|Home | About | Journals | Submit | Contact Us | Français|
We have developed a murine model expressing the rhesus macaque (RM) Mamu-A*01 MHC allele to characterize immune responses and vaccines based on antigens of importance to human disease processes. Towards that goal, transgenic (Tg) mice expressing chimeric RM (α1 and α2 Mamu-A*01 domains) and murine (α3, transmembrane, and cytoplasmic H-2Kb domains) MHC Class I molecules were derived by transgenesis of the H-2KbDb double MHC Class I knockout strain. After immunization of Mamu-A*01/Kb Tg mice with rVV-SIVGag-Pol, the mice generated CD8+ T-cell IFN-γ responses to several known Mamu-A*01 restricted epitopes from the SIV Gag and Pol antigen sequence. Fusion peptides of highly recognized CTL epitopes from SIV Pol and Gag and a strong T-help epitope were shown to be immunogenic and capable of limiting an rVV-SIVGag-Pol challenge. Mamu-A*01/Kb Tg mice provide a model system to study the Mamu-A*01 restricted T-cell response for various infectious diseases which are applicable to study in RM.
Homologues of the human major histocompatibility complex (MHC) Class I (Bontrop et al., 1995), Class II (Slierendregt et al., 1992) and TCR genes (Levinson et al., 1992) can be identified in rhesus macaque (RM) and are remarkably similar to human counterparts. The close evolutionary similarity between RM and humans makes them suitable animal models to study the immunology of human immunodeficiency virus (HIV) infection using the RM infectious homologue simian immunodeficiency virus (SIV) (Letvin et al., 1983;Letvin et al., 1984) and other examples of prominent infectious disease models with clinical applications (Barry et al., 2006;Belshe et al., 1977;Desrosiers et al., 1989;Felsenfeld & Schmidt, 1975;Polozov et al., 1978;Raengsakulrach et al., 1999;Rivailler et al., 2004;Sun et al., 2006;Wang, 2001;Yang et al., 1999;Yue et al., 2007;Zhu et al., 1997). Specifically, both SIV and HIV are primate lentiviruses which use the CD4 protein as a primary receptor and chemokine receptors as coreceptors for cell entry (Boyson et al., 1996;Klatzmann et al., 1984). After infection with SIV, most RM develop a disease similar to HIV-1-induced AIDS (King et al., 1990), consequently SIV infection of the RM is currently considered the best animal model for HIV infection of humans. RM are considered the pre-clinical standard for HIV/AIDS vaccine development utilizing SIV challenge.
Several vaccine studies in RM have suggested that strong immune responses to SIV protect against the development of AIDS in some monkeys (Daniel et al., 1985;Day et al., 2006;Lehner et al., 1996;Mossman et al., 1996). CD8+ T cells contribute to this protective immunity based on accumulated evidence implicating T cell responses in the control of HIV/SIV replication (Brodie et al., 1999;Charini et al., 2001;Hirsch et al., 1994;Matano et al., 1998;Schmitz et al., 1999). The Mamu-A*01 Class I MHC allele is present in about 25% of RM of Indian origin (Knapp et al., 1997;Vogel et al., 1995). Tetramers for the single Mamu-A*01-restricted cytotoxic T lymphocyte (CTL) epitope Gag-CM9 (CTPYDINQM) have been developed to analyze Gag-specific CTL in the peripheral blood of SIV infected RM (Allen et al., 1998;Charini et al., 2001;Kuroda et al., 1998).
SIV disease progression was significantly delayed in Mamu-A*01 positive RM infected with SIV compared to monkey’s whom did not express this allele (Evans et al., 1999;Evans et al., 2000;Muhl et al., 2002;Pal et al., 2002;Zhang et al., 2002). The studies also showed that RM that express the Class I Mamu-A*01 molecule control viremia after infection by a chimeric lentivirus composed of elements from both HIV and SIV, referred to as SHIV, more effectively than Mamu-A*01 negative animals (Mothe et al., 2003). This improved control of viral replication correlated with the development of a high-frequency dominant epitope Gag-specific CTL response following infection (Seaman et al., 2005). Similarly, the Mamu B*17 alleles also has been associated with slower disease progression in SIV-infected RM (Allen et al., 1998;Evans et al., 2000;Loffredo et al., 2007;Loffredo et al., 2008). A comparable set of observations was more recently made for RM with the B*08 allele (Loffredo et al., 2007;Loffredo et al., 2008). The term elite controllers (EC) has been coined to distinguish RM which durably control SIV viral load, and several different Mamu alleles (Mamu-A*02, B*08, B*17) have been implicated (Friedrich et al., 2007;Loffredo et al., 2007). It is an important question whether the mechanisms for control of SIV replication are synonymous with HIV-infected humans also designated as EC and who are characterized by unique HLA alleles.
