Characterization of neutralizing versus binding antibody and T cell responses to varicella-zoster virus in the elderly

Abstract

Age-related immune changes increase the risk of herpes zoster (HZ) caused by varicella-zoster virus (VZV) reactivation. Understanding immune responses to VZV is crucial for reducing the burden of HZ in aging populations. Due to the limited availability of data regarding the VZV immune profiles of elderly individuals, particularly in developing countries, more comprehensive immunological investigations are warranted. A total of 213 participants aged ≥ 60 years were included in this study. VZV-neutralizing antibodies (NAb) and glycoprotein-binding antibodies (BAb) were quantified. Furthermore, VZV-specific T cell subsets and their functionality were evaluated using flow cytometry. Elderly individuals demonstrated a high VZV seropositivity rate of 98.6%, exceeding that of the younger adults. Interestingly, VZV-BAb increased, whereas the proportion of NAb decreased with age, with a significantly lower proportion in the elderly aged ≥ 70 years. The elderly showed decreased naïve T cells and accumulated VZV-specific aged T cells; central memory and effector memory CD4⁺ and CD8⁺ T cells, and terminal effector memory CD8⁺ T cells; with elevated expression of senescence and exhaustion markers, indicating functional impairment. Nonetheless, VZV-specific functional T cells; percentages of VZV-specific interferon-γ-secreting CD4⁺ and CD8⁺ T cells; were not diminished. These findings provide insights into aging VZV immune profiles, which will facilitate the development of age-specific HZ vaccination policies.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-98107-8.

Introduction

Varicella-zoster virus (VZV), the causative agent of childhood chickenpox, establishes latency in the body and can later reactivate, leading to herpes zoster (HZ). HZ is more prevalent and tends to be more severe in immunocompromised and elderly individuals^1,2. Although the mechanism of VZV reactivation remains unclear, factors such as genetics, mechanical trauma, psychological stress, aging, and immune status have been associated with the occurrence of HZ³. Among these factors, age and immune status have been extensively studied. Emerging evidence indicates that the cell-mediated immune response (CMIR) is linked to VZV reactivation with age^4,5. However, the role of the humoral immune response (HIR) remains controversial^6,7. In the elderly, the decline in VZV-specific CMIR is attributed to the accumulation of aged T cells expressing senescent and/or exhausted phenotype(s), accompanied by a reduction in VZV-specific interferon-gamma (IFN-γ)-secreting T cells^8,9. This immune profile may account for the reduction in VZV reactivation control and HZ vaccine efficacy in the elderly population. The decline in CMIR may be attributed to a gradual waning of immunity from the primary VZV infection and immunosenescence. Additionally, latent cytomegalovirus (CMV) infection, which is prevalent in aging populations, can accelerate immunosenescence¹⁰. This may influence the course of HZ in the elderly, as studies have shown an association between CMV infection and VZV reactivation in this age group^11,12. Consequently, understanding the immune characteristics of VZV in the elderly is crucial for preventing the onset of HZ and mitigating its severity in the aging population. Thailand, which is recognized as an aged society, has observed an increased incidence of HZ, with rates increasing with age. Specifically, the incidence rate reaches 121.11 per 100,000 individuals aged ≥ 65 years¹³. The Infectious Diseases Association of Thailand has recommended HZ vaccines for adults and the elderly to reduce the incidence and severity of HZ¹⁴. Nevertheless, HZ vaccines have not been incorporated into the Thai Expanded Program on Immunization, and there is a scarcity of data regarding the immune profiles of the Thai elderly against VZV. To address this knowledge gap, we investigated immune profiles against VZV, encompassing both HIR and CMIR. In this study, we measured VZV-neutralizing and glycoprotein-binding antibodies and assessed T cell phenotype and function following stimulation with VZV-specific peptides by flow cytometry. Furthermore, we explored the associations between HIR and CMIR, as well as the relationships between immune parameters and various factors including age, gender, frailty status, history of HZ, and HZ vaccination. These insights could offer valuable information on immune profiles against VZV in the Thai elderly, which is critical for developing age-specific vaccine policies, especially for preventing HZ in developing countries where vaccine coverage is inadequate.

Results

Participant characteristics

A total of 213 elderly individuals aged 60–90 years from 26 high-density populated communities in Bangkok, Thailand, participated in this study (Table 1). The participants maintained a healthy weight range with a median body mass index (BMI) of 24.6, and the majority reported no history of HZ (85.4%) or HZ vaccination (96.7%). Among the participants, 81.7% had at least one underlying disease, and 50.2% were classified as prefrail. Furthermore, frailty was more prevalent in older age groups (Supplementary Table S1.1). The cohort comprised 54.5% females, with a median age of 70 years, which was higher than the 67 years observed in males (Supplementary Table S1.2). Frail status was more frequent in females (31.0%) than in males (12.4%) (Supplementary Table S1.2). Hematological results reflecting immune and inflammation characteristics were consistent across age and frailty status groups (Supplementary Table S2.1 and S2.3). However, certain hematological parameters differed between genders (Supplementary Table S2.2), and all values remained within the reference ranges. For the control groups, anonymous plasma (N = 22) and peripheral blood mononuclear cells (PBMCs; N = 31) from adults aged < 60 years were used.

