Introduction

Lyme disease is a tick-borne illness caused by the spirochete Borrelia burgdorferi sensu lato (Bbsl), (Borreliella genus novum under consideration1). The most effective way to avoid this disease is to avoid tick-infested areas. Additional disease control strategies are necessary to reduce incidence of Lyme disease including development of safe vaccines for human use.

Outer surface protein A (OspA) is the only immunogen proven to provide 76–92% protection2,3 against tick-transmitted B. burgdorferi in fully vaccinated human subjects. Mechanistically, antibody to OspA produced upon vaccination of a host is ingested with the bloodmeal by a feeding tick and it blocks transmission of B. burgdorferi from the tick to the mammalian host4,5,6,7. The titer of anti-OspA antibody in the host’ blood must be at or above a certain level to be effective2,8. In the first clinical trials2,3, the immunization protocols required 2 shots within 1 month and a third shot 12 months after prime. After a 20-year hiatus, a re-engineered vaccine based on the C-terminus sequence of OspA from several Bbsl genospecies (VLA15)9 is undergoing Phase III clinical trials by Pfizer/Valneva. To maintain sufficient protective antibody levels over time, the VLA15 Phase I immunization protocol (NCT03010228) required 3 intramuscular shots at days 1, 29 and 5710. Other immunization protocols are being tested that require a 4th shot 12 months after prime (NCT05477524). These immunization protocols further substantiate the initial finding that parenteral OspA-based vaccines do not induce prolonged immunity.

Parainfluenza virus 5 (PIV5) is a nonsegmented, negative-strand, RNA virus of the family Paramyxoviridae11. A distinctive trait of PIV5 is its ability to infect most mammalian cell types through sialic acid receptors without causing cytopathic effect. This allows for active replication of PIV5 in the respiratory tract after intranasal immunization, leading to the induction of mucosal immunity via the generation of antigen-specific IgA antibodies and long-lived IgA plasma cells, in addition to systemic humoral IgG and cell-mediated immune responses12,13. Live PIV5, not replication deficient, has been a component of the kennel cough vaccine administered to dogs intranasally for over 50 years. Due to potential shedding of the vaccine up to 5 days post-administration14, humans have been exposed to PIV5 since the vaccine came to market. However, no disease has been reported in exposed individuals, highlighting PIV5 excellent safety record. Previously, live PIV5-vectored vaccines for bacterial pathogens like Mycobacterium tuberculosis and Burkholderia mallei and B. pseudomallei have been tested and proven efficacious in a mouse model15,16,17. Furthermore, two PIV5-vectored vaccines for COVID-19 and respiratory syncytial virus (RSV) (CVXGA1 and BLB201, respectively) have undergone phase I clinical trials. Here, we describe the construction of a live PIV5-vectored vaccine expressing OspA from B. burgdorferi sensu lato. We show longer-lasting protection from challenge with B. burgdorferi-infected ticks in mice vaccinated intranasally with 2 doses of the vaccine than in mice vaccinated parenterally with 2 doses of OspA protein.

Results

Generation and characterization of PIV5-based vaccines expressing the outer membrane protein A (OspA)

We generated the PIV5-AB31 and PIV5-ABPBPk vaccine candidates by inserting the ospA gene sequence in the PIV5 genome between the SH and HN genes (Fig. 1a). PIV5-AB31 carried the full-length OspA sequence from B. burgdorferi strain B31. PIV5-ABPBPk carried a full-length chimeric sequence of OspA in which the B. burgdorferi B31 amino acid sequences 165-189 and 219-273 were replaced with the respective sequences from B. afzelii strains (BPBPk)18. The N-terminus and transmembrane domain sequences from the PIV5 HN protein were also introduced directly upstream of the ospA gene in both vaccine candidates to improve incorporation of the bacterial protein into the PIV5 virion. The vaccine viruses were rescued as described19, and their genomes were confirmed through RT-PCR and sequencing. Expression of the OspA protein in PIV5-AB31- and PIV5-ABPBPk-infected cells was confirmed through Western blot assay (Fig. 1b).

Fig. 1: Generation and characterization of PIV5-AB31 and PIV5-ABPBPk vaccine candidates.
figure 1

a Schematic of vaccine candidates. PIV5 has seven genes encoding for eight proteins, 3’ leader, NP, V/P, M, F, SH, HN, L, 5’ trailer. The OspA B31 and BPBPk proteins contain the N-terminus (NT) and transmembrane domain (TM) sequence from PIV5 HN protein. b Detection of OspA expression. Vero cells were mock-infected or infected with PIV5 vector control, PIV5-AB31, or PIV5-ABPBPk at an MOI of 1. Forty-eight hours after infection the cells were lysed, and the lysates resolved on an SDS-PAGE gel and immunoblotted with anti-OspA (184.1) and anti-PIV5 NP (NP214mAb) monoclonal antibodies diluted 1:100. The gel and blot derived from the same experiment, and they were processed sequentially.

Full size image

Analysis of B. burgdorferi strain variability in I. scapularis used for tick challenge

Sequencing OspC genes from the 3 colonies of I. scapularis ticks maintained in the laboratory for tick challenges (Fig. 2) showed that multi-strain MS’08/NY ticks mostly carried 10 types of OspC (A, B, Ba, D, E, I, M, Q, W, X) with >1000 unique read counts, whereas the other OspC types had <900 unique reads. MS’21/MA ticks mostly carried 19 types of OspC (A, B, C, D, E, F, Fa, Fb, G, H, I, J, K, L, M, N, O, W, X) with >1000 unique read counts, whereas the other two OspC types had <900 unique reads. MS’21/NY ticks mostly carried 13 types of OspC (A, D, E, F, Fa, Fb, H, I, J, K, L, M, O) with >1000 unique read counts, whereas the other OspC types had <900 unique reads.

