Aging has severe effects on antibody production, which can impact vaccine efficacy in the elderly. In older individuals, the primary antibody response is very weak and short-lived with reduced affinity maturation and reduced memory B cell generation (1
). These effects are thought to be due to reduced GC formation in the aged, because GCs are required for all of these events to occur (13
). The goal of our study was to determine which lymphocyte populations are responsible for the observed age-related declines in antibody production. To accomplish this, we used an adoptive transfer model using CD4 T cells from TCR Tg mice. Our previous studies have shown that T cells from aged AND Tg mice express a naive phenotype, with respect to both cell surface molecules (CD44lo
) and in vitro function (16
). This model eliminates the differences that are found in polyclonal CD4 T cell populations from young and aged mice and allows us to directly compare homogeneous populations of naive CD4 T cells of identical antigenic specificity, with the only difference being that they were obtained from young or aged Tg mice. Furthermore, this model involves the transfer of young and aged Tg CD4 T cell populations into CD4KO hosts, which lack CD4 T cells and, therefore, have no endogenous cognate helper activity. This allows us to specifically examine the helper function of the young and aged donor CD4 T cells.
The results presented in this study show that when we transfer Tg CD4 T cells from aged donors into young hosts, we observe significant reductions in antigen-specific B cell expansion and differentiation as well as reduced IgG production. These differences were evident at time points over a 3-wk period and over a wide range of donor cell concentrations. As we found that there was no difference in the migration of the young and aged donor cells, we surmised that the aged donor cells exhibited defects in cognate helper activity once they had trafficked to the B cell follicle. Therefore, we examined cell surface molecules that have been shown to be involved in CD4 T cell cognate function (21
). Even though the aged donor cells expressed normal levels of CXCR5, CD134 and CD28, they did not up-regulate CD154 to the same extent as young donor cells. This could have a significant impact on their cognate helper activity and on GC formation in the young hosts, considering CD40-CD154 interactions are required for GC formation as well as antibody affinity maturation and class switching (33
). Interestingly, we have shown that these aged donor CD4 T cells exhibit reduced NFκB activation after TCR stimulation (20
), which may negatively impact up-regulation of CD154 as this transcription factor is important for appropriate CD154 expression by T cells (39
In this study, our model also shows that host components, such as B cells and FDCs, do not exhibit the same degree of age-related defects that we observe in aged CD4 T cells. When we transfer young donor CD4 T cells into aged CD4KO mice, we find no reduction in antigen-specific B cell expansion or differentiation or serum IgG production. This is in contrast to other reports that have demonstrated defects in both B cell populations and FDCs with increasing age. Whereas many age-related defects in B cell responses can be attributed to reduced cognate function of aged CD4 T cells, other defects have been hypothesized to be intrinsic to B cell populations. Most notably, an age-related decline in B-lineage precursors leading to reduced production of new mature B cells has been described. Interestingly, this defect does not result in a decline in the generation of new B cells or in the numbers of peripheral mature B cells (42
). Other studies have examined the antigen-specific responses of young versus aged B cells using splenic fragment cultures. Although the numbers of B cells responding to haptens such as NP was shown to decline with age, the amount of antibody produced per cell remained constant (44
). We do not find differences in the number of NP-specific B cells in aged hosts in our model, but our studies are quite different from the splenic fragment culture experiments. Most notably, in the splenic fragment culture assay, antigen-specific B cells were enumerated by their ability to produce antibodies, whereas our assay involves only the ability to bind NP. Therefore, it is not surprising that we observed somewhat different results.
In addition, other studies, using in vitro systems, have shown defects in the ability of aged B cells to undergo Ig class switching, independent of aging effects on CD4 cognate helper function (36
). We do not find this in our antigen-specific in vivo model, which is quite different from in vitro polyclonal stimulation of B cell populations. When we transfer young Tg CD4 T cells into young or aged hosts, we find no age-related differences in either serum IgG1 titers or in the proportion of NP-specific cells expressing surface IgG1 after immunization. It is possible that there are other undetected defects in affinity maturation or somatic mutation of the NP-specific B cells in the aged hosts. We are examining these parameters in ongoing experiments.
Our model is well suited for this sort of analysis because in both the young and aged CD4KO hosts there is no endogenous CD4 helper activity (). This allows us to compare predominantly naive B cell populations in both the young and aged hosts (something that is not possible in immunologically intact models, where memory B cells accumulate with increasing age; 45
). One other possibility to account for our results is that B cells in the CD4KO model do not undergo age-associated changes that are observed in normal mice. This is probable because, unlike normal wild-type mice, B cells in unimmunized aged CD4KO mice do not show a decrease in follicular (CD21int
) B cells or a decrease in B220 expression (unpublished data), both of which are associated with aging B cell populations (46
One other cell type that has been implicated in age-related reductions in GC formation is the FDC. It has been shown, using in vitro models, that FDCs from older animals bind fewer antigen–antibody immune complexes, thus reducing the amount of cross-linked antigen available to B cells. This is thought to lead to age-related reductions in GC formation, affinity maturation, and the development of memory B cells (27
). Although these studies definitively show age-related defects in FDCs, it is not clear that they might have an impact on in vivo GC formation. Other researchers have shown that FDCs are not absolutely necessary for GC formation (48
), and we see no evidence of decreased GC development in aged hosts using our model. It is probable that our immunization protocol, using NP conjugated to PCC at a 3 to 1 ratio along with alum precipitation, provides enough cross-linking to adequately stimulate NP-specific B cells, thus bypassing any age-related functional defects in FDC populations.
In summary, our results show that age-related reductions in humoral responses are likely the result of decreased CD4 T cell helper activity and that other cell types, such as B cells and FDCs, are less affected by age. In addition, we hypothesize that this defect is related to significantly reduced CD154 expression by the aged CD4 T cells, leading to decreased B cell expansion, GC formation and IgG production. By defining specific age-related defects in specific lymphocyte populations, we can then target strategies for enhancing humoral responses. Ultimately, this will allow us to improve vaccine efficacy in the elderly.