Selective clinical and immune response of the oligoclonal autoreactive T cells in Omenn patients after cyclosporin A treatment

Authors


Dr R. Somech, Pediatric Department B North, Pediatric Immunology Service, Safra Children's Hospital, Chaim Sheba Medical Center, Tel Hashomer, 52621, Israel. E-mail: rsomech@hotmail.com; raz.somech@sheba.health.gov.il

Summary

The immunological hallmark of Omenn syndrome (OS) is the expansion and activation of an oligoclonal population of autoreactive T cells. These cells should be controlled rapidly by immunosuppressive agents, such as cyclosporin A (CsA), to avoid tissue infiltration and to improve the general outcome of the patients. Here we studied the clinical and the immune response to CsA in two Omenn patients and also examined the gene expression profile associated with good clinical response to such therapy. T cell receptor diversity was studied in cells obtained from OS patients during CsA therapy. Characterization of gene expression in these cells was carried out by using the TaqMan low-density array. One patient showed complete resolution of his symptoms after CsA therapy. The other patient showed selective response of his oligoclonal T cell population and combination therapy was required to control his symptoms. Transcriptional profile associated with good clinical response to CsA therapy revealed significant changes in 26·6% of the tested genes when compared with the transcriptional profile of the cells before treatment. Different clinical response to CsA in two OS patients is correlated with their immunological response. Varying clonal expansions in OS patients can cause autoimmune features and can respond differently to immunosuppressive therapy; therefore, additional treatment is sometimes indicated. CsA for OS patients causes regulation of genes that are involved closely with self-tolerance and autoimmunity.

Introduction

Omenn syndrome (OS) is an autosomal recessive severe combined immunodeficiency (SCID) characterized by generalized scaly exudative erythrodermia, enlarged lymph nodes, hepatosplenomegaly, severe susceptibility to infections, activation of T helper type 2 lymphocytes, eosinophilia and hyper-immunoglobulin (Ig)E [1]. Regardless of the underlying genetic defect, patients with OS have a restricted T cell receptor (TCR) repertoire and exhibit peripheral expansion of self-reactive oligoclonal T cells [2]. OS is invariably fatal within the first months of life unless immune restoration is performed by haematopoietic stem cell transplantation (HSCT). Abnormal autoreactive T cells may infiltrate and expand into different organs (e.g. skin, gut, liver and spleen) and cause significant tissue damage [3]. Poor clinical status before the HSCT results in high transplantation-related mortality [4]. In the past, interferon (IFN) gamma was used to counteract the predominance of T cell activation and proliferation, to down-regulate interleukin (IL)-4 and IL-5 production, to modulate the inflammatory reaction by enhancing phagocytic functions and to improve clinical status [5]. Today, topical/systemic steroids or cyclosporin A (CsA) are the widely used medications to control the skin manifestations [6]. CsA, a known calcineurin inhibitor, seems to act on the IL-2 by inhibiting its production and repressing the activity of various transcription factors, thus leading to a decrease in the proliferation of the activated lymphocyte [7,8]. Moreover, it may interfere with specific signal transduction pathways which are important to the hypertrophic response [9]. Little is known about the immune modifications induced by CsA in OS patients. Such information will further improve our understanding the pathophysiology underlying OS and mechanisms of potential treatment modalities. Here we describe two OS patients and their clinical and immune response to CsA.

Methods

Patients

Two patients with recombinase activating gene (RAG)2 deficiency SCID and clinical and immunological features suggestive of the diagnosis of OS phenotype were reported. Significant transplacentally acquired maternal T lymphocyte was excluded in both patients by fluorescence in-situ hybridization (FISH). The study was approved by the Institutional Review Board and informed consent was obtained from all participants' parents.

Immune work-up

Cell surface markers of peripheral blood mononuclear cells (PBMCs) and lymphocyte proliferative responses to mitogens were performed as described previously. The amount of signal joint (sj) T cell receptor excision circles (Trecs) were determined by quantitative real-time reverse transcriptase – polymerase chain reaction (qRT–PCR). Reactions were performed using 0·25–0·5 µg genomic DNA extracted from the patients' PBMCs. The standard curve was constructed by using serial dilutions of a known Trec plasmid (generously provided by Dr Daniel Douek, Vaccine Research Center, National Institute of Allergy and Infectious Diseases, Bethesda, MD, USA). The number of Trecs in a given sample was calculated automatically by comparing the obtained Ct value of a patient's sample to the standard curve using an absolute quantification algorithm. Representatives of specific TCR-Vβ families were detected and quantified using flow cytometry (Becton Dickinson, Calibur, Franklin Lakes, NJ, USA) according to the manufacturer's (Beckman Coulter, Nyon, Switzerland) instructions and as described previously [10].