Fig. 2: Heat map analysis of B. burgdorferi (Bb) strain variability in laboratory maintained Ixodes scapularis nymphal ticks.
figure 2

DNA purified from flat nymphal ticks was used for amplification and sequencing of the ospC gene by the Proton Ion Torrent Instrument. Legend: I.s., Ixodes scapularis; MS’08 (NY), MS’21 (MA) and MS’21 (NY) - the Bb culture used to produce the ticks originated from tissues from mice infected with ticks flagged in NY 2008, in MA 2021 and in NY 2021.

Full size image

Immunization with intranasal PIV5-ABPBPk or PIV5-AB31 induces long-lasting neutralizing antibody responses

To assess humoral IgG immune responses induced by intranasal PIV5-ABPBPk, intranasal PIV5-AB31 and s.c. rOspAB31 vaccines (Fig. 3), blood samples collected before tick challenge from the three studies were used for determination of anti-OspA antibody (1:102) by ELISA against rOspAB31. In contrast to the controls, serum from all mice vaccinated with OspA collected before (D17) and after the boost (>D86) had anti-OspA IgG antibody OD450 > 3.5. To quantify anti-OspA IgG endpoint titers we used serum from Study 3 (15-month challenge) collected at day 17 (preboost), and at months 3, 6 and 12 post-prime (Fig. 4) diluted at 102-106 on ELISA. Mice from the s.c. rOspAB31+Alum, intranasal PIV5-ABPBPk, and intranasal PIV5-AB31 vaccine groups had peak serum IgG endpoint titers (EPT) at 3-month postprime, with geometric means of 5.5, 5.5 and 5.4 log10, respectively, and not statistically different between groups. These values were significantly higher than serum IgG EPTs from day 17 post-prime, showcasing the boosting effect by the second dose of the vaccine. These levels of anti-OspAB31 serum IgG antibodies decreased over the next 9-months, with mice from the s.c. rOspAB31+Alum group having the biggest decrease of 1.7 log10 EPT compared to 0.7 log10 EPT seen in mice from the intranasal PIV5-ABPBPk and PIV5-AB31 vaccine groups. Mice from the PBS+Alum and PIV5 control groups had low levels of cross-reactive IgG antibodies throughout the study. As expected, analyses between intranasally vaccinated groups as well as between subcutaneously vaccinated groups show significant EPT differences between the controls (PIV5, Alum) and the groups of mice vaccinated with OspA (PIV5-ABPBPk, PIV5-AB31, rOspAB31+Alum), p < 0.0001. Of note, between OspA vaccinated groups, EPT differences between intranasally delivered PIV5-ABPBPk and PIV5-AB31 are not significant at 6- and 12-months post prime, but the differences in EPTs between rOspAB31+Alum delivered subcutaneously (which are lower) and the intranasal vaccines (PIV5-ABPBPk and PIV5-AB31) are significant (p = 0.0116 and p = 0.0001 for 6-months, and p = 0.0008 and p = 0.0005 for 12-months, respectively). These data indicate that immunization with OspA-based intranasal PIV5 leads to longer-lasting production of anti-OspA antibody than a recombinant OspA protein-based vaccine given subcutaneously.

Fig. 3: Vaccination schedule and methods for evaluation of efficacy before and after tick challenge.
figure 3

a Study 1, comprised of 4 groups of 5 mice used for neutralization assays and for tick challenge at 4-months post prime; b Study 2, comprised of 5 groups of 10 mice, some of which were used for neutralization assays (n = 4–6) and for tick challenge (n = 3–4) at 9-months post prime; c Study 3, comprised of 5 groups of 10 mice, some of which were used for tick challenge (n = 3) at 15-months post prime; the remaining surviving mice (n = 3–4) were used for neutralization assays at 18-months postprime. Legend: PIV5x, intranasal droplet immunization with PIV5 delivered vaccines; SC subcutaneous inoculation, M month, d day after prime, Bb B. burgdorferi, X euthanasia, BL blood, WBlot western blot, nAb neutralization assay, DFM dark field microscopy, MS’08 tick colony made from bladder cultures obtained from mice infected with field ticks in 2008, MS’21 tick colony made from bladder cultures obtained from mice infected with field ticks in 2021.

Full size image
Fig. 4: Intranasal vaccination with PIV5-ABPBPk or PIV5-AB31 induces long-lasting IgG responses.
figure 4

Mice in Study 3 (15-month challenge) received two doses of alum alone or alum + 20 µg of rOspAB31 protein subcutaneously, or 106 PFU of PIV5 vector control, PIV5-ABPBPk, or PIV5-AB31 intranasally at 21 days interval. Serum from blood collected at day 17 and months 3, 6, 12 postprime was tested against purified rOspAB31 and anti-OspA IgG antibody titers were quantified by ELISA. Statistical significance was calculated by Repeated Measures 2-Way ANOVA for comparisons within routes of immunization, i.e. intranasal PIV5 v PIV5-ABPBPk v PIV5-AB31 and subcutaneous Alum v OspAB31+Alum; nonparametric multiple Mann-Whitney tests were used for comparisons between two OspA vaccinated groups at each timepoint. Error bars, geometric means with geometric SD. 3-month comparisons are not significant, ns, not significant, *p < 0.05, **p < 0.005, ***p < 0.0005, ****p < 0.0001.

