Exaggerated T Follicular Helper Cell Responses in LRBA Deficiency Due to Failure of CTLA4-Mediated Regulation

Abstract

Purpose

LRBA (lipopolysaccharide-responsive beige like anchor protein) and CTLA4 (cytotoxic T lymphocyte antigen 4) deficiencies give rise to overlapping phenotypes of immune dysregulation and autoimmunity, with dramatically increased frequencies of circulating T Follicular helper (cT_FH) cells. We sought to determine the mechanisms of cT_FH cell dysregulation in LRBA deficiency and the utility of monitoring cT_FH cells as a correlate of clinical response to CTLA4-Ig therapy.

Methods

cT_FH cells and other lymphocyte subpopulations were characterized. Functional analyses included in vitro T_FH cell differentiation and cT_FH/naïve B cell co-cultures. Serum soluble IL-2 receptor alpha chain (sIL-2Rα), and in vitro immunoglobulin production by cultured B cells were quantified by ELISA.

Results

cT_FH cell frequencies in patients with LRBA or CTLA4 deficiency sharply declined with CTLA4-Ig therapy, in parallel with other markers of immune dysregulation including sIL-2R α, CD45RO⁺CD4⁺ effector T cells and auto-antibodies, and predictive of favorable clinical responses. cT_FH cells in patients with LRBA deficiency were biased towards a T_H1-like cell phenotype, which was partially reversed by CTLA4-Ig therapy. LRBA-sufficient but not -deficient regulatory T (T_reg) cells suppressed in vitro T_FH cell differentiation in a CTLA4-dependent manner. LRBA deficient T_FH cells supported in vitro antibody production by naïve LRBA-sufficient B cells.

Conclusions

cT_FH cell dysregulation in LRBA deficiency reflects impaired control of T_FH differentiation due to profoundly decreased CTLA4 expression on T_reg cells, and probably contributes to autoimmunity in this disease. Serial monitoring of cT_FH cell frequencies is highly useful in gauging the clinical response of LRBA deficient patients to CTLA4-Ig therapy.

Graphical abstract

Introduction

Follicular helper T (T_FH) cells are a distinct subset of T cells that have a crucial role in humoral adaptive immunity ¹. This specific T helper (T_H) lineage has a unique phenotype, expressing high levels of CXCR5, PD-1, ICOS and CD40L^{2, 3}. T_FH cells constitutively express high levels of B cell lymphoma 6 (Bcl-6), whereas other T helper cell populations including T_H1, T_H2 and T_H17 cells express high levels of the antagonizing transcription factor Blimp-1 ⁴. Various signaling pathways have been implicated in T_FH cell differentiation including signals from dendritic cells, B cells, cytokines (IL-6 and IL-21 in mice; IL12, 1L23 and TGF-β in human) and surface molecules (ICOS, CD28, CD40L, PD-1, BTLA, and SAP) ¹. T_FH cells are essential for germinal center formation and B cell differentiation into long-lived memory B cells and plasma cells within secondary lymphoid tissues. Up-regulation of CXCR5 and down-regulation of CCR7 guide T_FH cell migration into the B cell follicle ¹. The interaction between T_FH cells and their cognate B cells provides fundamental signals for high-affinity antibodies production through affinity maturation and class switch recombination. The importance of the T_FH cell lineage in promoting antibody responses was established by the observation of defective antibodies class switching in mice lacking T_FH cells ⁵, whereas unbridled T_FH cell responses trigger humoral autoimmunity due to generation of autoantibodies by autoreactive B cells ^5–10. Circulating T_FH (cT_FH) cells were recently identified in human subjects as reflective of T_FH cells ^{3, 11}. Increased cT_FH cells have been reported in several human autoimmune diseases, including systemic lupus erythematosus (SLE), rheumatoid arthritis, Sjogren’s syndrome and autoimmune thyroid disease ^12–16. Reciprocally, decreased cT_FH have been found in several monogenic immunodeficiency disorders associated with humoral deficiency ¹⁷.

