Human Leukocyte Antigen Class II Alleles and Rubella-Specific Humoral and Cell-Mediated Immunity following Measles-Mumps-RubellaII Vaccinati
Mayo Vaccine Research Group, Department of Pediatric and Adolescent Medicine, Department of Health Sciences Research
Program in Translational Immunovirology and Biodefense, Mayo Clinic and Foundation, Rochester, Minnesota
We examined associations between human leukocyte antigen (HLA) class II genes and both rubella-specific immunoglobulin G antibodies and lymphoproliferative responses after measles-mumps-rubellaII (MMR-II) vaccination, in a population-based sample of 346 schoolchildren. The DPB1*0301, DPB1*0401, DPB1*1301, and DPB1*1501 alleles were significantly associated with rubella vaccineinduced antibodies. Alleles suggestive of being positively associated with rubella-specific lymphoproliferative stimulation indices were DPB1*0301, DQB1*0501, DRB1*0101, and DRB1*1104. Conversely, the DPB1*0401, DPB1*1001, DPB1*1101, DQB1*0202, and DRB1*0701 alleles were negatively associated with rubella-specific lymphoproliferation. This study of HLA class IIrestricted humoral and cellular immune responses to rubella provides significant insight into mechanisms of vaccine response and new vaccine development.
Rubella virus (RV) is a positive-stranded RNA virus with 3 structural proteins2 glycoproteins located on the virion surface (envelope E1 and E2) and 1 nonglycosylated capsid (C) protein that is closely associated with the viral RNA core [1]. This virus is the causative agent of the infectious disease known as "German measles." The availability of effective vaccines has resulted in public health recommendations for routine rubella vaccination of infants and of susceptible women of child-bearing age [2].
Rubella immunization provides long-lasting immunity, which is believed to be mediated by humoral (i.e., antibody) and cellular immune responses. Measurement of circulating IgG antibody and of lymphoproliferation to RV are considered standard methods of measuring humoral and cell-mediated immunity to rubella vaccine [3, 4]. Studies of rubella structural proteins demonstrate that the E1 glycoprotein is a superior immunogen, compared with E2 or C, and suggest that E1 may be the most important rubella antigen [5].
Although associations have been found between HLA class II genes and measles vaccine response, little information is available about the role of allele-specific HLA class II polymorphisms and rubella vaccineinduced immune responses. A number of studies have identified DR-restricted T cell responses to rubella antigens [6, 7], particularly DR3- and DR4-restricted rubella E1 peptides [8].
HLA class II molecules present viral antigens in the form of peptides derived from the extracellular processing of rubella proteins and play an important role in the immune response to many viral infections [9]. This response occurs by class II presentation of these peptides to CD4+ T helper cells. Allelic variation in these polymorphic genes creates a diversity of HLA molecules capable of binding and presenting these peptides to T cells.
In a previous study, we identified associations between specific HLA class II alleles and measles vaccineinduced antibody levels following 1 dose of the measles-mumps-rubellaII (MMR-II) vaccine [10]. To test whether these findings are generalizable to other viral vaccines, we conducted a large population-based study to assess associations between HLA class II genes and both antibody and lymphoproliferative responses to rubella vaccine in children, following 2 doses of the MMR-II vaccine.
Patients and methods.
Details of subject recruitment have been described elsewhere [11]. In brief, we conducted a large population-based study to assess associations between HLA genes and immune responses to the MMR-II vaccine (Merck Research Laboratories) in healthy children and young adults in Olmsted County, Minnesota. A total of 346 subjects, of ages 1218 years and from all socioeconomic strata, were recruited into the study. All participants had documentation of having received 2 doses of MMR-II vaccine containing the attenuated RA27/3 Wistar strain of RV. Informed consent was obtained from all study participants, and human-experimentation guidelines of the US Department of Health and Human Services were followed. The institutional review board of the Mayo Clinic approved the study.
Rubella-specific IgG antibody titers were determined by whole rubellaspecific EIA (Enzygnost Anti-Rubella-Virus/IgG EIA; Dade Behring), according to the manufacturer's instructions [11]. A negative cutoff of 0.100 absorbance unit (equivalent to 4 IU/mL) was used.
