Chronic granulomatous disease (CGD) is the most common primary immunodeficiency affecting phagocytes. It is due to genetic defects in nicotinamide adenine dinucleotide phosphate (NADPH) oxidase (1, 2), leading to impaired reactive oxygen species (ROS) production by monocytes and neutrophils, defective microorganism clearance, and chronic inflammation (1–4). Reduced T-cell numbers were reported in an American CGD cohort, although the underlying mechanisms remain unclear (5). The defective ROS production by myeloid cells indirectly contributes to the T-cell loss through the promotion of inflammation (1, 2). On the other hand, T-cells have been shown to harbor NADPH (6), and T-cell intrinsic defects in ROS production are associated with disturbances in T-cell effector differentiation and regulatory function (6–11). Notably, CGD patients feature expansion of the interleukin (IL)-17-producing T helper cell subset, which is thought to contribute to their heightened immune activation state (12). T-cell alterations in susceptibility to apoptosis have also been described (13), as well as impairments in autophagy (14), which may also impact on T-cell development and function.

We report here long-lasting severe CD4 lymphopenia in a 36-year-old patient with CGD that allowed us to investigate the pathways involved in T-cell production and peripheral homeostasis in the context of markedly impaired ROS production. Our findings are consistent with an overall scenario of immunological exhaustion and are of foremost clinical relevance in the context of the increased life expectancy of CGD patients, as well as to the ongoing debate regarding aggressive early therapeutic intervention (15–17).

X-linked CGD was only diagnosed at the age of 13 (G897A mutation at the extreme end of exon 8 of the CYBB gene), although the patient had a clinical history typical of this disorder, namely, neonatal Staphylococcus aureus sepsis, Salmonella enteritidis sepsis (at 7 years of age), multiple liver abscesses due to S. aureus requiring partial hepatectomy (at 10 years of age), and recurrent suppurative lymphadenitis (axillary and cervical), as well as subcutaneous abscesses. ROS production was undetectable using a flow cytometry-based oxidative burst assay (<1% positive monocytes or neutrophils; BD Biosciences). Prophylactic antimicrobials were started with an effective control of major infections, except Pseudomonas aeruginosa keratoconjunctivitis (at age 16).

The patient featured a progressive loss of T-cells with sustained CD4 T-cell counts <200 cells/μl (Figure 1A) and an inverse CD4/CD8 ratio (varying between 0.6 and 0.09).

In parallel with the T-cell depletion, the phenotypic analysis revealed marked loss of naïve cells within both CD4 (Figure 1B) and CD8 T-cells (6.2% of total CD8 T-cells, 100 cells/μl at 36 years of age). This was in agreement with an impaired replenishment of the T-cell pool by recently produced cells. Indeed, we found evidence of compromised thymopoiesis via the quantification of by-products of T-cell receptor (TCR) rearrangement that are generated during thymic T-cell development [signal joint (sj) and DβJβ TCR rearrangement circles, T-cell receptor rearrangement excision circle (TREC)] and progressively decline during age-associated thymus involution. Both sjTREC frequency and the sj/βTREC ratio, which are considered to reflect intra-thymic precursor T-cell proliferation and directly correlate with thymic output (18, 19), were markedly low for the patient’s age (Figure 2A). Of note, reduced thymic activity was observed despite the levels of IL-7, an essential homeostatic cytokine, being highly enhanced (Figure 2B), even in comparison with those of untreated HIV-1-infected individuals. Thus, impaired thymopoiesis seemed to significantly contribute to naïve T-cell loss. The risks inherent to the reduced thymic activity in conjunction with the patient’s age contributed to the decision to not undergo hematopoietic stem-cell transplantation (HSCT).

Naïve CD4 T-cell loss was accompanied by upregulation of activation markers and by increased frequency of cycling cells (Figure 3A). The analysis of cytokine production by T-cells revealed an effector differentiation profile with a significant production of pro-inflammatory cytokines, particularly IL-17 (Figure 3B), as has been previously reported (12, 20). Of note, the majority of IL-17-producing T-cells were CD4+ (96%). There was also an increased frequency of CD4 T-cells producing IL-22 (Figure 3B), of which 42% concomitantly produced IL-17. Notably, there was a parallel overrepresentation of regulatory T-cells (Treg) expressing high levels of CD25 and CD39, markers associated with suppressive function (Figures 3A,B). CD8 T-cells featured an activated and terminally differentiated phenotype, as illustrated by their high levels of perforin and IFN-γ production (Figure 3C). In agreement with these findings, both CD4 and CD8 T-cells, irrespective of their degree of differentiation, featured markedly reduced telomere length, further supporting persistent immune stimulation and increased cell cycling in parallel with reduced de novo T-cell replenishment (Figure 3D).

