Immune cells display an abnormal maturation and a proinflammatory profile in telomere biology disorders Open Access

Telomeres, the natural ends of linear chromosomes composed of repetitive hexanucleotide sequences, maintain the cell’s genomic integrity and prevent damage.¹ Pathogenic germ line variants in genes involved in telomere biology cause excessive telomere erosion and dysfunction, resulting in cell senescence, apoptosis, and chromosomal instability.^2,3 In humans, these cellular abnormalities may translate into medical conditions collectively called telomere biology disorders (TBDs) or telomeropathies, preferentially affecting organs with higher regeneration capacity.^2-4 

Bone marrow failure (BMF), or aplastic anemia, is a common clinical manifestation, resulting in cytopenias (anemia, neutropenia, thrombocytopenia), and is a leading cause of death in patients with TBDs. In these cases, marrow failure is caused by hematopoietic stem cell exhaustion and defective maturation due to critically short telomeres unable to maintain hematopoiesis appropriately.⁵ In addition to bacterial and fungal infections, patients are more vulnerable to viral infections and developing cancer, suggesting defects in immunosurveillance beyond neutropenia.^6-8 A proportion of patients develop idiopathic pulmonary fibrosis (IPF), and others present liver disorders, including cirrhosis.³ How fibrotic changes in organs occur is not fully understood. The primary tissue cells appear affected by telomere shortening, disrupting organ integrity, but a fibrotic-biased inflammatory response to tissue injury appears to dominate, especially in dyskeratosis congenita.

How immune and inflammatory cell function is affected by telomere shortening is not entirely clear. As the hematopoietic stem cells/progenitor cells are affected by excessive telomere attrition, it is plausible that cells from the immune system are similarly affected. Patients harboring telomere-biology pathogenic variants may present immune alterations typically found in older individuals, such as CD4⁺ T-cell deficiency with naïve T-cell decline and T-cell receptor (TCR) repertoire impairment.⁹ T-cell exhaustion is also observed in patients with TBD, which limits immunosurveillance against solid tumors.⁸ 

Here, we aimed to comprehensively identify how telomere shortening in human TBD affects the innate and adaptive immune cells by multiparameter mass cytometry (cytometry by time-of-flight [CyTOF]), and how it may contribute to disease.

Patients previously diagnosed with TBDs and seen at the Ribeirão Preto Blood Center, Ribeirão Preto Medical School, University of São Paulo, Brazil, were invited to join the study if they met the inclusion criteria: (1) short telomere length (TL) and/or presence of pathogenic/likely pathogenic germ line variants in telomere-related genes identified by next-generation sequencing or Sanger sequencing, and (2) clinical manifestations related to TBDs, such as BMF, IPF, liver cirrhosis, of both sexes, and with ages ranging from 12 to 90 years. Patients who received bone marrow transplants were excluded. Twenty patients joined the study. Clinical data were extracted from their existing electronic medical records. Additionally, 10 healthy subjects from the Ribeirão Preto area, of both sexes, and ages 12 to 90 years, with no history of familiar BMF, or pulmonary or hepatic fibrosis, were recruited as controls.

Peripheral blood was collected in tubes containing EDTA or clot activator (silica) to obtain peripheral blood mononuclear cells (PBMCs) and serum, respectively. PBMCs were separated from whole blood by density gradient over Histopaque 1077 (Sigma-Aldrich), and aliquoted for DNA extraction and immunophenotyping, the latter being frozen in fetal bovine serum with 10% dimethyl sulfoxide at −80°C until analysis. Genomic DNA was extracted with a DNeasy Blood and Tissue kit (Qiagen), quantified, diluted to 50 ng/μL, and stored at −20°C. Serum was aliquoted and frozen at −80°C until analysis.