The cellular immune status to rubella vaccine was assessed using an in vitro [3H]-thymidine incorporation assay [4]. Rubella-specific T cell responses were measured by proliferation of fresh peripheral blood leukocytes (2 × 105cells/well) incubated in RPMI 1640 medium, supplemented with 5% autologous sera, with live attenuated rubella vaccine virus (Meruvax II, Merck; Wistar RA 27/3 strain; 75 pfu/well). Phytohemagglutinin (5 g/mL) was used to assess cell vitality, and antigens including measles and mumps vaccine viruses were also used to measure T cell responses. Cells were pulsed with 1 Ci of [3H]-thymidine (Perkin-Elmer) for 18 h before harvest. Lymphoproliferation was measured, after 4 days, by assessment of [3H]-thymidine uptake. The background counts were subtracted. Results were expressed as antigen-specific stimulation indices (SIs), defined as the ratio of the median counts per minute of RV-stimulated wells to the median counts per minute of unstimulated wells.
Genomic DNA was extracted from blood samples by conventional techniques, using the Puregene extraction kit (Gentra Systems). For HLA typing, we utilized high-resolution DRB1 sequence-specific primers (SSPs) and reference-strand conformation analysis (RSCA), DQB1 SSP, and DPB1 SSP Unitray typing kits with the entire locus on a single tray (Dynal Biotech). Polymerase chain reaction (PCR) was followed by use AmbiSolv primer mixes to resolve ambiguities and was analyzed using MatchTools software. RSCA was performed on an ABI 377 (Applied Biosystems), and results were analyzed using RSCA Typer software (version 3.0).
Plots of immune response by assay date showed an upward trend of cellular proliferation values over time. We fit polynomial linear-regression models to evaluate this association and used the resulting models to recalibrate measures of cellular immune response. No recalibration was necessary for humoral immune response. We tested for deviation from Hardy-Weinberg equilibrium by use of the software HWE [12].
Associations between immune response and demographic and clinical variables were assessed using analysis of variance methods. Because of data skewness, all P values were calculated on the basis of log-transformed values. We compared the degree of association between antibody responses and T cell responses by use of Pearson correlation coefficients and log-transformed values.
Descriptive associations between immune response and HLA loci were obtained on an allelic level. Each subject contributed 2 observations to these summaries, 1 for each allele. Alleles were grouped for each locus by subtype and were summarized using medians and interquartile ranges.
After the descriptive evaluations, associations were more formally examined using linear-regression analyses. In contrast to the descriptive comparisons, each subject contributed 1 observation to these analyses, on the basis of an observed genotype. Regression variables were created for each allele and were coded as 0, 1, or 2, according to the number of copies of the allele that a subject carried. Rare alleles, defined as those with <5 occurrences overall, were pooled into a category labeled "other." Original response values were again replaced with corresponding log-transformed values. Global differences in immune response among all alleles within a given locus were evaluated by simultaneously including all but 1 of the allele variables in a linear-regression model. After conducting these global tests, we examined individual allele effects. This latter series of tests was performed in the spirit of Fisher's protected least significant difference test; individual associations were not considered statistically significant in the absence of global significance. Each allele variable was included in a separate linear-regression analysis, effectively comparing immune response for the allele of interest against all other alleles combined. All analyses described above were performed after adjusting for the following set of potential confounding variables: age, race, sex, age at first MMR-II vaccination, and age at second MMR-II vaccination.
Results and discussion.
The majority of subjects were white (94%), and the median ages at first and second vaccination were 15.6 months and 12.1 years, respectively. There were 185 (53%) boys and 161 (47%) girls in the study. Median values for rubella humoral responses (expressed as antibody level) and cellular responses (expressed as SI) were 38.63 IU/mL and 2.29, respectively. Girls demonstrated significantly higher antibody responses than boys (median antibody level, 42.4 IU/mL vs. 33.9 IU/mL; P = .02). However, cellular responses (lymphoproliferation) to rubella vaccination were not sex or age associated (median SI, 2.28 for girls vs. 2.30 for boys; P = .72). The correlation between antibody response and calibrated T cell response was .04 (P = .50).
We found no violations of Hardy-Weinberg equilibrium for HLA class II DPB1 and DRB1 loci. However, a comparison of allele distributions for the DQB1 locus revealed possible departures from equilibrium (P = .0002). As a result, statistical comparisons involving the DQB1 locus should be viewed with a certain level of caution.