TCR-γ genes were amplified by PCR using fluorescence-labelled Vγ primers, according to the standardized Biomed 2 protocol [11]. Fluorescence-labelled PCR products (1 µl of each) were added to a mixture of 8·5 µl deionized formamide and 0·5 µl GeneScan 500TM Rox internal lane standard (PE Applied Biosystems, Weiterstadt, Germany) and separated using the 3100 Genetic Analyzer (PE Applied Biosystems). Results were analysed using the GeneMapper software (PE Applied Biosystems).

mRNA gene analyses by qRT–PCR

RNA from total PBMC, obtained from age-matched healthy controls and patient 1 before and after CsA treatment, was prepared using the Rneasy mini kit (QIAGEN Inc., Valencia, CA, USA). cDNA was prepared from 1 µg RNA using the high-capacity cDNA reverse transcription kit (PE Applied Biosystems). Predesigned TaqMan low-density arrays (TDLA, 96 TaqMan® gene expression assay human immune panel, 384-wells format, PE Applied Biosystems, catalogue number 4370499) were used in qRT–PCR. Each of the samples was analysed in two separate TLDA cards, using an PE Applied Biosystems 7900 HT fast real-time PCR system as described previously [12]. For analysis, expression levels of target genes were normalized to β-glucoronidase (GUSB). This gene was found by us [12] and others [13] to be an accurate housekeeping gene to analyse the gene expression profile in lymphocytes. Gene expression values were calculated based on the ΔΔCt method, with data normalized to the cDNA obtained from the age-matched healthy controls. Results were analysed using DataAssist™ version 2·0 software (PE Applied Biosystems). Only genes whose expression was significant (>twofold) were analysed and presented.

Results

Clinical response to CsA in OS patients

Patient 1 has been described previously [12]. Briefly, this male patient of Palestinian descent was born after a normal pregnancy and delivery to parents who are first-degree cousins. His clinical features included failure to thrive, severe infections [Pneumocystis carinii pneumonia (PCP) and cytomegalovirus (CMV)], remarkable erythrodermia, alopecia, massive lymphadenopathy and hepatosplenomegaly. The patient had undetectable levels of immunoglobulins and slightly reduced numbers of circulating lymphocytes (1320 cells per µl) with remarkable eosinophilia (2960 cells per µl). The rest of his initial immune work-up is summarized in Table 1. His genetic work-up revealed a homozygous missense RAG2 mutation (G35V). The patient was commenced on CsA treatment and significant cutaneous improvement was noticed within 72 h. CsA was continued at 2–3 mg/kg/day, resulting in blood levels between 50 and 100 ng/ml with complete resolution of erythrodermia. This treatment was continued until a successful human leucocyte antigen (HLA)-matched HSCT was performed at the age of 6 months.

Table 1.  Clinical description, immune work-up and genetic defects in 2 severe combined immunodeficient (SCID) patients with Omenn phenotype.
 Patient 1Patient 2
  • *

    Counts per minute (cpm), patient/control; HLA-DR: human leucocyte antigen D-related; Ig: immunoglobulin; PHA: phytohaemagglutinin; RAG: recombinase-activating gene; Trec: T cell receptor excision circle; UD: undetectable.