Table 1.

Demographic characteristics of the studied elderly cohort.

Characteristic	All participants (N = 213)
Age (years)
Median (min–max)	69.0 (60.0–90.0)
Age group, N (%)
60–69 years	111 (52.1)
70–79 years	79 (37.1)
≥ 80 years	23 (10.8)
Gender, N (%)
Male	97 (45.5)
Female	116 (54.5)
BMI (kg/m2), Median (min–max)	24.6 (13.7–39.4)
Frailty status*, N (%)
Robust	58 (27.2)
Prefrail	107 (50.2)
Frail	48 (22.5)
HZ history, N (%)
No	182 (85.4)
Yes	30 (14.1)
Unknown	1 (0.5)
HZ vaccination, N (%)
No	206 (96.7)
Yes	3 (1.4)
Unknown	4 (1.9)
Underlying disease**, N (%)
No known	39 (18.3)
Known	174 (81.7)
1 disease	50 (23.5)
2 diseases	42 (19.7)
≥ 3 diseases	82 (38.5)

*Frailty status was classified using Fried’s criteria¹⁵.

**Underlying diseases included hypertension, dyslipidemia, heart disease, peripheral artery disease, transient ischemic attack, dementia, chronic lung disease, asthma, emphysema, connective tissue disease, diabetes mellitus, liver disease, chronic kidney disease, arthritis, cancers, acquired immunodeficiency syndrome, and others.

BMI body mass index, HZ herpes zoster.

Frequency of positive VZV-specific Immunoglobulin G (IgG) and neutralizing antibodies (NAb)

The seropositivity rate for VZV was investigated using a VZV glycoprotein (gp) IgG enzyme-linked immunosorbent assay (ELISA). The findings demonstrated that 97.4% of participants, comprising elderly individuals (N = 213) and adults (N = 22), exhibited positive results for VZV-specific IgG antibodies. Notably, the seropositivity rate was significantly higher in the elderly group (98.6%) than in the adult group (86.4%; Table 2). However, among the VZV-seropositive samples, only 63.0% demonstrated neutralizing activity. The observed rate of VZV-specific NAb appeared to be higher in the elderly population (66.7%) compared to the adult cohort (27.3%). However, due to the limited sample size of adult controls, this observation should be considered preliminary and may necessitate further validation with a larger control group before extrapolation to broader populations. Furthermore, there was no significant difference in the VZV-specific IgG titers between samples with detectable and undetectable NAb titers (Supplementary Fig. S1). Additionally, 98.7% of participants tested positive for cytomegalovirus (CMV)-specific IgG, with the elderly exhibiting a higher seropositive rate (99.5%) than adults (90.9%) (Supplementary Table S3). All VZV-seropositive participants were CMV-seropositive.

Table 2.

Frequency of positive VZV-specific IgG and neutralizing antibodies in the elderly and adult comparison group.

Serostatus	All participants (N = 235)	Adult (N = 22)	Elderly (N = 213)	p-value
VZV gp IgG ELISA, N (%)				< 0.001
Seropositive	229 (97.4)	19 (86.4)	210 (98.6)
Borderline	2 (0.9)	0 (0.0)	2 (0.9)
Seronegative	4 (1.7)	3 (13.6)	1 (0.5)
VZV live virus MNT, N (%)				< 0.001
Detectable NAb titer	148 (63.0)	6 (27.3)	142 (66.7)
Undetectable NAb titer	87 (37.0)	16 (72.7)	71 (33.3)

Undetectable NAb titer was no detectable inhibition at the lowest dilution (1:20).

Pearson Chi-Square test was used to test the differences between adult and elderly groups.

Bold p-values indicate statistical difference (p < 0.05).

MNT microneutralization test, Nab neutralizing antibody.

Age-dependency of VZV-specific binding (BAb) and neutralizing antibodies (NAb) in the elderly

We measured the magnitude of BAb (VZV gp-specific IgG and glycoprotein H/glycoprotein L (gH/gL)-specific IgG) for seroepidemiological purposes using ELISA and the functional antibodies inhibiting virus infectivity using the live virus microneutralization test (MNT). VZV gp-specific IgG and gH/gL-specific IgG levels exhibited modest increases with age in both genders (Fig. 1A and B). In contrast, the NAb titers did not follow this trend (Fig. 1C). Although absolute NAb levels were comparable across the older age groups, BAb levels were higher in older adults. This led us to calculate the ratios of binding to neutralizing antibodies, an emerging immunological metric used to assess antibody quality in studies of immune responses to viral infection and vaccination^16–18. We found that the ratio of gH/gL-specific IgG to NAb slightly increased with age in both genders (Fig. 1E), suggesting a decline in NAb proportion with advancing age. Further analysis revealed that the age-dependent decrease in NAb proportion showed a significant reduction in the elderly aged ≥ 70 years, with a 2.3-fold decrease in the group aged 70–79 years and a 2.9-fold decrease in the group aged ≥ 80 years compared to younger adults (aged 60–69 years) (Fig. 1F). This evidence indicates that the majority of antibodies in the elderly are non-NAb. Additionally, no correlation was observed between the ratio of gp-specific IgG to NAb and age (Fig. 1D).