To begin to address some of these questions in an easily manipulatable model, we developed a transgenic mouse expressing the Mamu-A*01 gene in a double knockout (H-2Kb/Db) background. Immunization of the mice with rVV expressing SIV genes successfully caused stimulation of CD8+ T cells recognizing previously known Mamu A*01-restricted CTL epitopes that we quantified using cytokine flow cytometry. Mamu-A*01/Kb mice provides a simple and inexpensive model that may allow further mechanistic insight into EC status (Herberts et al., 2006;Loffredo et al., 2007;Loffredo et al., 2008). Additional alleles could be used to generate other Tg mouse strains that might inform on the mechanistic basis for suppression of SIV. The similarity of outcome of SIV infection in EC RM and HIV infection in EC humans with selected MHC alleles makes the study of the biology and immunology of these alleles of great interest. A Tg mouse model may improve understanding of the role of specific HLA or RM alleles on infection outcome.
The transgene encoding chimeric Mamu-A*01/Kb was constructed, as illustrated in Figure 1A. It is comprised of the native H-2Kb gene including the genomic promoter DNA sequence (Figure 1A, i), followed by the leader sequence through to the third intron derived from Mamu-A*01 genomic DNA, replacing the corresponding murine sequence (Figure 1A, ii). The chimeric gene terminates with the α3 domain of the H-2Kb gene, followed by transmembrane (TM) and cytoplasmic tail (CT) exons derived from the murine gene (Figure 1A, iii). To test if the chimeric Mamu-A*01/Kb protein has the capacity for cell surface expression, the chimeric gene construct was introduced into TAP-deficient RMA-S cells using standard approaches described in Materials and Methods. These cells are deficient in cell surface MHC Class I expression, because of the defect in endogenous peptide generation that can be overcome by providing exogenous optimal peptide motifs corresponding to MHC alleles carried by RMA-S cells (Ljunggren et al., 1990). The results show that a high level of Mamu-A*01/Kb complex formation is detected by incubation with epitope CM9 (an optimal Mamu-A*01 motif peptide, see Table 1), but not an irrelevant epitope in RMA-S transfectants (Figure 1B, right panel), and not in untransfected RMA-S cells (Figure 1B, left panel). Four identified Mamu-A*01 specific SIV epitopes: LA9, LV10, GM10 and MI8 (see Table 1) were also evaluated for their ability to promote cell-surface expression of the Mamu-A*01/Kb MHC protein at 37° C (Figure 1C). A relatively high level expression of Mamu-A*01/Kb was observed in the presence of LA9 peptide, and to a lesser extent, expression of Mamu-A*01/Kb was stabilized by incubation with GM10 and MI8 peptides. However, LV10 peptide stabilized a low level of the expression of Mamu-A*01/Kb. These results are evidence that the Mamu-A*01/Kb chimeric protein was translated and processed correctly, and could be brought to the cell surface in the presence of peptides that bound unmodified Mamu-A*01 MHC molecules.
Transgene DNA was microinjected into a total of 230 fertilized eggs from DKO mice, however only 17 pups were born and survived because of the reduced viability of the eggs from the DKO strain (data not shown). The Mamu-A*01 gene was detected in three founders by PCR using Mamu-A*01 specific primers on tail DNA from newly weaned mice (Figure 2A). Tg mice were backcrossed with DKO mice to derive Mamu-A*01 (heterozygous) animals for preliminary immunologic testing and to derive stable homozygous lines of Tg mice. Freshly prepared splenocytes were obtained from heterozygous Mamu-A*01/Kb Tg mice, though only weak levels of expression of Mamu-A*01 molecules was documented by staining using MAb W6/32, followed by flow cytometry (Figure 2B, right). Under the same conditions, no signal could be detected on spleen cells from parental DKO mice (Figure 2B, left). Since the Mamu-A*01/Kb chimeric gene is under the control of H-2Kb promoter, we treated Tg mouse splenocytes with murine IFN-γfor 16 hrs to activate the promoter (Israel et al., 1986). Consequently the surface expression of Mamu-A*01/Kb increased dramatically in cells from Mamu-A*01/Kb Tg mice (Figure 2B, right). IFN-γ also enhanced the background, which was detected by MAb W6/32 in cells from parental DKO mice, but the elevated level was much lower than in cells from Mamu-A*01/Kb Tg mice (Figure 2B, left).
The central goal of this mouse derivation was to generate a model to evaluate vaccines that would be applicable to testing in RM. As a first step, we evaluated the immunogenicity and processing of an rVV expressing full-length SIVGag-Pol antigens in Mamu-A*01/Kb Tg mice and the parental DKO strain of which they were derived (Figure 3A). The rVV-SIVGag-Pol was derived by homologous recombination and the expression of the SIV Gag protein was verified by Western Blot. SIV Gag protein is expressed as a 58-kDa polyprotein precursor (p58), which is cleaved in a sequential manner by the SIV-encoded viral protease into matrix, capsid (p27), and nucleocapsid proteins. Western Blot of lysates prepared from Hu TK− cells infected with rVV-SIVGag-Pol by using anti-p27 antibody confirmed the SIV Gag protein was expressed and correctly proteolytically processed (Figure 3B).