The association between class II HLA alleles and antibody level was examined (table 1). The global tests revealed significant associations with the DPB1 locus (P = .032). The individual alleles with the strongest associations were DPB1*0301 (median antibody level, 22.79 IU/mL; P = .024), DPB1*0401 (median antibody level, 42.39 IU/mL; P = .016), DPB1*1301 (median antibody level, 25.85 IU/mL; P = .050), and DPB1*1501 (median antibody level, 66.10 IU/mL; P = .032), suggesting that DPB1 class II molecules can restrict rubella vaccine humoral immune responses. The DPB1*0202 allele (median antibody level, 16.05 IU/mL; P = .079) was marginally associated with antibody response among rubella-vaccinated individuals.
Global tests did not reveal a statistically significant association between antibody levels and the DQB1 and DRB1 loci (P = 0.817 and 0.993, respectively). When alleles were examined individually in an exploratory fashion, only alleles DQB1*0303 (median antibody level, 31.39 IU/mL; P = .117) and DRB1*1104 (median antibody level, 48.12 IU/mL; P = .068) approached statistical significance without correction for multiple comparisons.
The associations between HLA class II loci (DPB1, DQB1, and DRB1) and rubella SI were examined (table 2). The global test revealed no significant associations between DPB1 alleles and rubella-specific lymphoproliferative responses. Results of our subsequent exploratory analyses by allele for the DPB1 locus suggested a potential association with allele DPB1*1101 (median SI, 1.85; P = .012). We also found potential evidence for lower levels of rubella-specific lymphoproliferation associated with alleles DPB1*0401 (median SI, 2.08; P = .056) and DPB1*1001 (median SI, 1.47; P = .062) alleles. In addition, there was some evidence that allele DPB1*0301 (median SI, 2.71; P = .059) was associated with a higher rubella lymphoproliferative response. However, these associations should be interpreted with caution, because of the absence of a significant global test result.
The global tests revealed marginally significant associations between cellular immune responses and the DQB1 and DRB1 loci (P = .061 and .056, respectively). For the DQB1 locus, the alleles with the strongest evidence for association with lymphoproliferative responses were DQB1*0202 (median SI, 1.85; P = .002) and DQB1*0501 (median SI, 2.97; P = .018). DQB1*0202 was associated with lower lymphoproliferative responses to rubella antigens, whereas the DQB1*0501 allele was associated with higher lymphoproliferation.
For the DRB1 locus, the alleles with the strongest evidence for association were DRB1*0101 (median SI, 2.87; P = .040), DRB1*0701 (median SI, 2.04; P = .013), and DRB1*1104 (median SI, 3.86; P = .014). These results suggest that RV antigens could be recognized and restricted in association with different HLA class II molecules.
We found associations between HLA class II alleles and humoral (IgG antibody) and cellular (rubella-specific lymphoproliferation) responses to rubella antigens after 2 doses of MMR-II vaccine. Genetic associations between rubella antibody levels and HLA class II alleles were not as strong as those observed for cellular responses. Whole virusspecific EIA testing was used as a rapid and inexpensive method of assessing antibody levels that is well accepted [4]. No alleles at the DRB1 and DQB1 loci were significantly associated with rubella antibody levels. The only locus with >1 significant allele was the DPB1 locus (alleles DPB1*0301, DPB1*0401, DPB1*1301, and DPB1*1501).
Several studies have examined the HLA effect on rubella immunity, although literature on associations between rubella-vaccinespecific lymphoproliferative responses and HLA genes is lacking. DR4- and DR9-restricted T cell responses to rubella antigens have been reported [6], and DR3 and DR4 restriction of rubella E1 peptides has also been demonstrated [7,8]. We found associations between specific HLA class II DRB1 (DRB1*0101, DRB1*0701, and DRB1*1104), DQB1 (DQB1*0202 and DQB1*0501) and DPB1 (DPB1*0301, DPB1*0401, DPB1*1001, and DPB1*1101) alleles and rubella-specific lymphoproliferative responses.
The DPB1*0301 and DPB1*0401 alleles are associated with both antibody levels and lymphoproliferation. In addition, the DPB1*1301/*1501 and DPB1*1001/*1101 alleles demonstrate evidence of associations with rubella antibody and lymphoproliferative responses, respectively. Since DPB1*0401 is overrepresented in white populations, one would expect that naturally processed and presented RV and self-peptides can efficiently bind and be recognized in the context of DPB1 molecules, and peptides that bind to DPB1*0401 could have a great impact as peptide-based vaccines [1315]. To date, no human HLA class II DPB1-restricted epitopes in the RV have been described. Although DR and DQ molecules are believed to be the major restriction determinants for CD4+ T cells, there is increasing evidence suggesting that DP might play an important role in regulating immune function. Thus, the identification of specific DPB1-restricted viral epitopes can be useful for the development of novel peptide-based vaccines against infectious diseases, including rubella [13].