Fig. 1.

Age-dependency of VZV-specific BAb and NAb titers in the elderly. Correlation plots between age, gender, and gp IgG ELISA (A), gH/gL IgG ELISA (B), live virus MNT (C), ratio of gp IgG ELISA to live virus MNT (D), and ratio of gH/gL IgG ELISA to live virus MNT (E). Each open circle and solid triangle represent the serological results of individual elderly male and female participants, respectively. Solid and dashed lines with gray areas indicate linear regression lines with 95% confidence intervals for male and female data. Pearson’s correlation coefficient (r) was used to test for correlations. Bold r and p-values indicate a significant correlation (p < 0.05). Ratio of gH/gL IgG ELISA to live virus MNT in the different age groups (F). Each dot represents an individual subject and the horizontal lines indicate the medians. P-values were determined using the Kruskal-Wallis test with Dunn’s multiple comparisons. *p < 0.05 and **p < 0.01. gp glycoproteins, gH/gL, glycoprotein H/glycoprotein L complex, ID₅₀ 50% inhibitory dilution, mIU milli-international unit.

Variations in VZV-specific T cell phenotypes and interferon-gamma (IFN-γ) production across age groups

The analysis was limited to 121 VZV-seropositive elderly PBMCs samples owing to the limited availability of PBMCs. T cell subpopulations differed between the elderly and adult groups but not within the elderly age strata. (Fig. 2, Supplementary Fig. S2 and S3). In the elderly group, there was a notable decrease in the percentage of naïve cells (CCR7⁺CD45RA⁺) and an increase in the percentage of central memory (T_CM; CCR7⁺CD45RA⁻) and effector memory (T_EM; CCR7⁻CD45RA⁻) cells in both CD4⁺ and CD8⁺ T cell subsets. Additionally, there was an elevated percentage of terminal effector memory (T_EMRA; CCR7⁻CD45RA⁺) CD8⁺ T cells. Higher levels of exhausted CD4⁺ and senescent CD8⁺ T cells were also observed in the elderly group. Additionally, the percentages of VZV-specific IFN-γ-secreting CD4⁺ and CD8⁺ T cells were consistent across the elderly group (Fig. 3). However, CD4⁺ T cell responses to VZV and CMV peptide stimulation were higher in elderly individuals than in adults. The significant age-related decline in the proportion of NAb prompted us to investigate whether this decline was correlated with T cell phenotypes and IFN-γ production. We further analyzed and found no such correlation (Supplementary Fig. S4).

Fig. 2.

Percentage of T cell population after stimulation with VZV gE and CMV pp65 peptides. Gating strategy with representative flow cytometry plots identifying subpopulations of CD4⁺ and CD8⁺ T cells (A). VZV-specific CD4⁺ (B) and CD8⁺ (C) T cell subpopulations. CMV-specific CD4⁺ (D) and CD8⁺ (E) T cell subpopulations. The fractions of the total CD4⁺ and CD8⁺ T cell differentiation subpopulations are depicted in the graphs. The different percentages of specific subsets across all elderly groups (≥ 60 years, N = 121) compared to the adult group (< 60 years, N = 31) under each condition are shown in Supplementary Fig. S2 and S3. P-values were determined using the Kruskal–Wallis test with Dunn’s multiple comparisons. Abbreviations: T_CM, central memory; T_EM, effector memory; T_EMRA, terminal effector memory T cells.

Fig. 3.

VZV- and CMV-specific responses in participants with simulated IFN-γ-secreting T cells. Gating strategy with representative flow cytometry plots identifying CD4⁺ and CD8⁺ T cells secreting IFN-γ (A). Percentages of CD4⁺ (left) and CD8⁺ (right) T cells secreting IFN-γ in VZV-stimulated individuals (B) and CMV-stimulated individuals (C). Each dot represents a single donor subtracted from the negative control (< 60 years, N = 31; 60–69 years, N = 65; 70–79 years, N = 40; ≥ 80 years, N = 16), and the horizontal lines indicate the medians. P-values were determined using the Kruskal-Wallis test with Dunn’s multiple comparisons. **p < 0.01, ***p < 0.001, and ****p < 0.0001.

Correlations between HIR and CMIR to VZV

We employed gp IgG ELISA, gH/gL IgG ELISA, and live virus MNT for HIR assessment. Simultaneously, for CMIR, T cell phenotypes and intracellular IFN-γ production following stimulation with VZV glycoprotein E (gE) peptides were evaluated using flow cytometry. The VZV gp-specific IgG level (mIU/ml) showed a moderate positive correlation with the gH/gL-specific IgG level (OD₄₅₀) (Fig. 4A) but did not correlate with the neutralizing antibody (NAb) titer (ID₅₀) (Fig. 4B). In contrast, NAb levels exhibited a modest positive correlation with the gH/gL-specific IgG levels (Fig. 4C). The percentage of VZV-specific IFN-γ-secreting CD4⁺ T cells was negatively correlated with VZV-gp-specific IgG levels (Fig. 4D). However, no other HIR results correlated with the percentages of VZV-specific IFN-γ-secreting T cells (Fig. 4E–H).