Full size image

Similar to Study 3, mice from Study 2 (9-month challenge) were immunized with either two-doses of alum alone (s.c. PBS+Alum) or alum plus 20 µg of rOspA protein (s.c. rOspAB31+Alum) subcutaneously, or with two-doses of 106 PFU of PIV5, PIV5-ABPBPk, or PIV5-AB31 intranasally (Fig. 4). Study 1 (4-month challenge) contained the same groups, except for the s.c. PBS+Alum which was an additional control deemed redundant at that stage of vaccine development. Neutralization assays (Fig. 5) were conducted using blood collected before tick challenge at D117 (Study 1), and from unchallenged mice at D270 (Study 2) and D533 (Study 3). Due to insufficient volumes of blood from Study 1, serum was pooled by group, while individual mouse samples were used for Studies 2 and 3. Total motile bacteria were counted in five fields in a Petroff-Hausser chamber under a dark field microscope on days 0, 2, 5, and 7 for Study 1, and days 0, 3, and 6 for Studies 2 and 3. In contrast to the controls, at 4-months after vaccination (Study 1, Fig. 5a), the numbers of motile B. burgdorferi in cultures incubated with serum from the rOspAB31+Alum, PIV5-ABPBPk, and PIV5-AB31 groups decreased by 1.6, 0.6, and 1.0 log10, respectively, at day 2 compared to day 0. These values further decreased on day 5 until they reached 0 for all vaccine groups at day 7. At 9-months after vaccination (Study 2, Fig. 5b), in control groups, B. burgdorferi cultures in BSK-H media grew a modest 0.2 log10, and numbers of B. burgdorferi in cultures treated with PBS+Alum and PIV5 serum did not increase from day 0 until day 6 post-neutralization. However, numbers of motile bacteria in cultures treated with serum from mice that received OspA (rOspAB31+Alum, PIV5-ABPBPk, and PIV5-AB31) were reduced at day 3 (~2 to 7 log10) and day 6 postneutralization (~1 to 7 log10) compared to day 0. Differences between each OspA vaccinated group and the respective control are significant (p < 0.0001). Differences between the three OspA vaccinated groups in terms of bacterial motility were only significant between PIV5-ABPBPk and PIV5-AB31 at day 6 (p = 0.0238). At 18-months postvaccination (Study 3, Fig. 5c), the data continues to show a reduction in numbers of motile B. burgdorferi per milliliter of culture in samples from the rOspAB31+Alum, PIV5-ABPBPk, and PIV5-AB31 vaccine groups at days 3 and 6 post-neutralization compared to day 0, with the largest significant differences seen on day 6 with an average decrease of 0.8, 1.5 and 2.0 log10, respectively. As observed in Study 2 in the control groups, B. burgdorferi cultures in BSK-H media grew 0.2 log10, and numbers of B. burgdorferi in cultures treated with PBS+Alum and PIV5 serum did not increase from day 0 until day 6 post-neutralization. Differences between each OspA vaccinated group and the respective control are significant (p < 0.0001). Differences between the three OspA vaccinated groups in terms of bacterial motility were only significant between rOspAB31+Alum and PIV5-ABPBPk with less numbers of motile bacteria counted in the latter, at day 3 post-neutralization (p = 0.0285). These results show that an intranasal PIV5-based vaccine carrying OspA sequences from different B. burgdorferi sensu lato genospecies can generate neutralizing IgG immune responses that lasted until 18-months post vaccination.

Fig. 5: Intranasal vaccination with PIV5-ABPBPk and PIV5-AB31 induces neutralizing antibodies up until 18-months post vaccination.
figure 5

Cultures of B. burgdorferi were treated with serum collected from mice before tick challenge at D117 (Study 1, a), and from unchallenged mice collected at D270 (Study 2, b) and D533 (Study 3, c). Neutralization was determined by counting motile spirochetes under a dark field microscope. Serum from Study 1 (4-months, D117) was pooled per group (n = 5 mice/group). For the subsequent Study 2 (9-month, D270) and Study 3 (18-month, D533), serum from n = 3–6 mice per group were individually tested. Scatter plots represent the average numbers of motile B. burgdorferi in 5 fields (Petroff-Hausser) counted under a dark field microscope. Statistical significance was calculated by Repeated Measures or Mixed Model 2-Way ANOVA (with Geisser-Greenhouse correction) for comparisons within routes of immunization, i.e. intranasal PIV5 v PIV5-ABPBPk v PIV5-AB31 and subcutaneous Alum v OspAB31 +Alum. Differences between two OspA vaccinated groups were analyzed by multiple Mann-Whitney tests at each timepoint except day 0. The Kruskal-Wallis was used to test for differences between timepoints within each group (ns, not significant, *p < 0.05, **p < 0.005, ***p < 0.0005, ****p < 0.0001), Bb, Borrelia burgdorferi.

Full size image

PIV5-ABPBPk and PIV5-AB31 provide long-term protection against tick challenge with multiple strains of B. burgdorferi, up to 15-months post prime-boost immunization

IgG to B. burgdorferi protein was quantified in serum from tick-challenged mice by ELISA (Figs. 6a, b) and Western blot (Fig. 6c). In Study 1, three weeks after the 4-month challenge, none of the mice vaccinated with OspA vaccines (rOspAB31+Alum, PIV5-ABPBPk and PIV5-AB31) had anti-OspCB or anti-VlsE antibody OD450 values above the cutoff, in contrast to the PIV5 control. Furthermore, serum from all mice that received OspA vaccines was negative on Western blot (2-4/10 bands and negative for OspC), in contrast to the PIV5 control (8-9/10 bands and positive for OspC). In Study 2, three weeks after the 9-month challenge, none of the mice vaccinated with PIV5-ABPBPk and PIV5-AB31 had anti-OspCB or anti-VlsE IgG antibody OD450 values above the cutoff, whereas 1 mouse from the rOspAB31+Alum group (1/4) had anti-OspCB and anti-VlsE IgG well above the cutoff. As observed in Study 1, the OspCB and VlsE serologic results were largely confirmed by Western blot, except one mouse vaccinated with OspAB31+Alum that although positive on ELISA was negative on Western blot. Last, in Study 3, three weeks after the 15-month challenge, all mice from the rOspAB31+Alum group (3/3), and 1/3 mice from PIV5-ABPBPk and from the PIV5-AB31 groups had levels of anti-OspCB IgG above the cutoff. Regarding anti-VlsE IgG, all mice from the rOspAB31+Alum group (3/3), and 1/3 mice from PIV5-ABPBPk group had levels of IgG above the cutoff. In this study, the Western blot results largely supported the OspCB+VlsE IgG ELISA. Differences between controls and OspA vaccinated groups were significant in the three studies. Within the OspA vaccinated groups, significant differences in OspCB and VlsE IgG serology were observed after the 15-month challenge: rOspAB31+Alum and PIV5-AB31 for OspC (p = 0.0427); rOspAB31+Alum and PIV5-ABPBPk for VlsE (p < 0.0001); rOspAB31+Alum and PIV5-AB31 for VlsE (p < 0.0001).