DiagnosisSCID-RAG2SCID-RAG2
Age at diagnosis (months)4/123/12
Maternal cells3·5%2%
Autoimmune features++++++
Infections++
Lymphocyte count/mm3132010686
Eosinophil/mm329604030
CD3/mm34882871
CD3+CD4+/mm32442351
CD3+CD8+/mm3224855
CD19+/mm300
CD3CD56+/mm35003800
HLA-DR+ (in total lymph)30%53%
IgM (IU/ml)UD110
IgG (IU/ml)UDUD
IgA (IU/ml)UDUD
IgE (IU/ml)UDUD
PHA mitogenic response*4500/746002650/40480
αCD3 mitogenic response*2500/3700010100/32000
Trecs/0·5µg DNAUDUD

Patient 2 is a male of Jewish Ashkenazi descent born after a normal pregnancy and delivery to non-consanguineous parents. His clinical features included severe erythrodermia, infection (adenovirus), alopecia and lymphadenopathy. The patient had undetectable levels of IgG, IgA and IgM and normal numbers of circulating lymphocytes (10 686 cells per µl) with remarkable eosinophilia (4030 cells per µl). The rest of his initial immune work-up is summarized in Table 1. Genetic work-up revealed a compound heterozygous RAG2 defect (G95V+E480X). The patient was commenced on CsA treatment; however, his cutaneous symptoms did not improve despite maintaining a high CsA trough level (100–150 ng/ml). Therefore, methylprednisone (2 mg/kg/day) was added and slow resolution of his cutaneous symptoms was observed. The patient was kept on both CsA and methylprednisone treatments until a successful HLA-matched cord blood transplantation was performed at the age of 6 months. In both patients, transplantations were successful and they have been currently followed for 2 years (patient 1) and 1 year (patient 2), with complete recovery of their symptoms and full reconstitution of their immune system.

Immune response to CsA in OS patients

TCR repertoire.  Examination of TCR-Vβ at presentation revealed peripheral expansion of oligoclonal T cells with dominant specific receptors. In patient 1, the dominant clone was TCR-Vβ 20, while in patient 2, TCR-Vβ 17 and TCR-Vβ 7·2 were dominant (Fig. 1a,b). Clonal patterns were also seen in the examined TCR-Vγ repertoire in both patients (Fig. 2a,b). These results suggest abnormal thymocyte selection and peripheral expansion, as expected in Omenn patients. Patient 1 showed a significant clinical improvement during CsA therapy; therefore, a follow-up analysis of his TCR repertoire was not indicated. However, in order to show that the patient did not have any expanded peripheral T cells, prior to the HSCT procedure, analysis of his TCR-Vγ repertoire was performed. The analysis revealed complete lymphopenia and no TCR expansion (Fig. 2c). In contrast, patient 2 did not respond completely to the initial treatment with CsA and remained symptomatic, therefore a follow-up analysis of his TCR repertoire was performed (Fig. 1c–e). Surprisingly, while the expression of the dominant TCR-Vβ 17 clone was reduced, the TCR-Vβ 7·2 clone did not respond to CsA therapy. Moreover, a few other TCRs, such as TCR-Vβ 14 and TCR-Vβ 5·1, started to appear (Fig. 1c). Only the addition of methylprednisone treatment resulted in suppression of these clones (Fig. 1d). However, even before the transplant, the patient still suffered mild skin symptoms, which were probably attributed to the presence of the TCR-Vβ 14 clone (Fig. 1e). Changes in the relevant TCRs during the treatment are presented in Fig. 3. During that time the patient was clinically stable apart from his skin symptoms and had no overt infection or other reason to explain clonal expansion.

Figure 1.

Figure 1.

Relative expression (black bar) of the various Vβ families in peripheral blood mononuclear cells (PBMCs) compared with normal controls (white bars). (a) Patient 1 upon diagnosis, (b) patient 2 upon diagnosis, (c) patient 2 during CsA treatment, (d) patient 2 after the addition of methylprednisone treatment, (e) patient 2 before haematopoietic stem cell transplantation (HSCT).

Figure 1.

Figure 1.

Relative expression (black bar) of the various Vβ families in peripheral blood mononuclear cells (PBMCs) compared with normal controls (white bars). (a) Patient 1 upon diagnosis, (b) patient 2 upon diagnosis, (c) patient 2 during CsA treatment, (d) patient 2 after the addition of methylprednisone treatment, (e) patient 2 before haematopoietic stem cell transplantation (HSCT).

Figure 2.

T cell receptor (TCR)-γ spectratyping of fluorescently labelled polymerase chain reaction (PCR) products using four consensus Vγ primers (vg9/2, vg11, vgf1, vg10/2) for the characterization of TCR-γ. (a) Patient 1 (upon diagnosis), (b) patient 2 (upon diagnosis), (c) patient 1 (before HSCT), (d) healthy control.