TL was determined by southern blot analysis of terminal restriction fragment (TRF), according to the manufacturer’s instructions, with minor changes (TeloTAGGG TL assay, Roche Applied Science).¹⁰ The mean TRF length was determined according to the formula TRF = Ʃ(ODi)/Ʃ(ODi/Li), where ODi is the chemiluminescent signal, and Li is the length of the fragment at a given position. The mean TRF of a known reference sample was included in every experiment to validate the results. The normal curves for TL plots were generated based on the data of 301 healthy donors (aged 0-90 years) from our own previously published database, using the same southern blot-based methodology.¹⁰ 

For immunophenotyping, PBMCs were thawed in RPMI-1640 media containing 10% fetal bovine serum, 1% penicillin/streptomycin, and 1 μL/mL benzonase, washed and resuspended in complete media for counting. Per sample, 5 × 10⁶ cells were transferred to V-bottom wells in a 96-well deep-well plate. Each sample was stained with a combination of 3 metal-tagged B2M antibodies for barcoding, and up to 20 samples were pooled in a single 5 mL tube. Six different metal isotopes were used to tag B2M for barcoding combinations. Due to the limited barcode combinations, samples were processed in 2 batches, and batch-control samples were used to control for batch variations. Cells were washed and resuspended in Maxpar Cell Staining Buffer. A total of 30 × 10⁶ cells was transferred to a new tube, and stained for surface and intracellular markers. Thirty-nine markers of cell lineages or states were used, and the complete list of antibodies, clones, and channels/metals are described in supplemental Table 1. Cell viability was determined using 103Rh and 193Ir. After fixation with paraformaldehyde, cells were passed through a filter, counted, separated into multiple fluorescence-activated cell sorter tubes to a concentration of 0.5 × 10⁶ cells per mL, and loaded into a CyTOF XT equipment (Standard BioTools, Inc) for data acquisition. Flow cytometry standard files were concatenated into a single file for data processing and analysis.

Single, live cells were gated, and sample debarcoding was performed using FlowJo v.10.9.0 software (BD Biosciences). Next, we gated cells based on CD45 and CD3 expression to generate flow cytometry standard files for CD45⁺CD3^– and CD45⁺CD3⁺ cells for each sample. Each file was then used for automated clustering by self-organizing maps (Flow-SOM),^11,12 using the web-based software Cytobank (Cytobank, Inc). We first generated 20 metaclusters for CD3^– cells and another 20 for CD3⁺ cells and, following annotation of each population based on their markers, we excluded nonspecific clusters and combined clusters with close similarity. Dimensionality reduction and visualization were conducted using viSNE maps in Cytobank.¹³ For viSNE, all samples from each group were concatenated, and 155 000 events from each group were randomly used for the map creation. Finally, the following major populations were manually gated using FlowJo, following previously published gating guidelines:^14,15 CD4⁺ and CD8⁺ T cells, B cells, T helpers (Th) 1, 2, 17, and 17.1, recent thymic emigrants (RTEs), CD4⁺CD95⁺PD-1⁺, CD8⁺CD95⁺PD-1⁺, and C19⁺CD95⁺ cells.

For mucosal-associated invariant T (MAIT) cell analysis, we used the following antibodies: anti-CD45 APC-H7 (no. 641399; BD Pharmingen), anti-CD3 PE (no. 555333; BD Pharmingen), anti-CD161 Brilliant Violet 510 (no. 339922; BioLegend), anti-TCR Vα7.2 Brilliant Violet 421 (no. 351716; BioLegend), anti-CD38 PE-Cy7 (no. 335790; BD Pharmingen), anti-HLA-DR FITC (no. 347400; BD Pharmingen), and 7-AAD (no. 559925; BD Pharmingen). Data were analyzed using FlowJo v.10.9.0 software (BD Biosciences), and MAIT cells were defined as CD45⁺CD3⁺CD161^hiVα7.2⁺.

We designed a panel to assess the serum levels of 32 analytes by Luminex xMAP technology in a MAGPIX instrument (Luminex; supplemental Table 2). The analytes were chosen after extensive literature review, based on their importance on bone marrow function, development of fibrosis, lung and liver diseases, and general inflammation. The panel was manufactured by Thermo Fisher, and the web-based software ProcartaPlex Analysis App (Thermo Fisher) was used for data analysis. Additionally, transforming growth factor-beta 1 (TGF-β1) was determined by conventional enzyme-linked immunosorbent assay (ELISA) using the TGF-β1 Pre-Coated Human ELISA kit (no. BGK01137; PeproTech).