Fig. 6: Serologic evidence of B. burgdorferi infection after tick challenge at 4-months, 9-months and 15-months post vaccination.
figure 6

IgG to B. burgdorferi was determined by assessment of serum antibody to recombinant OspCB and VlsE by ELISA, as well as Western blot. Scatter dot plots (OD450) are shown for serum IgG to rOspCB (a) and to rVlsE (b). Western blot data (c) is represented by heat maps of the enumerated bands: > 5 out of 10 bands (red, positive) and < 5 out of 10 bands (green, negative), m, mouse. Statistical significance was calculated by One-way ANOVA followed by multiple comparisons with uncorrected Fisher’s LSD (ns, not significant, *p < 0.05, **p < 0.005, ***p < 0.0005, ****p < 0.0001).

Full size image

Mice from the 3 studies were challenged with ticks infected with multiple strains of B. burgdorferi at 4-months (Study 1), 9-months (Study 2), or 15-months (Study 3) post-prime. To evaluate B. burgdorferi dissemination to target tissues, two tissues per mouse (heart, bladder, or joint) were tested for B. burgdorferi flaB load by qPCR (Fig. 7a). To assess B. burgdorferi viability, one tissue per mouse (heart or bladder) was cultured in BSK-H medium to evaluate growth and motility of B. burgdorferi under a dark field microscope (Fig. 7b, inset in red), which was confirmed by flaB PCR (Fig. 7b). In Study 1, the challenge done 4-months after vaccination with subcutaneous rOspAB31, and intranasal PIV5-ABPBPk and PIV5-AB31 resulted in absence of B. burgdorferi flaB DNA in heart and bladder (Fig. 7a), as well as absence of viable B. burgdorferi in cultures from heart which was confirmed by flaB qPCR (Fig. 7b), in contrast to the controls that received intranasal PIV5. Differences between controls and OspA vaccinated groups are significant. In Study 2, the challenge done 9-months after vaccination with intranasal PIV5-ABPBPk and PIV5-AB31 resulted in absence of B. burgdorferi flaB DNA in heart and joint (Fig. 7a), as well as absence of viable B. burgdorferi in culture from bladder which was confirmed by flaB qPCR (Fig. 7b). In contrast, 1 mouse that received subcutaneous rOspAB31+Alum had a positive joint flaB PCR which was confirmed by a positive culture from bladder. All controls that received subcutaneous PBS+alum or intranasal PIV5 had positive flaB PCR from heart and joint tissues, as well as PCR confirmed bladder cultures. Differences between controls and OspA vaccinated groups were significant for B. burgdorferi flaB load in tissues, and for amplification of B. burgdorferi flaB from culture of bladder in mice vaccinated with rOspAB31+Alum and its respective control. In Study 3, the challenge done 15-months after vaccination with intranasal PIV5-AB31 resulted in absence of B. burgdorferi flaB DNA in heart and joint (Fig. 7a), as well as absence of viable B. burgdorferi in cultures from bladder which was confirmed by flaB qPCR (Fig. 7B). Furthermore, one PIV5-ABPBPk vaccinated mouse had ~ 3500 copies of flaB in joint as well as in heart tissue, which was not confirmed by growth and motility analysis from the bladder culture or flaB PCR from the same culture. In contrast, 5 tissues (3 joint, 2 heart) from the 3 mice vaccinated with rOspAB31+Alum had ~600-74,000 flaB DNA copies, PCR from the cultures of the 3 bladders had 77-800 flaB DNA copies, but only 1 bladder of the 3 mice produced a culture with motile B. burgdorferi. Regarding the controls, the 3 mice that received subcutaneous PBS+Alum and the 3 mice that received intranasal PIV5 had all joint and heart tissues positive for flaB PCR and all the bladders from each mouse produced cultures with motile B. burgdorferi, that were confirmed by PCR. In Study 3, differences between controls and OspA vaccinated groups were significant for B. burgdorferi flaB in culture of tissues.

Fig. 7: B. burgdorferi load and viability in mouse tissues after tick challenge at 4 months, 9 months and 15 months post vaccination.
figure 7

Tissues collected from vaccinated mice euthanized at 4-months (heart, bladder), and from mice euthanized at 9-months and 15-months (heart, joint) were processed for Bb DNA purification and flaB qPCR. Heart (4-months) and bladder (9- and 15-months) tissues were also cultured in BSK-H media for analysis of B. burgdorferi (Bb) growth and motility under a dark field microscope which was confirmed by flaB qPCR. Scatter dot plots are shown for Bb flaB load in tissues (a) and Bb flaB load in culture of tissues quantified (b) after assessment of B. burgdorferi motility from each culture by dark field microscopy - inset (b). Bb, Borrelia burgdorferi. Statistical analysis was calculated by multiple Mann-Whitney tests for 4 Months and by One-way ANOVA followed by multiple comparisons with Uncorrected Fisher’s LSD for 9 Months and 15 Months (ns, not significant, *p < 0.05, **p < 0.005, ****p < 0.0001).

Full size image

The vaccine efficacy data is summarized in Table 1 and includes anti-B. burgdorferi Western blot serology, B. burgdorferi flaB load in tissues, flaB load in culture of tissues and B. burgdorferi motility under dark field microscopy after tick challenge.

