The use of xenogeneic vaccinations with highly homologous proteins to overcome tolerance to “self” antigens and increase vaccine efficacy has been investigated for the treatment of melanoma and prostate cancer.12,13,16,17,20,21
We have previously reported that immunization of Lewis rats with a DNA vaccine encoding human PAP (xenoantigen) or rat PAP (autologous antigen) elicits an antigen-specific immune response and that multiple courses of a DNA vaccine encoding rat PAP elicit TH
1-polarized immune response to the native antigen.2
Similarly, we reported that the repeated immunization of prostate cancer patients with a DNA vaccine encoding human PAP elicited human PAP-specific T-cell proliferative responses and IFNγ secretion.4,5
Here, in a preclinical model, we explored whether a DNA vaccine encoding a xenoantigen could promote increased immunogenicity, potentially with fewer immunizations. Specifically, we investigated whether the immunization of Lewis rats with a DNA vaccine encoding human PAP would elicit a cross-reactive response to the native antigen, rat PAP. We identified a dominant human PAP-specific epitope, HP201–215
, that was recognized in all animals following immunization. A second peptide, HP371–386
, was recognized by some of the animals, but this peptide is located at the C-terminus of the protein, contains an extra four amino acids compared with the rat PAP, and was shown not to be naturally presented. The HP201–215
epitope was found to be RT1.Al
-restricted and no cross-reactive immune responses were elicited to the corresponding rat PAP peptide, which differs only by three amino acids. Immunization with modified PAP-encoding DNA vaccines that had the immunodominant human PAP-specific epitope inserted or ablated did not result in cross-reactive immune responses to rat PAP. Overall, our findings demonstrate that, at least in Lewis rats, immunization with a xenoantigen can result in an immune response against a foreign epitope without cross-reactivity to the native antigen.
Naturally occurring, immunological “hotspot” regions in proteins have been reported as regions of interest for vaccine targets. We have previously identified a region in the cancer-testis antigen SSX-2 as one containing nested HLA-A2 and HLA-DR1 binding sites.22
Such epitope-rich regions, containing both MHC Class I and II epitopes, are of interest as vaccine antigens to generate both antigen-specific CD4+
T cells. Using overlapping peptides, the minimal immunodominant PAP-specific epitope in Lewis rats was mapped to a 12 amino acid peptide, HP204–215
. The sequence similarity of a 9-mer peptide contained within this region, WSKVYDPLY, to the preferred RT1.Al
binding motif (X-A/S/V-F/Y-X-X-X-X-X-Y/F/L/M), the peptide-elicited production of IFNγ by CD8+ T cells, and the ability of anti-RT1.Al
antibodies to partially block peptide-specific immune responses, strongly suggest that this particular epitope is RT1.Al
The presence of proliferating CD4+
T cells after stimulation with the HP201–255
peptide pool and following direct immunization with the HP201–215
peptide suggests that an MHC Class II epitope is likely to be adjacent to or encompassing this epitope. However, we did not characterize this further. RT1.Bl
has been the best-characterized MHC Class II molecule in Lewis rats. Still, we did not identify a potential epitope corresponding to the preferred RT1.Bl
-binding motif (X-X-T/S/V-F/H/Q-X-A/S/V/T-X-X-E/D), and anti-RT1-B1
antibodies were unable to block peptide-specific immune responses. This does not exclude the possibility that an MHC Class II epitope restricted to other Lewis rat MHC Class II haplotypes exists in this region.