Fig. 4.

Correlation between VZV-specific HIR and CMIR in the elderly and adults. Correlation plots between gp IgG ELISA and gH/gL IgG ELISA (A), gp IgG ELISA and live virus MNT (B), gH/gL IgG ELISA and live virus MNT (C), percentage of IFN-γ-secreting cells and gp IgG ELISA (D), percentage of IFN-γ-secreting cells and gH/gL IgG ELISA (E), percentage of IFN-γ-secreting cells and live virus MNT (F), percentage of IFN-γ-secreting cells and ratio of gp IgG ELISA to live virus MNT (G), and percentage of IFN-γ-secreting cells and ratio of gH/gL IgG ELISA to live virus MNT (H). Each dot represents the immunological results of the individual elderly and adult participants. Straight lines with gray areas indicate linear regression lines with 95% confidence intervals. Pearson’s correlation coefficient (r) was used to test for correlations. Bold r and p-values indicate a significant correlation (p < 0.05). gp glycoproteins, gH/gL glycoprotein H/glycoprotein L complex, ID₅₀ 50% inhibitory dilution, mIU milli-international unit.

Factors associated with VZV-specific HIR and CMIR

The associations of factors (gender, frailty status, HZ history, and HZ vaccination) with the HIR results (VZV-gp specific IgG, gH/gL-specific IgG, NAb, and ratio of BAb to NAb) and CMIR results (intracellular IFN-γ production) were analyzed using ANCOVA models. Gender significantly associated with BAb levels for both VZV gp-specific IgG and gH/gL-specific IgG, as well as the percentage of VZV-specific IFN-γ-secreting CD4⁺ and CD8⁺ T cells (Table 3). Males exhibited significantly higher levels of VZV gp-specific IgG and gH/gL-specific IgG than females. In contrast, the percentage of VZV-specific IFN-γ-secreting CD4⁺ and CD8⁺ T cells was lower in males. No other factors showed a significant association between HIR, CMIR, and VZV.

Table 3.

Factors associated with HIR and CMIR against VZV.

Group	Number of subjects	gp IgG ELISA (mIU/mL)	gH/gL IgG ELISA (OD450)	NT ID50	Ratio of gp IgG to NT ID50	Ratio of gH/gL IgG to NT ID50	% IFN-γ+CD4+ T cells	% IFN-γ+CD8+ T cells
Gender; age, frailty status, HZ history, HZ vaccination, and CMV#-adjusted mean ± SE
Male	97/46	946.24 ± 1.08	0.495 ± 1.102	40.09 ± 1.13	23.60 ± 1.15	0.012 ± 1.161	0.27 ± 0.08	0.24 ± 0.08
Female	116/75	601.17 ± 1.07	0.316 ± 1.091	33.65 ± 1.11	17.91 ± 1.13	0.009 ± 1.146	0.54 ± 0.06	0.48 ± 0.06
p-value		< 0.001	0.001	0.283	0.146	0.190	0.007	0.012
Frailty status; age, gender, HZ history, HZ vaccination, and CMV#-adjusted mean ± SE
Robust	58/34	816.58 ± 1.10	0.414 ± 1.135	38.82 ± 1.17	21.04 ± 1.20	0.011 ± 1.219	0.35 ± 0.09	0.35 ± 0.09
Prefrail	107/56	737.90 ± 1.07	0.400 ± 1.096	34.91 ± 1.12	21.09 ± 1.14	0.011 ± 1.151	0.49 ± 0.07	0.44 ± 0.07
Frail	48/31	659.17 ± 1.12	0.335 ± 1.161	36.98 ± 1.20	17.78 ± 1.24	0.009 ± 1.259	0.43 ± 0.10	0.34 ± 0.10
p-value		0.385	0.546	0.849	0.788	0.694	0.454	0.600
HZ history; age, gender, frailty status, HZ vaccination, and CMV#-adjusted mean ± SE
No	182/104	737.90 ± 1.05	0.374 ± 1.072	36.73 ± 1.09	20.09 ± 1.10	0.010 ± 1.114	0.41 ± 0.05	0.36 ± 0.05
Yes	30/16	741.31 ± 1.14	0.489 ± 1.189	36.48 ± 1.23	20.32 ± 1.27	0.013 ± 1.306	0.60 ± 0.13	0.60 ± 0.13
Unknown	1/1	1145.51 ± 2.05	0.306 ± 2.559	8.85 ± 3.16	129.42 ± 3.78	0.035 ± 4.295	0.70 ± 0.51	0.47 ± 0.50
P-value		0.828	0.351	0.470	0.378	0.463	0.352	0.202
HZ vaccination; age, gender, frailty status, HZ history, and CMV#-adjusted mean ± SE
No	206/117	741.31 ± 1.05	0.381 ± 1.067	36.73 ± 1.08	20.18 ± 1.10	0.010 ± 1.107	0.43 ± 0.05	0.39 ± 0.05
Yes	3/1	584.79 ± 1.52	0.336 ± 1.738	18.20 ± 1.97	32.14 ± 2.19	0.018 ± 2.355	1.21 ± 0.51	0.01 ± 0.51
Unknown	4/3	751.62 ± 1.43	1.059 ± 1.596	41.69 ± 1.77	18.03 ± 1.94	0.025 ± 2.065	0.34 ± 0.29	0.32 ± 0.29
p-value		0.854	0.095	0.576	0.829	0.384	0.305	0.736

ANCOVA was used to test the differences.