Table 1 Vaccine efficacy data summary.
Full size table

We lost mice throughout our study in all groups subjected to the 9-month and 15-months protocols due to attrition. This impacted vaccine efficacy analysis of each PIV5-OspA vaccine group. To evaluate if the vehicle and route of immunization affected vaccine efficacy, we analyzed differences between groups of mice vaccinated subcutaneously with OspAB31+Alum and mice vaccinated intranasally with PIV5 carrying OspA (ABPBPk+AB31) (Fig. 8). We found that differences in anti-OspCB (Fig. 8a) and anti-VlsE (Fig. 8b) IgG between subcutaneous rOspAB31+Alum and the intranasal immunizations (PIV5-ABPBPk + PIV5-AB31) are significant after the 15-month challenge (p = 0.0374 and p < 0.0001, respectively). The same was observed for differences in B. burgdorferi flaB DNA in tissues (Fig. 8c) and in culture from tissue (Fig. 8d) between subcutaneous rOspAB31+Alum and the PIV5-ABPBPk + PIV5-AB31 intranasal immunizations (p = 0.0093 and p = 0.0238, respectively). Furthermore, a combined analysis of motile B. burgdorferi in culture from tissues in the 9 M and 15 M challenges (Fig. 7b inset, and Table 1) shows that 2/6 (33%) mice vaccinated subcutaneously with OspAB31+Alum produced positive culture results, which stands in stark contrast to 0/5 (0%) and 0/6 (0%) mice vaccinated intranasally with PIV5-ABPBPk and PIV5-AB31, respectively.

Fig. 8: Vaccine efficacy at 4-months, 9-months and 15-months, by route of administration.
figure 8

Anti-B. burgdorferi OspC (a) and VlsE (b) antibody in serum, as well as B. burgdorferi flaB DNA amplified from tissues (c) and from culture of tissues (d) was compared between subcutaneous (s.c.) and intranasal (i.n.) routes of immunization with OspA. Statistics by Welch’s t test (a, b) and Mann Whitney (c, d) between s.c. and i.n. groups, *p < 0.05, ** p < 0.005, ****p < 0.0001. M, months.

Full size image

Together, these data demonstrate that an intranasal PIV5-vectored OspA-based Lyme disease vaccine can provide substantial longer-lasting protection than a recombinant protein-based OspA vaccine given subcutaneously to mice.

Discussion

Currently, the most effective way to prevent Lyme disease is to avoid I. scapularis tick infested areas. This is unfeasible for those who work outside and for those who enjoy spending quality time outdoors in the Spring and Summer in endemic areas. Thus, there is a pressing need for development and commercialization of effective and acceptable vaccines to control Lyme disease. We developed a parainfluenza virus 5 (PIV5) viral-vectored vaccine for intranasal delivery of OspA that provides mice with long-lasting protection against tick-transmitted B. burgdorferi using a prime-boost scheme of immunization.

Outer surface protein A from B. burgdorferi sensu stricto (OspA) is the only immunogen proven to provide high (LYMErixTM)2 or very high (ImuLymeTM)3 protection against tick-transmitted B. burgdorferi in human subjects after 3 intramuscular injections. One of the two vaccines (LYMErixTM) was approved by the FDA in 1998. Although analysis of adverse effects performed in both clinical trials showed no significant increase in the frequency of arthritis events between vaccine and control groups2,3, some individuals who received LYMErix, reported developing arthritis after the trial period ended20. Furthermore, within weeks of the clinical trial reports, another study suggested that a cross-reactive autoimmune event between an epitope in B. burgdorferi OspA and a human integrin (hLFA-1) might drive the inflammatory response in the joints of some treatment-resistant Lyme arthritis patients21. Even though LYMErix was never linked to causing arthritis, demand for this vaccine decreased substantially and the product was taken off the market in 200222. Keeping all these factors in mind, we designed a chimeric sequence of OspA by replacing the hLFA-1 partially homologous epitope within B. burgdorferi OspA with the analogous sequence from B. afzelii (a nonarthritogenic Borrelia species) to generate the new OspABPBPk construct. This sequence was cloned into a Lactobacillus expression vector and was shown to prevent tick-transmitted B. burgdorferi infection in mice, thus providing an effective oral vaccine candidate for Lyme disease18.

Parainfluenza virus 5 (PIV5) is a nonsegmented, negative-strand, RNA virus and a member of the Rubulavirus genus of the family Paramyxoviridae. PIV5 is a promising safe viral vaccine vector. Live PIV5 has been part of the kennel cough vaccine for dogs for 50 years, yet no disease has ever been reported due to human exposure to vaccinated dogs23. We have not performed studies looking at transmission of PIV5 from vaccinated to naïve individuals. Nevertheless, we found that ~30% of humans have anti-PIV5 antibodies, suggesting that humans can get infected with PIV5 from PIV5-vaccinated dogs24. In that study we also show that the anti-PIV5 pre-existing antibody did not affect the immunogenicity of the PIV5-vectored vaccine24.

PIV5 does not have a DNA phase in its life cycle. Thus, its use avoids the unintended consequences of genetic modifications of host cell DNA through recombination or insertion. Various recombinant PIV5 viruses expressing GFP or immunogens have been generated and shown to be genetically stable25,26. Furthermore, a study performed with a PIV5-based influenza vaccine in nude mice showed no sign of illness or weight loss, as well as no enhanced pathology, indicating that PIV5 is safe and non-pathogenic in an immune deficient mouse model27. In addition, PIV5 can grow in Vero cells at titers greater than 108 PFU/mL, making vaccine production cost effective. PIV5-based influenza (IAV)19,27,28,29,30,31, respiratory syncytial virus (RSV)12,26,32, MERS-CoV33 and, most recently SARS-CoV-234, vaccines are efficacious in various preclinical animal models. RSV and COVID-19 vaccines have recently completed phase 1 clinical trials.

In this study, we demonstrate that both PIV5-based vaccines, PIV5-ABPBPk and PIV5-AB31, administered in a homologous prime-boost vaccination regimen intranasally, can induce robust and long-lasting humoral IgG immune responses in mice reaching over 5 log10 EPT for serum IgG antibodies at 3-months post-vaccination, with these values moderately decreasing up to 12-months post-immunization. Although the data shows a significant increase in anti-OspAB31 EPTs in serum from mice that received the intranasal vaccines (PIV5-ABPBPk and PIV5-AB31) compared to the subcutaneous control (rOspAB31+Alum) at 6-months and at 12-months postprime-boost (Fig. 4) the ultimate functionality of the vaccine could be related to other factors such as affinity maturation and glycosylation. However, we have not performed studies to determine if glycosylation and affinity maturation play a role in the immunogenicity of PIV5-based vaccines.

