Characterization of immune function and analysis of RAG gene mutations in Omenn syndrome and related disorders

Abstract

Omenn syndrome was recently found to be caused by missense mutations in RAG1 or RAG2 gene that result in partial V(D)J recombination activity. Although the clinical hallmarks of the disease are well defined, there have been several cases with clinical findings similar to, but distinct from Omenn syndrome. The data on immune functions and RAG gene mutations of such cases are limited. We described five Japanese infants from four unrelated families, including two cases of Omenn syndrome and three cases of related disorders. Sibling cases with typical Omenn phenotype were found to be compound heterozygotes of R396C and L885R mutations in RAG1. The former has been reported in European cases and may constitute a hot spot. The latter is a novel missense mutation. Infants with related disorders exhibited erythroderma, eosinophilia, hypogammaglobulinaemia, decreased number of B cells and skewing to Th2, and their lymph node specimens showed architectural effacement, lymphocyte depletion and histiocytic hyperplasia, each of which is seen characteristically in Omenn syndrome. However, in these cases serum IgE levels were low or undetectable. We found no mutation in RAG genes except for a K820R substitution in RAG1, which was regarded to be a functional polymorphism, in two of these cases. Our study suggests that RAG missense mutation may be a genetic abnormality unique to Omenn syndrome with characteristic clinical and laboratory findings. Variations of Omenn syndrome, or related disorders, may represent a different type of immunodeficiency, distinct from abnormalities in lymphoid-specific recombinase activity.

INTRODUCTION

Omenn syndrome is characterized by generalized erythematous skin rash, lymph node enlargement, hepatosplenomegaly, increased serum IgE levels and evidence of combined immune deficiency [1,2]. Most of these findings are observed soon after birth or during early infancy. Although the clinical and laboratory findings are well defined, and most cases of Omenn syndrome are diagnosed without difficulty, there exist distinct conditions simulating Omenn syndrome. Some of these cases are due to the engraftment of maternal T cells within the fetus, causing graft-versus-host illness in utero. These conditions are classified as Omenn-like syndrome and can be easily differentiated from true Omenn syndrome once the chimerism is identified within the patient. However, there have been several reported cases with clinical findings similar to Omenn syndrome, sharing some of the features with the illness [3–6]. It has been difficult to draw a clear distinction between Omenn syndrome and such variants in the absence of a known genetic background to explain the immunodeficiency.

Recently, it was found that at least some of the typical cases of Omenn syndrome are the result of missense mutations in RAG1 or RAG2 gene and partial recombinase activity [7]. Total defect of the enzyme function is known to lead to severe combined immune deficiency without mature lymphoid cells [8]. Partial defect with leaky recombinase activity may, on the other hand, lead to variable clinical features reflecting the actual degree of the enzyme defect. It would thus be intriguing to analyse the RAG1 or RAG2 gene mutation in cases with atypical Omenn syndrome, in addition to classical Omenn syndrome, and correlate the type of mutation and the clinical presentations, to delineate the functional significance of RAG1 or RAG2 gene mutation in immune systems.

PATIENTS AND METHODS

Patients

We studied five patients with a diagnosis of Omenn syndrome or related disorders. All patients were born to non-consanguineous parents. The clinical and immunological data of cases 1 and 2, two sibling cases, have already been published [9]. Table 1 summarizes the main clinical and laboratory features of the five cases at the time of diagnosis. Hypogammaglobulinaemia was a universal feature. Serum IgE levels were elevated in patients 1 and 2, but constantly low in patients 3, 4 and 5. Repeated bacterial or fungal infections were present in all patients. Patient 3 underwent bone marrow transplantation at the age of 8 months, and is alive (without evidence of disease) 8 months later. Patient 4 is now 12 months old and is being treated with corticosteroid, cyclosporin A and intravenous immunoglobulin. Patient 5 died of sepsis at 4 months. Heparinized blood samples were collected from patients 3 and 4 after parental informed consent was obtained. Lymph node specimens for histological examination were obtained by biopsy or at autopsy in patients 1, 2, 4 and 5. No blood chimerism could be demonstrated with HLA typing and chromosome analysis in any case.

Table 1.

Clinical and laboratory features of patients

*Control; data (means ± s.d.) from age-matched (1–5 months) controls.

†ND, Not done.

‡NA, Not available.

