Background
The human aging process is well described to not spare the immune system [
1,
2]. As part of the search for potentially mitigable or even reversible age-related immune impairments, the suboptimal zinc status of old persons came into focus early on [
3,
4].
Despite of the lack of a reliable biomarker to accurately assess the zinc status, it is undisputed that zinc status becomes inadequate with advancing age [
5,
6]. In terms of the overall population, the prevalence of zinc deficiency (ZD) appeared to be highest in the subgroup of hospitalized elderly, consistent with the poor immune defense of this special cohort [
7,
8]. Such an age-related decline in zinc status is mainly due to a mixture of low dietary intake, an age-related detriment in taste, unfavorable subcellular and anatomical changes in the digestive tract, and polypharmacy [
9,
10].
Zinc is an essential cofactor for thymulin, the thymic hormone required for T cell differentiation and proliferation, regulation of mature T cell function, and T cell cytokine response to antigenic/mitogenic stimulation [
11,
12]. In turn, ZD is causative of decreasing the ability of T cells to produce adequate levels of cytokines in response to stimulation [
13,
14]. While some proinflammatory cytokines, such as interleukin 6 (IL-6) are released in greater amounts (inflammaging), the production of other cytokines such as IL-2 decreases sharply in old age as well as in ZD [
15‐
17].
The unbalanced stimulation-induced cytokine production of peripheral lymphocytes subsequently contributes to seniors´ limited vaccination response, increased susceptibility to diseases such as infections and cancer, and to their higher overall mortality [
18‐
20].
Nevertheless, the precise molecular mechanisms that may underlie the impact of either optimized or suboptimal zinc status on late life cytokine regulation remain unthoroughly understood.
In the present report, we first evaluated the existence of a distinctive molecular mechanism of zinc deficiency-triggered IL-2 suppression in human peripheral blood mononuclear cells (PBMC).
In preliminary studies using zinc-deficient human T cell lines and porcine PBMC supplemented with moderate doses of zinc in vitro, our group was the first to demonstrate that the transcription factor CREM
\(\alpha\) is upregulated in a ZD-dependent manner, thereby actively silencing the IL-2 gene and subsequent IL-2 secretion. The molecular mechanism involves binding of CREM
\(\alpha\) to the -180 cAMP responsive element (CRE) site within the IL-2 promoter (-164 to -189 bp) upstream of the transcription start, thereby actively inhibiting the initiation of IL-2 gene transcription in a chromatin-remodeling manner [
21]. Since the transferability of those in vitro results remained questionable, a major aim of this project was to verify the proposed mechanism for aged and ZD humans. Additionally, we also wanted to check whether it is possible to counteract potential CREM
\(\alpha\) overexpression in ZD seniors via in vivo short-term zinc supplementation (ZS) and thus increase the IL-2 production to an appropriate level.
To address these questions, we isolated primary human PBMC from subjects of two distinctive cohorts: a healthy, young control cohort, likely to show proper IL-2 production based on a high zinc status, and a frail, elderly cohort, further subdivided into either zinc-adequate (ZA) or ZD subjects at risk of displaying impaired immune response indicated by low IL-2 production.
Discussion
The frequent presence of suboptimal zinc status associated with low IL-2 production capacity of peripheral immune cells has been shown to predispose the elderly to various age-related diseases and contribute to their increased overall mortality [
33,
34]. Hence, a thorough understanding of the molecular mechanisms underlying the zinc deficiency-induced IL-2 production impairment may help unravel important immune aging principles and develop potential health-promoting strategies.
In the work presented herein, we observed impaired IL-2 production in the elderly caused by zinc deficiency-dependent CREM
\(\alpha\) overexpression. Although CREM
\(\alpha\)-mediated IL-2 gene silencing has been described before in other contexts [
35] to the best of our knowledge, this is the first description of (i) the existence of zinc deficiency-induced CREM
\(\alpha\)-tiggered IL-2 suppression in (elderly) humans and (ii) zinc posing a potential treatment of suppressed IL-2 production in the elderly via rapid reversibility of CREM
\(\alpha\) overexpression through transient in vivo supplementation.
The first challenge of this study was to accurately screen each subject in terms of their baseline zinc status. As summarized in detail by Lowe et al., the list of zinc biomarkers includes more than 30 different candidates that vary widely in their accuracy [
25]. For this reason, we inferred the baseline zinc status of our study participants assessing not only one but three indicators, all of which aimed to measure different zinc pools.
We calculated a ZD prevalence of 68% among the elderly subjects which even slightly exceeded the values described in other studies [
24,
36]. This could be because zinc status is described to decrease in an age-dependent manner, and the average age of our elderly subjects of 85.4 years was higher than that in most comparable studies [
32]. Among the young controls, a single subject was classified as mildly ZD (serum zinc: 69 µg/dL), which is unusual but may reflect the low zinc intake from its vegan diet, as this diet form was recently shown to predispose to micronutrient deficiencies [
37]. We intentionally chose not to exclude this subject from the dataset, as this would not be consistent with our predefined exclusion criteria but emphasizing that exclusion would certainly have resulted in even stronger significances between the young and the older subjects.