The longevity of the immune response to the PIV5 delivered OspAs can be noticed in Fig. 5, where differences in neutralization of B. burgdorferi with mouse serum favors PIV5-ABPBPk at 18-months post-immunization. One of the advantages of using live replicating viral vectored vaccines is that IgG responses are more durable35 than responses induced by mRNA vaccines which usually wane within 6 months to 1 year36. For OspA recombinant proteins, others have shown that OspA-specific IgG antibody also starts to decline 3-6 months after the 2nd or 3rd dose within the first year of vaccination37. Other schedules of administration (3 shots within 2 months) produced anti-OspA antibody responses that could only be protective for one tick-season (about 4 months)38. Furthermore, Comstedt et al., reported that immunization with 3 doses of VLA15 in mice resulted in a robust serum IgG antibody response, but the levels of antibodies decreased more than 10-fold, five months after immunization9. A booster dose was administered 5 months after immunization to increase these levels. To accommodate these issues, the VLA15 vaccine undergoing phase 3 clinical trials requires a 3-shot intramuscular vaccination schedule on Day 1 (Month 0), Day 57 (Month 2) and Day 180 (Month 6). Here, we show that prime-boost immunization with a mucosal delivered live viral-vectored vehicle leverages a non-invasive administration route (intranasal) and generates a rapid, durable and neutralizing IgG response to the immunogen that lasts up to 18-months.

Longevity of protection was associated with PIV5 delivered OspA vaccinations in challenges performed at 9- and 15-months post-prime (Figs. 6, 7, 8, Table 1). Our comparative analysis of serologic anti-OspC and anti-VlsE IgG antibodies, as well as presence of B. burgdorferi flaB DNA in tissues and in culture from tissues shows significant differences between mice subcutaneously vaccinated with rOspAB31 and mice vaccinated with both PIV5 delivered OspABPBPk and OspAB31 in groups challenged at 15-months after vaccination. Furthermore, we determined that 33% of mice vaccinated subcutaneously with OspAB31+Alum had motile B. burgdorferi in culture from tissues, whereas none of the mice vaccinated intranasally with PIV5-ABPBPk and PIV5-AB31 produced cultures with motile bacteria after the 9 M and 15 M challenges. Taken together, our data indicates that longevity of vaccine efficacy is consistently higher in mice that received intranasal PIV5 delivered OspA vaccines compared to mice that received parenteral OspA vaccine.

Of note, in challenges done at 9-months post vaccination, the mouse vaccinated with rOspAB31+Alum that developed IgG antibodies to OspC and VlsE had a negative Western blot (Fig. 6). However, it also produced a positive flaB result in joint, a positive flaB bladder culture and motile B. burgdorferi in the same culture under dark field microscopy (Fig. 7). In challenges done at 15-months post vaccination, serologic IgG to OspC and VlsE in the mouse vaccinated with PIV5-ABPBPk was further supported by a positive Western blot result (Fig. 6). However, although this mouse produced positive tissues for flaB DNA, B. burgdorferi did not grow in culture (Fig. 7). Similar data was observed for mice vaccinated with OspAB31+Alum challenged at 15-months post vaccination, in that the 3 mice had antibodies to OspC and VlsE which were supported by Western blot. However, although the 3 mice produced positive flaB results in tissues, only 1 produced a culture with motile B. burgdorferi. These data suggest that a weaker immune counter activity at 15-months possibly due to lower or impaired function of myeloid and lymphoid cells may allow persistence of viable B. burgdorferi and B. burgdorferi debris in some target tissues. The data also lends credence to the difficulties with development of diagnostic assays for Lyme disease by currently available serologic or molecular methods.

We developed a PIV5-vectored OspA-based intranasal vaccine that prevents tick-transmitted B. burgdorferi infection in mice challenged at 4-, 9- and 15- months post prime-boost immunization. Our work advances the field of development of vaccines for Lyme disease in three ways: i) it reduces the number of immunizations, ii) it simplifies administration from the classic intramuscular injection to an intranasal mist with possibility for self-administration, iii) and most importantly, it extends protection until 15-months post-immunization.

Methods

Construction of PIV5-vectored vaccines expressing OspAB31 and OspABPBPk

Cells

BHK21 cells were maintained in Dulbecco’s modified Eagle medium (DMEM) containing 10% tryptose phosphate broth (TPB), 5% fetal bovine serum (FBS), 100 IU/mL penicillin, and 100 µg/mL streptomycin (1% P/S; Mediatech Inc., Manassas, VA). Vero E6 and MDBK cells were maintained in Dulbecco’s modified Eagle media (DMEM) supplemented with 5% fetal bovine serum (FBS) plus 100 IU/mL penicillin and 100ug/mL streptomycin (1% P/S; Mediatech Inc, Manassas, VA, USA). All cells were incubated at 37°C, 5% CO2.

Viruses

The PIV5-AB31 and PIV5-ABPBPk plasmids, encoding the full-length genome of PIV5 and a Borrelia burgdorferi OspA gene (strain B31) and well as a chimeric OspA gene comprised of sequences from B. burgdorferi B31, and B. afzelii PGau and Pko (B311–164PGau165–189B31190–218Pko219–273 - BPBPk)18 inserted between the PIV5 small hydrophobic (SH) and hemagglutinin-neuraminidase (HN) genes was constructed as previously described19. Briefly, the PIV5-AB31 and PIV5-ABPBPk plasmids and four helper plasmids—pPIV5-NP, pPIV5-P, pPIV5-L, and pT7-polymerase, encoding the NP, P, and L proteins and T7 RNA polymerase, respectively—were co-transfected into BHK21 cells at 90% confluence n 6-cm plates using JetPrime (Polyplus). Recovery of the virus is indicated by syncytia formation. The virus was then plaque-purified as a single plaque from BHK21 cells. The full-length genomes of the plaque-purified single clone of PIV5-AB31 and PIV5-ABPBPk viruses were sequenced as described19. Viruses were grown in MDBK cells for 5 to 7 days using DMEM containing 2% fetal bovine serum (FBS). Media were collected and pelleted at 500 xg to remove cell debris by using a Sorvall tabletop centrifuge for 10 min. Virus supernatant was supplemented with 10% sucrose-phosphate-glutamate buffer, snap-frozen in liquid nitrogen, and stored at −80°C immediately after collection.

