/5 of the each M type. Dose 2 was administered one week after Dose 1. Groups of 3–4 female BALB/c mice (6 to 8 weeks-old) were intranasally immunized with 10⁹ CFU (colony forming units) of each dose on days 0, 14 and 28. Unimmunized mice received 50 μl of PBS. Two independent experiments of each evaluation were performed, therefore data of 6–8 mice per group is shown. Twenty-one days after the third immunization, intranasal challenge in a total volume of 50 μl of PBS was performed (25 μl in each nostril). To avoid swallowing the vaccine or infective dose, mice were intraperitoneally anesthetized with a mixture of ketamine and xylazine (20 mg/kg and 1 mg/kg, respectively). For the individual vaccine evaluation, infective doses of S. pyogenes of each M type ranged from 2 × 10⁸ to 7 × 10⁸ CFU. For the evaluation of Mx10 vaccine, mice were challenged with 7 × 10⁸ CFU of S. pyogenes type M28 and 4 × 10⁸ CFU of type M75. Infective doses of each M type of S. pyogenes were determined before immunization-challenge. However, lethal doses were obtained only for vaccines, M1, M2, M4 and M28 (data not shown).

Blood samples were obtained from all animals at days 0, 28 and 49. Blood was incubated for 1 h at 37°C and centrifuged for 5 min at 5,000 rpm. Serum was transferred to a clean tube and stored at 4°C. Body weight was determined each day after challenge until sacrifice at day 6. Mouth washes were performed for three consecutive days after challenge: 50 μl of PBS was pipetted 5 times up and down into the mouth of mice and finally plated on blood agar supplemented with 2 μg/ml erythromycin. All of the S. pyogenes strains used to challenge, were erythromycin resistant. Six days after challenge, mice were sacrificed and bronchoalveolar lavages (BAL) were performed.

The clinical severity score was determined based in the most important signs of discomfort and illness: weight loss, hypothermia, general appearance, and inability to move and react upon stimulus (23). Balb/c mice that are ill are less reactive upon cage opening and look “wet” because they do not clean their fur (24, 25). We used a clinical score composed of behavior, weight and general appearance. These variables were scored with 0: unaltered (healthy); 1: slightly affected; 2: affected and 3: very affected (severe illness).

Immunogenicity evaluation

Each M type-specific IgG was determined by whole-bacterial cell enzyme-linked immunosorbent assay (ELISA). S. pyogenes strains were grown overnight at 37°C/5% CO₂ in blood agar and resuspended in PBS pH 7.4 until an optical density at 600 nm (OD₆₀₀) of 0.34 was reached. The suspension was inactivated at 65°C for 30 min and bound to flat-bottom 96-well plates (50 μl/well) overnight at room temperature. For the peptide ELISA, synthetic peptides (5 μg/ml) (Peptide 2.0, Inc., Chantilly, VA) were bound to microtiter plates in 0.1 M sodium carbonate-bicarbonate buffer, pH 9.6 (100 μl/well), overnight at room temperature. Excess bacteria or peptide were removed through five washes with PBS–0.05% Tween 20 (PBS-T). Wells were blocked with PBS–3% bovine serum albumin (BSA) for 2 h at room temperature and then washed 3 times with PBS-T. Mouse sera was serially diluted 1:2 with PBS–1% BSA starting from 1/100, and 100 μl of each dilution was added to the wells and incubated for 1 h at room temperature. Mouse BALs were serially diluted 1:2 starting from undiluted fluid. Saliva was used undiluted. The wells were washed 3 times with PBS-T, then 100 μl of enzyme conjugate (horseradish peroxidase conjugated with anti-mouse IgG) was added, and the mixture was incubated for 1 h at room temperature. Wells were washed 3 times with PBS-T, 100 μl of substrate solution (tetramethylbenzidine - TMB; Thermo Scientific) was added, and the mixture was incubated for 20 min in the dark at room temperature. The reaction was stopped with 100 μl H₂SO₄ 2N, and the plates were read at 450 nm. The antibody titer was defined as the reciprocal of the highest dilution of serum that yielded an absorbance equal to or higher than pre-immune sera (around 0.1 at 450 nm). For saliva, the absorbance measurement was reported.

