Background
It is well established that a shift from a predominantly naïve to a more diverse memory T cell phenotype occurs during aging [
1‐
3]. This coincides with an increase in overall inflammation [
4] and a reduced vaccination efficacy in older individuals [
5,
6]. Furthermore, post-translational modifications (PTMs), like acetylation and phosphorylation, not only affect T cell functions, but might also influence the incidence of diseases during aging [
7‐
10]. While phosphorylation is mostly involved in cellular signaling, acetylation has a wide range of substrates resulting in the regulation of gene expression, DNA binding, mRNA and protein stability, interactions and localization [
11‐
14]. Acetylation is regulated by acetyltransferases (i.e., p300), which transfer an acetyl group to lysine residues in proteins, and histone deacetylases (HDAC), which remove it. HDACs include the NAD
+-dependent sirtuins (SIRT 1–7) as well as the Zn-dependent deacetylases (HDAC 1–11) [
14]. Both enzyme classes are key epigenetic regulators of gene and protein expression and are involved in the dynamic changes of histone acetylation.
One role of Zn-dependent HDACs is the control of longevity and the regulation of physiologic processes in immune cells. For example, HDAC3 is crucial for the T cell development after hematopoietic stem cell commitment to the T cell lineage [
15], T cell post-thymic maturation [
16] and acts as an inhibitor for the cytotoxicity of CD8
+ T cells after infection clearance [
17]. HDAC11 regulates the inflammatory response of T cells, which is confirmed by HDAC11 deficient mice exerting a higher inflammatory activity [
18]. Furthermore, an increased expression of several HDACs was found in numerous diseases, which was associated with aging including different cancer types, such as e.g., acute lymphoblastic lymphoma [
19‐
21], prostate, breast and colorectal cancer [
22].
Unlike genetic changes, histone acetylation is reversible and thus represents an attractive therapeutic target. Indeed, HDAC inhibitors (HDACi) are one of the most promising agents in anti-aging research. In addition, different HDACi have been shown to have clinical potential and are currently used for the treatment of many chronic pathological conditions due to their anti-inflammatory and anti-cancerous properties [
23‐
26]. These include trichostatin A (TSA), a pan-inhibitor of Zn-dependent HDACs [
27], which also interacts with SIRT6 [
28], and EX-527, a specific inhibitor of SIRT1 [
29]. EX-527 has anti-inflammatory properties and impairs T cell activation [
30]. Another inhibitor with anti-inflammatory activity is nicotinamide (NAM), which is the physiologic inhibitor of sirtuins [
31,
32]. These HDACi have been used in vitro and in vivo in experimental models and/or in humans to modulate HDAC activity and to determine the role of acetylation in physiologic and pathophysiologic cellular processes. The modification of the acetylation and deacetylation status of signaling molecules and transcription factors by classical HDACs provides the rational for the treatment with HDACi, which have been already approved for clinical use [
33,
34]. However, there exists little information about the acetylation status of proteins in cancer patients during treatment.
Stimulation of CD8
+ T cells in vitro using antibodies directed against the T cell receptor (TCR) and CD28 resulted in an activation of multiple proteins involved in the TCR signaling cascade including the serine-threonine-kinases [
35], such as e.g. protein kinase B (AKT), which plays a major role in proliferation, apoptosis and glucose metabolism of activated T cells [
36]. After TCR activation and CD28 co-stimulation, AKT is phosphorylated either by the mTOR complex 2 (mTORC2) or by other proteins [
36] leading to its activation and entry into the nucleus, where it phosphorylates the fork head box O (FoxO) transcription factors resulting in their inactivation. The FoxO proteins are more active in quiescent naïve T cells promoting the expression of molecules like L-selectin (SELL or CD62L) and C–C chemokine 7 (CCR7) [
35]. AKT is also essential for changes in the metabolism of memory T cells with a transition to glycolysis during activation [
37] and for the early metabolic reprogramming following naïve T cell activation [
38].
Acetylation changes have broad effects on the cellular metabolism [
39,
40] by transcriptional control of certain genes and modification of key proteins, such as glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Acetylation of GAPDH results in its activation and regulation of its localization [
41,
42]. Furthermore, many proteins involved in the tricarboxylic acid cycle [
11] as well as cytoskeletal components, such as tubulin [
43], actin and cortactin [
44,
45], are acetylated. There exists also increasing evidence that HDACi decrease immune suppression and enhance immune cell function including that of T cells [
46].
However, the role of acetylation in age-dependent immune responses has not yet been investigated in detail. Since the interplay between both processes is insufficiently understood, the aim of this study was to determine the effect of acetylation on the proteome and function of CD8+ T cells in the context of aging.