Figure 3.

Changes of the relative expression of the relevant T cell receptor (TCR)-Vβ families (TCR-Vβ 5·1, TCR-Vβ 7·2, TCR-Vβ 14 and TCR-Vβ 17) in peripheral blood mononuclear cells (PBMCs) obtained from patient 2 during his treatments (CsA and methyl-prednisone) from diagnosis to the haematopoietic stem cell transplantation (HSCT) procedure.

Trec quantification.  The amount of recent thymic emigrant cells as determined by real-time PCR analysis of Trecs was undetectable in both patients compared to 40 healthy age-matched controls (Table 1).

Gene expression in OS circulating T cells pre- and post-CsA treatment

In order to understand more clearly the gene transcriptional profiles associated with CsA treatment in OS patients, 90 genes related to the immune system were examined by TLDA before and after successful treatment (patient 1). After treatment, 26·6% (24 of 90) of genes showed an expression level of more than twofold increase or decrease compared with the patient's baseline gene expression (Fig. 4). Of these, the expression of 11 genes (12·2%) was down-regulated (by a factor of 2·3–5·2, values of 0·44–0·19, respectively, in Fig. 4, and 13 genes (14·4%) were up-regulated (by a factor of 2·04–19). The expression of several genes that are known to be down-regulated by CsA therapy such as IL-2 and Fas ligand (FasL) were found to be low (0·197- and 0·32-fold decrease). Interestingly, several genes that are known to be involved in immune regulation and autoimmunity were found to be markedly up-regulated [e.g. IL-10, intercellular adhesion molecule (ICAM) 1 and transforming growth factor (TGF)-β] or down-regulated (e.g. CCR4 and CCR5).

Figure 4.

Gene expression profile in patient 1 circulating T cells after cyclosporin A (CsA) treatment. The TaqMan low-density array (TLDA) was used to determine the transcriptional profiles of 90 genes associated with the immune system. The relative change to initial transcript level is shown. Twenty-four genes whose expression were significantly (> twofold) up- (13 genes) or down- (11 genes) regulated in patient 1 after successful treatment with CsA are listed.

Discussion

The immunological hallmark of OS is the expansion and activation of an oligoclonal population of autoreactive T cells. We have already shown that, in OS, similar T cell expansions are found in peripheral blood and in target organs (e.g. skin) [12]. These cells should be controlled rapidly by immunosuppressive agents to avoid tissue infiltration and to improve the general outcome of OS patients [14]. Here we describe a selective immune response to such treatments in patients with Omenn phenotype.

Diverse topical and systemic immunosuppressive therapies have been shown to be useful in OS patients. Many use CsA as the gold standard treatment for these patients. Alternatively, tacrolimus (FK506) is used. Despite similarity in their accepted mode of action, they alter T cell receptor expression differentially in vivo, therefore can have different effects on OS patients [15]. Failure of treatment sometimes requires alternative or a combination of therapies [16]. Herein, we report on two patients; the first patient responded to CsA treatment while the second patient did not. Surprisingly, the initial expanded oligoclonal autoreactive clone (TCR-Vβ 17) in the latter patient responded well to CsA treatment, but other TCR-Vβs had started to expand, probably causing the patient's unremitting autoimmune symptoms. Many unknown environmental and/or host factors can produce expanded lymphocytes in OS. In some cases the trigger (e.g. infection) that exacerbates the autoreactive process is found [17]. Patient 2 underwent a thorough infectious work-up, which was found to be negative, and no other obvious factor to trigger his symptoms could be detected to explain the presence of new TCR clones. However, expansion of certain new TCR-Vβ clones may also represent not only pathogen exposure, but also skewing towards self-antigens and autoimmunity [9]. Alternatively, we speculate that distinct lymphocyte subsets respond differently to the CsA effect, as suggested previously [18]; therefore, partial host resistance exists. Another possibility for the different levels of responsiveness to CsA among the reported patients might be the differences in the initial number of lymphocytes requiring suppression. As both patients also differed in their specific genetic defect (homozygosity versus compound heterozygosity), we can also hypothesize that in patient 2, the ongoing autoimmune process and resistance to the standard therapy might be secondary to his primary defect. This speculation regarding the severity of compound genetic defect has been described previously in patients with non-immunodeficiency diseases [19,20] and in patients with immunodeficiency diseases, including RAG defect [21,22]. The fact that patient 2 harbours two different mutations in the RAG2 gene, one resulting in a premature termination codon, reinforces this speculation. Recently, it was shown that the autoimmune regulator (AIRE) protein plays a critical role in eliminating self-reactive T cells and in the maintenance of tolerance. AIRE mRNA and protein deficiency in patients with OS suggests its participation in the development of the autoimmune features associated with this condition [12]. Therefore, we can also suggest that a lower level of AIRE mRNA transcript or abnormal protein function determines the severity of the autoimmune symptoms, enabling clones' leak that matures in the process to form autoreactive cells.