In summary, our study provides evidence that HLA genes have important immunogenetic associations with circulating antibody levels and rubella-specific lymphoproliferative responses in children, following 2 doses of MMR-II vaccine. Results demonstrate that DPB1, DQB1, and DRB1 molecules may act as restriction elements for the recognition of the RV peptides by human T cells. Our study of HLA allele associations with immune responses to rubella should provide the experimental framework for the selection of immunodominant epitopes that trigger immune responses in the human population and, eventually, for the development of synthetic peptide vaccines.
Acknowledgments
We thank the parents and children who participated in this study. We acknowledge the efforts of the fellows, technicians, and nurses from the Vaccine Research Group. We gratefully acknowledge Kim S. Zabel for her editorial assistance in preparing this article.
References
1. Pettersson RF, Oker-Blom C, Kalkkinen N, et al. Molecular and antigenic characteristics and synthesis of rubella virus structural proteins. Rev Infect Dis 1985; 7:S1409. First citation in article
2. Frey T. Report of an international meeting on rubella vaccines and vaccination, 9 August 1993, Glasgow, United Kingdom. J Infect Dis 1994; 170:5079. First citation in article
3. Balfour HH Jr, Groth KE, Edelman CK, Amren DP, Best JM, Banatvala JE. Rubella viraemia and antibody responses after rubella vaccination and reimmunisation. Lancet 1981; 1:107880. First citation in article
4. Mitchell LA, Tingle AJ, Décarie D, Shukin R. Identification of rubella virus T-cell epitopes recognized in anamnestic response to RA27/3 vaccine: associations with boost in neutralizing antibody titer. Vaccine 1999; 17:235665. First citation in article
5. Chaye HH, Mauracher CA, Tingle AJ, Gillam S. Cellular and humoral immune responses to rubella virus structural proteins E1, E2, and C. J Clin Microbiol 1992; 30:23239. First citation in article
6. Ou D, Mitchell LA, Décarie D, Tingle AJ, Nepom GT. Promiscuous T-cell recognition of a rubella capsid protein epitope restricted by DRB1*0403 and DRB1*0901 molecules sharing an HLA DR supertype. Hum Immunol 1998; 59:14957. First citation in article
7. Ou D, Mitchell LA, Ho M, et al. Analysis of overlapping T- and B-cell antigenic sites on rubella virus E1 envelope protein. Influence of HLA-DR4 polymorphism on T-cell clonal recognition. Hum Immunol 1994; 39:17787. First citation in article
8. Nepom GT, Domeier ME, Ou D, Kovats S, Mitchell LA, Tingle AJ. Recognition of contiguous allele-specific peptide elements in the rubella virus E1 envelope protein. Vaccine 1997; 15:64852. First citation in article
9. Pamer EG. Antigen presentation in the immune response to infectious diseases. Clin Infect Dis 1999; 28:7146. First citation in article
10. Poland GA, Ovsyannikova IG, Jacobson RM, et al. Identification of an association between HLA class II alleles and low antibody levels after measles immunization. Vaccine 2001; 20:4308. First citation in article
11. Ovsyannikova IG, Jacobson RM, Vierkant RA, Jacobsen SJ, Pankratz VS, Poland GA. The contribution of HLA class I antigens in immune status following two doses of rubella vaccination. Hum Immunol 2004; 65:150615. First citation in article
12. Guo SW, Thompson EA. Performing the exact test of Hardy-Weinberg proportion for multiple alleles. Biometrics 1992; 48:36172. First citation in article
13. Castelli FA, Buhot C, Sanson A, et al. HLA-DP4, the most frequent HLA II molecule, defines a new supertype of peptide-binding specificity. J Immunol 2002; 169:692834. First citation in article
14. al Daccak R, Wang FQ, Theophille D, Lethielleux P, Colombani J, Loiseau P. Gene polymorphism of HLA-DPB1 and DPA1 loci in caucasoid population: frequencies and DPB1-DPA1 associations. Hum Immunol 1991; 31:27785. First citation in article
15. Baselmans PJ, Pollabauer E, van Reijsen FC, et al. IgE production after antigen-specific and cognate activation of HLA-DPw4-restricted T-cell clones, by 78% of randomly selected B-cell donors. Hum Immunol 2000; 61:78998. First citation in article, 百拇医药(Inna G. Ovsyannikova, Rob)
Program in Translational Immunovirology and Biodefense, Mayo Clinic and Foundation, Rochester, Minnesota
We examined associations between human leukocyte antigen (HLA) class II genes and both rubella-specific immunoglobulin G antibodies and lymphoproliferative responses after measles-mumps-rubellaII (MMR-II) vaccination, in a population-based sample of 346 schoolchildren. The DPB1*0301, DPB1*0401, DPB1*1301, and DPB1*1501 alleles were significantly associated with rubella vaccineinduced antibodies. Alleles suggestive of being positively associated with rubella-specific lymphoproliferative stimulation indices were DPB1*0301, DQB1*0501, DRB1*0101, and DRB1*1104. Conversely, the DPB1*0401, DPB1*1001, DPB1*1101, DQB1*0202, and DRB1*0701 alleles were negatively associated with rubella-specific lymphoproliferation. This study of HLA class IIrestricted humoral and cellular immune responses to rubella provides significant insight into mechanisms of vaccine response and new vaccine development.