As indicated above, the amino acid sequence of the human immunodominant peptide (GIWSKVYDPLYC) and corresponding rat peptide (EIWSRLYDPLYC) differ only by three amino acids, yet the corresponding rat peptide was not immunogenic in Lewis rats. Based on the predicted RT1.Al
-binding motifs, the anchor residue required for MHC Class I molecule binding was not present in the corresponding rat peptide.19
It is perhaps not surprising that the sequence of the presented epitope, and not the overall homology of the protein, was most important for the development of a cross-reactive immune response. However, this particular case illustrates that, depending on the MHC type of the host, xenoantigen vaccination may not be advantageous and only elicit immune responses to “foreign” epitopes. Fong and colleagues have previously reported that viral vectors encoding hPAP can elicit cross-reactive immune responses in Copenhagen rats, as demonstrated by the generation of CTL responses detectable in vitro and prostate tissue inflammation in vivo.16
Thus, while the generation of cross-reactive immune responses may certainly occur with xenoantigen vaccines, our results in Lewis rats demonstrate that this is not a generalizable phenomenon. Such a strain-specific cross-reactive immune response following immunization with a xenoantigen is most relevant to human immunization of diverse MHC types, and highlights the disadvantage of generalized conclusions drawn from studies based on inbred rodent strains.
Prostate cancer patients immunized with vaccines targeting PAP have developed PAP-specific immune responses. In a phase I clinical trial, Fong, et al. demonstrated that prostate cancer patients treated with mouse PAP (xenoantigen)-loaded DCs developed T-cell responses to mouse PAP, and 11 out of 21 of these patients also developed cross-reactive human PAP-specific TH
We showed that 10 out of 22 prostate cancer patients immunized with a DNA vaccine encoding human PAP developed a TH
1-polarized/CTL response to the native PAP with long-term memory to PAP in some patients.4,5
In addition, the development of human PAP-specific T cells and antibodies has been documented in 28% patients treated with sipuleucel-T.1
Although PAP was targeted using different immunization methods, patients generated PAP-specific immune responses regardless of whether they were immunized with a xenoantigen or with the native antigen. Thus, even though xenogeneic immunization has been shown to elicit cross-reactive immune responses, there is not an obvious advantage of using a xenoantigen approach targeting this antigen in humans. To determine whether xenoantigen immunization can improve the immunogenicity of a particular vaccine approach in humans, it will be important to determine whether antigen-specific epitopes recognized in patients following xenoantigen immunization are identical or not to those of the native antigen.
We have previously reported that, in both a preclinical model and a clinical trial, multiple immunizations (at least six) with a DNA vaccine encoding an autologous antigen were necessary to overcome tolerance and elicit a TH
1-polarized/CTL immune response to native PAP.2,4
While cumbersome, multiple immunizations with a DNA vaccine encoding an autologous antigen may be preferable as compared with a xenoantigen-coding DNA vaccine. Other potential means to increase the immunogenicity of this approach perhaps include different prime/boost strategies based on alternating immunization schedules of DNA vaccines encoding native and xenogeneic PAP. Such prime/boost strategies using DNA vaccines encoding xenoantigens and autologous antigens have been explored in melanoma and prostate cancer models.10,11,14
A prime/boost approach using xenogeneic DNA vaccines encoding mouse or human PSMA has specifically been explored in patients with prostate cancer.14
A priori, however, our data would suggest that priming or boosting the immune system with a DNA vaccine coding for a xenoantigen may promote an immune response against a dominant foreign epitope regardless of priming or boosting with a DNA vaccine encoding the autologous antigen. Alternatively, approaches to specifically modify known epitopes might be advantageous. This is notably feasible in the case of PAP, for which multiple HLA-A2, HLA-A24, HLA-A3 and HLA-DR1 epitopes have already been identified.23-28
In conclusion, we identified a RT1.Al-restricted immunodominant human PAP-specific peptide that is naturally processed, presented and recognized in Lewis rats following a xenogeneic immunization with a DNA vaccine encoding human PAP. However, no cross-reactive immune response was elicited to the corresponding peptide from the native antigen. Overall, these results indicate that immunization with a DNA vaccine encoding a xenoantigen may result in immune responses toward specific foreign epitopes, with no cross-reactive responses to the autologous antigen. For the specific targeting of PAP, our results suggest that a preferred vaccination strategy may be to immunize with vaccines encoding the autologous antigen. Future directions could be to modify DNA vaccines at specific epitopes, in order to increase MHC Class I and II molecule binding and/or T-cell receptor recognition, in order to directly target known MHC Class I and II epitopes rather than using foreign homologous proteins, a setting in which random epitope differences may not be advantageous.