|Home | About | Journals | Submit | Contact Us | Français|
Immune senescence in the elderly results in decreased immunity with a concomitant increase in susceptibility to infection and diminished efficacy of vaccination. Nonhuman primate (NHP) models have proven critical for testing of vaccines and therapeutics in the general population, but a model using old animals has not been established. Towards that end, immunity to LcrV, a protective antigen from Yersinia pestis, was tested in young and old baboons. Surprisingly, there was no age-associated loss in immune competence; LcrV elicited high-titer, protective antibody responses in the older individuals. The primary responses in the younger baboons were lower, but they did show boosting upon secondary immunization to the levels achieved in the old animals. The LcrV antigen was also tested in mice and, as expected, age-associated loss of immunity was seen; older animals responded with lower titer antibodies and as a result, were more susceptible to Yersinia challenge. Thus, although age-related loss in immune function has been observed in humans, rodents and some nonhuman primates, baboons appear to be unusual; they age without losing immune competence.
Elderly individuals show diminished immune responses, making them significantly more susceptible to infections and cancer (reviewed in 1–4). In addition, vaccination protocols are typically less efficacious in the elderly and although higher doses of immunogen may enhance the response, it is typically still lower than the one elicited in younger individuals (5–8). Deficits in the ability of older subjects to generate immune responses, particularly to “new” antigens that they have not previously encountered, have been widely reported. On the other hand, immune memory to antigens encountered in one’s youth does survive aging and can be recalled in old age (9–10). Given the current demographic composition in the United States, the numbers of aging individuals will continue to grow and they already comprise a significant “at risk” population. Thus, it is critically important to develop and test protocols for enhancing immunity particularly to “new” antigens in old, as well as in young, individuals.
Most of the research into the effects of aging on immunity has incorporated rodent models and for the most part, analogous age-associated deficiencies of cellular and humoral immunity have been seen (11–16). For example, the ability to generate an immune response to a “new” antigen or epitope not previously encountered is significantly diminished in older animals. On the other hand, memory immunity to antigens encountered in one’s youth appears largely intact (17–20). Similarly, the age-associated loss of immunity can be overcome by giving multiple immunizations or higher doses of the antigen. This further substantiates the imperative to test vaccine protocols for their effectiveness in both old and young subjects.
Although many vaccines are first tested in rodents, this may not be ideal for protection studies since mice are resistant to many human pathogens (like HIV) due to sequence difference in their cellular receptors. Thus, many vaccine protocols for use in humans have been tested in nonhuman primate (NHP) models (21–26). However, the vast majority of these studies have been undertaken in young or middle-aged NHPs and none of these primate models has been validated for use in testing vaccines for efficacy in older individuals. Thus, in the current study, the ability of young and old NHPs to respond to a “new” antigen has been assessed. We chose to focus on the baboon, Papio hamadryas, for several reasons. It is an excellent primate model system due both to its close genetic relatedness (≈ 96% DNA homology) and the similarity of its immune system to humans (27). For example, unlike macaques and some other monkeys, baboons resemble humans and chimpanzees in exhibiting four IgG subclasses (28). Moreover, since baboons breed well in captivity, they are more readily available that some other NHPs. Baboons are being used extensively in infectious disease and vaccine studies (21,24–25,27) so it will be important to assess the effects of aging on this NHP model. It has been reported that serum autoantibodies in baboons increase with age, analogous to humans (29), but there are no studies that assess the effects of aging on humoral immunity. Fortunately for this study, the Southwest National Primate Research Center (SNPRC) in San Antonio maintains the largest colony of baboons worldwide; it consists of more than 3700 individual animals, including a geriatric cohort.
Given that aging most dramatically affects immune responses to antigens not previously encountered by the subject, it was imperative to select an immunogen that would elicit a primary response in the baboon colony. Thus, we chose LcrV, a protein antigen from Yersinia pestis, the causative agent of bubonic plague. Y. pestis is the most virulent bacterial pathogen currently known and in geographic areas where it is endemic in rodent populations, including the southwestern United States, humans remain at risk. Any baboon that had come into contact with Y. pestis would most likely have succumbed as the infection is typically fatal. Thus, none of the subjects used in this study were likely to have had a prior exposure to this virulent bacterium and therefore they should respond to LcrV as a “new” antigen.
