Introduction
Common variable immunodeficiency (CVID) is characterized by recurrent infections caused by low IgG, and IgA or IgM [
1], for which patients are treated with immunoglobulin G replacement therapy (IgRT) [
2]. IgRT has significantly decreased the risk of infectious complications in CVID, but nonetheless, over a third of patients develop immune dysregulation complications [
3,
4], resulting in significant morbidity and mortality [
5,
6]. A wide range of immune dysregulation phenomena can be observed in CVID, including granulomatous-lymphocytic interstitial lung disease (GLILD), enteritis, autoimmune cytopenias, lymphoproliferation, and hematological malignancies [
4,
7]. The underlying pathophysiology of immune dysregulation in CVID is currently poorly understood, which complicates diagnostics and treatment [
8,
9].
While the defining hypogammaglobinemia in CVID is considered to be primarily the result of B-cell dysfunction, several lines of evidence suggest an additional role for T-cells in CVID with immune dysregulation (CVIDid). Biopsies of lung granulomas in CVIDid show predominance of CD4 + T helper (Th) cells [
10,
11], while regulatory T-cells (Tregs) are often absent [
10]. In peripheral blood of patients with CVIDid, a decreased CD4/CD8 ratio was observed with decrease of naïve T-cells [
12], Tregs, Th17 cells, and follicular helper T (Tfh) cells [
13,
14]. Moreover, there are indications that CVID T-cells may be functionally exhausted, including reduced capacity to respond to bacterial antigens and increased expression of PD1 [
15,
16]. Our group and others have previously demonstrated that serum cytokines in CVIDid [
17,
18] are shifted towards a Th1 phenotype, and we observed an upregulation of proteins associated with immune regulation—IL10, LAG3, and 4-1BB. Monogenic primary immunodeficiencies caused by mutations in immune regulation genes such as CTLA4 [
19] and ICOS [
20] often result in a CVIDid phenotype. However, how the interplay between immune regulation and immune activation results in CVIDid remains poorly understood.
To further study the balance between immune activation and immune regulation in CVIDid, we used flow cytometry to evaluate naïve T-cell subsets, T-cell activation and cytokine production, exhaustion, negative regulators of immune activation, regulatory T-cells, and T-cell homing markers.
Discussion
In this study, we showed that T-cells in patients with CVIDid were more often activated, proliferating, and Th1-skewed than those of patients with CVIDio. In addition, more CVIDid T-cells expressed immune co-inhibitory receptors PD1, LAG3, CTLA4, ICOS, and TIGIT, and these cells retained their inflammatory properties. Chronic activation was observed in both the CD4 + and CD8 + compartment. In the Treg compartment, we observed low CTLA4 expression in CVIDid, while ICOS expression remained intact. Finally, CVIDid T-cells showed increased migratory capacities to mucosal tissues.
High T-cell activation and Th-1 skewing in CVIDid were consistent with previous studies [
12,
18]. In addition, we did not observe increased frequencies of IL17-producing T-cells, despite our previous finding of increased IL-17a in serum of the same patients sampled at the same time. This supports the hypothesis for an alternative source of IL17a production in CVIDid, such as type-3 ILCs [
30].
Previous studies have also reported increased PD1 expression in CVID, and interpreted this as a sign of functional exhaustion and impaired T-cell function [
15,
18,
31]. However, we observed that the PD1- and LAG3-expressing cells in CVIDid retained the capacity to produce pro-inflammatory cytokines and to proliferate, and thus were not functionally exhausted. Upregulation of negative regulators of co-stimulation has been suggested to be a mechanism to limit inflammation-related damage to tissues in settings of chronic inflammation, while maintaining the ability to respond to pathogens [
32]. In CVIDid, however, it is possible that this compensatory response is insufficient in severe states of immune dysregulation and that these chronically activated cells still contribute to immune dysregulation-related pathology.
In addition to this chronically activated T-cell state, we observed a decreased ability of Tregs to upregulate CTLA4, while CTLA4 expression was increased in the whole CD4 + population. Expression of CTLA4 by Tregs is an important mechanism by which Tregs mediate their suppressive function [
33]. Clinical CTLA4 haploinsufficiency often results in a CVIDid phenotype with hypogammaglobulinemia and autoimmune disease, and functional Treg dysfunction has been described [
19]. In this study, the T-cell profile of the patient with CTLA4 haploinsufficiency was often not very different from the other non-genetic CVIDid patients. Therefore, the expression of CTLA4 in Tregs of non-genetic CVIDid patients may be relevant to the overall underlying pathophysiology of CVIDid and warrants further research. A recent study shows that abatacept, a CTLA4 fusion protein, was safe and effective in the treatment of CVIDid with interstitial lung disease [
34]. As the population of CVIDid patients with low CTLA4 + CD27 + Tregs represented a mix of patients with pulmonary inflammation but also other organ-specific autoimmunity (data not shown), abatacept may be efficacious in other CVIDid patients as well. In addition, longitudinal monitoring of CTLA4 Treg expression in CVIDid may indicate whether it can be used as a biomarker for disease exacerbation and/or therapeutic response.
Despite these overall differences between CVIDid and CVIDio, the heterogeneity within the CVIDid group was substantial. Subgroup analyses of patients with organ-specific autoimmunity did not yield insightful patterns, except that GLILD patients were often more extreme in all observed differences (data not shown). In addition, disease severity did not always reflect immune activation. For example, the patient with CTLA4 haploinsufficiency did not show an inflammatory state in peripheral blood, while the patient that had the strongest IFNγ signature combined with low immune regulation markers (Supplementary Fig.
11) was clinically stable and did not require immunosuppressive therapy. It is possible that the peripheral blood immune phenotype only gives limited information that is clinically relevant, as T-cells may have migrated to the inflamed tissues in the sickest patients. In addition, this study may be limited by sample size to detect differences between subgroups of CVIDid. One other aspect that this study does not address is differences in absolute T-cell numbers. A recent study showed that while absolute T-cells are often lower in CVID, they do not differ between CVIDio and CVIDid [
13].
To conclude, these results indicate that CVIDid Th cells are highly activated, and that, unlike in classically exhausted cells such as originally described in chronic viral infection and cancer [
32], PD1 and LAG3 expression in CVIDid CD4 + T-cells reflects chronic activation with preserved inflammatory potential rather than functional exhaustion, similar to previous findings in human auto-immune inflammation [
24]. Combined studies of T-cell dysfunction and circulating inflammatory proteins in peripheral blood may help predict response to T-cell-targeted therapies in individual CVIDid patients. Moreover, loss of CTLA4 upregulation in activated CVIDid Tregs may be mechanistically important in maintaining the inflammatory loop in CVIDid, and warrants further research.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.