Peripheral blood was collected on Guthrie cards, and a 3.2-mm dried blood spot punch was used for T-cell excision circle (TREC) and kappa-deleting recombinant excision circle (KREC) analysis by quantitative polymerase chain reaction, as described elsewhere.^16,17 

The frequencies of each Flow-SOM-generated metacluster were compared using the Mann-Whitney U test. Manually gated populations were compared by Mann-Whitney or unpaired Student t test (with Welch’s correction when applicable), depending on their normality, assessed by the Shapiro-Wilk test. Cytokines, chemokines, and growth factors presented highly skewed distribution across samples; therefore, logarithmic transformation was used. In this case, geometric means with interquartile ranges (Q1-Q3) are presented as better central tendency and dispersion measures than arithmetic means and standard deviations. Due to the small sample size, outliers, and possible deviations from normality, bootstrap t-tests were used to compare the means of 2 groups (or geometric means). The “boot.t.test” function of the CRAN “MKinfer” package, available in R software, was used to obtain the bootstrapped P values. When 3 groups were compared, a bootstrap test based on the Welch-James statistic and trimmed means was used.¹⁸ Corresponding P values were obtained using the “welchADF.test” function of the CRAN “welchADF” package of the R software.¹⁹ Spearman’s correlation coefficient was used to analyze associations between TL z scores and the variables of interest. TL z scores were calculated based on the TLs from 301 healthy donors present in a database from our group. All P values were corrected by the Benjamini-Hochberg false-discovery rate, and values ≤0.05 were considered significant. The R software version 4.1.1 and GraphPad Prism v.9.5.1 (GraphPad Software, Boston, MA) were used to analyze the data.

The study was approved by the local Ethics Committee, and conducted in accordance with the ethical standards established by the Declaration of Helsinki.

Table 1 describes the characteristics of patients and healthy individuals in the study. Eighteen patients (90%) had a TL below the 10th percentile (short); 8 of them were below the 1st percentile (very short; Figure 1A). Two patients (10%) had a TL above the 10th percentile, despite having pathogenic variants identified in TERT and RTEL1. Both patients, however, presented TBD-related clinical manifestations, including liver disease with portal hypertension and lung disease, and the patient harboring a mutation in RTEL1 also had BMF (supplemental Table 3). All healthy individuals had normal age-adjusted TLs between the 20th and 95th percentiles (Table 1; Figure 1A). Liver and/or lung disease was present in 60% of patients (supplemental Table 3). Past infection history for each patient is presented in supplemental Table 3. The population includes, at a minimum, a representation of the genetic and phenotypic diversity of TBDs observed in our clinic.

Using Flow-SOM algorithm for high-dimensional analysis, 20 clusters were initially created for each CD3 population (positive and negative), followed by merging similar clusters or deleting unspecific subsets, resulting in 14 clusters for CD3⁺ and 15 for CD3^– cells (Figure 1B-C). Expression intensities for all markers in each cluster and channel-colored viSNE maps for the primary lineage markers are shown in supplemental Figures 1-3.

TBD recapitulates phenotypic manifestations acquired with aging; thus, we expected immune alterations typically observed in the older individuals. Patients with TBD had lower CD4⁺ T-cell frequency (33% ± 14% vs 47% ± 16% of CD3⁺ in controls; P = .049), whereas they accumulated CD8⁺ T cells (54% ± 14% vs 41% ± 11% in controls; P = .043; Figure 2A-B). The CD4/CD8 ratio was <1.0 in 80% of patients (Figure 2C). Patients also had fewer B cells (11% ± 8.1% vs 17.5% ± 7.2% of CD3^– in controls; P = .046; Figure 2D). Naïve T (CD45RA⁺CCR7⁺) CD4⁺ and CD8⁺, and B (CD19⁺CD27^–IgD⁺) cells were markedly reduced in TBD, with low frequencies being observed for all 3 populations (Figure 2E-H). A subset of CD8⁺ T effector memory/terminally differentiated effector memory cells (CD27^loCD28^–) was increased in patients (39% ± 17% vs 20.5% ± 12.5% of CD3⁺ in controls; P = .036; Figure 2I). Among B cells, immunoglobulin M or nonclass switched memory B cells (CD19⁺CD27⁺IgD⁺) were significantly reduced (0.57% ± 0.67% of CD3^– vs 2.6% ± 1.8% of CD3^– in controls; P = .0074; Figure 2J-K). We also investigated whether the affected gene could influence the immune profile, but we did not find any difference among mutation carriers in TERT and other genes (supplemental Figure 4).