Mamu-A*01/Kb Tg (N=2) and parental DKO mice (N=2) were infected with rVV-SIVGag-Pol by i.p. injection for two weeks. Spleens were removed, and cell suspensions were incubated for 7 days with peptide CM9 (Table 1), using peptide-pulsed RMA-S/Mamu-A*01/Kb as APC. Analysis of cultured splenocytes by flow cytometry to detect intracellular staining of IFN-γ was performed. Remarkably, the immunodominant CM9 epitope stimulated 67.25% of CD8+ T lymphocytes to produce IFN-γfrom the splenocyte culture, shown as one example of the Mamu-A*01/Kb Tg mice we immunized, while only background levels were detected in the absence of the SIV epitope restimulation (Figure 3C, iii and iv). As expected, splenocytes from parental DKO mice showed minimal recognition of the CM9 epitope (Fig 3C, i and ii).
We further investigated the breadth of epitope recognition after rVV-SIVGag-Pol immunization of Mamu-A*01/Kb Tg mice, by examining the immune recognition of 4 previously characterized Mamu-A*01 restricted epitopes (LA9, LV10, GM10 and MI8, Table 1) after IVS with these 4 peptides. All of the peptides which were found to stabilize Mamu-A*01 surface staining in RMA-S/Mamu-A*01/Kb cells (Figure 1C) also induced CD8+ T lymphocytes to produce IFN-γ (Figure 3D). LA9 which stabilized a high level of the Mamu-A*01 surface staining in RMA-S/Mamu-A*01/Kb cells (Figure 1C), also induced a high level of IFN-γ producing CD8+ T lymphocytes (9.52%) in Mamu-A*01/Kb splenocyte cultures (Figure 3D, i). LV10 which minimally stabilized Mamu-A*01 surface staining in RMA-S/Mamu-A*01/Kb cells also induced a similar level of IFN-γ production as the other two peptides GM10 and MI8 which moderately stabilized Mamu-A*01 surface staining. These results are supporting evidence that the Mamu-A*01/Kb Tg mice contain a T cell repertoire that faithfully reproduces T cell recognition comparable to Mamu-A*01 RM, and will be a useful tool for the evaluation of virus-specific, Mamu-A*01 restricted T cell response.
We tested whether Mamu-A*01/Kb Tg mice could be used to distinguish and finely map Mamu-A*01 restricted epitopes to the level of minimal peptides starting with a library of 125 15-mer peptides overlapping by 11 amino acids that encompasses the entire SIVmac239 Gag amino acid sequence. Two Mamu-A*01/Kb Tg mice (Tg line 4) which were previously immunized with rVV-SIVGag-Pol were used for this approach. Immune splenocytes were stimulated with the complete library or 8 peptide subpools, each containing 16 component peptides from the whole library (Figure 4A). T cell responses induced against each pool of peptides were analyzed by ICC for IFN-γ. Peptide pool 33–48 reproducibly induced high frequencies of IFN-γ expressing CD8+ T cells (46% and 69%), while peptide pools 81–96 and 97–112 also induced moderate levels of IFN-γ expressing CD8+ T cells ranging from 5.45% to 31.8 % (Figure 4A). Each of the three responding pools was further analyzed by testing each of the component peptides individually. Peptides 45 and 46 containing the CM9 epitope sequence (Table 1) induced IFN-γ expressing T cells very efficiently (81.5% and 84.5%, Figure 4B), similar to the levels stimulated by the minimal 9-mer CM9 epitope (Figure 3C). In addition, peptides 38 and 39 also induced IFN-γ producing T cells, although at a lower level (16.2% and 9.32%). Peptide 38 contains a previously identified Mamu-A*01 restricted epitope LW9 (Table 1). In addition, as shown in Figure 4C, both peptides 92 and 93 induced a moderate frequency of IFN-γ producing CD8+ T cells (15.9% and 16.8%). These two 15-mer peptides contain an identified Mamu-A*01 restricted epitope, LA9 (Table 1).
In SIV or SHIV infected RM, Pol specific CTL responses were observed at low frequencies (Allen et al., 2001;Egan et al., 1999). These observations led us to investigate the processing of the Pol protein sequence in Mamu-A*01/Kb Tg mice. In rVV-SIVGag-Pol infected lysates, SIV Gag polyprotein was proteolytically cleaved precisely according to expected MW, suggesting that the protease encoded in the Pol sequence is active, and that SIV Pol protein was also expressed (Figure 3B). IVS was conducted on splenocytes from two rVV-SIVGag-Pol immunized Mamu-A*01/Kb Tg mice with eleven peptide subpools, each consisting of 24 15-mer peptides spanning the entire SIVmac239 Pol protein. 7 days later, the IVS-stimulated splenocytes were analyzed for expression of the intracellular cytokine IFN-γ in CD8 + T cells (Figure 5A). A moderate level of IFN-γ expressing CD8+ T cells were induced by peptide pool 145–168 (7.7% and 16.6%). In addition, low levels of IFN-γ expressing CD8+ T cells were reproducibly induced by peptide pools 1–24, 25–48, 49–72, 169–192 and 217–240 with frequencies ranging from 2.5% to 3.9%. We assessed individual component peptides from pool 145–168 for their ability to induce IFN-γ expressing CD8+ T cells (Figures 5B). We selected pool 145–168 for further analysis because it induced the highest level of IFN-γ producing CD8+ T cells among all the 11 subpools. Two peptides among the 24 separately evaluated from pool 145–168 showed equally high levels of response using the ICC assay (Figure 5B). Both peptides 156 and 157 contain the known Mamu-A*01 restricted epitope SV9 were found to induce IFN-γ producing T cells at equivalent levels (12.2% and 12.3%, Fig. 5B) and (Table 1).