Rubella virus (RV) is a positive-stranded RNA virus with 3 structural proteins2 glycoproteins located on the virion surface (envelope E1 and E2) and 1 nonglycosylated capsid (C) protein that is closely associated with the viral RNA core [1]. This virus is the causative agent of the infectious disease known as "German measles." The availability of effective vaccines has resulted in public health recommendations for routine rubella vaccination of infants and of susceptible women of child-bearing age [2].
Rubella immunization provides long-lasting immunity, which is believed to be mediated by humoral (i.e., antibody) and cellular immune responses. Measurement of circulating IgG antibody and of lymphoproliferation to RV are considered standard methods of measuring humoral and cell-mediated immunity to rubella vaccine [3, 4]. Studies of rubella structural proteins demonstrate that the E1 glycoprotein is a superior immunogen, compared with E2 or C, and suggest that E1 may be the most important rubella antigen [5].
Although associations have been found between HLA class II genes and measles vaccine response, little information is available about the role of allele-specific HLA class II polymorphisms and rubella vaccineinduced immune responses. A number of studies have identified DR-restricted T cell responses to rubella antigens [6, 7], particularly DR3- and DR4-restricted rubella E1 peptides [8].
HLA class II molecules present viral antigens in the form of peptides derived from the extracellular processing of rubella proteins and play an important role in the immune response to many viral infections [9]. This response occurs by class II presentation of these peptides to CD4+ T helper cells. Allelic variation in these polymorphic genes creates a diversity of HLA molecules capable of binding and presenting these peptides to T cells.
In a previous study, we identified associations between specific HLA class II alleles and measles vaccineinduced antibody levels following 1 dose of the measles-mumps-rubellaII (MMR-II) vaccine [10]. To test whether these findings are generalizable to other viral vaccines, we conducted a large population-based study to assess associations between HLA class II genes and both antibody and lymphoproliferative responses to rubella vaccine in children, following 2 doses of the MMR-II vaccine.
Patients and methods.
Details of subject recruitment have been described elsewhere [11]. In brief, we conducted a large population-based study to assess associations between HLA genes and immune responses to the MMR-II vaccine (Merck Research Laboratories) in healthy children and young adults in Olmsted County, Minnesota. A total of 346 subjects, of ages 1218 years and from all socioeconomic strata, were recruited into the study. All participants had documentation of having received 2 doses of MMR-II vaccine containing the attenuated RA27/3 Wistar strain of RV. Informed consent was obtained from all study participants, and human-experimentation guidelines of the US Department of Health and Human Services were followed. The institutional review board of the Mayo Clinic approved the study.
Rubella-specific IgG antibody titers were determined by whole rubellaspecific EIA (Enzygnost Anti-Rubella-Virus/IgG EIA; Dade Behring), according to the manufacturer's instructions [11]. A negative cutoff of 0.100 absorbance unit (equivalent to 4 IU/mL) was used.