CsA is a potent immunosuppressant that has been used extensively to attenuate autoimmune symptoms. The molecular biological mechanism of CsA has been investigated extensively in human T cells, and it has been shown to involve modulation of the intracellular calcineurin pathway [23]. The cDNA microarray method showed that CsA-treated PBMCs displayed significant induction of genes involved in the control of cell-cycle regulation, apoptosis/DNA repair, DNA metabolism/response to DNA damage stimulus, transcription and cell proliferation [24]. In order to understand more clearly the gene transcriptional profiles associated with CsA treatment for OS, genes related to the immune system were examined by the TLDA assay. This assay has already been used successfully by us to demonstrate that dysregulated genes in OS patients are involved closely with self-tolerance and autoimmunity. Endothelin 1 (EDN1) and P-selectin (SELP), which were reported previously to be regulated by CsA therapy [25,26], were found by us to have the highest mRNA expression change after CsA therapy. The high expression of these genes is an acceptable explanation for the renal toxicity induced by CsA [27]. CsA is known to inhibit IL-2 induction, to decrease the expression of Fas and FasL and to increase the production of IL-10 [28,29]. CsA is not a general inducer of TGF-β biosynthesis but can cause different effects on TGF-β, depending on the cell type and concentrations used [30]. All these anticipated mRNA expression changes were observed in our reported patient. In contrast, while both CCR4 and CCR5 chemokine receptors were down-regulated after CsA therapy in our studied patient, only the CCR5 chemokine receptor was found to be affected by combined CsA and prednisone treatment in patients with Behçet uveitis, a different form of autoimmune disease [31]. Other cytokines, such as IFN-γ and TNF-α, have been shown previously to be affected by CsA treatment [7]. Interestingly, we observed such an affect only on the expression of IFN-γ, but not on TNF-α. This might suggest that a more selective immune suppressive medication is sufficient to control the autoimmune features of Omenn. The mRNA expression levels of several genes, such as ICAM 1 adhesion molecule and IL-13–T helper type 2 (Th2) lymphocyte activator, which are known to be expressed highly in various autoimmune diseases [8], were found to be high even after successful CsA therapy, suggesting that their contribution to the autoimmune feature associated with OS is minimal [32,33].

In both patients, large eosinophilia was detected before the immunosuppressive therapy. This is a typical finding in patients with OS and is related to the expanded T cell clones that are found consistently to be predominantly of Th2 type and to secrete IL-4 and IL-13 (which promote immunoglobulin class-switching to IgE) as well as IL-5 (which activates eosinophils) and IL-9 (which activates mast cells) [34]. Interestingly, a recent study [35] showed that despite the prominent eosinophilia, marked activation of eosinophils is not always observed.

It is worth noting that the interpretation of our results may be limited because only one patient was studied, and the low number of his T cells may partially affect the gene expression profile.

In summary, we observed different clinical responses to CsA in two OS patients, which was correlated with the immunological response. Varying clonal expansions in OS patients can cause the autoimmune features and can respond differently to the immunosuppressive therapy; therefore, additional therapy is sometimes indicated. Monitoring the clinical response in OS patients can also be supported by follow-up analysis of the TCR repertoire. The gene expression profile associated with good clinical outcome after CsA in OS may be used to identify a more selective immunosuppressive therapy for such patients.

Acknowledgements

The authors thank the Jeffery Modell Foundation, the Israeli Science Foundation and the Israeli Ministry of Health for their support of Dr Somech. Esther Eshkol is thanked for editorial assistance.

Disclosure

The authors declare no competing financial interests.

Ancillary