Remarkably, concerning the length of hospital stay, we found that the ZD seniors without ZS were hospitalized significantly longer as their ZA counterparts (mean ZA: 13.3 days, mean ZD w/o ZS: 28.3 days). Considering the well described role of zinc in proper immunity it is not surprising that subjects with pre-existing ZD face a longer hospitalization period than others with adequate zinc status. This is supported by recent data demonstrating that old COVID-19 patients with suboptimal zinc status are hospitalized significantly longer than controls with improved zinc status [
38,
39].
We next aimed to evaluate how the current zinc status of each subject affected their basal CREM
\(\alpha\) expression. All zinc parameters correlated in a significant negative manner with CREM
\(\alpha\) protein and RNA expression, thereby confirming the reciprocal trend between overall zinc status and CREM
\(\alpha\) expression, which was first proposed by our laboratory [
21]. Given this ZD-induced CREM
\(\alpha\) overexpression, we subsequently wanted to examine whether the elderly suffered from reduced activation-induced IL-2 production as a result. Consistent with previous findings we observed lower IL-2 production in the elderly, which was exacerbated by pre-existing ZD [
29,
31]. Many groups have aimed to mitigate or even reverse such restriction of IL-2 production. Regarding nutrient-based interventions, there are reports that vitamin E supply of aged mice and humans improves IL-2 production capacity [
40,
41]. However, this effect was at least in parts considered to be a side effect of a reduction in macrophage-produced T cell suppressive prostaglandin (PG) E2 [
42]. The initial description of zinc having a beneficial effect on IL-2 gene expression in human PBMC, dates to in vitro work by Prasad and coworkers [
29]. In contrast, there have also been some rare descriptions of zinc suppressing IL-2 production by T cells [
43]. This supposed contradiction is most likely due to the differences in experimental protocols such as amounts of zinc used in those studies. Indeed, supraphysiological amounts of zinc were shown to promote regulatory T cell (Treg) induction, which in turn can suppress the T cell response due to IL-2 scavenging via soluble and membrane bound CD25 [
44,
45].
Because comparability is an important feature of nutrient supplementation studies, we decided to administer a zinc dose (and formula) that has been used in similar studies and is also in line with the U.S. National Institute of Health's (NIH) recommended daily allowance (RDA) for zinc [
46]. ZS was conducted with 10 mg of zinc aspartate daily, an amount that is considered moderate and, most importantly, not described to suppress proper T cell function [
14,
47]. Since we wanted to gain new insights into the question of how quickly the zinc status improves in vivo we opted for a shorter period of ZS than in common ZS studies, in which zinc was usually given for several weeks or months [
32]. The literature on the topic of short-term zinc supplementation is inhomogeneous. Most studies on this have been conducted in neonates, while studies in seniors are scarce. Even the term itself is inconsistent, leading short-term ZS to vary from a few days to several weeks [
48,
49]. We showed, that an average ZS of 5.7 days seems to be sufficient to compensate for suboptimal zinc status. This was reflected by an 20% increase in serum zinc (73.5 ± 12.3 µg/dL, by definition ZA) and an even greater 81% increase of labile intracellular zinc. The fact that suboptimal zinc status can be corrected in such a short period of time may be of great importance for medical care. However, an earlier study carried out with 19 elderly people using the same zinc formula and dose for a mean period of 48 days, found an increase of only 11% and 61% for serum zinc and intracellular free zinc, respectively [
47]. Therefore, it may be assumed that the “driving force” of aging at least somewhat counteracts the initial improvement in zinc status in the longer term. This kinetics should be given attention in future studies. At this point, the question may arise whether it is more advisable to supplement with a lower dose of zinc for a longer period or with a higher zinc dose for a shorter period. Based on our data, this question cannot be answered. Although one should always consider the individual case, it can be deduced from current recommendations and risk assessments of the NIH, European Food Safety Authority (EFSA), and other health authorities that it seems more beneficial for health to constantly administer zinc in a low dose [
46]. Weißenborn et al. calculated maximum levels for minerals in food supplements, which could be supplied without the risk of adverse effects in addition the regular diet [
50]. According to this report, 6.5 mg of zinc per day in addition to regular diet, containing the recommended daily amount of zinc, should be harmless even for individuals with an adequate zinc status. On the other hand, the NOAEL (no observed adverse effect level) for zinc is 0.43 mg/kg body weight, resulting in 25.8 mg per day for an adult of just 60 kg body weight [
51]. Therefore, long-term, low dose zinc supplementation should be better for health promotion than short-term, high dose supplementation. Furthermore, we can conclude that early screening for suboptimal zinc status on hospital admission, e.g. by means of a clinically established zinc-specific FFQ, appears to be useful, easy to perform, cost effective and potentially even of higher diagnostic value than serum zinc measurements in non-specialized routine laboratories as shown by Trame and coworkers [
26]. Our results suggest that identifying those individuals who might benefit from immediate zinc supply is of clinical importance.