Immunophenotyping of lymphocyte subset

Surface analysis of lymphocytes was performed by indirect or direct immunofluorescence using the following MoAbs; anti-CD3, anti-CD4, anti-CD8, anti-CD16, anti-CD20, anti-TCRαβ, anti-CD25, anti-HLA-DR, anti-CD45RO and anti-CD95 as previously described [10,11]. The cells were analysed with a Cytoron Absolute flow cytometer (Ortho Clinical Diagnostic Systems, Tokyo, Japan).

Mutation analysis

DNA samples were prepared from paraffin-embedded lymph nodes in cases 1, 2 and 5 with DEXPAT (Takara Shuzo Co., Tokyo, Japan) according to the manufacturer's instructions, and from leucocytes in cases 3 and 4 with a standard method. RAG1 and RAG2 coding sequence, each located on one exon [12], were amplified from genomic DNA using specific primers (Table 2). Because of poor amplification of DNA from paraffin-embedded tissue, RAG1 was amplified in 17 different overlapping segments and RAG2 in 10 segments. The polymerase chain reaction (PCR) conditions were: denaturation at 95°C for 30 s, annealing at 54°C for 30 s and extension at 72°C for 1 min for a total 40 cycles. Sequencing was performed directly on the PCR products purified from the gel with ABI Prism BigDye Terminator Cycle sequencing kit (Perkin Elmer Applied Biosystems, Foster City, CA).

Table 2.

Polymerase chain reaction primers used for RAG gene analysis

Histological and immunohistochemical analysis

Paraffin-embedded lymph node specimens were examined by standard technique as described [9,13]. After blocking with normal goat serum, the sections were stained with appropriate dilutions of anti-CD3 MoAb (YLEM, Rome, Italy), anti-CD45RO (Dako Japan Co. Ltd, Kyoto, Japan), anti-CD79α (Dako) or anti-CD30 (Ber-H2; Dako) for 1 h at room temperature. After washing the slides in Tris buffer, Envision polymer reagent (Dako), alkaline phosphatase and goat anti-rabbit/mouse IgG-conjugated dextran polymer were reacted for 30 min at room temperature. Alkaline phosphatase activity was visualized using Fast Red TR salt (Sigma Chemical Co., St Louis, MO) after further washing of the slides in Tris buffer.

Detection of intracellular cytokine

Peripheral blood mononuclear cells (PBMC) were obtained by Ficoll–Hypaque density gradient centrifugation. The culture medium consisted of RPMI 1640 medium containing 10% fetal calf serum (FCS), 20 mm HEPES, 2 × 10⁻⁵ m 2-mercaptoethanol (2-ME), 0.3 mg/ml l-glutamine, 200 U#x002F;ml penicillin G and 10 mg#x002F;ml gentamycin. PBMC were cultured in 24-well plates (1 × 10⁶/ml) for 4 h at 37°C and 5% CO₂ with or without phorbol myristate acetate (PMA; 10 ng/ml) and ionomycin 0.5 μg#x002F;ml. GolgiStop (PharMingen, San Diego, CA) was also added to the culture. After washing twice in PBS, the cells were fixed and permeabilized with CytoStain Kit (PharMingen) according to the manufacturer's instructions. Then the cells were incubated with PE-conjugated anti-IL-4 and FITC-conjugated anti-interferon-gamma (IFN-γ) MoAbs (PharMingen). After washing, the cells were analysed with a flow cytometer.

Detection of apoptosis

PBMC were cultured for 24 h with or without 1 μg#x002F;ml of anti-Fas MoAb (7C11; Coulter, Hialeah, FL). Cell viability was analysed using Annexin V binding after the culture, as previously described [13]. Briefly, after washing in PBS, the cultured cells were incubated with PE-conjugated anti-CD3 MoAb (Becton Dickinson Immunocytometry Systems, San Jose, CA) and further stained with FITC-labelled Annexin V (Bender MedSystems, Vienna, Austria) for 15 min. Then the percentage of apoptotic cells among CD3⁺ T cells was determined with a flow cytometer.

RESULTS

RAG mutations

As shown in Table 3, patients 1 and 2 were found to be compound heterozygotes of C1298T and T2766G in RAG1 gene, causing R396C and L885R changes, respectively. The former has been reported in two European cases [7] and occurs in the homeodomain of RAG1. The latter is a novel missense mutation and localized within the active core of RAG1. A K820R substitution resulting from A2571G in RAG1 was identified in homozygous state in patient 3, and heterozygous state in patient 4. When we screened for the mutation in the DNA of 10 normal controls, four children were found to be homozygous and two heterozygous. A R249H substitution was found in patient 4. No other mutations in RAG genes were found in patients 3 and 4. Patient 5 showed no mutation in RAG genes either.