Furthermore, although there is no currently licensed plague vaccine for use in the US, a new subunit vaccine, which includes LcrV as one of its components, is showing promise (30–32). By incorporating LcrV in this study, we could assess both the titer of reactive antibody produced and its ability to protect against infection. This was considered a significant advantage as there is growing concern that plague may re-emerge as a significant danger to human health due to the recent identification of multi-drug resistant strains of Y. pestis; thus making the development of an effective vaccine a priority (33–34). Yet, in no case has a protein-based vaccine for plague been tested in older animals whose immune responses are likely compromised. Thus, these studies will provide the first data on the effects of aging on the humoral immune response to LcrV in two different species, mice and baboons. Moreover, although the Y. pestis LcrV antigen was chosen for this study, our findings should be generalizable to all protein-based and subunit vaccines and should also provide important insights into the use of NHP models for testing vaccine efficacy in the elderly.
In order to produce recombinant immunogens, wild-type CO92 genomic DNA, containing the pMT and pCD1 virulence plasmids, was used as the template in PCR reactions to amplify the caf1 (F1) and lcrV genes. The primers used for F1 were: 5′-GAC GAC GAC GAC AAG GAT TTA ACT GCA AGC (forward) and 5′-TCA CTA TCA TCA TTA TTG GTT AGA TAC GGT TAC (reverse). The primers designed to amplify the LcrV coding sequence were: 5′-GAC GAC GAC GAC AAG ATG ATT AGA GCC TAC GAA CAA AAC CCA CAA CAT (forward) and 5′-AAG ACC TTG TGA GCA TCC TCG (reverse). In addition, each of the forward primers included sequences added at the 5′ end to introduce an enterokinase cleavage site; this is shown underlined in the primer sequences above. The LcrV and F1 PCR products were TA cloned into the bacterial expression vector pQE-30UA vector (Qiagen) which provides an amino terminal histidine tag to facilitate protein purification. The chimeric antigen, LcrV::TTFC was created by subcloning the TTFC (tetanus toxoid fragment C) coding sequence in-frame downstream from LcrV at the BamH I site in pQE-30; the original TTFC clone was provided by Dr. Robert Ulrich and is described elsewhere (35). All three inserts were sequenced by the UTHSCSA DNA Core Facility to ensure that the proper sequence had been cloned as a functional translational fusion.
The recombinant clones were transformed into E. coli BL21 (DE3) pLys S and expression was induced by the addition of IPTG. Cell pellets were collected and disrupted by sonication and the lysate was cleared by centrifugation at 20000 X G for 20 min. Recombinant proteins were purified from the soluble fraction using affinity chromatography on nickel chelating resin (Pharmacia) and elution with 500 mM imidazole. The fractions containing recombinant protein were further purified by gel filtration followed by ion exchange chromatography on a resource Q column before being eluted with 500mM NaCl. Positive fractions were pooled and the buffer was exchanged to a phosphate buffered saline solution pH 7.4 on a gel filtration column. Contaminating endotoxin was removed with polymixin agarose (Sigma) and the protein was stored in aliquots at −80°C prior to use in immunization protocols.
To demonstrate purity of the recombinant proteins, a sample of each was fractionated by size on an SDS-polyacrylamide gel and stained (Figure 1). In addition, a duplicate gel was electrophoretically transferred to Duralon Membrane (Millipore) and the resulting Western blot was incubated with an antibody against the vector-encoded “his” tag (Qiagen) and developed using an alkaline phosphatase conjugated secondary antibody.
The purified recombinant proteins were absorbed to 25% alum (Vol/Vol) and used to immunize young (2½ years) and old (19–24 years) baboons from the pedigreed colony at the Southwest National Primate Research Center. Each animal was vaccinated by intramuscular injection with 100 μg of LcrV or the chimeric antigen, LcrV::TTFC. Sera were obtained from each animal prior to immunization (pre-immune) and at intervals following the primary immunization (4 and 8 weeks). All of the baboons received a second immunization with 100 μg of recombinant LcrV in Alum six months subsequent to the primary exposure. The animals were bled 2 weeks and 6 weeks after the secondary inoculation.
