Relationship between functional ability in older people, immune system status, and intensity of response to CMV

Study population

One hundred elderly (76 women and 24 men) living at the Santa Teresa nursing home (Oviedo, Spain) were enrolled in the study. Blood for the hematological and immunological analyses were drawn from the 100 individuals. All subjects received a physical examination and answered standardized questionnaires to assess clinical history, current disease, and medication. Inclusion criteria were elders who were older than 69 years and who were classified according to the Barthel index as a measure of their functional status. Exclusion criteria were conditions with possible influence on the immune system such as recent or current infection, inflammation, autoimmune or malignant disease, malnutrition, abnormal laboratory data (hemoglobin <12 mg/dL, leukopenia <3,500 cells/μL, neutropenia <1,500 cells/μL, leukocytosis >15,000 cells/μL, platelets <10⁵ cells/μL, and PCR >5 mg/dL), and pharmacological interference (steroids, nonsteroidal anti-inflammatory agents, and immunosuppressive drugs). Informed consent was obtained from the elders prior to participation in the study. The study was approved by the Hospital Central de Asturias (Oviedo, Spain) ethics committee.

The functional abilities of the subjects were assessed by using the Barthel Index of Activities of Daily Living (BI) (Mahoney and Barthel 1965). Each person was evaluated at 10 tasks that measured daily functioning for various activities of daily living and mobility. The highest BI score (100) meant that the person needed no assistance with any part of the tasks. The BI scores were used to divide the elderly into four groups from total independence to total dependence: group 0 = 100 (n = 25), group 1 = 95–80 (n = 25), group 2 = 75–40 (n = 27), and group 3 = 35–0 (n = 23), as previously described (Saxena et al. 2006; Supervia et al. 2008).

Hematological analysis and immunological phenotyping

The hematological parameters were determined by a Sysmex XT-2000i (Syxmex, Hamburg-Norderstedt, Germany). Cytometric studies were acquired and analyzed in the FACSCalibur Cytometer using CellQuest software (BD Biosciences, San José, CA, USA). CaliBRITE Beads (BD Biosciences) were used to adjust instrument settings, set fluorescence compensation, and check instrument sensitivity. Surface staining of EDTA peripheral blood was performed with Multiset CD3-FITC/CD16+56-PE/CD45-PerCP/CD19-APC Reagent, anti-CD4 (APC), anti-CD8 (PE), anti-CD8 (PerCP), anti-CD45RA (FITC), anti-CD27 (PE) (Immunostep, Salamanca, Spain), anti-CD4 (PerCP), anti-CD28 (PerCP), anti-CCR7 (Alexa Fluor 647), anti-CD3 (FITC), anti-CD45RO (FITC), anti-CD25 (APC) (BD Biosciences), anti-NKG2D (PE), and anti-CD127 (PE) (eBioscience, San Diego, CA, USA). One hundred microliters of whole blood from elderly were stained with different combinations of labeled monoclonal antibodies for 20 min at room temperature. Samples were red blood lysed with FACS Lysing Solution (BD Biosciences), washed in PBS, and analyzed with CellQuest software in the FACSCalibur Cytometer. Appropriate isotype control mAbs were used for marker settings.

To analyze the proliferation status of CD4+ and CD8+ T cells, peripheral blood mononuclear cells (PBMC) were isolated by centrifugation on Ficoll-Hypaque gradients (Lymphoprep; Nycomed, Oslo, Norway). CD4+ and CD8+ T cells from elderly were isolated with magnetic beads (Myltenyi Biotec GmbH, Bergisch Gladbach, Germany). Cells were lysed and fixed with Fixation/Permeabilization Solution (eBioscience), permeabilized with Permeabilizing Buffer (eBioscience), and stained with anti-Ki-67-PE (eBioscience) for 30 min at room temperature. Cells were washed and resuspended in 1% paraformaldehyde until FACS analysis. Frequencies of positive cells for Ki-67 from groups 0 and 1 (n = 10) and groups 2 and 3 (n = 10) were compared.