Discussion
Epigenetic processes including DNA methylation and histone modifications have been implicated in the progression of age-related disorders and cancers. Therapeutic targeting of these processes led to novel therapeutic strategies for the prevention and/or treatment of diseases. Indeed, HDACi have been successfully implemented in the treatment of diverse tumor entities [
68]. Despite HDACi have been shown to rewire cancer-associated transcriptional programs and affect adaptive immunity [
69], little information exists on the age-dependent role of HDACi on immune cells. Therefore, this study analyzed the interplay between acetylation and aging. First, CD8
+ T cells were magnetically sorted from PBMC of young and old healthy blood donors and activated in vitro in the presence or absence of a HDACi cocktail [
31,
70‐
73]. As shown in Fig.
1A–C, the relative protein and mRNA expression of HDAC in CD8
+ T cells from the two age groups did not significantly differ. While SIRT1 has been shown to be decreased in aged tissues [
74‐
76] and SIRT3 and 6 have been linked to longevity [
77‐
80], no differences in the expression pattern of these genes were detected between the age groups within the current cohort. HDAC3 [
16,
81] and HDAC4 [
82,
83], which are required for post-thymic T cell maturation, have displayed statistical, but no biological significant differences in their gene expression between the age groups (Fig.
1A). HDAC11, a known negative regulator of immune cell functions [
18], neither exhibited an age-dependent difference at the mRNA level, nor at a protein level (Fig.
1A and B). Furthermore, the acetylation profiles observed with WB were also similar for both age groups, despite one outlier within the old group (Fig.
1E). In order to determine whether there exist differences in the protein expression of activated CD8
+ T cells from young and old donors, 2DE-based proteomics approach was implemented. Despite the fact that this method has specific technical limitations regarding the number of spots identified using a total, unfractionated lysate, more than 80% of the identified proteins had a decreased expression indicating a downregulation of the most abundant proteins in CD8
+ T cells.
HDACi treatment resulted in the modification of several proteins in only one age group (2 for the old and 12 for the young donors—Additional file
1: Table S2). Only in old donors, TRUB1 and EF2 had a significantly increased or decreased expression, respectively. These two proteins are regulators of the mRNA processing [
84,
85], with EF2 known as a target for cancer therapy [
86].
Moreover, the proteins identified were not a random group, but were interconnected and participated in shared functions as demonstrated by the high protein–protein interaction enrichment found upon STRING analysis (< 10
–13). As expected, most GO terms were associated to the immune system (Fig.
3) including terms specifically related to the inflammatory activity of CD8
+ T cells.
A higher secretion of pro-inflammatory factors from CD8
+ T cells of old donors is well-documented [
4,
63‐
66]. In the current study, the secretion of CD8
+ T cells from old donors in the absence of HDACi treatment is higher for all soluble proteins tested, including anti-inflammatory cytokines, such as IL-10 and IL-4. In some studies IL-10 was found to be increased with aging [
87]. Although IL-10 is primarily secreted by CD4
+ T cells, a subpopulation of CD8
+ T cells with regulatory functions has been also shown to secrete high amounts of IL-10 [
88‐
90]. In the current study, CD8
+ T cells of old donors produced significantly more IL-10 than that of young donors suggesting that within the CD8
+ T cell population of the old donors an increased frequency of cells with regulatory properties exists.
The HDACi treatment of CD8
+ T cells induces a reduction of cytokine secretion in both age group, but to a different extent. While HDACi treatment induces a shut down in the secretion of the cytokines analyzed in the young group, the secretion of pro-inflammatory and cytotoxic factors is maintained in the old group. Since the transcript levels of these soluble proteins are mostly down-regulated (Fig.
6F) HDACi treatment blocked the production of these proteins and not only their exocytosis, which is probably a result of the reduced availability of the euchromatin due to an increased cellular acetylation [
91]. In most cases, the decreased secretion was higher in CD8
+ T cells from young donors. Thus, the anti-inflammatory effect of HDACi on CD8
+ T cells is stronger in young individuals.
Additional to the cytokine secretion, several GO terms also referred to the activation and proliferation of the CD8
+ T cells (Fig.
3 and Additional file
1: Table S3). So far, the evidence that the surface expression of activation markers is different between CD8
+ T cells from individuals of different ages is limited [
92,
93]. Therefore, the expression of CD69, CD25 and CD71 was determined on CD8
+ T cells from old and young donors with and without HDACi treatment. A heterogeneous CD69 expression was found in CD8
+ T cells in both age groups regardless of HDACi treatment, which is not in accordance with the literature [
92,
93]. In contrast, both CD25 and CD71 have an age-dependent expression, which was severely decreased after HDACi treatment (Fig.
5A). The reduced activation caused by the HDACi treatment was associated with a decreased expression of p-AKT (Fig.