T follicular regulatory (T_FR) cells, that originate from FOXP3⁺ T cells and express high levels of CXCR5, play a pivotal role in controlling immune dysregulation by regulating T_FH and activated B cell responses through the inhibitory effect of cytotoxic T lymphocyte associated antigen 4 (CTLA4) ^18–22. CTLA4 deletion in mice results in exaggerated B cell response and increased T_FH cells frequencies ^21–24. A number of inherited disorders of immune dysregulation that potentially impact T_FR function result in high cT_FH cell frequencies, which positively correlate with autoantibodies production. The best characterized of these are monogenic defects involving the LRBA-CTLA4 pathway ^25–30. LPS-responsive beige-like anchor (LRBA) deficiency is a primary immunodeficiency characterized by recurrent infections with hypogammaglobulinemia, giving rise to a common variable immunodeficiency-like phenotype, and immunodysregulation with autoimmunity, including inflammatory bowel disease, autoimmune endocrinopathies and cytopenias ^{25, 26, 29–31}. LRBA regulates the intracellular trafficking of CTLA4 ³⁰. Its deficiency results in dysregulation of the T_FH cell response and is associated with intense immune dysregulation and auto-antibody production ²⁹. The precise mechanism for T_FH cell dysregulation in LRBA deficiency, whether it involves enhanced T_FH cell differentiation, impaired T_FR regulation or both, remains unclear. CTLA4-Ig therapy has emerged as an effective treatment for the immune dysregulation associated with both gene defects ³⁰. The capacity to rapidly and sensitively monitor the immune status of these patients and their response to CTLA4-Ig therapy would be of great clinical utility in optimizing their care.

In this report, we examined the mechanisms of T_FH dysregulation in human subjects with LRBA deficiency and the utility of using cT_FH cell frequencies as a clinical indicator of response to treatment with CTLA4-Ig. Serial cT_FH cell analysis was carried out on LRBA deficient patients before and after starting CTLA4-Ig therapy. Our results show that T_FH dysregulation in LRBA deficiency was not due to an intrinsic abnormality in T_FH cells. Rather, it reflects failure of LRBA-deficient T_reg cells to control T_FH cell differentiation. cT_FH cells were found to be a sensitive marker of the clinical response to CTLA4-Ig therapy. Our finding supports the use of cT_FH cell frequencies as a marker for monitoring the clinical status of LRBA deficient patients and their response to therapy.

Materials and Methods

Patients

Patients P1 and P2 are two previously described Saudi Arabian siblings with LRBA deficiency due to a homozygous deletion in the BEACH domain of LRBA that abolished protein expression (patients P5 and P6; Family C, in our original report) ²⁹. Patient P3 is an 11 year old girl with chronic immune dysregulation who was diagnosed with LRBA deficiency due to exon 57 deletion as confirmed by genomic analysis, cDNA sequencing and absent LRBA protein on flow cytometry and immunoblotting (Fig. E1 in this article’s Online Repository). Her clinical presentation is detailed in the Online Repository Text. Patient P4 is a 6 year old Saudi Arabian boy who developed severe Crohn’s-like illness and type 1 diabetes and was found on whole exome sequencing to have a 4 base pair (bp) deletion in LRBA (c.4757_4760del, p.L1586fs). LRBA deficiency was confirmed by absent protein expression on flow cytometry (Fig E2 in the online repository). Patients P5 and P6 are newly diagnosed Turkish brothers with LRBA deficiency due to a single bp deletion in exon 54 of LRBA, as confirmed by genomic analysis, leading to a frame shift and premature stop codon (c.7885delA, p.R2629fs). LRBA deficiency was confirmed by near absent protein expression on flow cytometry (Fig E3 in the online repository). Patients P7–P9 were found to have heterozygous mutations in CTLA4 (Fig E4 in the online repository). Their clinical presentations are detailed in the Online Repository Text. A patient with IPEX due to an A384T amino acid substitution in FOXP3 was identified by exome analysis followed by Sanger sequencing of the mutation site in the FOXP3 gene. Subject with the diagnosis of SLE was identified at the Rheumatology clinic at the Boston children’s Hospital. Control subjects were age group-matched. All study participants were recruited using written informed consent approved by the local Institutional Review Boards. Studies at the Boston Children’s Hospital were conducted under approved protocol #04-09-113R.