Based on our previous success with the PADRE:CTL fusion peptide vaccine model for HIV vaccines in HLA A2/Kb mice (Daftarian et al., 2003), we synthesized an analogous structure for 3 different CTL epitopes from SIV Gag and Pol shown in Table 2 (MI8, LW9, CM9). The vaccination approach we employed for the studies shown in Figure 6 combined 100 nmole fusion peptide and 25μg CPG 7909 DNA (Hartmann et al., 2000), administered s.c., similar to our published studies for HIV-specific fusion peptides (Daftarian et al., 2003). Shown are results from two identically immunized mice for each peptide fusion, analyzed for IFN-γ producing T-cells using ICC (Figure 6A). ICC evaluations of splenocytes were conducted with a non-specific control peptide (grey bars) and the CTL epitope that is part of the fusion peptide (black bars) for each fusion peptide that was studied. Of the three CTL epitopes fused to the PADRE T-help peptide, the MI8 octamer was recognized most inefficiently after rVV-SIVGag-Pol infection, but its fusion peptide was the best at eliciting antigen-specific CD8+ T cells after immunization (Table 1 and Figure 6A). The LW9 nonamer was recognized with intermediate intensity, but its fusion peptide was poorly immunogenic, whereas the CM9 nonamer that was efficiently recognized after rVV-SIVGag-Pol infection also elicited strong immunogencity as a fusion peptide (Table 1 and Figure 6A). Interestingly, the MI8 fusion peptide immunized without CpG 7909, showed an average of only 8.0% IFN-γ+ CD8+ T-cells in Mamu-A*01 Tg mice (N=2), a >90% reduction from the immunization containing CPG 7909 DNA(Figure 6A).
The MI8, LW9, and CM9 fusion peptides represent a large range of immunogenicities. All 3 fusion peptides were administered to Mamu-A*01/Kb Tg mice as a series of two immunizations that were spaced two weeks apart, conducted as a limited dose titration between 1 and 100 nanomoles plus 25 μg of CpG 7909 ODN. Similar to the studies of Berzofsky and collaborators, fusion peptides were introduced s.c. and i.p. for both immunizations (Ahlers et al., 2001). Seven days after the second immunization, immunized mice and a control unimmunized litter-mate were i.p. administered 10 million infectious units (pfu) of rVV-SIVGag-Pol or rVV-GFP. Five days later, ovaries were removed and lysates were prepared, and stored in aliquots at −80°C. Infection was carried by a series of dilutions of the ovary lysate on CV-1 cell monolayers for a period of 18 hours by standard approaches. Results with accompanying p-values are presented in Figure 6B. Plaques were counted by immunostain using a monoclonal antibody to VV as a primary, and an HRP-conjugate as secondary (Wang et al., 2008). The left (L) and right (R) ovaries were separately analyzed, but results of both ovaries were consistent for both immunized and unimmunized mice. At the lowest dilution tested of 103 of the ovary lysate from PADRE:MI8-immunized mice (100 or 10 nmole), no plaques were visible in lysates made of ovaries from rVV-SIVGag-Pol challenged mice (N=8). In addition, PADRE:CM9 (10 nmole) immunized mice (N=5) also showed no evidence of viral plaques at 103 dilution of the ovary lysate, similar to the PADRE:MI8 immunized mice. However, the dose titration down to 1 nmole fusion peptide was only partially effective at controlling rVV-SIVGag-Pol infection for both MI8 (N=8) and CM9 (N=5) fusion peptides. The weakly immunogenic PADRE:LW9 (N=6) immunized mice (100 nmole) only gave partial protection to challenge with rVV-SIVGag-Pol, similar to mice immunized with 1 nmole of either MI8 or CM9 fusion peptides. As expected, non-immunized mice had substantial visible plaques at 107–108 dilutions of the ovary lysates from either challenge virus (rVV-GFP (N=5) or rVV-SIVGag-Pol (N=5), Figure 6B). These results are informative for exploring immunization strategies in RM.