C57Bl/6 mice were obtained from the NIA contract colony (Harlan) and were used at either 3 months of age (young) or 19–21 months (old). The old and young mice were immunized with 10 μg of LcrV, LcrV::TTFC, or F1 antigen in alum. The mice were bled from the retroorbital sinus prior to immunization and at intervals post primary (4 and 9 weeks) and secondary (2.4 weeks and 5.4 weeks) exposures. The sera were frozen at −20°C for subsequent analyses.
Direct ELISAs were developed to measure the levels of serum antibodies specific for LcrV and F1 using the recombinant proteins generated above. For assessing reactivity to TTFC, a recombinant protein lacking the “his” tag was obtained from commercial sources (Roche). In all cases, Immunosorb 96-well plates (Nunc) were coated with the appropriate antigen (5 μg/ml for LcrV and F1, 10 μg/ml for TTFC). A dilution series of each baboon or murine sera was prepared and incubated with the antigen. Specific antibody binding was detected with HRP-conjugated anti-monkey IgG (KPL) or rabbit anti-mouse IgG (Sigma). The ELISAs were developed by the addition of a chromogenic substrate ABTS and the absorbance at 410 nm was determined. Titers were fitted to a sigmoidal curve (GraphPad Prism 5) and the end-point titers at 0.1 optical density above background were determined by interpolation.
Statistically significant differences between young and old animals were determined by one-way ANOVA (p value ≤0.05). In order to satisfy the assumption of variance equivalence among treatment groups, a log base ten transformation of the data was performed. The statistical significance of differences in responses between the young and old animals in a given treatment group was then assessed by one way analysis of variance on the transformed data.
ELISAs were also performed using HRP-conjugated sheep antibodies specific for human IgG1 (AP006), IgG2 (AP007), IgG3 (AP008), and IgG4 (AP009) from The Binding Site, Birmingham, U.K. For this purpose, ELISA plates were coated with recombinant LcrV as described above and then incubated with baboon and mouse sera diluted 1:500; this dilution was within the linear range for total IgG. The secondary antibodies were then tested in duplicate wells at three different dilutions (1:1000, 1:3000, and 1:9000). The plates were developed and read as described above.
To assess the protective capacity of the baboon sera, pools of individual sera from a given time point were tested for the ability to protect mice from Y. pestis challenge; this is the USPH approved method for measuring protective immune responses to this virulent pathogen (30). Briefly, 0.5 ml of each pool (from a given time point) was used to passively immunize female CD-1 mice (5/group). As a negative control, one group of mice was given 0.5 ml phosphate buffered saline (PBS) and tested in parallel. One day after serum transfer, the mice were sedated with avertin 0.5 mg/kg IP and challenged intradermally (ID) with 119 cfu of CO92 in the ear as described (36). The bacteria were originally obtained from the Centers for Disease Control and Prevention Select Agent Distribution Activity (CDC SADA Fort Collins, CO) and were grown overnight at 37°C in heart infusion broth supplemented with 0.2% xylose. The actual challenge dose delivered to the mice was determined by plating serial dilutions of the bacteria on Congo red plates and enumerating the number of colony forming units. In our hands the ID LD-50 for CO92 is between 1–10 cfu (Dube, unpublished). The mice were monitored daily and scored for percent survival.
Young and old mice were immunized twice with Alum (negative control) or with one of the test immunogens in Alum, Y. pestis F1 antigen, LcrV, or a chimeric molecule LcrV::TTFC, as described above. They were then challenged intradermally with 8000 cfu Y. pestis CO92. The percent survival (10 mice/group) was scored daily.