RESULTS

Vaccine construction and M protein expression

Vaccine strains expressing M1, M2, M4, M9, M12 and M28 were previously constructed (19), whereas vaccine strains expressing M3, M6, M22 and M77 were constructed in the present study. As shown in Figure 1A, a protein of the expected size was overexpressed by each of the 10 vaccine strains in the presence of 20 ng/ml of nisin. The amount of protein was estimated using a standard curve of known BSA concentrations. Different levels of protein expression were obtained as shown in Figure 1B. Vaccine types M9, M22, M28 and M77 showed the highest expression levels.

Efficacy evaluation of each 10 vaccine strains individually

Nine of the 10 vaccines were evaluated individually in Balb/c mice which were immunized 3 times every 14 days. Twenty-one days after the last immunization, mice were intranasally challenged. Blood and saliva samples were obtained on days 28 and 49. Antibody titers were determined in a whole bacterial cell ELISA assay according to Araya and coworkers (26) (Figure 2). All vaccine types induced serum IgG titers that ranged from 2 to 44-fold higher than those of unimmunized mice (Figure 2A). Mean values of IgG titers in BAL were 2–15 fold higher compared to unimmunized mice, and the highest values were obtained for M2 and M4 vaccine types (Figure 2B). The individual evaluation of vaccine types demonstrated that high serum antibody response of individually evaluated vaccines is not always associated to high BAL antibody response and vice versa, as observed for vaccine strains M2 and M77 (Figure 2A and B).

Figure 2. Immunogenicity of vaccine strains evaluated individually.

A: IgG antibody titers against vaccine M types of S. pyogenes in sera from mice immunized with the L. lactis-M or unimmunized. Serial two-fold dilutions starting from 1/100 were analyzed. B: IgG titers in BALs (Bronchoalveolar Lavages) from mice 6 days post-challenge were analyzed using two-fold dilutions starting from undiluted BAL. The titer is defined as the reciprical of the end-point dilution. Black sqares: L. lactis-M; White sqares: Unimmunized. Each bar represents the mean ± standard error of measurements from 6 to 8 mice in each group. Vaccines M3, M6 and M77 were evaluated with 4 mice per group. *, P value < 0.05; **, P value < 0.005, by Student t test between values of mice immunized with the L. lactis-M vaccine and unimmunized mice. ND, Not Determined.

Body weight was determined for 6 days after challenge and percentage of body weight loss is shown in Figure 3A. Weight losses of 10% are considered acceptable according to evaluation of other vaccines in Balb/c mice (27). Mice immunized with most of the vaccines were protected against challenge with virulent S. pyogenes as evidenced by a significantly lower body weight loss than unimmunized mice. Antibody response is not associated with protection against challenge as observed for vaccine type M22, which elicited low levels of antibodies (Figure 2A) but protection against infection was achieved (Figure 3A). Mice immunized with M3, M4, M9, M12, M22 and M28 had fewer CFU of S. pyogenes in mouth washes performed 1–3 days post-infection compared to unimmunized mice (Figure 3B).

Figure 3. Efficacy evaluation of vaccine strains tested individually.

A: Body weight loss 3–5 days after intanasal challenge. B: Number of S. pyogenes CFU (Colony Forming Units) recovered from oropharyngeal washes performed 2 days after challenge. Black squares: L. lactis-M; White sqares: Unimmunized. Each bar represents the mean ± standard error of measurements from 6 to 8 mice in each group. Vaccines M3, M6 and M77 were evaluated with 4 mice per group. *, P value < 0.05; **, P value < 0.005, by Student t test between values of mice immunized with the L. lactis-M vaccine and unimmunized mice.

Efficacy evaluation of the combined 10-valent vaccine (Mx10)

The Mx10 vaccine induced serum IgG titers that ranged from 3 to 10-fold higher compared to unimmunized mice. Among the 10 M types used for mice immunization, M2, M6, M12 and M77 vaccine types elicited the highest serum antibody titers (Figure 4A). Vaccine types M6 and M2 produced low antibody titers in the individual evaluation but antibody levels were higher in response to the Mx10 vaccine. In contrast, serum antibody titers for M4 and M28 were lower in Mx10 immunized mice compared to mice that received the individual vaccine components. Serum IgG titers were evaluated against two non-vaccine types, M81 and M11 in order to test the specificity of the antibodies. Serum from mice immunized with Mx10 elicited IgG titers against M81 and M11 that were 2 to 10-fold lower compared to titers against most M types included in the vaccine (Figure 4A). These results indicate that the antibodies were for the most part directed against the M peptides expressed by the L. lactis strains. To further assess the specificity of the antibodies, serum was assayed against purified peptides. IgG titers in serum of Mx10-immunized mice were significantly higher than titers of unimmunized mice (Figure 4B). IgG antibodies were measured in BAL and they were 3 to 7-fold higher compared to unimmunized mice (Figure 4C). IgG antibody levels in saliva of Mx10-immunized mice were significantly higher than in unimmunized mice for types M2 and M12 (Figure 4C lower table). The volume of saliva recovered was insufficient to determine antibody levels against all vaccine M types. IgA was not detected in saliva or BAL.