At 34 years of age, colonoscopy was performed due to an episode of prolonged diarrhea, with watery stool four to five times a day without mucus or blood, with no apparent microbial cause, accompanied by hypoalbuminemia without other evidence of exudative enteropathy or significant malabsorption. The gastrointestinal symptoms subsided spontaneously after 4 weeks. The mucosa was macroscopically normal with an overall preserved structure, with areas of mild inflammatory infiltrates with lymphoid aggregates in the gut biopsies. Cell suspensions from sigmoid biopsies were analyzed by flow cytometry, and patient data were compared with those of healthy subjects submitted to routine colon cancer screening colonoscopy. The CGD patient featured a decreased frequency of CD4+ cells within total T-lymphocytes in the sigmoid mucosa (Figure 4A) and a very low CD4/CD8 ratio (0.14), although we were unable to quantify the absolute counts in the histological material to confirm the CD4 loss. Importantly, in contrast to the circulating CD4 subset, the relative proportions of cells producing IL-17, IL-22, or IFN-γ, as well as FoxP3-expressing Treg, within sigmoid CD4 cells were within the range of healthy subjects (Figure 4B). Moreover, there were no alterations in the levels of expression of genes implicated in the regulation of these populations, namely, IL-17, IL-22, IL-23, IDO1, and AHR, in the sigmoid biopsies (Figure 4C). To the best of our knowledge, this is the first report of T-cell imbalances illustrating the occurrence of CD4 depletion at the gut mucosa in CGD. The apparent lack of Th17 expansion at the mucosal level warrants further investigation, given the known role of IL-17 in inflammatory bowel disorders (21, 22) and the importance of gut pathology in CGD patients.

Our finding of sustained severe long-term CD4 T-cell lymphopenia in X-linked CGD calls attention to the occurrence of major T-cell imbalances in the context of burst oxidative defects and raises the possibility of T-cell immunological exhaustion. To our knowledge, this is the first report of CD4 depletion at the gut mucosal level in CGD, in agreement with peripheral blood data. Therefore, increasing effort should be devoted to the understanding of the mechanisms underlying T-cell loss in CGD and their clinical implications.

This immune senescent profile has been considered of immunological risk for infections and shown to be an independent predictor of death in aged subjects (29). Moreover, it resembles the profile of immunosenescence documented in individuals thymectomized early in life during corrective cardiac surgery (26, 30, 31).

Our case suggests that CGD-associated T-cell depletion may worsen with age, which further supports early therapy potentially curative interventions such as HSCT and gene therapy (2, 4, 15–17, 32).

Of note, the persistent immune activation of our patient featured an AIDS-like immunological profile (33–35). CGD is typically associated with an inflammatory and hyper-activated state (3, 32, 36, 37). In CGD mouse models, inflammation has been linked to a defective production of the immunosuppressive molecule l-kynurenine (20), a tryptophan catabolism intermediate generated by the enzyme IDO, leading to increased IL-17 production. However, recent data showed that CGD patients have preserved IDO function in circulating leukocytes (38, 39). Nevertheless, our patient featured a relative expansion of IL-17-producing CD4 T-cells as compared to other CD4 T-cell subsets, in agreement with previous reports (3, 12, 20), which likely contributed to the hyperactivated state. Notably, the Th17 overrepresentation was not found at the gut mucosa level, a finding that warrants further investigation given the importance of gut pathology in CGD and the known role of IL-17 in inflammatory bowel disorders (21, 22).

The observed massive IL-7 levels may also contribute to the exaggerated immune stimulation. The increase in IL-7 levels in lymphopenic settings is thought to be mostly related to lack of consumption due to the reduction in available targets (25, 40). However, in our case, they were much above those found in untreated HIV-infected individuals with comparable T-cell depletion (25). This finding raises the possibility of IL-7 being partially produced as an acute-phase response in the liver, as reported upon toll-like receptors signaling in mouse models (41).

Of note, in spite of the high levels of IL-7, there was marked impairment of thymopoiesis and a consequent defective replenishment of the naïve compartment, as attested by the reduced sj/βTREC ratio and T-cell telomere length.

Interestingly, it has also been suggested that T-cells have intrinsic defects in ROS production in CGD, which may compromise their thymic development as well as their peripheral survival and function (6–11).

Long-lasting severe CD4 lymphopenia may occur in CGD, likely due to both persistent immune activation and impaired T-cell production. This profile of early immune senescence represents a further argument in favor of timely curative therapeutic interventions in CGD, namely, stem-cell transplantation or gene therapy. Additionally, in a broader perspective, our results open new areas of research regarding the interplay of ROS production and T-cell homeostasis and immune senescence.

The study was approved by the ethical board of the Faculty of Medicine of University of Lisbon. All individuals gave written informed consent.

AA, RV, and AS designed the study; AA, SF, RT, RC, and SS performed research; ML, SS, and RV collected clinical data; AS supervised the study; and AA and AS wrote the paper.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.