The cellular immune status to rubella vaccine was assessed using an in vitro [3H]-thymidine incorporation assay [4]. Rubella-specific T cell responses were measured by proliferation of fresh peripheral blood leukocytes (2 × 105cells/well) incubated in RPMI 1640 medium, supplemented with 5% autologous sera, with live attenuated rubella vaccine virus (Meruvax II, Merck; Wistar RA 27/3 strain; 75 pfu/well). Phytohemagglutinin (5 g/mL) was used to assess cell vitality, and antigens including measles and mumps vaccine viruses were also used to measure T cell responses. Cells were pulsed with 1 Ci of [3H]-thymidine (Perkin-Elmer) for 18 h before harvest. Lymphoproliferation was measured, after 4 days, by assessment of [3H]-thymidine uptake. The background counts were subtracted. Results were expressed as antigen-specific stimulation indices (SIs), defined as the ratio of the median counts per minute of RV-stimulated wells to the median counts per minute of unstimulated wells.
Genomic DNA was extracted from blood samples by conventional techniques, using the Puregene extraction kit (Gentra Systems). For HLA typing, we utilized high-resolution DRB1 sequence-specific primers (SSPs) and reference-strand conformation analysis (RSCA), DQB1 SSP, and DPB1 SSP Unitray typing kits with the entire locus on a single tray (Dynal Biotech). Polymerase chain reaction (PCR) was followed by use AmbiSolv primer mixes to resolve ambiguities and was analyzed using MatchTools software. RSCA was performed on an ABI 377 (Applied Biosystems), and results were analyzed using RSCA Typer software (version 3.0).
Plots of immune response by assay date showed an upward trend of cellular proliferation values over time. We fit polynomial linear-regression models to evaluate this association and used the resulting models to recalibrate measures of cellular immune response. No recalibration was necessary for humoral immune response. We tested for deviation from Hardy-Weinberg equilibrium by use of the software HWE [12].
Associations between immune response and demographic and clinical variables were assessed using analysis of variance methods. Because of data skewness, all P values were calculated on the basis of log-transformed values. We compared the degree of association between antibody responses and T cell responses by use of Pearson correlation coefficients and log-transformed values.
Descriptive associations between immune response and HLA loci were obtained on an allelic level. Each subject contributed 2 observations to these summaries, 1 for each allele. Alleles were grouped for each locus by subtype and were summarized using medians and interquartile ranges.
After the descriptive evaluations, associations were more formally examined using linear-regression analyses. In contrast to the descriptive comparisons, each subject contributed 1 observation to these analyses, on the basis of an observed genotype. Regression variables were created for each allele and were coded as 0, 1, or 2, according to the number of copies of the allele that a subject carried. Rare alleles, defined as those with <5 occurrences overall, were pooled into a category labeled "other." Original response values were again replaced with corresponding log-transformed values. Global differences in immune response among all alleles within a given locus were evaluated by simultaneously including all but 1 of the allele variables in a linear-regression model. After conducting these global tests, we examined individual allele effects. This latter series of tests was performed in the spirit of Fisher's protected least significant difference test; individual associations were not considered statistically significant in the absence of global significance. Each allele variable was included in a separate linear-regression analysis, effectively comparing immune response for the allele of interest against all other alleles combined. All analyses described above were performed after adjusting for the following set of potential confounding variables: age, race, sex, age at first MMR-II vaccination, and age at second MMR-II vaccination.
Results and discussion.
The majority of subjects were white (94%), and the median ages at first and second vaccination were 15.6 months and 12.1 years, respectively. There were 185 (53%) boys and 161 (47%) girls in the study. Median values for rubella humoral responses (expressed as antibody level) and cellular responses (expressed as SI) were 38.63 IU/mL and 2.29, respectively. Girls demonstrated significantly higher antibody responses than boys (median antibody level, 42.4 IU/mL vs. 33.9 IU/mL; P = .02). However, cellular responses (lymphoproliferation) to rubella vaccination were not sex or age associated (median SI, 2.28 for girls vs. 2.30 for boys; P = .72). The correlation between antibody response and calibrated T cell response was .04 (P = .50).
We found no violations of Hardy-Weinberg equilibrium for HLA class II DPB1 and DRB1 loci. However, a comparison of allele distributions for the DQB1 locus revealed possible departures from equilibrium (P = .0002). As a result, statistical comparisons involving the DQB1 locus should be viewed with a certain level of caution.
The association between class II HLA alleles and antibody level was examined (table 1). The global tests revealed significant associations with the DPB1 locus (P = .032). The individual alleles with the strongest associations were DPB1*0301 (median antibody level, 22.79 IU/mL; P = .024), DPB1*0401 (median antibody level, 42.39 IU/mL; P = .016), DPB1*1301 (median antibody level, 25.85 IU/mL; P = .050), and DPB1*1501 (median antibody level, 66.10 IU/mL; P = .032), suggesting that DPB1 class II molecules can restrict rubella vaccine humoral immune responses. The DPB1*0202 allele (median antibody level, 16.05 IU/mL; P = .079) was marginally associated with antibody response among rubella-vaccinated individuals.