Figure 4. Immunogenicity of Mx10 vaccine.

A and B: IgG antibody titers against whole S. pyogenes of the 10 M types (A) and against synthetic M peptides (B) in sera from mice immunized with Mx10, L. lactis-WT or unimmunized. C: IgG antibody titers in BAL of mice 6 days post-challenge were analyzed using two-fold dilutions starting from undiluted BAL. IgG antibody levels in saliva of mice obtained 14 days after the last immunization using undiluted saliva is shown and are expressed as Absorbance at 450nm x 100 (C lower table). Black squares: L. lactis–M; Grey squares: L. lactis–WT; White squares: Unimmunized. Each bar represents the mean ± standard error of measurements from 6 to 8 mice in each group. *, P value < 0.05; **, P value < 0.005, by Student t test between values of mice immunized with the L. lactis-M vaccine and unimmunized mice. ND, Not Determined.

One group of Mx10-immunized mice was challenged with S. pyogenes M28 (Figure 5, left panel) and another group was challenged with S. pyogenes M75 (Figure 5, right panel) (M type not included in the vaccine). Oropharyngeal colonization with S. pyogenes M28 was significantly different between Mx10-immunized and unimmunized mice 2 days post-infection: Mx10-immunized mice were only mildly or no colonized and none of them was heavily colonized (Figure 5A). In contrast, around 30% of unimmunized mice were heavily colonized with S. pyogenes (Figure 5A). Mice immunized with Mx10 and challenged with S. pyogenes M75 were heavily colonized by S. pyogenes (Figure 5B). By day 3 most mice were mildly or not colonized. The clinical severity score of mice was determined after challenge and ranged from 0 (healthy) to 3 (severe illness). Figure 5C and 5D show that mice immunized with Mx10 and challenged with M28 reached a score of 2 two days after challenge but immediately returned to a healthy score of 0. In contrast, unimmunized mice had a score of 3 until the end of the experiment. Mice immunized with L. lactis-WT reached a score of 3 at day 2 post-challenge and improved the following days but did not return to normality (score of 0) (Figure 5C). The clinical score of unimmunized and Mx10-immunized mice were significantly different throughout the experiment course (Figure 5C). Mice immunized with Mx10 and challenged with M75 reached a score higher than 2 and did not return to normal score, similar to mice immunized with L. lactis-WT (Figure 5D). After challenge with S. pyogenes M28, survival of Mx10-immunized mice is significantly higher from that of unimmunized mice (p=0.0014) whereas survival of L. lactis WT-immunized mice was not significantly different from survial of unimmunized mice (p=0.2657) (Figure 5E). In contrast, when mice were challenged with S. pyogenes M75, survival of unimmunized mice was not significantly different from Mx10-immunized mice (p>0.999), and from L. lactis WT-immunized mice (p>0.999) (Figure 5F).

Figure 5. Protective immunogenicity of Mx10 vaccine.

A and B: Number of Mx10-immunized, L. lactis-WT-immunized or unimmunized mice colonized with S. pyogenes assayed through culture of mouth washes performed 2 and 3 days after challenge with S. pyogenes M28 (A) or S. pyogenes M75 (B). *, P value < 0.05 by Fisher’s exact test between number of mice immunized with the L. lactis-Mx10 vaccine and unimmunized mice. C and D: Clinical score of mice immunized with Mx10, L. lactis-WT or unimmunized after challenge with S. pyogenes M28 (C) or S. pyogenes M75 (D). **, P value < 0.005 by Student t test between values of mice immunized with the L. lactis-Mx10 vaccine and unimmunized mice. E and F: Survival of mice during the 6 days after challenge with S. pyogenes M28 (E) or with S. pyogenes M75 (F). **, P value < 0.005 by Fisher’s exact test between final survival values of mice immunized with the L. lactis-Mx10 vaccine and unimmunized mice. Black squares: L. lactis–Mx10; Grey squares: L. lactis-WT; White squares: Unimmunized; White circles in E and F: Uninfected mice. Each symbol represents the mean value of measurements from 6 to 8 mice in each group. Colonization assay in A was done with 12 mice per group. Bars represent standard error.