Given the reduced number of naïve T cells, we sought to explore the thymic output by looking for RTEs, defined as CD31⁺CD45RA⁺ cells. Decreased RTE fractions were observed in CD4⁺ and CD8⁺ subsets in comparison to healthy controls (Figure 3A), suggesting thymus hypofunction. To further scrutinize this hypothesis, we collected new samples from 10 patients and 10 age-matched controls (supplemental Table 4) to quantify TRECs and KRECs (for B cells) by quantitative polymerase chain reaction, with TRECs being significantly lower in patients with TBD (Figure 3B). Four patients (40%) had undetectable copies of TRECs. One pediatric patient (12.3 years old) presented only 1 TREC copy per microliter of blood. Moreover, TRECs were inversely correlated with age in controls (r_s= −0.66; P = .043), but not in patients (r_s = −0.25; P = .48), further suggesting that the premature shortening of telomeres is sufficient to impair thymic function and T-cell development (Figure 3C-D). Although no statistical difference was observed for KRECs, 4 patients (40%) showed <5 copies per μL of blood, with 1 individual (46.2 years old) having undetectable KRECs (Figure 3C). The affected gene did not influence thymic output (supplemental Figure 5).

Our initial data showed that the CD4⁺ T-cell maturation was dysregulated in TBD. We then assessed the 5 main Th cell subsets, Th1, Th2, Th17, Th17.1, and T regulatory (T_regs), to investigate whether and how these cells were affected. The Th subsets were gated on effector cells according to their expression of surface markers: CD25^–CD127⁺CCR6^–CXCR3⁺CCR4^– for Th1, CD25^–CD127⁺CCR6^–CXCR3^–CCR4⁺ for Th2, CD25^–CD127⁺CCR6⁺CXCR3^–CCR4^– for Th17, and CD25^–CD127⁺CCR6⁺CXCR3⁺CCR4^– for Th17.1. T_regs (CD127^–CD25⁺FOXP3⁺) were identified by Flow-SOM clustering. Gating strategies for Th cells and identification of T_regs by Flow-SOM clustering are shown in Figure 4A-B.

Th1, Th17, and Th17.1 subsets were significantly decreased in patients as compared with controls, whereas the Th2/Th1 ratio was significantly higher (1.6 ± 0.84 vs 0.82 ± 0.39; P = .0063). The Th2 and T_reg frequencies did not differ between groups (Figure 4C-D). In aggregate, these findings indicate a poor Th cell diversity in patients with TBD.

Automated clustering identified a range of subpopulations expressing proliferation, activation, and exhaustion markers. A subset of CD4⁺ proliferating effector cells, defined as CD4⁺CD45RA^–CCR7^–CD27^loCD28⁺CD38⁺HLA-DR⁺Ki67⁺CD95⁺PD-1⁺CTLA-4⁺, was almost twice as high in patients than in controls (0.85% ± 0.83% of CD3⁺ vs 0.43% ± 0.51% of CD3⁺; P = .017; Figure 5A-B). Fas (CD95) and programmed cell death-1 marker (PD-1 or CD279), both present in this cell population, are expressed following antigen-mediated T-cell activation, and function as regulators of apoptosis to maintain immune homeostasis.¹⁴ The frequency of CD4⁺CD95⁺PD-1⁺ cells was significantly higher in patients (13% ± 6.6% of CD4⁺ vs 8.3% ± 2.5% of CD4⁺ in controls; P = .020), and inversely correlated with CD4⁺ T cells (r_s = −0.54; P = .0030) and CD4/CD8 ratio (r_s = −0.53; P = .0040; Figure 4C; supplemental Figure 6A-B). In contrast, CD8⁺ T cells did not follow the same pattern (Figure 4D; supplemental Figure 6C).