Western blot

Immunoblotting was performed on Vero cells in 6-well plates that were infected with PIV5, PIV5-AB31, or PIV5-ABPBPk at an MOI of 1. At 48 hours post-infection (hpi), Laemmli sample buffer (Bio-Rad, catalog no. 1610737) with 5% β-mercaptoethanol was used to lyse cells. The lysates were separated on an SDS–polyacrylamide gel electrophoresis (SDS-PAGE) gel and immunoblotted with a mouse anti–OspA monoclonal antibody (184.1, 1:100) and an anti–PIV5-NP monoclonal antibody (NP214mAb, 1:100, a kind gift from Dr. Randall to Dr. He).

All blots were processed in parallel and derive from the same experiments.

Immunization, immunogenicity, and vaccine efficacy analyses

All experiments were performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals under protocols approved by the Institutional Animal Care and Use Committee (IACUC) of the University of Georgia (A2023 01-021-Y1-A0) and the University of Tennessee Health Science Center (19-0103). A graphic representation on the experimental design is shown in Fig. 3.

Immunization

Briefly, groups of 5 to 8-week-old female C3H-HeN mice (Envigo) were anesthetized and intranasally inoculated with 50 µL of 106 PFU of PIV5 vector, PIV5-AB31, or PIV5-ABPBPk, or with 100 µL of 20 µg of rOspAB31+Alum subcutaneously (Fig. 3). Twenty-one days after prime immunization, the mice were boosted with the same preparations. Blood was collected before tick challenge for determination of anti-OspA antibody on the following days: d17, d86, d100 and d117 for Study 1 (4-month challenge, D117 serum was used for neutralization assays); d17, d88, d103, d118 and monthly thereafter until d270 for Study 2 (9-month challenge, D270 serum was used for neutralization assays). For Study 3 (15-month challenge and 18-month neutralization), groups of mice were bled on d17, d86, d100 and d117 and monthly thereafter until D455 for 15-month challenge; a subset of these mice were not challenged and were kept until 18-months post prime-boost for an additional collection of blood at D533 for analysis of anti-B. burgdorferi neutralization activity.

Tick challenge

Three colonies of infected I. scapularis were maintained in the laboratory. MS’08/NY derived from cultures of tissue from Peromyscus leucopus infected with field caught NY ticks between 2005 and 2008 and frozen at -80C; MS’21/MA derived from frozen cultures obtained from C3H-HeN mice inflected with ticks collected in MA parks in 2021; MS’21/NY derived from frozen cultures from C3H-HeN mice inflected with ticks collected in NY parks in 2021. Cultures were used to infect C3H-HeN mice with 50,000 B. burgdorferi in 100-200 µl; uninfected larval ticks were allowed to engorge in the infected mice and were allowed to molt to the nymphal stage before they were used for challenge. Challenge of experimental animals with B. burgdorferi was performed as described39. Ticks from the MS’08/NY colony were used for the 4-month challenge experiment; equal numbers of ticks from each 2021 colony (MS’21/MA and MS/21/NY) were used for the 9-month and 15-month challenges. The B. burgdorferi infection prevalence of each tick colony was determined to be >80% by flaB qPCR. Briefly, 8-10 flat nymphal Ixodes scapularis ticks were placed between the ears of mice that were caged separately and held in FIC-2 isolators for a week. Engorged ticks that naturally fell off after taking a bloodmeal were collected from the bottom of the cage, counted, labeled, and stored at -20°C. Three weeks after the last day of challenge, mice were euthanized and blood, heart, joint and bladder were collected. Tissues were placed in BSK-H culture or RNAlater (Invitrogen, MA) for determination of B. burgdorferi growth and motility (counting live spirochetes under a dark field microscope) and B. burgdorferi presence in tissues and in culture of tissues was done by flaB qPCR. Blood was used for analysis of anti-B. burgdorferi antibodies by ELISA.

Experimental endpoints

1) euthanasia of vaccinated mice for collection of blood for neutralization assays on D270 for Study 2 and on D533 for Study 3; 2) euthanasia of mice previously vaccinated and challenged, 3 weeks after the last day of challenge, for collection of blood and tissues for analyses of vaccine efficacy on D146 for Study 1, on D333 for Study 2 and on D494 for Study 3. Methods of anesthesia (inhalation of 3% isoflurane) and euthanasia (inhalation of 3-4% isoflurane, followed by exsanguination and thoracotomy for collection of tissues) were consistent with the recommendations of the American Veterinary Medical Association (AVMA) Guidelines.

Enzyme-linked immunosorbent assay (ELISA)

Anti-OspA, anti-OspC type B (OspCB) and anti-VlsE antibody was determined by ELISA40,41. Briefly, 5-10 µg/ml of purified recombinant OspA, OspCB or VlsE were coated on flat-bottom ELISA plates (Thermo Fisher Nunc MaxiSorp) and incubated at 4°C overnight. The following day, the plates were washed, blocked, and incubated with primary antibodies, i.e., serum (1:100 or 1:1000), followed by the secondary antibody HRP-conjugated goat anti-mouse IgG (Jackson ImmunoResearch, Inc, cat. no.115-035-146) diluted at 1:10000. Endpoint titers of anti-OspA antibodies were calculated using serum (1:102 to 1:106) from study 3 (15-month challenge) collected at pre-boost, 3-months, 6-months and 12-months post-prime vaccination. Blood used for OspCB and VlsE ELISA was collected at euthanasia on D146 (4-month challenge), D333 (9-month challenge) and D494 (15-month challenge). The ELISA cutoff was established at 5 standard deviations above the mean of 4 samples confirmed negative by Western blot.