TRECs quantification

The DNA of isolated CD4+ (purity > 90%) from PBMCs was extracted using a QIAamp DNA Mini Kit (Qiagen GmbH, Hilden, Germany) according to the manufacturer’s instructions. Quantification of signal-joint (sj) TREC was performed by using SYBR Green real-time quantitative PCR and an iCycler thermocycler (Bio-Rad; Life Science Research Group, Hercules, CA, USA). The sequences of the utilized primers were the following: forward primer 5′-CCATGCTGACACCTCTGGTT-3′, reverse primer 5′-TCGTGAGAACG GTGAATGAAG-3′. As an internal control measurement to normalize for input DNA, the Cα constant region that remains present on TCR genes despite rearrangement processes was amplified in every sample tested (forward primer 5′-CCTGATCCTCTTGTCCCACAG-3′, reverse primer 5′-GGATTT AGAGTCTCTCAGCTGGTACA-3′). Thermal cycling conditions began with 50°C for 2 min and 95°C for 10 min, followed by 40 cycles of 95°C for 15 s and 60°C for 1 min. Experimental samples were run in duplicate and the replicate average value recorded as the sample result.

Cytomegalovirus (CMV) and influenza virus serology

Immunoglobulin G levels of CMV-specific antibodies were determined by enzyme-linked immunosorbent assay Vir-ELISA Anti-CMV-IgG (Viro-Immun Labor-Diagnostika GmbH, Oberursel, Germany) according to the manufacturer’s specifications. Patient samples were quantified and interpreted by means of the calculation of the ratio (cutoff index = OD value of sample/cut-off value), whereby a ratio of 1.0 is equivalent to the cut-off value. Cutoff indexes >1.1 were considered positive, and the result of this ratio is a semi-quantitative titer.

Anti-influenza virus antibodies in serum obtained from elderly individuals were measured by ELISA as previously described (Ohishi et al. 2002), with some modifications (Kang et al. 2004). The OD values of individual samples were compared against a calibration curve made by the OD values of serial dilutions of the same internal control serum from a healthy vaccinated young person throughout the experiments. Because antibody titer decreased significantly with time since vaccination, we normalized this value by dividing the titer by the elapsed time since immunization.

Activation studies

The activation of heparinized whole blood by anti-CD3 or CMV antigens was assessed by surface staining with anti-CD69 (eBioscience). Briefly, heparinized whole blood (250 μL) was stimulated with soluble anti-CD3 (10 ng/mL) (eBioscience) or with a CMV supernatant (10⁴ PFU/mL) in 15 mL conical polypropylene tubes for 18 h at 37°C and 5% carbon dioxide. CMV-infected cell lysate was prepared by infecting human embryonic lung fibroblasts with the AD169 CMV strain, and viral titers in the supernatant were determined by standard plaque assays. The virus was inactivated by five repeated freeze–thaw cycles. The cells were also stained with anti-CD4 and anti-CD8 mononuclear antibodies.

Proliferation cultures

PBMCs were resuspended in PBS at a final concentration of 5–10 × 10⁶ cells/mL and incubated with 1.5 μM CFSE (Invitrogen, Paisley, Scotland, UK) for 10 min at 37°C, washed with RPMI 1640 medium containing 2 × 10³ M l-glutamine and Hepes (BioWhitaker, Verviers, Belgium) twice, and cultured at 1 × 10⁶ cells/mL in the presence of soluble anti-CD3 (10 ng/mL) and CMV extract (10⁴ PFU/mL). The proliferative responses of CD4+ and CD8+ T cells were analyzed on day 7 by FACSCalibur after staining with anti-CD4 and anti-CD8.

