Rapamycin Controls Lymphoproliferation and Reverses T-Cell Responses in a Patient with a Novel STIM1 Loss-of-Function Deletion

The local ethics committee from Marmara University approved the clinical and research studies protocol, and written informed consent was obtained from the family. We documented the clinical and demographic features of the patient and provided data regarding long-term follow-up.

T lymphoblasts were generated from the ex vivo expansion of peripheral blood mononuclear cells (PBMCs) with Phytohemagglutinin (PHA, 1 μg/mL) and IL-2 (100 U/mL) in RPMI-1640 media containing 5% human serum, HEPES, Penicillin–Streptomycin, non-essential amino acids and sodium pyruvate and co-cultured with irradiated feeder cells (PBMCs from healthy donors) in a ratio of 1:1. T lymphoblasts were split every 2–3 days with fresh IL-2 (50 U/mL) added. T lymphoblasts were used after 9–13 days of expansion. EBV-transformed B cell lines were generated by incubating PBMCs (~ 5 × 10⁶ cells) with EBV-containing supernatant at a volume ratio of 1:1 in RMPI-1640 media supplemented with 10% FCS, 1% Penicillin–Streptomycin, and 1% HEPES. After at least 2 h, cyclosporine A (5 μg/mL) was added to inhibit T cells before prolonged cultivation until the EBV-immortalized B cells were established. Subsequent experiments using these cell lines were performed without prior treatments or stimulations other than those required for their expansion and maintenance.

Peripheral blood lymphocyte subset analyses, intracellular cytokine staining, upregulation, and proliferation assays were performed by flow cytometry, as described previously [20‐25]. Age-matched healthy donors were used for comparisons. Stained cells were acquired by Navios EX cytometer (Beckman Coulter) and analyzed by FlowJo software (TreeStar, Ashland, Ore). The details are presented in the Supplementary file.

T lymphoblasts (1 × 10⁶) were washed in Ca²⁺-free PBS and stained with Calcium Sensor Dye eFluor™ 514 (eBiosciences) for 30 min at 37℃. Cells were washed and resuspended in PBS containing 1 mM EGTA. Cells were first incubated with 1 μg/mL anti-CD3 (OKT3, eBioscience) for 5 min at 37C before a 30 s baseline measurement was recorded. Then, 20 μg/mL of AffiniPure goat anti-Mouse IgG (Jackson ImmunoResearch) was added to induce T-cell receptor (TCR) crosslinking, followed by 3 min of measurement. Finally, CaCl (2 mM final concentration) to induce SOCE was added, and a further 2 min was recorded. The assay was performed using an LSRFortessa™ (BD) and analyzed using FlowJo (Ver. 10). Results, recorded as intensity per unit time, were normalized to the mean basal intensity and plotted against time.

T- lymphoblasts and EBV-transformed B-cells were washed in PBS, pelleted by centrifugation, and lysed in Frackeltons buffer containing protease and phosphatase inhibitors. Lysate protein concentration was determined using a DC reagent (Bio-Rad). Samples (25 μg) in Laemmli Buffer were separated on acrylamide gels (10%) at 75 V for 45 min followed by 120 V and transferred to PVDF membranes at 110 V for 90 min, or 40 V overnight at 4℃. Membranes were washed in TBST, blocked in 5% milk, and incubated in primary antibody overnight at 4℃. The following primary (STIM1—1:1000, Cell Signaling Technology, #5668; β-actin—1:5000, Sigma Aldrich, #A1978; HSP90α/β—1:5000, Santa Cruz Biotechnology, #sc-13119) and secondary (anti-Mouse HRP—BD Biosciences, #554002; anti-Rabbit HRP—Bio-Rad, #172–1019) antibodies were used.

Intracellular cytokine production was assessed in T-lymphoblasts by stimulating 0.2 × 10⁶ cells for 5 h with Phorbol 12-myristate 13-acetate (PMA, 0.2 mM) and Ionomycin (1 μg/mL). Brefeldin A was added during the final 4 h of the stimulation. Cells were washed and stained for T-cell surface markers (CD4, CD8) on ice for 30 min. Subsequently, cells were fixed, permeabilized, and stained with IL2. An LSRFortessa™ (BD) was used for acquisition, and plots were analyzed using FlowJo (Ver. 10).