^#CMV-specific IgG level, and CMV-specific IFN-γ⁺CD4⁺ T cells and CMV-specific IFN-γ⁺CD8⁺ T cells were used as covariates for VZV-specific HIR and CMIR, respectively.

Number of subjects: number of subjects for HIR results/number of subjects for CMIR results.

Bold p-values indicate a significant difference (p < 0.05).

ANCOVA analysis of covariance, gp glycoproteins, gH/gL glycoprotein H/glycoprotein L complex, NT ID₅₀ neutralization titer at 50% inhibitory dilution, SE standard error, HZ herpes zoster.

Correlations between antibodies against VZV and CMV

A correlation was observed between VZV- and CMV-BAb. Specifically, VZV gp-specific IgG levels demonstrated a modest positive correlation with the CMV-specific IgG levels. (Fig. 5A). Nevertheless, no significant correlation was observed between CMV-specific IgG and VZV gH/gL-specific IgG levels (Fig. 5B). Similarly, no significant correlation was observed between CMV-specific IgG levels and VZV-NAb, indicating that NAb responses to VZV operate independently of immune responses directed toward CMV (Fig. 5C).

Fig. 5.

Correlation between VZV- and CMV-specific antibodies in the elderly and adults. Correlation plots of CMV IgG ELISA and VZV gp IgG ELISA (A), VZV gH/gL IgG ELISA (B), and VZV live virus MNT (C). Each dot represents the immunological results of the individual elderly and adult participants. Straight lines with gray areas indicate linear regression lines with 95% data confidence intervals. Pearson’s correlation coefficient (r) was used to test for correlations. Bold r and p-values indicate a significant correlation (p < 0.05). gp glycoproteins, gH/gL glycoprotein H/glycoprotein L complex, ID₅₀ 50% inhibitory dilution, mIU milli-international unit, PEI-U Paul Ehrlich Institute unit.

Discussion

The study on varicella-zoster virus (VZV) immune response in the elderly is critical due to the heightened risk of severe complications associated with VZV reactivation in this age group. As individuals age, their immune system undergoes immunosenescence, reducing the ability to maintain immunity against latent infections. This makes older adults more susceptible to herpes zoster (HZ), a painful condition caused by VZV reactivation, and its complications, including postherpetic neuralgia and neurological sequelae. The necessity for such research is particularly acute in developing nations, where immunization coverage for VZV and other preventable diseases is frequently inadequate. Limited access to healthcare, lack of public health infrastructure, and financial constraints contribute to lower vaccination rates, leaving many elderly individuals unprotected. Understanding the immune mechanisms involved in VZV reactivation and evaluating vaccine efficacy in resource-limited settings can inform the development of cost-effective interventions tailored to these populations. In this study, we examined the immune profile of elderly individuals residing in high-density populated areas in Thailand, employing multiple immunological assays to evaluate humoral immune response (HIR) and cell-mediated immune response (CMIR) to VZV infection.

In this cohort, 98.6% of the elderly population exhibited seropositivity for VZV, and all VZV-seropositive individuals demonstrated seropositivity for CMV. The VZV seropositivity rate was higher in elders than in adults, with binding antibody (BAb; VZV gp-specific IgG and gH/gL-specific IgG) levels increasing with age. This concurs with other studies showing that the percentage of VZV seropositivity and antibody levels increase with age^19,20. A recent study in Thailand reported a VZV seropositivity rate of 95.9% in individuals aged 50–59 years and reached 100% in those aged ≥ 70 years, with higher antibody levels in older adults²¹. We found that males had higher VZV-BAb levels than females, agreeing with a study on healthcare workers in Taiwan²⁰. However, the reason for increased BAb levels in the elderly remains unclear and requires further investigation.

Although BAb is useful for VZV seroepidemiology, it may not reflect the antibody functionality. This study measured neutralizing antibody (NAb) levels and found that only a subset of VZV-seropositive individuals exhibited neutralizing activity. While BAb levels increased with age, absolute NAb levels remained stable, suggesting that the majority of antibodies in the elderly were BAb. This indicated an increase in the proportion of non-NAb in older individuals, coinciding with the increased ratio of BAb to NAb, particularly for VZV gH/gL-specific IgG to NAb. This study revealed an age-dependent decline in the proportion of NAb against VZV, indicating that the elderly had a response less focused on VZV-neutralizing domains, which may have led to diminished protective immunity and contributed to a higher incidence of VZV reactivation in the elderly. Further research is required to investigate the role of non-NAb in controlling VZV reactivation. This could provide deeper insights into the mechanisms through which the immune system regulates viral latency and reactivation. Furthermore, the ratio of binding to neutralizing antibodies may potentially be utilized to indicate whether an older individual requires vaccination.