Following in vivo ZS, we observed a significant decrease in CREM
\(\alpha\) expression paralleled by an approximately threefold increase in IL-2 production. Considering the preceding study demonstrating this mechanism in zinc-deficient and in vitro zinc-supplemented T-cell lines and PBMC, we strongly attribute this zinc effect to the abolished CREM
\(\alpha\)-mediated inhibition of IL-2 gene transcription [
21]. The finding of elevated IL-2 levels following zinc supplementation fits preliminary results of a study conducted by Prasad et al. in which 6 initially ZD seniors received a zinc dose equivalent to our study for a period of 6 months. While IL-2 mRNA levels of ZD PBMC were significantly decreased at baseline (
p = 0.0001), they experienced a significant increase (
p < 0.05) after ZS compared with a group of 12 ZA, age-matched controls [
19]. Together, these findings strongly suggest that zinc status correction may contribute to a substantial improvement of the diminished IL-2 production capacity at old age.
Given these results, future studies conducting short-term ZS may aim to include larger participants numbers and recruit different cohorts and might also evaluate other surrogate parameters for improved T cell response. For instance, in order to exclude any effect of hospitalization it would be interesting to investigate whether the same mechanisms also apply to non-hospitalized elderly. And to further elucidate any specific age-related effect it would be also interesting to pursue supplemention of young hospitalized patients with ZD. However, also in non-hospitalized elderly it would be against good clinical practice to withhold supplemention despite the diagnostic knowledge about ZD. Therefore, a large cohort of elderly ought to be investigated with random zinc or placebo supplementation, where the zinc status at beginning will would have to be unblinded only in hindsight at the end of the supplementation. Due to new regulatory publications, suggesting no more than 6.5 mg of zinc as a dietary supplement to regular diet, it may be critical to obtain approval for such a study in Germany [
50]. Although zinc deficiency screening in young, hospitalized patients should be possible in principle, it becomes much more difficult to include a critical number of patients in this age group. This is because (i) the number of young ZD patients is half that of the elderly [
52], (ii) the number of patients between 18 and 60 years of age is only a quarter of those over 65 years of age [
53], and (iii) length of hospitalization increases with age while most young patients stay in hospital for less than 6 days [
54]. Therefore, the number of young patients to be screened must be around 16 times higher than in geriatric medicine.
Future studies may include other known risk groups for ZD and examine them for similar effects. Unfortunately, the small number of subjects on vegan or vegetarian diets included in our study is not sufficient for a powerful subgroup analysis. Furthermore, it should be sought to investigate whether a rapidly improved zinc status is also sufficient to improve T cell proliferation which is often reduced in old age and therefore regarded as another characteristic of immunosenescence.
Despite working in all conscience, our study still has some limitations. As a transcription factor, CREM
\(\mathrm{\alpha }\)'s main area of activity is expected to be in the cell nucleus. Nevertheless, we did not separately assess the nuclear localization of CREM
\(\mathrm{\alpha }\), which is necessary for transcriptional activity. However, we are certain that such observed CREMα overexpression is also reflected at the nucleus level, as has been demonstrated previously [
55].
We are also aware, that it cannot fully be excluded that the study enrollment immediately after hospitalization may have had a confounding effect on some zinc and immune measures in the seniors. However, to avoid this best as possible, we did not include subjects in the study who suffered from common conditions known to affect zinc homeostasis, the immune system, and their measurable parameters. In addition, we were dependent on hospitalized seniors, because, as mentioned earlier, they represent one of the few and easily accessible in vivo collectives for ZD.
Moreover, it unfortunately was not possible to establish a placebo-receiving control group of ZD seniors. As mentioned above, a diagnosed ZD is an obligatory indication for ZS. Leaving such a condition untreated would therefore be negligent and unethical. Although we could not conclusively exclude confounding factors by the selection criteria, we still consider our study design suitable to claim that the measured effects regarding CREMα and IL-2 are primarily due to the observed improvement in zinc status. We believe so for several reasons: On the one hand, our results are fully in line with well-controlled observations in previous in vitro studies [
21]. On the other hand, previous placebo-controlled studies including large numbers of subjects with similar baseline characteristics have already shown that zinc status can be improved by moderate zinc supplementation in a dose-dependent manner [
19,
56]. Thus, this life-like study approach may even be of clinical importance, as it has already been shown that ZS improves immune status in healthy elderly [
47], and, in addition to current knowledge, we hereby provide evidence that the effect appears to be much greater in hospitalized elderly with ZD, since their impairment of the immune response is stronger than in their ZA counterparts. Against this background, and especially considering the large interindividual heterogeneity in the fourth stage of life, we therefore consider this approach reasonable.