The genetic analysis was performed using whole-exome sequencing (WES). Briefly, genomic DNA was extracted from peripheral blood samples, and 1 μg of DNA was used for library preparation using the Twist Exome 2.0 kit. Paired-end (100 bp) sequencing was performed on an Illumina NovaSeq 6000 device. The homozygous STIM1 deletion was called with the ExomeDepth software [26], validated by long-range PCR followed by Sanger sequencing, and segregated in all available family members.

A 3-year-old girl (PII.2), born to consanguineous parents, was admitted to our clinic at 16 months of age with bilateral axillary lumps and occasional episodes of fever. She was born prematurely and had esophageal atresia. She underwent repair surgery and stayed in the intensive care unit for one month. Her medical history was remarkable for recurrent hospitalizations due to pneumonia with high erythrocyte sedimentation rates (range: 74-120 mm/hr) and C-reactive protein levels (range: 6-47 mg/L). She also had anhidrosis with some suspected associated fevers. She also experienced an impetigo-like infectious cutaneous lesion due to Staphylococcus aureus infection. She received standard childhood vaccinations without any complications. Upon physical examination, the liver and spleen were palpable, 3 cm and 2 cm below the edge of the ribs, respectively. Additionally, multiple enlarged lymph node conglomerates were detected in the cervical, supraclavicular, axillary, mediastinal, abdominal, and inguinal regions (Fig. 1A and B). Her skin exhibited xerosis and pruritus without eczema. She had a kyphotic posture with mild axial hypotonia. Ophthalmologic examination showed mydriasis, iris hypoplasia, and persistent pupillary membranes over the lens (Fig. 1C). The pupils were non-responsive to light. Additionally, her teeth had brownish enamel with hypoplasia and increased notches (Fig. 1D).

Laboratory results revealed severe leukocytosis, eosinophilia, and elevated serum IgG levels (Table 1). She had adequate IgG antibodies against protein antigens and sufficient isohemagglutinin responses. She also had significantly high serum total IgE level and specific IgE to egg white without clinical symptoms upon exposure to egg. Her creatinine kinase level was in a normal range. The direct Coombs test was positive in the sera without signs of autoimmune cytopenia. At the same time, we did not detect positivity for other rheumatological autoantibodies, except for antinuclear antibody (1/320). Blood Epstein-Barr virus (EBV) and cytomegalovirus (CMV) PCR were negative. We performed bone marrow and axillary lymph node biopsies to exclude hematological and lymphoid malignancies. While the bone marrow examination was normal, axillary lymph node biopsy revealed atypical lymphoid proliferation. In the lymph node, pronounced T zone hyperplasia was identified, resulting in the regression of lymphoid follicles and disruption of the characteristic nodal architecture. Regressed follicles, as evidenced by CD20 and CD23 staining, were observed. Notably, the T cells exhibited diffuse positivity for CD4, while CD8 positivity was scattered. There was very sparse staining in CD4⁺ T cells with PD1 and Bcl6, delineating reduced resident T_FH cells in follicular centers (Fig. 1E). Additionally, kappa and lambda-positive polyclonal plasma cells were present, predominantly localized beneath the capsule. The biopsy material was negative for HHV-8, EBV, CMV, Mycobacterium tuberculosis, and Bartonella henselae. Due to the severe lymphoproliferation without malignancy, we considered a probability of ALPS-like disease.