Several studies have shown a decline in VZV-specific CMIR with age^4,8,22,23; however, our findings indicated no significant differences in VZV-specific T cell phenotype and function across elderly age groups, although differences were noted between elderly and adult individuals. We observed a marginal increase in the percentage of VZV-specific IFN-γ-secreting CD4⁺ T cells in the elderly compared with in adults. This inconsistency might be due to variations in the study populations or antigens used for stimulation. Similarly, a study conducted in Japan revealed no age-related decrease in VZV-specific CMIR as measured by IFN-γ ELISPOT assay; however, it observed a decline when assessed by skin test²⁴.

We noted a reduction in the naïve subset and an increase in T_CM and T_EM cells among CD4⁺ and CD8⁺ T cells, along with an increase in T_EMRA CD8⁺ T cells, in the elderly. These cells exhibited higher dysfunction levels, as evidenced by increased expression of an exhaustion marker (PD-1) in CD4⁺ T cells and a senescent marker (CD57) in CD8⁺ T cells, supporting previous studies^8,9. This highlights the onset of immunosenescence, suggesting that older adults may be more vulnerable to severe HZ and may have a weaker response to HZ vaccination. Notably, males had a lower percentage of VZV-specific IFN-γ-secreting CD4⁺ and CD8⁺ T cells than females. This observation suggests that elderly males may exhibit a higher susceptibility to HZ reactivation compared to females in our cohort; however, this finding contradicts previous reports of higher HZ incidence in females²⁵. This discrepancy may be attributed to factors beyond immunological considerations²⁶.

While CCR7 is a valuable and widely used marker for studying T cell phenotypes^27–30, our findings suggest potential limitations in its application for T cell subset analysis. Specifically, our observations indicated that CCR7 expression exhibited sensitivity to in vitro stimulation, demonstrating a decreasing trend in the CD4⁺ T cell subsets and an increasing trend in the CD8⁺ T cell subsets (Supplementary Fig. S5). This phenomenon may potentially result in an underestimation of CCR7⁺ cells within the CD4⁺ T cell subsets and a corresponding overestimation within the CD8⁺ T cell subsets. To mitigate potential bias, it is advisable to incorporate additional markers, such as CCR7 with CD27 following stimulation, or to consider utilizing non-stimulation-based assays, such as the MHC-tetramer assay, for a more comprehensive assessment in subsequent studies.

At present, Thailand offers two types of HZ vaccines; the live-attenuated zoster vaccine (ZVL), recommended for individuals aged ≥ 60 years, and the recombinant zoster vaccine (RZV) for those aged ≥ 50 years¹⁴. Both vaccines are intended to reduce the incidence and severity of HZ in the Thai elderly population. However, these vaccines remain optional and are associated with significant costs. Our analysis revealed a significant reduction in the proportion of NAb in individuals aged ≥ 70 years. In conjunction with the existing evidence, these findings demonstrated consistent levels of functional T cells across the elderly population. This observation may suggest considering the inclusion of the HZ vaccination in the Thai Expanded Program on Immunization (EPI) for individuals aged ≥ 70 years, which could potentially enhance cost-effectiveness and more precisely target the population at greatest risk in Thailand.

Our results showed a correlation between VZV gp-specific IgG and gH/gL-specific IgG results, with only gH/gL IgG levels correlated with the NAb titer. VZV contains several glycoproteins (gp), with gH/gL being the third most abundant and highly immunodominant³¹ which could explain the correlation between VZV gH/gL-specific IgG and gp-specific IgG levels. The correlation between VZV gH/gL-specific IgG and NAb is consistent with studies suggesting that antibodies targeting the gH/gL domain can inhibit VZV infection^32–34. Antibodies targeting the VZV gH/gL domain exhibited the strongest neutralizing capacity³², explaining the correlation between the gH/gL-specific IgG level and NAb titer. In contrast, antibodies targeting other gp domains demonstrate minimal or no neutralizing capacity³², which may account for the lack of correlation between gp-specific IgG levels and NAb titers. Therefore, VZV gH/gL-specific IgG may be a good adjunct marker for VZV-NAb. Additionally, we observed an inverse correlation between the percentage of VZV-specific IFN-γ-secreting CD4⁺ T cells and VZV gp-specific IgG levels. This is consistent with previous findings showing a declining trend in VZV-specific IFN-γ ELISPOT response as VZV gp-specific IgG increases in the elderly²⁴.