Antibodies and flow cytometry

Information on the antibodies employed is provided in the Online Repository Text. Whole blood was incubated with mAbs against surface markers for 30 min on ice. Intracellular staining with FOXP3 and CTLA4, was performed using eBioscience Fixation/Permeabilization according to the manufacturer’s instructions. Indirect intracellular staining for LRBA was performed on freshly isolated peripheral blood mononuclear cells (PBMCs) using BD Biosciences Fixation/Permeabilization buffer with polyclonal rabbit anti-LRBA antibodies (Sigma-Aldrich, St. Louis, MO) or Rabbit IgG XP (R) isotype control (cell signaling), followed by secondary detection with Brilliant Violet 421^™ Donkey anti-Rabbit IgG (Biolegend). For chemokine receptor staining in cT_FH cells, PBMCs were isolated by Ficoll-Paque Plus gradient and surface staining for PD1, CXCR5, CXCR3 and CCR6 was performed for 30 min on ice. For cytokine detection among T_FH cells, PBMCs were stimulated in vitro for 4h in complete RPMI medium with Golgi plug (1/1000), PMA (50ng/mL) and ionomycin (500 ng/mL). Cells were washed and incubated with mAbs against surface markers for 30 min on ice, permeabilized using BD Biosciences Fixation/Permeabilization buffer and intracellular staining with mAbs against cytokines was performed. For characterization of circulating B cell subsets, previously frozen PBMCs were stained for surface antigens CD19, CD27 and IgD. Data were collected with a Fortessa cytometer (BD Biosciences) and analyzed with FlowJo software (TreeStar).

Whole exome sequencing (WES)

WES was performed on genomic DNA of the probands P3, P4, P5 and P9 as described in earlier reports ^{25, 29}.

Sanger sequencing analysis

LRBA and CTLA4 sequences were amplified from genomic and complimentary (c)DNA by the polymerase chain reaction and sequenced bidirectionally using dye-terminator chemistry ²⁹. Primers used in amplification reactions are available upon request.

Immunoblotting

Protein lysates derived from lymphocytes were resolved on SDS-PAGE gels and transferred to nitrocellulose filters as described. Immunoblots were carried out using polyclonal rabbit anti-LRBA antibody (Sigma-Aldrich, St. Louis, MO). The blots were reprobed with polyclonal rabbit anti-DOCK8 antibody (Sigma-Aldrich, St. Louis, MO).

T_FH and naïve B cell co-culture

cT_FH and naïve B cells were isolated from the peripheral blood of patient P3 and her human leukocyte antigen (HLA) fully matched healthy sister by fluorescent activated cell sorting (FACS). The cells were mixed as indicated and cultured for 12 days in the presence of endotoxin-reduced staphylococcal enterotoxin B (SEB) at 1μg/ml in RPMI 1640 complete medium supplemented with 10% heat-inactivated fetal bovine serum ³². At the end of the incubation period, the IgM and IgG concentrations in the culture supernatants were measured by ELISA.

In vitro generation of induced T_FH

CD4⁺ T cells were isolated from PBMCs by negative selection using magnetic beads (Miltenyi). Naïve CD3⁺CD4⁺CCR7⁺CD45RA⁺ cells were purified from CD4⁺ T cells by cell sorting on a FACS ARIA (purity > 98%). Naïve CD4 T cells were seeded at a concentration of 5×10⁴ per well of a 96-well plate and stimulated with recombinant human IL-12, 1L-23 and TGF-β1 in the presence or absence of T_reg cells, CTLA4-Ig and anti-CTLA4 ³³. The cells were cultured for 4 days, at the end of which they were stained for T_FH cell markers.

Autoantibody assays

For autoantibody detection, plasma aliquots from patient and control subjects were analyzed using microarrays spotted with 84 autoantigens (University of Texas Southwestern Medical Center, Genomic and Microarray Core Facility), as described ³⁴. Data was normalized to healthy controls.