Mamu-A*01/Kb Tg derived from DKO mice were successfully generated and employed to investigate recognition of Mamu-A*01 restricted epitopes produced by a model rVV-SIV. We constructed the Mamu-A*01 transgene based on results of previous studies that showed increased immunologic function of chimeric MHC molecules composed of α1α2 domains from human HLA and H-2Kb α3 from mouse (Kalinke et al., 1990;Vitiello et al., 1991). This strategy was used to enable TCR-CDR3 and MHC-α1α2 interactions to more efficiently activate T cells based on Mamu-A*01 restricted peptide specificity. Complex stability between CD8+ T cells and professional APC was enhanced by the interaction of endogenous murine CD8 molecules with the H-2Kb α3 mouse partner as a means to stabilize the tripartite complex (Vitiello et al., 1991). The first construct we developed was composed of the promoter and three first exons of Mamu-A*01, and theα3, transmembrane and cytoplasmic domains from H-2Kb genomic DNA (Figure 1A and not shown). Following transfection into murine fibroblast NIH3T3 cells, the cell surface expression of Mamu-A*01/Kb monochains was detected by indirect immunofluorescence. However, the cell surface expression of Mamu-A*01/Kb was minimal (data not shown). We attempted to improve the transcriptional efficiency in mouse by replacing the Mamu-A*01 promoter with the H-2Kb promoter (Figure 1A). This substitution proved successful, because of improved surface expression that could be dramatically enhanced by treatment of splenocytes derived from Mamu-A*01/Kb Tg mice with murine IFN-γ (Figure 2). This is the first example of a Mamu MHC Tg mouse which may serve as a paradigm for construction of other Mamu Class I allele strains.
We were concerned that differences in TAP specificity of RM and murine Class I MHC would reduce the efficiency of endogenous murine antigen-processing, resulting in reduced levels of peptide-loaded and surface expressed Mamu A*01 MHC molecules. The reduction of surface Mamu A*01, might negatively impact the frequency and antigen specificity of the Mamu A*01-restricted peripheral T cell repertoire in the Mamu-A*01/Kb Tg mice, rendering them not useful for RM-specific vaccine studies. Others have found that human β2-microglobulin (β2m) introduced into Tg mice by covalent linkage to the N-terminus of the transgenic Class I heavy chain constructed analogously, increased the cell surface expression of the chimeric molecule in RMA cells (Pascolo et al., 1997). However, we demonstrated that a relatively high level of Mamu-A*01/Kb expression could be stabilized by exogenous peptides in TAP-deficient mouse RMA-S lymphoma cells expressing Mamu-A*01/Kb without introducing covalently added human or RM β2m. Consistent with previous reports describing HHD Tg mice derived from H-2Db β2m knockout mice, the expression of Mamu-A*01/Kb molecules on T and B peripheral cells was low(Pascolo et al., 1997). Nevertheless, our results showed efficient usage of the Mamu-A*01/Kb monochain to generate a potent immune response induced by CD8+ T lymphocytes. Once homozygous animals are isolated, greater levels of CD8+ T lymphocytes may be observed. Similar to the case of HHD Tg mice, the expression of Mamu-A*01/Kb partially restores to normal levels, the low levels of CD8+ T lymphocytes found in the periphery before immunization or viral infection, potentially because of defective positive selection in the absence of a classical MHC Class I gene in the DKO mice (data not shown). Our results suggest that Mamu-A*01/Kb monochain is able to replace the missing endogenous Kb or Db MHC molecules to stabilize CD8 surface expression on circulating T lymphocytes in the Tg mice.
We examined the usefulness of the Mamu-A*01/Kb mice to aid in RM-specific CTL epitope identification. Our strategy was to immunize Mamu-A*01/Kb Tg mice with rVV-SIVGag-Pol that we constructed from SIV genomic sequence containing plasmid DNA. CD8+ T cell reactivity was screened using overlapping 15-mer peptide libraries spanning SIV Gag and Pol proteins. Four CTL epitopes located within the SIV Gag protein were identified. Among them, the previously characterized CM9 peptide, and two other epitopes called LA9 and LW9 (Allen et al., 2001) were identified, because they stimulated IFN-γ expression in splenic CD8+ T cells. A previously described epitope located within SIV Pol, SV9 (Egan et al., 1999) was also identified. Although the Mamu-A*01/Kb Tg mice showed dramatic CD8+ T cell responses to SIV immunodominant epitope CM9, they also recognized a broad distribution of epitopes in SIV Gag and Pol proteins. Both high (CM9) and moderate-low (LA9, LW9, SV9, LV10, GM10 and MI8, Table 1) affinity peptides elicited immunologic recognition in Mamu-A*01/Kb Tg mice. The immunogenicity of the Mamu-A*01 restricted peptides identified through rVV infection and peptide library screening suggests that the CD8+ T cell repertoire in the Mamu-A*01/Kb Tg mice faithfully represents at least a portion of the RM T cell repertoire. We also identified and fine mapped two novel Mamu-A*01-restricted CTL epitopes (data not shown): the SIV Gag-derived epitope RW9 (RAPRRQGCW) and Pol-derived epitope NE8 (NFPIAKVE). To confirm and further characterize these two newly defined epitopes with respect to their stage of appearance after SIV infection, we will infect Mamu-A*01 positive RM with SIV and enumerate antigen-specific CTL response in infected animals.