ELISPOT assay

PBMCs were resuspended to 2 × 10⁶/mL in RPMI 1640 medium containing 2 × 10⁻³ M l-glutamine and Hepes and supplemented with 10% FCS (ICN Flow, Costa Mesa, CA, USA) and antibiotics. PBMC were stimulated in triplicate with anti-CD3 (10 ng/mL) and with the same extract of CMV used in the activation and proliferation cultures. PBMC (100 μL, 1.5 × 10⁵/well) were placed into each well of a 96-well filter plate (Millipore, Billerica, MA, USA) that was coated with anti-IFN-γ Ab (BD Biosciences), and the cells were cultured for 20 h. IFN-γ captured by the plate-bound Ab was detected by biotinylated anti-IFN-γ Ab (BD Biosciences), followed by streptavidin-conjugated alkaline phosphatase (BD Biosciences). Spots were developed using the red color-forming substrate 3-amino-9-ethylcarbazole (BD Biosciences). Spots were counted with the ELISPOT reader system ELR02 (Autoimmun Diagnostika GmbH, Straβberg, Germany). A sample of cultured cells was stained with anti-CD3 (FITC) and this percentage of CD3⁺ cells was used to calculate the number of T cells in each well of the assay.

Statistical analysis

Two groups were compared with the Mann–Whitney U non-parametric method for data that were not normally distributed, or with the Student’s t test. More than two groups were compared by using the non-parametric Kruskal–Wallis test or by ANOVA analysis for data that were normally distributed. Results were expressed as median and range or mean and standard deviation. In some graphs, mean and standard error of the mean were displayed. Correlations between variables were assessed by the nonparametric Spearman test (ρ). The χ² test was used to compare dichotomous variables and multiple linear regression was used in multivariate analysis. Analyses were performed using the SPSS 15.0 statistical software package program (SPSS Inc. Chicago, IL, USA) and p values of 0.05 or less were considered significant.

Demographic and hematologic characteristics of the study population

The characteristics of the 100 individuals enrolled in the study are shown in Table 1. People were placed into four groups according to their BI. The female/male ratio in each group was 2.1:1 in group 0, 4.0:1 in group 1, 2.7:1 in group 2, and 2.8:1 in group 3. There were no significant differences in the age of donors belonging to the four groups. The average number of drugs taken by members included in the study and the most frequent pathologies associated with each group were listed.

Blood cell counts and an immune phenotype of majority populations were performed in all individuals included in the study. The groups did not show significant differences in absolute numbers or percentages in the blood cell counts (Table 2). In contrast, the groups with worse functional status showed a gradual higher percentage of NK cells (CD16+56+) and reduction of B cells (CD19+) (Fig. 1a). Although the differences in T lymphocytes between the four groups were not significant, the populations of CD4+ and CD8+ T lymphocytes were significantly different between groups (ANOVA test, p = 0.02 and p = 0.027, respectively) (Fig. 1b). Elders in group 3 had the lowest percentage of CD4+ and the highest percentage of CD8+ T cells. The four groups presented similar frequencies of CD4+ regulatory T cells (data not shown).

Loss of CD28 is a typical feature of senescence in T cells (Borthwick et al. 1996). CD28 expression on CD8+ T cells did not differ significantly between groups, whereas the CD4+CD28− T-cell population showed a tendency to increase with worse functional status (data not shown). The double positive CD4+CD8+ and CD4+NKG2D+ are two other subsets of T cells that have been related to aging (Pawelec 1995; Colombatti et al. 1998; Alonso-Arias et al. 2009). The proportion of CD4+CD8+ and CD4+NKG2D+ in the peripheral blood increased in elders with worse functional status (Fig. 1c). Donors in groups 2 and 3 had significantly higher levels than those in group 0. In summary, these results demonstrated that many of the age-related changes in immune parameters were dependent on the functional status of the elderly.

IRP parameters and functional status

Parameters that define the IRP are the inverted ratio CD4/CD8 (ratio < 1.0) and CMV infection (Olsson et al. 2000; Hadrup et al. 2006). Low CD4+ numbers and elevated CD8+ T cells were associated with a worse functional status. Accordingly, the CD4/CD8 ratio were significantly different between groups (Kruskal–Wallis non-parametric test, p = 0.022) (Fig. 2a). The proportion of elders with an inverted ratio also showed a gradual and significant increase with reduced functional status (χ² test, p = 0.041) (Table 3).