We also tested whether B cells overexpressed CD95 in TBDs, which was confirmed by an elevated frequency of CD19⁺CD95⁺ cells (27% ± 9.8% of CD19⁺ vs 11% ± 3.5% of CD19⁺ in controls; P < .0001). Similar to T cells, the relative frequency of CD19⁺CD95⁺ cells was negatively correlated with total B cells (r_s = −0.70; P < .0001; Figure 4E; supplemental Figure 6D).

Classical infiltrating monocytes expressing CXCR3 (CD14⁺CD16^–CD31⁺CXCR3⁺CCR4⁺CD38⁺HLA-DR⁺CD39⁺)^20,21 were significantly increased in patients with TBD (2.6% ± 1.7% of CD3^– vs 0.8% ± 0.26% of CD3^– in controls; P = .023; Figure 5F-G).

A cluster of cytotoxic double-negative T cells, defined as CD3⁺CD4^–CD8^–Vδ2^–CD45RA⁺CCR7^–Tbet⁺GZMB⁺, was significantly expanded in patients with TBD (5.7% ± 5.8% of CD3⁺ vs 1.2% ± 0.98% of CD3⁺ in controls; P = .017; Figure 6A-B). This CD3⁺ cell subset has been closely linked with inflammatory conditions, especially autoimmune diseases.^22,23 

Unconventional T cells, which have limited TCR diversity,²⁴ were also differentially abundant in patients with TBD. Our panel could detect 2 types of unconventional T cells, namely Vδ2 γδ T cells (CD3⁺CD4^–CD8^–Vδ2⁺) and MAIT cells (CD3⁺CD161^hiVα7.2⁺), and both were reduced in patients with TBD. First, Vδ2 γδ T cells were identified as a very scarce population on patients’ viSNE plots (Figure 6B). However, the difference between patients with TBD and controls was only marginally significant following false-discovery rate correction (3.9% ± 8.5% of CD3⁺ vs 7.4% ± 8.2% of CD3⁺ in controls; P = .080). It is worth noting, however, that while healthy individuals showed a wide range from 0.71% to 27% of Vδ2 γδ T cells, 10 patients (50%) had <0.5% of these cells in circulation (Figure 6C).

Circulating MAIT cells were markedly decreased in patients (0.12% ± 0.15% of CD3⁺ vs 2.64% ± 2.7% of CD3⁺ in controls; P = .0002; Figure 6D). We collected new samples from healthy volunteers and patients with TBD to better understand the reasons for such decline. We performed additional conventional flow cytometry to analyze the expression of activation and exhaustion markers in the residual cells (Figure 6E-H). The frequency of MAIT cells expressing the activation marker CD38 was higher in patients, as was the expression of HLA-DR (P = .0020 for both). Concomitant expression of both markers was found in 26% ± 9.7% of MAIT cells in patients (vs 10% ± 9.4% of MAIT in controls; P = .0063). The frequency of CD38⁺HLA-DR⁺ MAIT cells was inversely correlated with total MAIT cells (r_s = −0.56; P = .026), suggesting that overactivation of MAIT cells might lead to their decline in peripheral blood (Figure 6I).

Independent experiments involving samples from both patients and healthy controls were conducted using conventional flow cytometry to analyze CD4 and CD8 T-cell populations, including effector/naïve subsets, RTEs, and MAIT cells. These experiments validated our findings obtained through CyTOF, as presented in supplemental Figure 7. A statistically significant correlation was observed between the results from CyTOF and conventional flow cytometry.

The main cell populations whose frequencies were markedly different between healthy subjects and patients with TBD were selected for hierarchical clustering. Using this strategy, patients with TBD and controls were clustered separately with >90% accuracy (supplemental Figure 8).

Levels of serum cytokines, chemokines, and growth factors are positively correlated with TL

One of the main TBD clinical manifestations is pancytopenia, and our data also reveal immune alterations compatible with those seen in inflammatory processes. Therefore, we aimed to assess the serum levels of cytokines, chemokines, and growth factors.