Statistical Analysis

Comparison between groups was carried out using Student’s unpaired two tailed t test and 2-way ANOVA with Bonferroni post-test analysis, as indicated. Differences in mean values were considered significant at a p < .05.

Results

Three patients with definitive LRBA deficiency were studied for their response to CTLA4 therapy. Details of the clinical and immunological findings of the three patients are shown in Table E1 and Table E2, respectively, in the Online Repository. The three patients exhibited severe immune dysregulation with autoimmune cytopenias, chronic end organ inflammation and damage, especially affecting the lungs (P1, P3) and the gut (P1, P2). All patients showed marked clinical response to CTLA4-Ig therapy with decreased disease symptomatology, weight gain and resolution of their thrombocytopenia (Fig 1, A and B). Patients P1 and P2, who suffered from colitis, showed good clinical response with decreased diarrhea. Patient P2, who developed type 1 diabetes shortly before initiation of therapy, had resolution of her insulin dependence and normalization of her blood glucose (Fig 1, C). Patient P3, who suffered from severe lung disease and was oxygen dependent, responded very favorably to CTLA4-Ig therapy with marked improvement in her lung imaging and function and resolution of her oxygen dependency (Fig 1, D and E).

Figure 1. Clinical response of LRBA-deficient subjects to CTLA4-Ig therapy.

A and B, Body mass index (Fig 1, A) and platelet counts (Fig 1, B) of LRBA deficient subjects pre- and post-treatment with CTLA4-Ig. C, Serum glucose levels and insulin use in patient P2 status post CTLA4-Ig therapy. D and E, High-resolution chest CT scans (Fig 1, D) and Pulmonary function tests (Fig 1, E) of patient P3 pre and post treatment with CTLA4-Ig; FVC, forced vital capacity; FEV1, forced expiratory volume; TLC, total lung capacity. ****P < .0001, 2-way ANOVA with post-test analysis.

We have previously documented that patients with LRBA deficiency exhibit dysregulated cT_FH²⁹, a phenotype related to the profound deficiency of CTLA4 expression by LRBA-deficient T_reg cells ^{21, 22, 29}. Flow cytometric analysis of peripheral blood T lymphocyte cells of the three patients demonstrated a markedly increased number of CD4⁺FOXP3⁻PD1⁺CXCR5⁺ cT_FH cells (Fig 2, A). The frequency of cT_FH and CTLA4 expression of T_reg cells of three other patients with LRBA deficiency (P4–P6) were also analyzed. They also exhibited high frequency of cT_FH cells in association with low CTLA4 expression on their T_reg cells (Fig E2 and E3). Importantly, the frequency of cT_FH cells significantly declined after CTLA4-Ig therapy (Fig 2, A and B). CTLA4-Ig therapy also resulted in the decline of other markers of inflammation, including serum soluble CD25 (sCD25) (Fig 2, C). The frequency of circulating naïve CD4⁺ T cells (CD4⁺CD45RA⁺CCR7⁺) increased following CTLA4-Ig therapy, indicative of more effective immunoregulation (Fig 2, D). The decline in cT_FH cells upon CTLA4-Ig therapy correlated well with that of sCD25, validating the monitoring of these cells as a proxy for the immune dysregulatory status of the patients (Fig 2, E). In addition to their increased frequency, cT_FH cells of LRBA-deficient subjects also had increased expression of the cT_FH cell markers ICOS and PD1 as compared to those of control subjects, consistent with them being activated ^{11, 35}. The expression levels of both markers were normalized following therapy with CTLA4-Ig (Fig 2, F and G).

Figure 2. T_FH cell frequency in LRBA-deficient patients correlates with other markers of immune dysregulation.