Tg mice expressing full length HLA Class I molecules were the first examples used to study HLA class I restricted CTL responses (Arnold & Hammerling, 1991). Either through virus infection or allo-stimulation by other HLA Class I alleles, these mice preferentially or exclusively developed H-2–restricted CD8+ T cell responses, rather than the desired HLA-restricted T cell responses (Barra et al., 1989;Engelhard et al., 1991;La Rosa et al., 2001). Consequently, H-2Db and β2m knockout mice expressing the HHD chimeric Class I molecule were derived to force the CD8+ T cell repertoire to be educated by Tg HLA Class I molecules (Pascolo et al., 1997). As a result, virus infection of these Tg mice generated predominantly HLA-A2.1–restricted CD8+ T cell responses against influenza A and vaccinia viruses as representative examples (Pascolo et al., 1997). We sought to simplify and speed up the process of deriving a Tg Class I strain in the DKO background. The process for deriving HHD mice required several time-consuming steps, and we sought to reduce the process to a single step. By doing so, it reduced the number of additional crosses to just one that would be needed to generate homozygotes. This work represents the first example of a new Tg mouse strain derived by direct transgenesis of a DKO strain. If we can improve the viability of the eggs from the DKO mice during microinjection, it may be the best way to generate Tg mice for other primate and human MHC Class I alleles.
There are many areas of study where these mice will prove useful. Recently, we developed a vaccine consisting of rhesus cytomegalovirus (RhCMV) pp65, gB and IE1 expressed via modified vaccinia Ankara (MVA), and evaluated it in RM, including some which are Mamu-A*01 positive (Yue et al., 2008). Rhesus pp65 and IE1-specific cellular immune responses were detected. To track the evolution of these cellular immune responses using tetramers, especially in CMV-negative RM, Mamu-A*01 restricted CTL epitopes need to be identified. The Mamu-A*01/Kb Tg mice, represent a facile system to preliminarily identify potential CTL epitopes that can be later verified in the Mamu A*01 RM. Identified epitopes would allow detailed studies including those with tetramers to determine avidity and other functional parameters (Egan et al., 1999). While we found no effect on specificity of epitope detection, all of our studies utilized an intermediate step of IVS to amplify the epitope-specific splenocyte populations prior to conducting ICC. Our approach of employing IVS when using transgenic mice is consistent with other studies from a variety of laboratories (Pascolo et al., 1997;Vitiello et al., 1997). While not optimal, the fact that protection studies correlate with the levels of immunogenicity measured in Figure 6A suggests minimal disruption of functional properties to distort the good correlation with the challenge results (Figure 6B).
CTL epitopes can be quickly identified using the mice, and verified using appropriately infected RM with Mamu-A*01 allele. Another area of investigation that will benefit is the testing of vaccine concepts in the relatively inexpensive Tg mice prior to evaluation in the costly RM model. Since all HIV vaccine concepts must pass a macaque feasibility test, using SIV or SHIV as a challenge virus, we envision preliminary testing in this mouse model. Viruses, peptides, or DNA products that elicit T cell response could be first tested for immunogenicity in mice prior to advanced testing in monkeys. Epitope based vaccines are candidates for preliminary evaluation in mice prior to monkey testing, including constructs similar to what we developed for HIV (Daftarian et al., 2005). Our preliminary investigation of 3 Mamu-A*01 epitopes incorporated into fusion peptide vaccines was informative, since the immunogenicity characteristics of CTL epitopes processed from rVV-SIVGag-Pol differed from the ranking of immunogenicity of the peptide fusion vaccines. Whether the immunogenicity ranking predicted by the mice is also similar in RM needs to be evaluated in a separate study. The results of the challenge study reassured and followed the results of the immunogenicity ranking, as LW9 fusion peptide performed poorly in both tests. In contrast, the MI8 and CM9 fusion peptides performed well in both immunogenicity and challenge tests. Many other antigens from pathogens that infect RM could be intensively studied immunologically in these mice to rapidly identify epitopes that could be used to study disease pathogenesis or vaccine strategies in the more costly primate model.
Continuously growing cell lines were maintained in RPMI 1640 (Cellgro, Herndon, CA) containing 10% FBS (Omega, Tarzana, CA), 4 mM L-glutamine (GIBCO, Rockville, MD), 0.1 mM 2-mercaptoethanol, and 100 U/ml penicillin and 100 μg/ml streptomycin (GIBCO). 2 × 107 RMA-S cells (Ljunggren et al., 1990) (transporter associated with antigen presentation [TAP] negative, a kind gift from Dr. Martin Kast, USC Keck School of Medicine) in 0.5 ml of PBS were electroporated with 15 μg of linearized pH2KbMamu-A*01α1α2 DNA at 250 V and 2000 μF using a BTX ECM 600 Electroporation System (BTX, San Diego, USA). After 24h, cells were transferred in selective medium containing 1.0 mg/ml G418 (Invitrogen, Carlsbard, CA) and cloned by limiting dilution in 96 well plates. Neomycin-resistant stable clones expressing the Mamu-A*01/Kb gene were identified by PCR using the Mamu-A*01 specific primers 5- GACAGCGACGCCGCGAGCCAA-3 and 5- CGCTGCAGCGTCTCCTTCCCC-3. These cells were expanded in the absence of G418 selection, but were periodically examined for expression of Mamu-A*01/Kb protein by indirect immunofluorescence using fluorescein isothiocyanate (FITC)-conjugated W6/32 MAb (Maziarz et al., 1986) (eBioscience, San Diego, USA). Cells were evaluated by incubation overnight at 26 °C with a Mamu-A*01 specific peptide and subsequent at 37° C, followed by flow cytometry analysis. C57BL/6 H-2Kb/Db knockout (DKO) mice were obtained from the Institute Pasteur (Paris, France) through an agreement with NIAID and Taconic Laboratories (Germantown, NY). Mice used throughout this study were bred and maintained under standard pathogen-free conditions in the American Association for Laboratory Animal Care approved animal care facility at City of Hope.