A recent report identified an association between VZV-specific antibodies and chickenpox history in a Thai population²¹. However, our study did not observe a similar association with the history of HZ or HZ vaccination. This discrepancy may be attributed to recall bias, as participants self-reported their history of HZ and HZ vaccination. Furthermore, the small number of individuals who experienced HZ and received the HZ vaccine may have limited our ability to detect a significant association. Our investigation demonstrated that VZV-specific BAb, NAb, and IFN-γ-secreting T cells were not associated with frailty status in Thai elderly individuals. These findings are consistent with those of previous studies, which similarly did not observe an association between these immune parameters and frailty^35,36. It is expected that VZV-specific CMIR and CMV-specific CMIR will produce similar results, as both viruses belong to the same Herpesviridae family and may induce similar T cell changes through similar mechanisms. We observed a correlation between VZV gp-specific IgG and CMV-specific IgG levels in this cohort. This finding supports previous studies indicating that CMV infection is associated with VZV reactivation^11,12. These studies observed a co-elevated level of VZV-specific antibodies and CMV-specific antibodies in patients with HZ. CMV infection contributes to immunosenescence, resulting in less T cell responsiveness to VZV compared to CMV-seronegative individuals³⁷. This impairs VZV control and leads to VZV reactivation, which prompts the production of antibodies and leads to increased levels of antibodies against both VZV and CMV.

Our study had some limitations. First, we focused solely on NAb as a functional antibody without exploring the roles of non-NAb, such as those involved in antibody-dependent cellular cytotoxicity (ADCC). Investigating these functions would provide a more comprehensive understanding of the antibody-mediated immune response to VZV³⁸. Second, our analysis relied exclusively on VZV-specific T cells secreting IFN-γ, whereas assessing polyfunctional T cell responses involving a broader range of Th1 cytokines could offer deeper insights into CMIR to VZV^9,39. Third, the study included a limited number of VZV- and CMV-seronegative samples, and unequal sample sizes between comparison groups may have reduced the statistical power of our analysis. Future studies with equal sample sizes and greater representation of seronegative individuals are needed to validate these findings. Lastly, further exploration of the immune response to VZV vaccination among the Thai elderly could provide valuable insights.

In conclusion, our study revealed that VZV-BAb levels increased with age, while the proportion of NAb declined, which was significantly lower in elderly aged ≥ 70 years. In the elderly population, these antibodies are predominantly BAb rather than NAb, reflecting non-neutralizing domain-focused antibody responses. T cell phenotype and function against VZV were consistent across the elderly age groups; however, there was a notable decline in naïve T cells and an accumulation of aged T cells expressing markers of exhaustion and senescence, indicative of immune dysfunction. This immune profile suggests a progressive decline in immunity to VZV in older adults, potentially contributing to an increased susceptibility to VZV reactivation and more severe cases of HZ. These insights could guide the development of age-specific HZ vaccination campaigns.

Methods

Participants

Individuals aged ≥ 60 years residing in Bangkok-Noi, Bangkok, Thailand, were enrolled in the study conducted between December 2022 and November 2023. Informed consent was obtained from all participants before conducting the interviews, frailty assessments using Fried’s criteria¹⁵, and blood sample collection. This study was conducted in accordance with the Declaration of Helsinki. The study protocol was reviewed and approved by the Siriraj Institutional Review Board, Faculty of Medicine Siriraj Hospital, Mahidol University, Thailand (approval no. Si 213/2021).

Plasma and peripheral blood mononuclear cells (PBMCs) preparation

Plasma samples were separated from heparinized blood samples and stored at -80 °C for serology assays. PBMCs were isolated from heparinized blood samples using the Lymphoprep™ medium and SepMate™ tubes (STEMCELL Technologies). PBMCs were collected after repeated washing with phosphate-buffered saline (PBS), cryopreserved in fetal bovine serum (FBS) containing 10% dimethyl sulfoxide (DMSO), and stored in liquid nitrogen for CMIR assessment.

Virus propagation and cell-free virus Preparation

VZV strain Ellen (ATCC VR-1367) was propagated in MRC-5 cells (ATCC CCL-171)⁴⁰. VZV-infected cells were mixed with uninfected cells and incubated at 37 °C in a 5% CO₂ incubator. The infected MRC-5 monolayer was monitored, and once the cytopathic effect reached 75–100%, the monolayer was harvested for cell-free VZV preparation⁴¹. The infected monolayer was washed and scraped. The scraped cells were subjected to three freeze-thaw cycles and sonication. The supernatant containing cell-free VZV was collected and stored at -80 °C for the live virus microneutralization test (MNT).

Serology assays

VZV glycoprotein (gp) Immunoglobulin G (IgG) enzyme-linked immunosorbent assay (ELISA)

VZV gp-specific IgG titer was quantified using a SERION ELISA classic VZV IgG kit (Virion\Serion). The seropositive cutoff (mIU/ml) was > 100, the borderline range was 50–100, and the seronegative cutoff was < 50.