A, Flow cytometric analyses of CXCR5 and PD1 expression in CD4⁺ T cells in LRBA- deficient subjects pre- and post-treatment with CTLA4-Ig. B–D, cT_FH cell frequencies (Fig 2, B), serum sCD25 levels (Fig 2, C) and naïve CD4⁺CD45RA⁺CCR7⁺ T cell frequencies (Fig 2, D) in LRBA deficient patients pre and post-treatment with CTLA4-Ig. E, Correlation between cT_FH cell frequencies and sCD25 levels in LRBA-deficient patients. F, Flow cytometric analysis of ICOS and PD1 expression on control cT_FH cells and on patient cT_FH cells before and after CTLA4-Ig therapy. G, Relative mean fluorescence intensity of ICOS and PD1 expression on patient versus control cT_FH cells. ***P < .001 and ****P < .0001, by student’s two tailed t test.

To further characterize the phenotype of cT_FH cells in LRBA-deficient patients and also the impact of CTLA4-Ig treatment, we evaluated the expression of chemokine receptors and the capacity of cT_FH to secrete cytokines. cT_FH cells of LRBA deficient patients highly expressed the T_H1-associated chemokine receptor CXCR3 and IFN-γ, markers that define T_H1-biased T_FH cells (or T_FH-1)^{11, 35}. In contrast, expression of the T_H17-associated chemokine receptor CCR6 and IL-17, markers that define T_H17-biased T_FH cells (T_FH17) ^{11, 35}, was markedly decreased in patient as compared to control cT_FH cells (Fig 3, A–D). Some of these abnormalities, such as increased CXCR3 expression, persisted despite CTLA4-Ig therapy, and probably reflected either attributes intrinsic to LRBA-deficient T_FH cells or ongoing T_H1 cell inflammation in the hosts. In contrast, expression of IL-21 and IL-4 were similar in patient and control cT_FH cells and appeared unaffected by CTLA4-Ig therapy.

Figure 3. cT_FH cells of LRBA-deficient patients are skewed towards a T_H1-like cell phenotype.

A and B, Flow cytometric analyses of CXCR3 and CCR6 expression in cT_FH cells of LRBA-sufficient and LRBA- deficient subjects pre- and post-treatment with CTLA4-Ig. C and D, Flow cytometric analyses of IL-17 and IL-21 expression (Fig 3, C, upper panels) or IFN-γ and IL-4 (Fig 3, C, lower panels) and the respective scatter plot representation (Fig 3, D) in cT_FH cells of LRBA-sufficient and LRBA- deficient subjects pre- and post-treatment with CTLA4-Ig. *P < .05, **P < .01 by 1-way ANOVA with post-test analysis.

We have previously reported that patients with LRBA deficiency have profound T_reg cell abnormalities including decreased T_reg cell frequencies and reduced expression of several T_reg cell markers, including CTLA4 ²⁹. In view of this immune dysregulation, we analyzed the correlation of the magnitude of CTLA4 expression on T_reg cells and the frequency of cT_FH cells. To that end, we examined the frequency of cT_FH cells in three patients with heterozygous loss of function mutations in CTLA4. Their mutational analysis is shown in Fig E4 in the online Repository, and their clinical and immunological findings are detailed in Table E1 and Table E3, respectively, in the Online Repository. Consistent with the role of CTLA4 deficiency in dysregulated cT_FH cells in LRBA deficiency, patients with heterozygous loss of function mutations in CTLA4 also manifested increased cT_FH cell frequencies (Fig 4, A and B). Overall, cT_FH cells frequencies were inversely correlated with the CTLA4 MFI on T_reg cells of LRBA and CTLA4-deficient subjects, reflecting the role of CTLA4 in controlling T_FH cell differentiation. Subjects with complete LRBA deficiency, associated with near complete absence of CTLA4 expression on T_reg cells, manifested the highest cT_FH frequencies, while those with heterozygous loss of function CTLA4 mutations had a more moderate increase in cT_FH frequencies. (Fig 4, C). Similar to the case of CTLA4-Ig-treated LRBA-deficient subjects, treatment of a patient with CTLA4 deficiency with CTLA4-Ig resulted in the decline of cT_FH cells, consistent with the role of CTLA4 deficiency in cT_FH dysregulation in both disorders (Fig 4, D and E).

Figure 4. T_FH cell frequency in CTLA4 deficiency and its response to CTLA4-Ig therapy.