Genomic DNA was isolated from C57BL/6 mouse tails by using DNAzol (Molecular Research Center Inc., Cincinnati, OH). 5- CGCGGATCCTGTGCTGAGGGGACCAGC-3 and 5- ATTTGCGGCCGCGTCTGTCACCAAGTCCACTCCAGGC-3 primers were used to amplify a DNA fragment which contains the H-2Kb α3, transmembrane and cytoplasmic domains. The 1.3 kb PCR product was cloned into BamH I – Not I sites of vector pEGFP-N3 (Clontech, Palo Alto, CA). 5-CGCATTAATGTCGACGGTCAGGGAGGGAGGGAGAGGGTC-3 and 5-CCCAAGCTTGTCGGCGATTCGCGACTTCTGAG-3 primers were used to amplify the H-2Kb promoter. The 1.0 kb PCR product was cloned into Ase I-Hind III sites of pEGFP-N3 vector which we designated as pH2KbΔα1α2 (Figure 1A). The DNA fragment containing the leader, αl, and α2 domains of Mamu-A*01 was amplified from genomic DNA of Mamu-A*01 positive RM by PCR with primers 5-CCCTCGCGACAGGATGGCGGTCATGGCGC-3 and 5-CGCGGATCCCATTTGTCTCCCCTCC-3. The 1.1 kb PCR product was cloned into Nru I-BamH I sites of pH2KbΔα1α2 vector, and the completed plasmid was designated as pH2KbMamu-A*01α1α2 (Figure 1A). DNA sequence of all constructs was confirmed by standard automated DNA sequencing methods at the City of Hope Core Sequencing Facility.
A 3.4 kb Sal I-Not I chimeric monkey-mouse DNA fragment containing the murine H-2Kb promoter, the leader, αl, and α2 domains of rhesus macaque Mamu-A*01 and theα3, transmembrane, and cytoplasmic domains of murine H-2Kb was excised from pH2KbMamu-A*01α1α2 (Figure 1A) and purified by sucrose density gradient fractionation, and injected into the pronuclei of fertilized eggs from C57BL/6 H-2KbDb double MHC knockout (DKO) mice according to established protocols (Mann & McMahon, 1993). The microinjected eggs were transferred surgically to the uteri of 0.5 day-post-coitus pseudo-pregnant recipient female mice (Mann, 1993). Tg mouse founder lines were established by PCR detection of a 300 bp genomic DNA fragment using the Mamu-A*01 specific primers 5-CCCCCACGGACGGCCCGCGTCG-3 and 5-CCGTCGTAGGCGTACTGTTCAT-3. Heterozygous Tg mice from a founder line designated Mamu-A*01/Kb Tg mice line #4 (Figure 2A) was selected for these studies. Cell surface expression of Mamu-A*01 was detected in spleen cells by staining with mAb W6/32 and analyzing by FACS immediately, or following overnight culture in RPMI-10% FBS with 15ng/ml murine IFN-γ (PeproTech, Rocky Hill, NJ). All studies involving live animals were conducted following review and approval of all protocols by the City of Hope Institutional Animal Use and Care Committee, and in accordance with all applicable guidelines.
The SIVmac239 Gag-Pol genes were amplified using primers: 5-TAAGAATGCGGCCGCGAGATGGGCGTGAGAAACTCC-3 and 5- GACGTCGACCTATGCCACCTCTCTAGCC-3 from plasmid pSHIV-89.6 5′ (NIH AIDS Research and Reference Reagent Program). The resulting 4.4 kb DNA was cloned into pSC11-GFP using Not I and Sal I sites. pSC11-GFP is a modified pSC11 vector in which the β− galactosidase gene is replaced with the GFP gene (Chakrabarti et al., 1985) (kind gift of Dr. Bernard Moss, LVD, NIAID, NIH). The SIVGag-Pol recombinant vaccinia virus (rVV-SIVGag-Pol) was generated by transfecting the plasmid into WR strain VV-infected Hu TK− cells. rVV-SIVGag-Pol infected GFP plaques were detected using a Olympus CK40 fluorescence microscope (Olympus, Tokyo, Japan) and simultaneously selected with 0.1mg/ml bromodeoxyuridine (Sigma-Aldrich, St Louis, MO) for 5 rounds. The expression of SIV Gag protein was detected from cell lysates prepared from rVV-SIVGag-Pol infected Hu TK− cells by Western Blot analysis with anti-p27 MAb (NIH AIDS Research and Reference Reagent Program).