The relationship between CMV infection and functional status was also analyzed. Although the largest percentage of CMV-seronegative elders was in group 0 (16%), no significant differences between groups were observed (Table 3). However, we detected a gradual increase in CMV antibody titer concomitant with deteriorating functional status (Fig. 2b).

We have also examined the relationship between the median BI score and CMV serological status. The median BI score was not significantly related to CMV serological status: the median in seronegative individuals was 95 points (IR = 53.75 points) and in CMV seropositive was 75 points (IR = 50 points). However, comparison of the CMV antibody titer with the BI score across all individuals revealed a negative correlation (Spearman Rho test; rho −0.346; p = 0.0005). Only CMV serological status was independently associated with the BI score in a multivariate analysis (multiple linear regression, p < 0.001) with age and gender as confounding factors.

Comparison between the BI score and CMV titer separately in men and women showed that the correlation remained significant among women (Spearman Rho test; rho −0.379; p = 0.001) (Fig. 2c) but not among men (Fig. 2d). The lack of correlation in men may be due to the low number of male volunteers in the study. In fact, there is a clear trend, which likely would be significant with a higher number of volunteers.

T-cell differentiation subsets

One of the most widely accepted models in immunosenescence is that the T-cell compartment is progressively deteriorating with advancing age. T cells can be separated into functionally different populations using combinations of cell surface markers such as the tyrosine phosphatase isoform CD45RA and the chemokine receptor CCR7. With these markers, we subdivided the T cells into naïve (NAÏVE; CD45RA+CCR7+), central memory (CM; CD45RA−CCR7+), effector memory (EM; CD45RA−CCR7−), and effector memory RA (EMRA; CD45RA+CCR7−) (Sallusto et al. 1999). To detect an association between functional status and the differentiation degree of T-cell subsets, we compared the distribution of the distinct T-cell subpopulations in young and elderly individuals, as shown by representative samples (Fig. 3a). The analysis of these subsets of CD4+ T cells revealed that the functional deterioration in elderly was related to the reduced population of undifferentiated subsets (Fig. 3b). In fact, the frequency of both the NAÏVE CD4+ cells and the CM populations in elders from groups 2 and 3 were significantly decreased compared to those from groups 0 and 1. On the contrary, the more differentiated subset EM had a significant and progressive increase from group 0 to group 3. Although the proportions of EMRA cells increased with age, they were not significantly different between the groups (Fig. 3b). CD8+ T cells also did not differ significantly among groups: the level of the four populations was similar in all groups. Most CD8+ T cells were in the EM and EMRA subsets, which were the last stages of differentiation (Fig. 3b).

EM and EMRA are heterogeneous populations, and the staining of two additional markers, CD27 and CD28, has proven useful in identifying less differentiated (CD27+ and/or CD28+) or more differentiated (CD27^nullCD28^null) cells (Koch et al. 2008) (Fig. 3c). Differentiating CD4+ T cells lose expression of CD27 first and subsequently lose CD28 in a later phase (Amyes et al. 2003; van Leeuwen et al. 2004). In contrast, CD8+ T cells lose expression of CD28 first and then CD27 (Gamadia et al. 2003). Despite the existence of a different degree of differentiation defined by the expression of CCR7 and CD45RA markers, we found no significant differences in CD27 and CD28 expression between groups (Fig. 3d).

Taken together, these results indicated that a higher differentiation degree of CD4+ T cells, but not CD8+ T cells, was related to worse functional status. Thus, the maturation stage and number of past episodes of activation and cell cycling of CD4+ T cells differentiated elders with different motor abilities.

TREC quantification and basal proliferation

To further examine the differences found in the differentiated status of CD4+ subsets, we assessed the replicative history of these cells by quantifying the content of TREC in CD4+ T cells belonging to the four elder groups. TREC is a traceable molecular marker produced in newly naïve T cells; the content of TREC in peripheral T cells is an indicator of the number of divisions that the cell has undergone (Douek et al. 1998). Since sample volumes were insufficient to perform CD4+ T-cell isolation in all donors and differences were not found between groups 0 and 1 and between groups 2 and 3, we grouped samples: group 0 + 1 and group 2 + 3. The analysis of the groups displayed significant differences in TREC content. The TREC was lower in CD4+ T cells from elders with worse functional status and more differentiated subsets (Mann–Whitney U test, p = 0.016) (Fig. 4a).