In order to test the effects of aging on immune responses to a protein antigen in a NHP model, the Y. pestis LcrV gene was cloned into a prokaryotic expression vector and protein immunogen was prepared. To demonstrate purity of the recombinant LcrV, it was subjected to SDS-polyacrylamide gel electrophoresis and Western blot analysis using an antibody against the vector-encoded “his” tag. As shown in Figure 1, there was one major protein species detected and it was of the size predicted for the “his”-tagged LcrV. The purified LcrV was then absorbed to alum, an adjuvant approved for use in humans, and used to immunize young (2½ years) and old (19–24 years) baboons from the pedigreed colony at the Southwest National Primate Research Center. Sera were obtained from each baboon prior to immunization (pre-immune) and at 4 and 8 weeks following the primary inoculation. The titer of LcrV reactive antibodies in each sample was determined by direct ELISA and the end-point titers were calculated using the GraphPad Prism 5 program; the results are shown in Figure 2. Since end-point titers could not be determined for the preimmune sera, all individual baboon samples were re-tested at one dilution point within the linear range (1/1200) to allow direct comparison across samples; these data are shown in Table I. As expected, the preimmune titers to LcrV were relatively low, although one of the young baboons did have a slightly elevated level (Table I). Upon immunization, the titers increased in all animals relative to the pre-immune levels, but unexpectedly, the elderly baboons responded at least as well as the younger animals; there was no age-associated loss in immune competence (Figure 2A).
The vigorous immune responses in the old baboons immunized with LcrV were unexpected and we considered that the antigen itself, LcrV, could be skewing the results (possibly by suppressing the responses in the young animals). In order to address this unlikely possibility, young (3 months of age) and old (19–21 months) C57Bl/6 mice were immunized with the same antigenic preparation, recombinant LcrV in alum. The mice were bled prior to immunization and at the prescribed intervals after exposure; the sera were then tested for anti-LcrV titers by direct ELISA. As shown in Figure 2B, the mice responded as expected; old mice show lower titers upon LcrV immunization than do the young animals. Thus, the antigen used did not have a negative effect in young mice. It appears more likely that baboons do not undergo the expected age-associated decline in immune responsiveness that has been reported for humans and numerous animal models.
To pursue this question further, old and young animals were immunized with a second immunogen, a chimeric molecule composed of LcrV and TTFC (tetanus toxoid fragment C); the purity of this immunogen was similarly demonstrated (Figure 1). Since they were immunized with a chimeric molecule, the baboons and mice should produce antibodies to each of the individual components, LcrV and TTFC. Therefore, ELISAs were performed independently with these two antigens (Figure 3). In both cases, old baboons responded as well, if not better, than the young animals. In mice, as expected, the pattern was reversed; young responded better than old.
In mice and humans, boosting of the immune response is expected with subsequent exposures to the antigen due to the prior activation of memory B and T cells. Since the primary immune responses in old baboons had been surprisingly vigorous, in comparison to young baboons, we asked whether memory T and B cell responses had been generated. Briefly, the mice and baboons previously immunized with LcrV were given a secondary immunization and the titers of the resulting responses were measured by ELISA. The secondary responses were higher for both young and old mice, as expected (Figure 2). In baboons, the secondary responses were boosted for the young animals, but were not enhanced over the primary responses for the older animals (Figure 2 and Table I). Thus, in both cases, memory immune responses were generated but again, the baboon profile does not look like the one typically seen in mice or humans.
Even though the elderly baboons showed higher antibody responses to LcrV immunization when compared to the young, it was possible that the antibodies produced by the aged animals would not be as capable of neutralizing virulent bacteria. To test this possibility, mice were passively immunized with baboon sera pooled from a given time point and were then challenged with virulent Y. pestis. The mice were scored daily; a baboon serum was considered protective if at least 4/5 recipient mice (≥80%) survived 10 days. Protection was not seen in the mice receiving young preimmune sera (0% survival, Figure 4A), old preimmune sera (20% survival, Figure 4A), or PBS (0% survival, data not shown). Thus, the infectious dose delivered was sufficient to cause terminal disease and the baboons did not have significant pre-existing immunity to Y. pestis.
The kinetics of survival (protection) is shown for the serum samples taken 4 weeks after the primary immunization (second panel in Figure 4A). As seen with the pre-immune samples, the young immune sera provide little protection. However, transfer of the sera from old baboons, even after a single immunization, provided high level protection from Y. pestis challenge. The survival data for all time points taken after primary immunization are summarized (table on right, Figure 4A). The protective capability of the old baboon sera had diminished by 8 weeks after immunization. This was somewhat surprising since there were still relatively high titers at this time. However, the protection assay is more physiologically relevant; it measures the functional capability of the antibodies elicited.