Global tests did not reveal a statistically significant association between antibody levels and the DQB1 and DRB1 loci (P = 0.817 and 0.993, respectively). When alleles were examined individually in an exploratory fashion, only alleles DQB1*0303 (median antibody level, 31.39 IU/mL; P = .117) and DRB1*1104 (median antibody level, 48.12 IU/mL; P = .068) approached statistical significance without correction for multiple comparisons.
The associations between HLA class II loci (DPB1, DQB1, and DRB1) and rubella SI were examined (table 2). The global test revealed no significant associations between DPB1 alleles and rubella-specific lymphoproliferative responses. Results of our subsequent exploratory analyses by allele for the DPB1 locus suggested a potential association with allele DPB1*1101 (median SI, 1.85; P = .012). We also found potential evidence for lower levels of rubella-specific lymphoproliferation associated with alleles DPB1*0401 (median SI, 2.08; P = .056) and DPB1*1001 (median SI, 1.47; P = .062) alleles. In addition, there was some evidence that allele DPB1*0301 (median SI, 2.71; P = .059) was associated with a higher rubella lymphoproliferative response. However, these associations should be interpreted with caution, because of the absence of a significant global test result.
The global tests revealed marginally significant associations between cellular immune responses and the DQB1 and DRB1 loci (P = .061 and .056, respectively). For the DQB1 locus, the alleles with the strongest evidence for association with lymphoproliferative responses were DQB1*0202 (median SI, 1.85; P = .002) and DQB1*0501 (median SI, 2.97; P = .018). DQB1*0202 was associated with lower lymphoproliferative responses to rubella antigens, whereas the DQB1*0501 allele was associated with higher lymphoproliferation.
For the DRB1 locus, the alleles with the strongest evidence for association were DRB1*0101 (median SI, 2.87; P = .040), DRB1*0701 (median SI, 2.04; P = .013), and DRB1*1104 (median SI, 3.86; P = .014). These results suggest that RV antigens could be recognized and restricted in association with different HLA class II molecules.
We found associations between HLA class II alleles and humoral (IgG antibody) and cellular (rubella-specific lymphoproliferation) responses to rubella antigens after 2 doses of MMR-II vaccine. Genetic associations between rubella antibody levels and HLA class II alleles were not as strong as those observed for cellular responses. Whole virusspecific EIA testing was used as a rapid and inexpensive method of assessing antibody levels that is well accepted [4]. No alleles at the DRB1 and DQB1 loci were significantly associated with rubella antibody levels. The only locus with >1 significant allele was the DPB1 locus (alleles DPB1*0301, DPB1*0401, DPB1*1301, and DPB1*1501).
Several studies have examined the HLA effect on rubella immunity, although literature on associations between rubella-vaccinespecific lymphoproliferative responses and HLA genes is lacking. DR4- and DR9-restricted T cell responses to rubella antigens have been reported [6], and DR3 and DR4 restriction of rubella E1 peptides has also been demonstrated [7,8]. We found associations between specific HLA class II DRB1 (DRB1*0101, DRB1*0701, and DRB1*1104), DQB1 (DQB1*0202 and DQB1*0501) and DPB1 (DPB1*0301, DPB1*0401, DPB1*1001, and DPB1*1101) alleles and rubella-specific lymphoproliferative responses.
The DPB1*0301 and DPB1*0401 alleles are associated with both antibody levels and lymphoproliferation. In addition, the DPB1*1301/*1501 and DPB1*1001/*1101 alleles demonstrate evidence of associations with rubella antibody and lymphoproliferative responses, respectively. Since DPB1*0401 is overrepresented in white populations, one would expect that naturally processed and presented RV and self-peptides can efficiently bind and be recognized in the context of DPB1 molecules, and peptides that bind to DPB1*0401 could have a great impact as peptide-based vaccines [1315]. To date, no human HLA class II DPB1-restricted epitopes in the RV have been described. Although DR and DQ molecules are believed to be the major restriction determinants for CD4+ T cells, there is increasing evidence suggesting that DP might play an important role in regulating immune function. Thus, the identification of specific DPB1-restricted viral epitopes can be useful for the development of novel peptide-based vaccines against infectious diseases, including rubella [13].