Western blot

the B. burgdorferi IgG Virablot kit (Viramed Biotech AG, cat. no. V-BBSGUS) was used according to the manufacturer recommendations. A positive result was determined by observation of >5 out of 10 bands; a negative result was determined by observation of <5 out of 10 bands.

Neutralization antibody (nAb) assay

Briefly, blood was collected from groups of mice the day before challenge (d117) for the 4-month challenge experiment, and on days 270 and 533 from groups of mice that were not subjected to challenge for the 9-months and 18-months experiments, respectively. Neutralization of motility of multi-strain cultures of B. burgdorferi by fresh serum was performed as described40. Briefly, 8 µl of the bacterial culture was mixed with 4 µl of heat-inactivated mouse serum obtained from vaccinated and control mice and with 4 µl guinea pig complement (MP Biomedicals™) in a 0.2 ml sterile PCR microtube (VWR, LLC Radnor, PA). The positive control group consisted of 8 µl BSK media (Sigma-Aldrich, Saint Louis, MO) with 8 µl B. burgdorferi culture. Samples were incubated at 34°C for 6 to 7 days. The cultures were counted in five fields on days 0, 2, 5, 7 (4-month) or 0, 3, and 6 (9- and 18-month) for motile B. burgdorferi using a Petroff-Hausser chamber under a dark-filed microscope (Zeiss USA, Hawthorne, NY) and averaged to get the total number of motile bacteria.

Amplification of OspC by conventional PCR

Five flat nymphal Ixodes scapularis ticks were crushed and DNA was extracted using the DNAeasy tissue kit as per manufacturer’ recommendation (Qiagen, Valencia CA). OspC specific amplicons were prepared using primers (F)- AATAAAAAGGAGGCACAAATTAATG and (R)-GTAACTGGAAAAATAAAGTCAATAT by conventional PCR. Each 20 µl PCR reaction mixture contained 200 µM deoxynucleoside triphosphate (dNTP) (Thermo Scientific), 1 U Taq DNA polymerase (Thermo Scientific), 2 µl of 10X Taq Buffer (Thermo Scientific), 2.5 mM MgCl2 and 0.4 µM of each OspC primer, and 1 µl of purified genomic DNA (gDNA). The reaction mixture was heated at 95°C for 4 min, amplified for 36 cycles at 95°C for 30 s, 58°C for 30 s, and 72°C for 60 s, and finally incubated at 72°C for 5 min. The PCR products (2-5 ul) were electrophoresed on a 2% agarose gel, pre-stained with 1:10,000 DNA SafeStain (Lambda Biotech-C138) and imaged under a UV gel doc system. AM-Pure beads were used to clean unused primers, dNTPs, and other reagents. The amplicon quantity was measured on a Qubit 4 fluorometer (Thermo Fisher Scientific, Waltham, MA, USA) using the Qubit dsDNA HS assay kit to ensure at least 10 ng/µl for library preparation as per manufacturer’ recommendation. Amplicon quality was measured using a nanodrop instrument to confirm gDNA purity with A260/230 and A260/280 absorbance ratio of at least 1.8 for all samples. Library preparation was done using NEBNext® Fast DNA Fragmentation & Library Prep Set for Ion Torrent™. Samples were amplified and barcoded with barcode adapters from the Ion Xpress™ Barcode Adapters 1-16 Kit. Sequencing was done with the Proton Ion Torrent Sequencing instrument (ThermoFisher) at the UTHSC Molecular Resource Center.

OspC sequencing data analysis

Fastq files were retrieved from the Ion Torrent Sequencing instrument. The individual runs were combined to create the merged fastq files. FastQC was used to check the quality of the reads. Reads were trimmed if needed using FASTX-Trimmer. Data was aligned against the fasta file containing OspC variants provided. SAM files were mined for total read count tables. SAM files were then filtered to obtain only uniquely aligned fragments. Filtered SAM files were mined for unique read count tables. Graphs were created using R42 and Graphpad Prism.

Quantitative PCR to enumerate B. burgdorferi

Quantitative PCR (qPCR) was used to enumerate B. burgdorferi in BSK cultures from glycerol stock (used to infect mice), cultures from tissues from experimental mice and B. burgdorferi load in tissues and ticks used for challenge. Ticks and tissues (bladder, heart, joint) were processed for DNA extraction using the DNAeasy tissue kit as per the manufacturer’s recommendation (Qiagen, Valencia CA). The eluted DNA was stored at -20°C. qPCR was done on QuantStudio 3 (Applied Biosystems) using B. burgdorferi flaB primers, a known conserved gene of B. burgdorferi43. The following are the sequences: forward GCAGCTAATGTTGCAAATCTTTTC, reverse GCAGGTGCTGGCTGTTGA, and probe [6 ~ FAM]-AAACTGCTCAGGCTGCACCGG-[Tamra~Q]. For the standard curve, DNA from a B. burgdorferi culture from a stock of 106 (determined by counting all visible B. burgdorferi under a dark field microscope) was purified and serially diluted from 105 to 1. The PCR reaction was performed using the fast advance master mix (Applied Biosystems™ Taqman™) in a final 20 µl volume which contained 25 µM of each primer, 250 nM of the specific probe, and 2 µl of DNA.

Statistical analysis

Data are represented in scatter dot plots. Statistical analysis was done by Repeated Measures or Mixed Model 2-Way ANOVA (with Geisser-Greenhouse correction) for comparisons within routes of immunization (Fig. 4, Fig. 5). Differences between two groups were analyzed by nonparametric multiple Mann-Whitney tests (Fig. 4, Fig. 5, Fig. 7, Figs. 8C, 8D) or unpaired Welch’s t test (Figs. 8A, 8B). The Kruskal-Wallis was used to test for differences between timepoints within each group (Fig. 5). One-way ANOVA followed by multiple comparisons with Uncorrected Fisher’s LSD was used to compare pairs of groups (Fig. 6, Fig. 7). GraphPad Prism was used for statistical analysis and plotting the graphs.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.