A and B, Flow cytometric analysis of CTLA4 expression on CD4⁺FOXP3⁺ T_reg cells (Fig 3, A) and cT_FH (Fig 3, B) of control (Ctrl) subjects and patients P7 and P8 with a heterozygous CTLA4 mutation. C, Correlation between cT_FH cell frequencies and the mean fluorescence intensity (MFI) of CTLA4 expression in T_reg cells of Ctrl (n=8), LRBA-deficient (n=6) and heterozygous CTLA4-mutant subjects (n=3). D, Flow cytometric analysis of CTLA4 expression on CD4⁺FOXP3⁺ cells in patient P9 with a heterozygous CTLA4 mutation. E, cT_FH cells in patient P9 before and after CTLA4-Ig therapy.

To determine whether LRBA deficiency impaired T_FH differentiation, we tested the capacity of LRBA-sufficient and deficient naive CD4⁺ T cells to differentiate into induced follicular helper (iT_FH) cells upon stimulation with IL12, IL23 and TGF-β1 in the presence or absence of CTLA4-Ig ³³. CD4⁺ T cells from patients and control subjects were enriched by magnetic bead separation, and naive CD4⁺ T cells were purified to greater than 99% purity by cell sorting (Fig E5 in the Online Repository). LRBA-sufficient and -deficient T cells were found to be equivalent in their capacity to differentiate into iT_FH cells that expressed similar levels of Inducible T cell Costimulator (ICOS) regardless of the CTLA4 status of naïve CD4⁺ cells (Fig 5, A–C). However, iT_FH cells generated from LRBA-deficient naïve T cells exhibited higher expression of PD1, as compared to those from LRBA-sufficient naïve T cells (Fig 5, D and E). Both iT_FH differentiation and expression of ICOS and PD1 were partially inhibited upon treatment of the differentiating T cell cultures with CTLA4-Ig (Fig 5, B–E). Given the normal differentiation of LRBA deficient T cell into iT_FH cells, we further analyzed the capability of LRBA-sufficient and deficient T_reg cells to control iT_FH cells differentiation. T_reg cells were isolated from patient and healthy control by means of cell sorting of CD4⁺CD25⁺CD127low T_reg cells. Equal numbers of patient and control T_reg cells were added to an equal number of control naïve CD4⁺cells that were stimulated with IL12, IL23 and TGF-β in the presence or absence of anti-CTLA4. LRBA-deficient T_reg cells failed to suppress iT_FH differentiation compared with control T_reg cells, indicative of their impaired function (Fig 5, F and G).

Figure 5. Ineffective T_reg cells control of T_FH Cell differentiation in LRBA deficient subjects.

A, Flow cytometric analysis of CXCR5 and ICOS expression in in vitro-differentiated induced (i)T_FH-like cells of control and LRBA-deficient subjects. B and C, MFI of ICOS expression in iT_FH-like cells (Fig 5, B), and frequencies of in vitro-differentiated iT_FH-like cells (Fig 5, C) derived from control and LRBA-deficient naïve CD4⁺ T cells in the absence or presence of CTLA4-Ig. D and E, Flow cytometric analysis of PD1 expression (Fig 5, D) and bar graph representation (Fig 5, E) of PD1 expression in iT_FH cells of control and LRBA-deficient subjects. F and G, Flow cytometric analysis (Fig 5, F) and frequencies (Fig 5, G) of in vitro-differentiated iT_FH-like cells derived from control and LRBA-deficient naïve CD4⁺ T cells in the absence or presence of anti-CTLA4 mAb and/or patient or control T_reg cells. Results are representative of 2 independent experiments. *P < .05, **P < .01 and ***P < .001 2-way ANOVA with post-test analysis.