In summary, our study provides evidence that HLA genes have important immunogenetic associations with circulating antibody levels and rubella-specific lymphoproliferative responses in children, following 2 doses of MMR-II vaccine. Results demonstrate that DPB1, DQB1, and DRB1 molecules may act as restriction elements for the recognition of the RV peptides by human T cells. Our study of HLA allele associations with immune responses to rubella should provide the experimental framework for the selection of immunodominant epitopes that trigger immune responses in the human population and, eventually, for the development of synthetic peptide vaccines.
Acknowledgments
We thank the parents and children who participated in this study. We acknowledge the efforts of the fellows, technicians, and nurses from the Vaccine Research Group. We gratefully acknowledge Kim S. Zabel for her editorial assistance in preparing this article.
References
1. Pettersson RF, Oker-Blom C, Kalkkinen N, et al. Molecular and antigenic characteristics and synthesis of rubella virus structural proteins. Rev Infect Dis 1985; 7:S1409. First citation in article
2. Frey T. Report of an international meeting on rubella vaccines and vaccination, 9 August 1993, Glasgow, United Kingdom. J Infect Dis 1994; 170:5079. First citation in article
3. Balfour HH Jr, Groth KE, Edelman CK, Amren DP, Best JM, Banatvala JE. Rubella viraemia and antibody responses after rubella vaccination and reimmunisation. Lancet 1981; 1:107880. First citation in article
4. Mitchell LA, Tingle AJ, Décarie D, Shukin R. Identification of rubella virus T-cell epitopes recognized in anamnestic response to RA27/3 vaccine: associations with boost in neutralizing antibody titer. Vaccine 1999; 17:235665. First citation in article
5. Chaye HH, Mauracher CA, Tingle AJ, Gillam S. Cellular and humoral immune responses to rubella virus structural proteins E1, E2, and C. J Clin Microbiol 1992; 30:23239. First citation in article
6. Ou D, Mitchell LA, Décarie D, Tingle AJ, Nepom GT. Promiscuous T-cell recognition of a rubella capsid protein epitope restricted by DRB1*0403 and DRB1*0901 molecules sharing an HLA DR supertype. Hum Immunol 1998; 59:14957. First citation in article
7. Ou D, Mitchell LA, Ho M, et al. Analysis of overlapping T- and B-cell antigenic sites on rubella virus E1 envelope protein. Influence of HLA-DR4 polymorphism on T-cell clonal recognition. Hum Immunol 1994; 39:17787. First citation in article
8. Nepom GT, Domeier ME, Ou D, Kovats S, Mitchell LA, Tingle AJ. Recognition of contiguous allele-specific peptide elements in the rubella virus E1 envelope protein. Vaccine 1997; 15:64852. First citation in article
9. Pamer EG. Antigen presentation in the immune response to infectious diseases. Clin Infect Dis 1999; 28:7146. First citation in article
10. Poland GA, Ovsyannikova IG, Jacobson RM, et al. Identification of an association between HLA class II alleles and low antibody levels after measles immunization. Vaccine 2001; 20:4308. First citation in article
11. Ovsyannikova IG, Jacobson RM, Vierkant RA, Jacobsen SJ, Pankratz VS, Poland GA. The contribution of HLA class I antigens in immune status following two doses of rubella vaccination. Hum Immunol 2004; 65:150615. First citation in article
12. Guo SW, Thompson EA. Performing the exact test of Hardy-Weinberg proportion for multiple alleles. Biometrics 1992; 48:36172. First citation in article
13. Castelli FA, Buhot C, Sanson A, et al. HLA-DP4, the most frequent HLA II molecule, defines a new supertype of peptide-binding specificity. J Immunol 2002; 169:692834. First citation in article
14. al Daccak R, Wang FQ, Theophille D, Lethielleux P, Colombani J, Loiseau P. Gene polymorphism of HLA-DPB1 and DPA1 loci in caucasoid population: frequencies and DPB1-DPA1 associations. Hum Immunol 1991; 31:27785. First citation in article
15. Baselmans PJ, Pollabauer E, van Reijsen FC, et al. IgE production after antigen-specific and cognate activation of HLA-DPw4-restricted T-cell clones, by 78% of randomly selected B-cell donors. Hum Immunol 2000; 61:78998. First citation in article, 百拇医药(Inna G. Ovsyannikova, Rob)