The ELISA titers were also generally predictive of protection after secondary exposure. For both young and old baboons, relatively high titers were seen within 2 weeks after immunization and these antibodies were protective (third panel of figure 4A). However, as seen with the primary immunization, protection declined relatively quickly and was below the 80% cut-off by 6 weeks post-secondary immunization (Figure 4A, summary table). Thus, these data confirm that the old baboons generate antibodies to LcrV that are both high titer and protective in vivo.
For comparison, the generation of protective responses in mice was tested directly by immunization and subsequent challenge with virulent Y. pestis, following the secondary exposure. As shown in Figure 4B, protection correlated with antibody titer, for the time points tested. As expected, old mice were much more susceptible to challenge, presumably due to the lower titers of anti-LcrV antibodies elicited. This finding was generalizable to three different immunogens, LcrV, LcrV::TTFC, and another plague vaccine component, F1. Thus, as expected, aging negatively impacts immunity in mice. On the other hand, older baboons appeared to have a healthier immune system than the younger animals.
There have been well documented changes in cytokine expression with aging. This could impact the class of antibody elicited in response to certain immunogens or pathogens. To assess whether age is affecting heavy chain switching in the primates, we determined the proportions of IgG subclass antibodies obtained in response to LcrV immunization. Like humans, baboons have four subclasses of IgG and the secondary reagents sold by The Binding Site have been reported to work with either of these primate species (28). Thus, we performed an ELISA to compare the isotype profiles of young and old baboons responding to LcrV immunization. The levels of IgG3 and IgG4 were below detection in all samples (data not shown), while IgG1 and IgG2 dominated the responses (Figure 5). There was no significant difference in the subclass distribution seen with age. Thus, these data confirm that the old baboons generate antibodies to LcrV that are high titer, protective in vivo, and of similar immunoglobulin subclasses to those seen in younger animals.
It has been generally accepted that all animals undergo immunosenescence associated with a decline in the ability of older individuals to mount a protective immune response to antigens not previously encountered. This report demonstrating that older baboons are at least as good as young animals at generating a protective humoral response to LcrV challenges that age-old paradigm. We considered several factors that might have contributed to this surprising finding. First, it was possible, although highly unlikely, that the specific baboon ages chosen for the analysis were not appropriate. The young baboons were 2 ½ – 3 years old when first immunized which approximates 7½ – 9 year old humans. As juveniles, they would have attained immunological maturity, but not sexual maturity. This age group was chosen based on extensive baboon data demonstrating that they would be immunologically competent (much like school-age children). For example, Attanasio and colleagues (29) demonstrated that the levels of serum immunoglobulin are close to adult levels by 1 year of age in baboons. In addition, in a West Nile virus vaccine study, a 3 year old juvenile baboon was immunized and generated IgG levels as high or higher than those seen in young adults (5.5 – 9 years of age) (24). In fact, even fetal baboons are capable of responding with IgG responses upon immunization with HBV antigens (37). The older animals (19–24 years of age) were chosen to be ≥2/3 of the average live span; which would be roughly equivalent to 57–72 year old humans and to 24 month old mice. If immunosenescence occurred equivalently in baboons, these animals should have shown a significant decrease in their antibody titers upon immunization, relative to the younger animals. Instead, they exhibited just the opposite; old baboons responded at very high levels to immunization.
In mice, the decline in immune capability has been shown to be a gradual one; middle-aged animals produce a humoral response that is intermediate, falling between the very low levels seen in older mice and the much high levels produced by younger ones (14,17). Given that the old baboons used in this study were not extremely old or geriatric, we expected detectable induced titers, but these should, if the paradigm were correct, have been lower than those achieved in younger animals. Clearly, this was not the case; aging appears to have less of an effect on humoral immunity in baboons than in other species examined to date.