Patients with LRBA deficiency frequently present with hypogammaglobulinemia and defective specific antibody titers, with the majority exhibiting B cell dysfunction and reduced class-switched memory B-cells ^{31, 36}. Nevertheless, most of these patients suffer from autoimmunity with increased circulating autoantibodies ^{29, 31, 36}. To determine whether the production of circulating autoantibodies responded to CTLA4-Ig therapy, we screened our patients for serum autoantibodies before and two time points after CTLA4-Ig therapy using an array of 84 auto-antigens ³⁴. Also included in the analysis were four healthy subjects whose sera were used as negative controls and two positive control sera including one patient with IPEX and another patient with SLE. LRBA patients exhibited a number of circulating IgG autoantibodies whose relative abundance significantly decreased following CTLA4-Ig therapy (Fig 6, A). We next employed in vitro co-culture studies to examine the functional capacity of LRBA-deficient cT_FH to support IgM and IgG antibody production by naïve B cells. CD4⁺CD25⁻CD127^HighPD1⁺CXCR5⁺ cT_FH cells and CD19⁺IgD⁺CD27⁻ naïve B cells were isolated from LRBA-deficient patient P3 and her fully HLA-matched healthy sister by cell sorting and co-cultured in the presence of staphylococcal enterotoxin B (SEB). Results showed that while cT_FH cells of both the LRBA-deficient patient and her healthy sibling supported in vitro IgM and IgG antibody production by the sibling’s naïve LRBA-sufficient B cells (Fig 6, B), the LRBA-deficient cT_FH cells were more effective in that regard, consistent with their heightened activation state (Fig 2, F and G) ^{11, 35}. In contrast, the patient’s cT_FH supported IgM but minimally IgG production by B cells, in agreement with the previous report of defective IgG isotype switching in LRBA-deficient B cells ²⁶. We also evaluated the impact of CTLA4-Ig treatment on B cell phenotype. Whereas LRBA-deficient patients exhibited increased frequencies of naïve (IgD⁺CD27^low) and correspondingly decreased frequencies of switched memory (IgD⁺CD27^low) B cells, CTLA4-Ig treatment did not change the ratio of naïve/memory B cells in circulation (Fig 6, C). Given that expansion of T_FH cells promotes auto-antibody production ^5–10, these findings indicated that the dysregulated T_FH expansion in LRBA deficiency is functionally relevant to the excess auto-antibody production in this disorder.

Figure 6. Autoantibody production and cT_FH function in LRBA-deficient subjects.

A, Autoantibody production in LRBA deficient patients before and after CTLA4-Ig therapy. Heat map showing IgG autoantibodies against self-antigens in sera of LRBA-deficient patients, healthy control subjects, patient with IPEX, and a patient with SLE. A value of 1 (black) is equal to the control average + 1 SD. Autoantibody responses affected by CTLA4-Ig therapy are boxed in red. B, IgM and IgG production in co-cultures of cell-sorted cT_FH and naïve B cells derived from patient P3 and her HLA fully matched sister in the presence of Staphylococcal enterotoxin B (SEB). C. Flow cytometric analysis (Fig 6, C, upper panels) and scatter plot representation (Fig 6, C, lower panels) of CD27 and IgD expression on circulating B cells of control subjects and LRBA deficient subjects before and after treatment with CTLA4-Ig. *p<0.05 and **p<0.01, Student unpaired 2-tailed t test; ****P < .0001 by 1-way ANOVA with post-test analysis.

Discussion

Patients with deleterious mutations in LRBA gene have dysregulated T_FH cell responses, as reflected by the high frequency of cT_FH cells, which may play a causative role in disease-related autoimmunity. In this report, we examined the mechanisms of T_FH dysregulation in LRBA deficiency and the usefulness of monitoring cT_FH cell frequencies as a measure of disease activity and response to therapy. We found that T_FH cell dysregulation involved failure of CTLA4-dependent T_reg cell to control T_FH cell differentiation. Furthermore, the elevated cT_FH cell frequencies in LRBA deficiency and the related CTLA4 deficiency dramatically declined following CTLA4-Ig therapy, in concordance with the decline in other markers of disease activity and improved clinical outcome. While cT_FH cell frequencies in LRBA-deficient subjects tightly correlated with other markers of immune dysregulation such as sCD25, the former offers the advantage of ease of its measurement by flow cytometry and its sensitive and fast response to CTLA4-Ig therapy. Thus, monitoring cT_FH cell frequencies in LRBA and CTLA4 deficiencies maybe particularly useful in tracking disease activity and response to different therapies.