8 to 12 week old (N varies, see Figure Legends) mice (Tg line 4) were i.p. injected with 2 × 107 PFU (plaque forming units) rVV-SIVGag-Pol using a 1 ml syringe in a volume of 100 μL of PBS. After 14 days, the mice were killed, spleens were aseptically removed and splenocyte suspensions were produced by teasing the spleen through a sterile nylon mesh. RMA-S/Mamu-A*01/Kb as antigen-presenting cells (APC) were loaded with 100 μM of the SIV Gag or Pol peptides for 3 hours in a 37°C and 5% CO2 incubator with occasional shaking. The IVS was initiated by co-culture of 106 irradiated (5000 rad, Isomedix Model 19 Gammator; Nuclear Canada, Parsippany, NJ) peptide-loaded RMA-S/Mamu-A*01/Kb cells with splenocytes (3 × 106) from VV-infected or control mice for 7 days in 2 ml medium containing 10% T-Stim Culture Supplement (BD Biosciences, San Diego, CA).
A total of 1 × 106 cells from 7 day IVS spleen cell cultures were incubated at 37°C for 1h with 1 to 10 μM SIV peptides, or a mock stimulation without peptide. Cells were then treated with 10 μg of brefeldin A per ml to inhibit protein trafficking and incubated overnight at 37°C. Cells were then washed twice with PBS plus 0.1% BSA (Bovine Serum Albumin) and stained with FITC–conjugated CD8 MAb. After fixation and permeabilization with the Cytofix/Cytoperm Plus Kit (BD Biosciences) according to the manufacturer’s protocol, cells were washed twice and then incubated with Allophycocyanin-conjugated antibody to IFN-γ (BD Biosciences), and analyzed by FACS on a two laser-6 color FACS-Canto™ (BD-Immunocytometry systems, San Jose, CA). The number of double-positive cells is expressed as a percentage of the CD8+ T-cell population.
Groups of Mamu-A*01 Tg female mice (8- to 12-weeks) were immunized twice in a two week period on days 0 and 14 by a combined s.c. at the base of the tail and i.p. approach with various amounts of peptide vaccine plus 25μg synthetic oligodeoxynucleotide (ODN) 7909 with CpG motifs in a final volume of 100 μl saline as previously described. Seven days after the second immunization, the mice were challenged i.p. with 1×107 PFU of rVV-SIVGag-Pol or rVV-GFP. Five days post-challenge, mice were sacrificed and both ovaries were removed. Following homogenization and three rounds of freeze-thaw, ovary lysates (total volume was about 10μl per ovary) were assayed for VV titer by a modification of a standard approach. The titer of VV in ovary lysates was determined by plating serial 10-fold dilutions on duplicate six-well plates of CV-1 cells (~106 cells per well). Following overnight incubation, monolayers were immuno-stained with mAb to VV as primary, and an HRP-conjugate as secondary, and plaques were counted (Wang et al., 2008).
Individual nonamer and fusion vaccine peptides were prepared by the standard solid phase F-moc procedure using Symphony Peptide Synthesizer (Protein Technologies, Inc., Tucson, AZ). Peptides were purified by standard high performance liquid chromatography methods using Agilent 1200 series (Agilent Technologies, Santa Clara, CA), and molecular weight of peptides was confirmed by matrix assisted laser desorption/ionization (MALDI; Kratos, Chestnut Ridge, NY) as previously described (La Rosa et al., 2002). The complete SIVmac239 Gag and Pol (15-mer) peptides sets were obtained from the NIH AIDS Research and Reference Reagent Program.
Results of challenge studies shown in Figure 6B were analyzed using the Wilcoxon rank-sum test. All P-values were calculated as two sided, with significance defined as P<0.05. The statistical package Systat 12™ (Cranes Software Int. Ltd) was used for all calculations.
The authors wish to thank the staff of the Animal Resource Center at COH for their expert animal handling and assistance in the husbandry of the founder transgenic strains. Ms. Aparna Krishnan is acknowledged for guiding us and performing some early examples of IVS with mouse splenocytes. PCR-based tail typing to monitor the breeding colony was carried out with the help of Dr. Zhongqi Li and Ms. Pooja Manchada. We are grateful to Professor Lemonnier of the Pasteur Institute for agreeing to transfer the DKO mouse strain to the NIAID, which made them available to investigators through a cooperative agreement with Taconic Laboratories (Tarrytown, NY). We are grateful to the staff of the California National Primate Research Center for providing rhesus macaque DNA of the Mamu A*01 genotype. We thank the NIH AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH for supplying us with SIVmac239 gag and pol peptide libraries, pSHIV-89.6 and the anti-p27 mAb. This work was partially supported by AI062496, CA077544, and CA030206Prj3 to DJD, AI063356 to PAB and RR000169 to the California National Primate Research Center. The COH Cancer Center is supported by CA33572.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.