Next, to determine whether the lower TREC number observed in elders with worse functional capabilities correlated to a lower frequency of naïve CD4+ T cells, we compared the TREC content to levels of naïve CD4+ T cells and detected a positive correlation (Spearman Rho test; rho 0.421; p = 0.041) (Fig. 4b). The elders with a lower TREC content had a worse functional status and lower frequency of naïve CD4+ T cells. Thus, cells from the elderly groups 2 and 3 have undergone more cell divisions since they migrated from the thymus than cells from groups 0 and 1.

To further examine the proliferative capacity, we measured the level of the cellular marker for proliferation Ki-67 on CD4+ and CD8+ populations isolated from 10 elderly in groups 0 and 1 and from 10 volunteers from groups 2 and 3 (Fig. 4c). No significant differences in Ki-67 levels on both cell populations between the two groups of elderly were detected, and thus the basal proliferation did not significantly differ between the groups (Fig. 4d).

Functional immunoresponse in vitro

To evaluate whether phenotypic changes that were observed with the deteriorating functional ability were also associated with the reduced immune responsiveness associated with aging in previous reports, we measured the activation capability by comparing CD69 expression in CD4+ and CD8+ T cells after anti-CD3 stimulation. Elders with better functional status (group 0 + 1) showed a higher CD69 expression level in both CD4+ and CD8+ T cells than group 2 + 3 did (Student’s t test, p = 0.025 and p = 0.015, respectively) (Fig. 5a). Group 0 + 1 also exhibited a higher anti-CD3 proliferative response in the CD4+ and CD8+ T-cell subsets than group 2 + 3 (Student’s t test, p = 0.0005 and p = 0.037, respectively) (Fig. 5b).

CMV antibody titer may reflect the number of previous CMV reactivations. Since elders who had the highest levels of CMV antibodies also mounted the worse response to anti-CD3, we postulated that episodes of viral activation could reflect this deficient cellular immune response or, by the contrary, could boost CMV-specific T-cell responses. The cellular response against CMV was measured by stimulating whole-blood cultures with CMV antigens. The magnitude of the CD4+ T-cell immune response to CMV was significantly higher in the elders with the worst functional status, as detected by CD69 expression (Mann–Whitney U test, p = 0.025) (Fig. 5c) and by specific proliferation (Mann–Whitney U test, p = 0.001) (Fig. 5d). Differences in CD8+ T-cell activation were not found (Fig. 5c), and proliferative responses in both groups were very limited (Fig. 5d).

Accordingly to the above results, the frequency of IFN-γ production by anti-CD3 stimulated T cells in group 0 + 1 was 123.8 (IR = 85.6) and in group 2 + 3 was 85.8 (IR = 61.1) (Fig. 6; Mann–Whitney U test, p = 0.013). The median frequency of CMV stimulated group 0 + 1 T cells was 2.9 (IR = 9.2) and that from group 2 + 3 was 11.3 (IR = 17.05) (Mann–Whitney U test, p = 0.005). These results indicated that despite an impaired cellular immune response, elders with worse functional status presented the highest activation, proliferation, and IFN-γ production against CMV antigens.

Response to vaccination and functional capacity

Next, to determine whether the lower immune response observed in vitro in elders with worse functional capabilities correlated to poor ability to respond to immunization in vivo, we measured the specific antibodies produced against influenza virus vaccination in the four groups of elders. The production of specific antibodies to vaccine was significantly lower in the group 3 than in the other groups (Fig. 7a). Comparison of the CMV antibody titer with influenza vaccination response revealed a negative correlation (Spearman Rho test; rho −0.303; p = 0.002) (Fig. 7b).

Therefore, we concluded that those elders with a higher CMV antibody titer had worse functional status and may indicate a causal effect between CMV reactivation and reduced immune responses.