Recent studies identified the critical role of CTLA4 in T_reg cells in controlling humoral immunity. In vivo T_reg/T_FR cells depletion or selective deletion or blockage of CTLA4 in T_reg compartment resulted in unrestrained T_FH cell differentiation and profoundly increased auto-antibody production ^{21, 22}. In concordance with these studies, we demonstrated that LRBA-deficient naïve CD4⁺ T cells effectively differentiate into T_FH cells. However, LRBA-deficient T_reg cells, like normal T_reg cells treated with anti-CTLA4 mAb, failed to suppress the expansion of in vitro differentiated T_FH cells, pointing to the reduced CTLA4 expression on LRBA-deficient T_reg cells as a key mechanism underlying the dysregulated T_FH response in LRBA deficiency.

Many patients with LRBA deficiency present with hypogammaglobulinemia and low numbers of switched memory B cells ^{31, 36}. LRBA deficient naïve B cells have an intrinsic defect in their capacity to differentiate into IgM and IgG producing plasma cells ²⁶. Despite their B cell switch defect, LRBA-deficient subjects mount an intense autoantibody response. cT_FH cells of LRBA-deficient subjects were activated, as evidenced by their heightened expression of ICOS and PD1, a phenotype that was normalized upon CTLA4-Ig therapy. They were more effective in supporting immunoglobulin production by LRBA-sufficient B cells as compared to control cT_FH cells, indicative of their augmented functional capacity. In vitro differentiated iT_FH cells also exhibited increased PD1 expression that was reversed by CTLA4-Ig treatment, reflecting the role of CTLA4 deficiency in dysregulating LRBA-deficient T_FH cells both in vitro and in vivo. We suggest that the profound dysregulation of the T_FH cell compartment in LRBA deficient subjects, precipitated by T_FR cell depletion and virtually absent CTLA4 expression, coupled with defective B cell isotype switching and its associated somatic hypermutation, may propel humoral autoimmunity by limiting the pruning of auto-reactive antibodies normally achieved with somatic hypermutation ³⁷. Such a mechanism would be consistent with the reversal of T_FH dysregulation and down-regulation of auto-antibody production by CTLA4-Ig therapy. Because CD80 signaling favors B cell isotype switching and immunoglobulin production, we cannot exclude an additional direct effect of CTLA4-Ig treatment on B cells resulting in the reduction of auto-antibody production in LRBA-deficient patients ³⁸. Further studies would be required to validate these predictions.

In summary, these findings support the monitoring of cT_FH cells as a sensitive marker for monitoring the response of LRBA and CTLA4-deficient patients to CTLA4-Ig therapy. Its use may help optimize the management of LRBA-deficient subjects to reach sustained clinical improvement and optimal preparation for hematopoietic stem cell transplantation. More broadly, monitoring the T_FH/T_FR axis may also aid in the diagnosis and treatment of primary immune dysregulatory disorders.

Key Messages.

• LRBA and CTLA4 deficiencies result in highly elevated frequencies of circulating T follicular helper (cT_FH) cells.

• cT_FH dysregulation in LRBA deficiency reflects impaired CTLA4-dependent suppression of T_FH cell differentiation by T regulatory (T_reg) cells.

• Monitoring cT_FH frequencies in LRBA and CTLA4 deficiencies is useful in gauging the clinical response to CTLA4-Ig therapy.

Abbreviations

CTLA4

cytotoxic T lymphocyte associated antigen 4

CVID

Common variable immunodeficiency

ELISA

enzyme-linked immunosorbant assay

FACS

fluorescence-activated cell sorting

HLA

human leukocyte antigens

IPEX

Immune dysregulation, polyendocrinopathy, enteropathy, X-linked

LRBA

LPS-responsive beige-like anchor

MFI

mean fluorescence intensity

SLE

systemic lupus erythematosus

SEB

staphylococcal enterotoxin B

T_reg

T regulatory

T_FH

T Follicular Helper

T_FR

T Follicular Regulatory

T_H1

T helper type 1

T_H17

T helper Type 17

WES

Whole exome sequencing