The immunological evaluation was notable for CD3⁺ T lymphocytosis with skewing of CD4⁺ and CD8⁺ T cells towards a memory phenotype. In addition, recent thymic emigrant cells were decreased compared with the healthy control. Increased CD21^low CD38^low activated B cells were also detected (Table 1). The immune system was further evaluated by lymphocyte proliferation, CD69 and CD25 activation. The patient showed blunted CD69 expression after stimulation with αCD3/CD28 compared with healthy control (Fig. 1F). Furthermore, the patient’s CD4⁺ and CD8⁺ T cells exhibited drastically impaired CD25 expression and proliferation following stimulation with αCD3, αCD3/CD28, and PHA compared with healthy controls. The T cells showed higher proliferation and activation after αCD3/IL-2 stimulation but still lower than controls (Fig. 1G). Based on these results, at 17 months, the patient was commenced on prophylactic antibiotic therapy and intravenous immunoglobulin replacement, which resulted in the control of infections. Her acute phase reactants were also effectively normalized, suggesting the absence of non-infectious inflammatory processes contributing to elevated levels.

The WES and copy number variation analysis revealed a novel homozygous deletion in the STIM1 gene. The deletion (11:3967441_3972019del, hg38) comprises 4579 bp, including exon 2 and parts of its flanking introns (NM_001382567.1). This deletion was confirmed by Sanger sequencing (Fig. 2A and B). The parents and healthy sister were carriers for the genetic deletion, as confirmed by variant segregation using long-range PCR (Fig. 2C). This novel mutation and previously reported mutations are depicted in Fig. 2D. The list of other rare mutations detected in the patient is presented in Table S1. When variants with high combined annotation dependent depletion scores, identified by WES, were examined in detail, they did not explain the patient's ALPS-like phenotype and accompanying esophageal atresia findings. However, some variants with functions not fully explored may still contribute to the clinical presentation.

The STIM1^Ex2del variant caused a total abolishment of STIM1 protein expression in the patient’s derived T lymphoblasts and EBV-cell lines (Fig. 2E and F). Consistent with the identified STIM1 defect, the SOCE assay showed a complete loss of extracellular Ca²⁺ uptake following ER calcium depletion compared with the healthy control (Fig. 2G), as in previously reported patients [5, 12]. We also measured cytokine expression in isolated peripheral blood mononuclear cells by gating on CD4⁺CD45RO⁺ memory T cells after stimulation with PMA and ionomycin in vitro. The production of IFN-γ, IL-4, IL-10, and IL-17A was strongly reduced in the patient compared with healthy controls (Fig. 2H and I). Furthermore, stimulation-induced IL-2 expression was absent in T lymphoblast cells (Fig. 2J). Overall, these analyses strongly suggest that the STIM1^Ex2del variant is deleterious and acts as a LOF mutation, leading to impaired activation, proliferation, and cytokine production of T cells.

Mice deficient in both Stim1 and Stim2 in T cells develop a lymphoproliferative disorder and dermatitis, characterized by eosinophilia and augmented IgE and IgG1 responses. Furthermore, these mice display diminished percentages of regulatory T (Treg) and follicular regulatory T (T_FR) cells with increased numbers of T_FH in splenocytes, underscoring the role of STIM1/STIM2 in controlling T cell responses [27, 28]. Interestingly, the development of T_FH cells in mice with T cell-specific deletion of Stim1/2 was severely impaired after acute viral infection or immunization [29]. Based on these findings in mice, we investigated for the first time the frequencies of circulating T-helper cell subtypes in human STIM1 deficiency. These included cT_FH (CD4⁺CXCR5⁺CD45RA^– and CD4⁺CXCR5⁺PD1⁺), non-cT_FH memory T-helper cells (CD4⁺CXCR5^–CD45RA^–), Treg (CD4⁺CD25^hiCD127^lo) and circulating (c)T_FR (CD4⁺CXCR5⁺CD45RA^–CD25^hiCD127^lo) cells. The gating strategy of these analysis are presented in Fig.S1 and Fig.S2. The patient with the STIM1^Ex2del variant exhibited an increased frequency of circulating memory CD4⁺/CD8⁺ T cells (CD3⁺CD4⁺/CD8⁺CD45RA^–CD45RO⁺) and a decreased frequency of circulating naïve CD4⁺/CD8⁺ T cells (CD3⁺CD4⁺/CD8⁺CD45RA⁺CD45RO^–) (Fig. 3A-D). A more detailed analysis of the CD4⁺ T cells compartment showed that the patient had a decreased percentage of cT_FH cells with slightly increased PD-1 expression (Fig. 3E-G). The increased PD-1 expression was more prominent in total CD4⁺ T cells, indicating their activation (Fig. 3H). Further, when we evaluated the subtypes of cT_FH, they displayed a higher percentage of the T_H2-cell-like phenotype (CD4⁺CXCR5⁺CD45RA^–CD25^loCD127^hiCXCR3^–CCR6^–) compared with healthy controls and a trend for fewer of the T_H17-cell-like (CD4⁺CXCR5⁺CD45RA^–CD25^loCD127^hiCXCR3^–CCR6⁺) phenotype. The T_H1-cell-like (CD4⁺CXCR5⁺CD45RA^–CD25^loCD127^hiCXCR3⁺CCR6^–) percentage was comparable with healthy controls (Fig. 3I and J). On the other hand, patient non-cT_FH memory T-helper effector cells exhibited greater skewing towards T_H1- and T_H2-cell-like phenotypes (Fig. 3K and L).