We next explored the possibility that the old baboons had been previously exposed to LcrV. If so, the immunization for this study would have boosted an existing memory response which would potentially account for the unexpectedly high responses in the elderly baboons. However, we consider it highly unlikely for several reasons. First, we chose the antigen, LcrV, specifically because the animals should have been naïve. Y. pestis has not been reported at the Southwest National Primate Research Center and had any animal been infected, it would likely have died previously. Moreover, due to the extreme susceptibility of NHP to Yersinia spp, it is also unlikely that these animals were exposed to the related enteropathogenic species of Yersinia. Second, none of our animals had been previously enrolled in a study involving any antigen from this bacterium. Third, another antigen, TTFC, as part of the chimeric molecule, also elicited high responses from the older baboons. We can not unconditionally eliminate the possibility that the old animals had been exposed to some other bacterium or antigen that elicited a response that would be cross-reactive with LcrV and/or TTFC. However, we consider it highly unlikely as the pre-immune sera titers were uniformly low. Moreover, for this to explain the unexpectedly high humoral responses seen for the older baboons, they would all have to have had the same prior exposure and none of the younger animals would have been similarly affected. In other words, these genetically heterogeneous baboons would have all been responding in the same way; this is highly improbable. Thus, it appears most likely that the response to LcrV is a primary one in both age groups and that the higher titers of protective antibodies seen in the older baboons is due to a difference in the manner in which aging affects immune responsiveness in this NHP species.
In humans and rodents, immunosenescence is accompanied by decreased antibody affinity, due to diminished somatic mutation in older individuals (38). A decrease in affinity for antigen would be expected to negatively impact its ability to function optimally in response to a pathogen and may well compromise its neutralizing capabilities. Thus, we might have expected that the antisera from old baboons, in spite of their relatively high titer antibody content, would have proven less efficacious in the passive transfer protection assay. However, this was not the case; the sera from old baboons, even after a single immunization, were significantly more potent in protecting from Y. pestis challenge than were the sera from younger individuals.
Other aspects of the aging immune system in baboon have been studied, with mixed results. For example, old baboons show a decrease in the number of B cells and an increase in T cell numbers with age (39); this is unlike humans. However, some aspects of the immune dysregulation seen in humans and rodents may also occur in baboons since the levels of serum autoantibodies do increase with age (29), suggesting that tolerance mechanisms may be impacted by aging. By comparing the aging human immune system, which shows profound deficiencies in immunity, to the baboon immune system, which is less impacted, we may gain important insights into immunosenescence and novel mechanisms to slow the process down. Importantly, given that the expected decrease in immunity to newly encountered antigens does not occur in aging baboons, other NHP models should be considered for testing vaccine compositions and immune-based therapeutics, particularly for their use in the elderly. There have been only a few studies in NHP to assess the effects of aging on the immune system. Thus far, the results have been quite mixed, but there are parallels seen to humans in several of the NHP species tested. For example, thymus involution does occur with aging in macaques (40). Moreover, lymphocyte proliferative responses and humoral immunity have been shown to decline with age in these monkeys, but the changes did not entirely mimic humans (41). Additional studies in rhesus macaques have shown that CD28 expression declines with age as it does in humans but signaling and cell cycle regulation appear to differ (42,43). Similarly, in aging cynomolgus monkeys, an increase in double-positive (CD4+CD8+) T cells was demonstrated (44), suggesting a possible loss of thymic selection. Prior to selection of a particular NHP for use in testing vaccines or therapeutics, the responses to standard immunization protocols should be tested in young and old individuals in order to assess whether they show immunosenescence, like humans, or fail to show age-associated losses in humoral immunity, like baboons.
We would like to express our gratitude to Dr. Karen Rice and Ms. Sabrina Chatman for providing expertise and for coordinating the nonhuman primate protocols and procedures. We would also like to thank Dr. William Morgan for his very helpful advice in the statistical analysis of these data. In addition, we are grateful to Dr. Robert Ulrich for kindly providing the TTFC clone. Lastly, we would like to express our appreciation to Dr. Philip LoVerde whose laboratory screened the Binding Site secondary antibodies for reactivity with their baboon sera prior to our using them for this study.
1This work was supported by funding from the Southwest National Primate Center [pilot study grant P51 RR13986], the National Institute on Aging [R03AG22675 (SS)], and a UTHSCSA Presidential Research Enhancement Fund grant.