VZV glycoprotein H/glycoprotein L (gH/gL) IgG ELISA

VZV gH/gL-specific IgG levels (OD₄₅₀ units) were determined using an in-house method modified from Schmid et al.⁴². MaxiSorp™ ELISA plates (Nunc) were coated with 100 ng/well of gH/gL antigen (SinoBiological) at 4°C for 24 h and then blocked with 5% skim milk in Tween 20-PBS. Plasma samples (1:100 dilution) were incubated for 30 min, followed by incubation with a goat anti-human IgG-peroxidase conjugate (KPL SeraCare) for 30 min. TMB substrate (3,3’,5,5’ tetramethylbenzidine) solution (Cell Signaling Technology) was incubated for 10 min, and the reaction was stopped with 1 N sulfuric acid. Optical densities (OD) were measured at 450 nm (OD₄₅₀) and 630 nm (OD₆₃₀). The mean OD was calculated from duplicates, and the mean test OD₄₅₀ was subtracted from the mean control OD₄₅₀.

VZV live virus microneutralization test (MNT)

The MNT was performed as previously described⁴². Briefly, plasma samples were heat-inactivated at 56 °C for 30 min and 2-fold serially diluted from 1:20 to 1:640. Diluted plasma samples were incubated with an equal volume of diluted cell-free VZV (1 unit of OD₄₅₀ at 3 days post-infection in MRC-5 cells) for 30 min, followed by guinea pig complement (Sigma-Aldrich) at 1.5 U/well for 30 min. Sixty microliters of plasma virus-complement mixtures were incubated with MRC-5 monolayers for 1 h. The mixtures were replaced with 10% FBS-supplemented EMEM and incubated at 37 °C in a 5% CO₂ incubator for 3 days. The cells were then fixed with 80% methanol. Mouse anti-VZV gI mAb (Sigma-Aldrich) was added for 30 min, followed by incubation with goat anti-mouse IgG (KPL SeraCare) for another 30 min. TMB solution (Cell Signaling Technology) was added for 10 min, followed by the addition of 1 N sulfuric acid. OD₄₅₀ and OD₆₃₀ were measured. The mean OD was calculated from duplicates, and the mean test OD₄₅₀ was subtracted from the mean control OD₄₅₀ (no virus control wells). Percent inhibition was calculated as the percent reduction in OD₄₅₀ compared to the no plasma control wells. A nonlinear regression analysis between log-transformed reciprocals of the plasma dilutions and percent inhibitions was performed to estimate the 50% inhibitory dilution (ID₅₀)¹⁶.

Cytomegalovirus (CMV) IgG ELISA

CMV-specific IgG titers were measured using a SERION ELISA classic CMV IgG kit (Virion\Serion). The cutoff (PEI-U/ml) for seropositivity was > 40, the borderline range was 25–40, and < 40 was considered seronegative.

T cell phenotyping and intracellular cytokine staining using flow cytometry

Frozen PBMCs were thawed and rested overnight in RPMI 1640 medium supplemented with 10% FBS. PBMCs (3 × 10⁵ cells) were stimulated for 24 h with 0.25 µg/ml PepMix™ VZV glycoprotein E (gE), CMV phosphoprotein 65 (pp65) (JPT Innovative Peptide Solutions), 50 ng/ml phorbol myristate acetate (PMA) with 250 ng/ml ionomycin (positive control), or RPMI (negative control). GolgiPlug™ and GolgiStop™ (2.5:10 ratio) (BD Biosciences) were added under all conditions. The cells were stained for 30 min with the following antibodies; anti-CD45RA-PE-Cy7, anti-CCR7-APC, anti-CD3-BV605, anti-CD4-BV510, anti-CD57-BV421 (BD Biosciences), anti-CD8-FITC, and anti-PD-1-PE (BioLegend). After permeabilization with Cytofix/Cytoperm™ (BD Biosciences), the cells were stained with anti-IFN-γ-BV711 (BD Biosciences) for 30 min, resuspended in 2% paraformaldehyde, and analyzed using a FACSymphony™ A1 cell analyzer (BD Biosciences). Analyses were conducted using the FlowJo™ v10.9.0 software (BD Life Sciences), and the gating strategy is shown in Figs. 2 and 3.

Statistical analyses

Data distribution was assessed using the Shapiro-Wilk test. The differences between categorical and numerical variables were tested using Pearson’s chi-squared test and the Kruskal-Wallis test with Dunn’s multiple comparisons, respectively. The effects of gender, frailty status, HZ history, and HZ vaccination on HIR and CMIR results were evaluated using analysis of covariance (ANCOVA). Correlations were determined using Pearson’s correlation coefficient (r). Statistical significance was set at p-value < 0.05. Analyses were conducted utilizing Prism 8 (GraphPad) and PASW Statistics 18 software.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

K. B. designed and supervised the study, validated the data, and revised the manuscript. S. M. collected the samples and data, performed ELISA, neutralization tests, data analysis and interpretation, and drafted the manuscript. S. K. performed ELISA, flow cytometry, and data analysis. S. P. performed flow cytometry, data analysis and interpretation, and drafted the manuscript. S. S., O. L., and N. K. collected the samples and data, and prepared the samples. R. A. C. prepared the samples and drafted the manuscript. K. S. and C. C. provided resources. T. R., J. S., and P. W. collected the samples and data, and provided resources. P. A., W. M., and S. I. supervised the study participants’ selection and provided resources. All authors have reviewed and approved the final version of the manuscript submitted for publication.

Data availability

The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.