The Luminex panel was customized for 32 analytes. However, fibroblast growth factor 2 and interferon alpha were out of range and could not be measured, resulting in 30 assessed proteins apart from TGF-β1, quantified by conventional ELISA. At first, none of the quantified proteins was found to have different mean concentrations in patients compared with controls (data not shown). Further analyses revealed that several cytokines, chemokines, and growth factors were positively correlated with TL (z score transformed), which shows that the shorter the telomere, the lower the serum biomarkers levels (Figure 7A). Three members of the interleukin 1 (IL-1) superfamily, IL-1α, IL-1β, and IL-1RA, were correlated with the z score, and total leukocytes, neutrophils, monocytes, and platelets were also positively correlated to IL-1α and IL-1RA, which prompted us to infer that the decline of such cytokines is linked to cytopenia in patients. Similar associations were found for IL-4, IL-7, macrophage inflammatory protein 3 alpha, and angiopoietin, as all correlated with TL z score. Other analytes were also correlated to cell counts, and full associations are shown in Figure 7B.

In this study, deep immunophenotyping of peripheral blood leukocytes revealed that patients with TBD have an immune signature that represents a combination of aging, exhaustion, and inflammation. We also observed a significant correlation between cytokine serum levels, TL, and cytopenias, further supporting the hypothesis of an abnormal immune system in these patients.

Many alterations observed in TBD lymphocytes are typical of aging, as illustrated by the decrease in naïve T- and B-cell subsets, low thymic output, and effector memory cell accumulation. Wagner et al⁹ have previously shown that patients with TBD display decreased naïve T cells, and undetectable or low TREC numbers (but no decreased RTEs), increased T-cell apoptosis and exhaustion markers, and their immunophenotype pattern was comparable to individuals 5 decades older. Our data confirm their findings but expand them to decreased RTEs, consistent with low TREC numbers, and other immune cells, including Th cells, unconventional T cells, B cells, and monocytes.

The key to naïve T-cell decline is thymic involution.^25-27 Thymus-suppressive cytokines increase with age, including leukemia inhibitory factor, IL-6, and oncostatin M,^28,29 whereas IL-7, a key cytokine required for thymopoiesis, decreases.^30,31 We found correlation between TL and IL-7 serum levels, suggesting that patients with very short telomeres might undergo a more intense decrease in IL-7 levels and consequently thymic involution.

B-cell production is directly affected by bone marrow aging.^32-34 Aging is associated with an overall B-cell reduction and a shift in B-cell subsets.³⁵ The naïve (CD19⁺CD27^–IgD⁺) and unswitched memory (CD19⁺CD27⁺IgD⁺) B-cell pools are usually unchanged or decreased with age, at the expense of increased dominance of antigen-experienced and IgG⁺ B cells.^36-39 Here, both populations were decreased and the KREC copy numbers were very reduced in some subjects, further indicating an aged B-cell compartment in patients with TBD.

We speculate that, as telomeres shorten to a critical length, maturating T cells in the thymus and B cells in the bone marrow or peripheral blood are no longer able to proliferate and give rise to their subsequent subsets. For B cells, it is known that the dynamics from naïve to effector memory cells lead to the gradual decrease in TL,³⁹ so it is plausible to infer that a premature telomere erosion will disrupt this chain of events. The reduction of important lymphopoietic cytokines, such as IL-7, contributes to their developmental impairment.

We also found aberrant Th subset proportions. It is unclear, however, whether this results from a differentiation stage bias, increased apoptosis, or a mixture of both. Matthe et al⁴⁰ reported a preferential Th1 in vitro differentiation in studies with telomerase-knockout mice, but they did not evaluate the differentiation toward other Th subsets. Also, they studied C57BL6/J mice, known for being naturally skewed toward Th1 phenotype.⁴¹ Th2 lymphocytes are notably enriched in cirrhotic liver and closer in proximity with collagen-producing hepatic stellate cells, also being critically important for the development of fibrosis in lungs due to their capacity of secreting IL-4 and IL-13, both profibrotic cytokines.^42-44 As patients with TBD are predisposed to liver cirrhosis and IPF, the relationship between the imbalanced Th2/Th1 ratio we found here deserves further investigation.