Treg cell frequencies were normal at baseline; however, this cell population had reduced expression of canonical markers, including CD25, FOXP3, and CTLA4, especially after stimulation (Fig. 4A-F). Like other T-helper cell phenotypes, expanded percentages of T_H1-(CD4⁺CD25^hiCD127^loCXCR3⁺CCR6^–) and T_H2-like (CD4⁺CD25^hiCD127^loCXCR3^–CCR6^–) Treg cells were observed in the patient, indicating their abnormal reprogramming. In contrast, percentages of T_H17-cell-like Treg cells (CD4⁺CD25^hiCD127^loCXCR3^–CCR6⁺) were lower compared with healthy controls (Fig. 4G and H). Interestingly, the frequencies of cT_FR increased in the patient compared with healthy controls (Fig. 4I and J). These results revealed abnormal T-cell subtypes in STIM1 deficiency, characterized by increased T_H1- and T_H2-cell-like responses.

Since the patient displayed an ALPS-like phenotype, we initiated oral rapamycin (1.3 mg/m²/once daily) to control the lymphoproliferation. After one month of the therapy (plasma rapamycin level: 8.3 ng/mL, therapeutic reference range: 5 – 15 ng/mL), the enlarged lymphadenopathies started to reduce in size and resolved within 3 months of treatment (Fig. 5A). Her blood leukocyte, lymphocyte, and eosinophil numbers and serum IgE levels all normalized within 3 months of treatment; however, non-immune defects, including myopathy and enamel hypoplasia had remained unchanged (Table 1). After 20 months of rapamycin therapy, she continues to do well without recurrence of lymphadenopathy. Rapamycin therapy was not associated with drug-related adverse effects or infections. At 3 years of age, she was transplanted from a full HLA-matched mother and is currently in the 2^nd month of the post-HSCT period.

Immunological studies were performed after 18 months of rapamycin. There was a marked restoration in the frequency of naïve T cells and decreased percentage of CD21^low CD38^low B cells (Fig. 5B and C). The percentage of cT_FH cells increased, and PD1 expression on CD4⁺ T cells declined with rapamycin (Fig. 5D and E). We further observed a decreased percentage of cT_FR cells (Fig. 5F). Together with the reversion of total cT_FH cells, we observed a slight reduction in the percentage of T_H2-cell-like phenotype in this population (Fig. 5G). Rapamycin also reduced T_H1-cell-like percentages in non-cT_FH memory T-helper and Treg cells (Fig. 5H and I). Similar changes were also detected in absolute numbers of these populations when compared with age-matched healthy controls (Fig.S3). The treatment partially restored the upregulation of activation markers and the proliferative capacity of T cells, particularly in CD8⁺ T cells, in response to stimulation with αCD3/IL-2, αCD3/CD28, and PHA (Fig. 6A and D). These results indicate that treatment with the mTOR inhibitor rapamycin improved the clinical and immunological findings associated with STIM1^Ex2del mutation.