DISCUSSION

In the current study, we have shown that L. lactis strains engineered to express truncated fragments of streptococcal M protein elicit immune responses in mice that confer protection against challenge infections delivered intranasally. Whether the major contribution to protection was serum or mucosal antibody responses is not clear. For example, the mean serum titer of anti-M28 antibodies in mice immunized with L. lactis-M28 was higher than that in mice immunized with Mx10. Additionally, the mean BAL fluid titer of anti-M28 was ~4-fold higher in mice immunized with Mx10 over those immunized with the L. lactis-M28 component. Yet both groups of mice showed 100% survival following challenge infections, suggesting that protection may be mediated by immune responses at either location. The amount of M protein in each vaccine dose of the individual evaluation was ≥ 10 μg. However, the amount of each M protein in the combined Mx10 vaccine is 2/5 of the amount used in individual tests. In spite of this, the immune response and the protective efficacy was maintained, at least for M28 type.

Another variable in assessing the contributions to protective efficacy of these vaccines is the observation that L. lactis alone appears to confer some level of protection against challenge infections with S. pyogenes. The protection observed in L. lactis-WT-immunized mice could be due to the fact that intranasal administration of live bacteria, despite being “food-grade”, produces a general state of inflammation that augments the innate immune response upon challenge. It is also possible that the L. lactis vector expresses surface antigens that evoke mucosal antibodies that cross-react with S. pyogenes antigens. In fact, sequence homology between these species reveals that L. lactis is closely related to S. pyogenes (28). These results suggest that L. lactis may have potential immunological value in vaccine design because their pathogen-associated molecular patterns may act as potent enhancers of innate immune responses. However, this is not enough to achieve full protection against S. pyogenes challenge because mice immunized with L. lactis-WT had a lower percent survival than Mx10-immunized mice. Therefore, specific anti-M protein antibodies appear necessary for optimal protection against lethal infection in this model.

It was recently demonstrated that parenteral immunization did not induce mucosal immunity and did not protect mice against intranasal challenge with S. pyogenes (29). When the same antigen was administered intranasally, mice induced specific IgG and IgA in BAL and were protected against intranasal challenge (29). Although we did not detect IgA in Mx10-immunized mice, we did detect significant levels of IgG in BAL. It may be that mucosal immunity plays an important role in protection against infection in this mouse model, independent of antibody type.

Previous studies have shown that multivalent M protein-based vaccines evoke functionally cross-reactive antibodies against a number of non-vaccine M types of S. pyogenes (30). This observation was consistent with the new cluster-based typing system of M types that is based on sequence similarities and similar patterns of binding plasma proteins (31). Subsequent studies showed that M types within a cluster could be cross-opsonized by antisera against different M peptides within the same cluster (32). In the current study, we found that mice immunized with Mx10 vaccine were not cross-protected against S. pyogenes M75, which is a non-vaccine type from cluster E6. The Mx10 vaccine does not contain any M proteins from cluster E6, yet does contain four M types from cluster E4 (M2, M22, M28, and M77). It is possible that antibodies evoked by any one of these E4 cluster M proteins contributed to protection against M28 challenge infections.

The vaccine presented here potentially covers around 85% of total infections (pharyngeal and invasive) in Chile (21) and in countries with similar M-type distributions like Mexico (33), Argentina (34), Brazil (35) and United States (36). L. lactis-based vaccines may provide an alternative solution to the development of broadly protective group A streptococcal vaccines.

LIST OF ABBREVIATIONS

ARF

Acute Rheumatic Fever

BAL

Bronchoalveolar Lavage

BSA

Bovine Serum Albumin

ELISA

Enzyme-linked Immunosorbent Assay

ND

Not Determined

PBS-T

PBS–0.05% Tween 20

PBS

Phosphate Buffered Saline

RHD

Rheumatic Heart Disease

SDS-PAGE

polyacrylamide gel electrophoresis with SDS

SDS

Sodium Dodecyl Sulfate

TMB

3,3′, 5,5;-tetramethylbenzidine

WT

Wild Type