A rare subset of CD4⁺ effector lymphocytes was elevated in patients with TBD. The presence of HLA-DR, Ki67, PD-1, and CD38 in this subset indicates that it is activated and undergoing proliferation. The expression of HLA-DR by CD4⁺ T cells has been attributed to inflammatory processes, including autoimmune and infectious diseases.^45-47 This subtype still expresses CD27 and CD28, suggesting that it is not yet affected by immunosenescence, as their progressive loss is directly correlated with immunological aging;⁴⁸ the presence of Ki67, a proliferation marker, corroborates this point. This finding suggests that, despite the aged immune system, remaining functional effector T cells appear chronically activated in patients with TBD. These findings could be a response to systemic inflammation caused by telomere attrition, senescence and senescence-associated secretory phenotype.³ As telomeres reach critical length, several signaling pathways related to DNA damage are activated, and culminate in cell cycle arrest and senescence, mainly through the induction of cyclin-dependent kinase inhibitors.⁴⁹ These cells secrete several cytokines in a persistent manner, and we hypothesize that, in TBDs, the high number of senescent cells potentially existing in several organs will lead to immune responses and accumulation of effector lymphocytes.

The low frequency of circulating MAIT cells in patients with TBD is also a potential indicator of inflammatory processes. Although aging appears to modulate MAIT cell decrease, it is relatively limited. Chen et al⁵⁰ found an increased MAIT cell activation in peripheral blood, and lower cell percentages in individuals aged >60 years. Other groups found similar results.^51,52 Our study, however, showed that patients with TBD have much lower MAIT cell percentages, with an average of 0.12% of MAIT cells in the CD3⁺ population, and a frequency of <0.05% in 58% of patients. Recently, it has been shown that MAIT cells in older adults display higher expression of granzyme-B and interferon-γ at baseline, suggesting a low-grade basal inflammatory activation.⁵² In aggregate, our findings suggest that the so-called “inflamm-aging” also occurs in TBD, but other hypotheses should be tested to clarify this phenomenon. The migration of these cells from peripheral blood to inflamed tissues has also been proposed to explain the reduction observed in circulating MAIT cells in inflammatory diseases. However, no consensus has been reached. The accumulation of MAIT cells in tissues has been observed for some disorders, such as bacterial pulmonary infection, ulcerative colitis, and multiple sclerosis.^53-56 Despite being naturally enriched in the liver, corresponding to ≤45% of T lymphocytes in healthy conditions,^57,58 MAIT cell expansion in the inflamed liver has rarely been found, with their levels frequently being decreased.^59-63 However, their role in hepatic fibrosis is regarded as necessary, as they accumulate in fibrotic septa, and interact with profibrotic macrophages.^60,64 

Altogether, the immune profile displayed by patients with telomeropathies indicates that these syndromes affect multiple cell populations, causing disturbances that may interfere with the clinical manifestations associated with TBDs. Although certain aberrant populations displayed some degree of variance among patients with TBD, we did not observe any correlation between the results and the mutated gene, age, androgen therapy, or TL (data not shown). Increased risks of developing infections, cancer, and fibrotic lesions are likely attached to the overly activated, exhausted, proinflammatory, and dysregulated immune system, which is chronically challenged by excessive telomere erosion from an early age.

This study was supported by São Paulo Research Foundation grants 2019/25002-2 and 2022/07634-4.

Contribution: R.T.C., G.J.M., and W.R.G. formulated the hypothesis and research design; W.R.G. and R.T.C. wrote the manuscript; W.R.G., G.N., and A.T. designed and optimized the mass cytometry panels; W.R.G., B.A.S., and V.S.C. performed experiments; L.F.B.C. and R.T.C. provided patient records; F.S.D. provided next-generation sequencing data; V.S.C. and W.R.G. generated the figures; S.H., E.Z.M., and M.M.K. performed the statistical analyses; A.C.-N. provided T-cell excision circle and kappa-deleting recombinant excision circle quantifications; and R.T.C. and G.J.M. supervised the study.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Rodrigo T. Calado, Department of Medical Imaging, Hematology, and Clinical Oncology, FMRP-USP, Av. Bandeirantes, 3900 Laboratório de Hematologia, bloco G, subsolo, HCRP, Ribeirão Preto, SP 14049-900, Brazil; email: rtcalado@usp.br.