4A), which is indirectly involved in the TCR-mediated T cell activation by controlling the phosphorylation of Foxo1 [
35]. Non-phosphorylated AKT was not altered by HDACi treatment in both age groups, while p-AKT had a lower expression in CD8
+ T cells from young donors. Furthermore, HSP90 had a decreased expression in the 2D protein expression profiles (Additional file
1: Table S2) and the WB data (Fig.
2C and D) after the HDACi treatment for both age groups. HSP90 plays a critical role in eliciting both CD4
+ and CD8
+ T cell responses [
94] by acting as a “docking site” for many kinases that are part of the TCR signaling pathway [
95]. Furthermore, treatment of T cells with geldanamycin, a specific HSP90 inhibitor, has been shown to reduce the surface expression of glycoproteins, such as CD3, CD4 and CD8 [
96].
The 14-3-3 ζ/δ protein, however, had an increased protein expression after HDACi treatment (Additional file
1: Table S2, Fig.
2C and D). The 14-3-3 protein family is involved in the TCR signaling pathway; they bind to phosphorylated FoxO transcription factors and block their activity, permitting the activation of T cells [
35,
97]. FoxO transcription factors are more active in quiescent naïve T cells [
36], which could explain the higher expression of the 14-3-3 ζ/δ protein in young individuals with a 1.62 difference in the RFs.
The higher expression levels of CD25 and CD71 on CD8
+ T cells of old donors coupled with an increased secretion of IL-2 (Fig.
6) suggest that CD8
+ T cells from older individuals might have an increased survival. However, no differences were found in the proliferation and apoptosis rates of CD8
+ T cells from young and old donors. Furthermore, the proliferation of CD8
+ T cells was similar between both age groups and equally inhibited by HDACi treatment (Fig.
5B). An effect of HDACi treatment on the apoptosis rate of CD8
+ T cells was significant in the first 2 days in culture but disappears after 5 days (Fig.
5C). This is in contrast to other studies demonstrating that HDACi negatively interferes with the proliferation and induces apoptosis in cancer cell lines [
31,
70‐
73]. In the current study, the effect on the apoptosis rate of CD8
+ T cells was limited, while the T cell activation and proliferation was directly inhibited by this treatment.
In the network analysis, there were no GO terms found to be associated with the differentiation of CD8
+ T cells, although the inhibitors used had effects on the cell differentiation in other studies [
70‐
72,
98]. Additionally, transcripts implicated in CD8
+ T cell differentiation [
56,
57] were analyzed by qPCR in both groups (Fig.
4C) demonstrating a similar distribution between both age groups with the exception of CNR2, which was shown to be involved in the secretion of inflammatory cytokines [
58]. Therefore, an effect of HDACi on the differentiation of CD8
+ T cells cannot be postulated.
One highlight of the identified proteins were changes in the cytoskeleton (Additional file
1: Table S2), which is important for the T cell activation [
99‐
102]. Numerous GO terms related to the cytoskeleton were found with most of the proteins identified localized in the cytoplasm, while some were distributed between “secretory granule lumen” and “intracellular non-membrane-bounded organelle” (Fig.
3F). The HDACi treatment reduced the expression of tubulin mRNA and protein (Fig.
4A) with a higher expression of acetyl-tubulin in CD8
+ T cells from old donors (Fig.
4B).
Conclusions
In our study, we have analyzed the effect of acetylation on the proteome of CD8+ T cells from young and old healthy blood donors by comparing the protein expression patterns of untreated and HDACi-treated CD8+ T cells. Using this approach, the intracellular acetylation levels were increased upon HDACi treatment. Furthermore, acetylation affects the proteome of CD8+ T cells in an organized manner, targeting primarily the cytoskeleton and cellular chaperones, which coincided with a significant reduction in secretion and proliferation after triggering the TCR of CD8+ T cells.
Upon comparison of the effect of HDACi on the activation of CD8+ T cells from young and old donors, CD8+ T cells from old donors had a stronger response to TCR-mediated activation as demonstrated by a higher surface expression of CD25 and CD71, a stable p-AKT expression, an increased secretion of IL-2 and an increase in acetyl-tubulin. Nevertheless, the HDACi treatment strongly affected T cell activation of both age groups.
It is noteworthy that HDACi compounds are currently used as anti-inflammatory and anti-cancerous agents in clinical settings. Our findings demonstrate a residual inflammatory activity for the CD8+ T cells from old donors after the HDACi treatment suggesting that the treatment efficacy is age-dependent, which must be considered for their clinical implementation. Since the data presented are exploratory and need to be confirmed in an independent cohort, the interplay of aging and HDACi treatment must be considered in the future and the role of acetylation in the aging immune system has to be investigated further.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.