Table 3.

RAG gene mutations of patients

Histological and immunohistochemical studies

Histological evaluation showed total effacement of the lymph node architecture with depletion of lymphocytes and accumulation of interdigitating reticulum cells (Fig. 1). Immunohistochemical studies were performed for cases 1, 2, 4 and 5. In all cases examined, eosinophil infiltration, markedly decreased T cells and absence of B cells were observed. Representative illustrations from case 1 are shown in Fig. 2. Areas of eosinophil infiltration were prominent (Fig. 2A,B). Multiple clusters of lymphocytes were seen in the periphery of the lymph node and most of these lymphocytes were CD3⁺CD45RO⁺ T cells (Fig. 2C,D). However, follicle formation was absent and CD3⁺ T cells were sparse in other areas of the lymph node (Fig. 2E). CD79α⁺ B cells were not detectable (Fig. 2F). CD30 expression was examined in these specimens. A lymph node specimen from a Hodgkin's disease was used as a positive control. Typical Hodgkin cells exhibited intense surface and cytoplasmic staining (Fig. 3A). In contrast, only few surface CD30⁺ lymphocytes were found within the lymph node of reactive hyperplasia (Fig. 3B). It was intriguing that localized clusters of CD30⁺ cells were detectable within the lymph node of case 1 at 1 month, but they were not detectable at 2 months when the lymphocyte depletion progressed further (Fig. 3C,D). Autopsy specimens of case 1 at 5 months or the lymph node of case 2, younger brother of case 1, did not reveal any detectable CD30⁺ cells. CD30⁺ cells were not detectable within the lymph nodes from case 4 or 5, either (data not shown).

Fig. 1.

Haematoxylin and eosin staining of lymph node specimens. (A) Control (reactive lymph node) (× 100). (B) Case 1 (× 100). (C) Case 4 (× 100). (D) Case 5 (× 100).

Fig. 2.

Immunohistological findings of lymph node specimen. A paraffin-embedded lymph node specimen from case 1 was stained with haematoxylin and eosin (H–E) (A,B), or with MoAb against CD45RO (C,D), CD3 (E) or CD79α (F). Lower magnifications (× 100, A,C) and higher magnifications (× 400, B,D) are shown for H–E staining and anti-CD45RO staining. Only higher magnifications (× 400) are shown for anti-CD3 and anti-CD79α staining.

Fig. 3.

Evaluation of CD30 expression by lymph node cells. Paraffin-embedded lymph node specimens were stained with anti-CD30 MoAb (Ber-H2). Sections were obtained from Hodgkin's disease (A), reactive hyperplasia (B), case 1 at 1 month (C) and at 2 months (D).

Lymphocyte phenotype and proliferative responses

Consistent with previous observations [14–17], the activated phenotypes, HLA-DR and CD25, were expressed on a large fraction of T cells. Moreover, expression of CD45RO and Fas#x002F;CD95 was observed on a large number of T cells from patients 3 and 4. Two-colour analysis showed that CD95 expression was found on the majority of CD3⁺ T cells (Table 1). The number of B cells was markedly decreased in all cases except case 4. In case 4 it was initially normal, but progressively declined below 1% after 3 months. Lymphocyte proliferative responses were markedly reduced in cases 1, 2 and 5.

Cytokine expression and apoptosis of T cells

As shown in Fig. 4, PBMC from patients 3 and 4 produced much higher levels of IL-4 and lower levels of IFN-γ than those from a normal control in response to PMA and ionomycin. These findings were similar to those seen in Omenn syndrome [18–20]. In these cases, marked IL-4 production was observed even when cells were examined without stimulation (data not shown).

Fig. 4.

Expression of IL-4 and IFN-γ. Peripheral blood mononuclear cells (PBMC) were cultured for 4 h at 37°C with phorbol myristate acetate (PMA) and ionomycin. After permeabilization, the cells were stained with PE-conjugated anti-IL-4 and FITC-conjugated anti-IFN-γ MoAbs. Intracellular cytokine profiles were determined with a flow cytometer. *Eosinophil count was 90#x002F;μl and serum IgE level was 145 U#x002F;ml. †There was marked eosinophilia (13 560#x002F;μl) and elevated serum IgE (7050 U#x002F;ml).

Accelerated apoptosis of T cells was seen in patients 3 and 4 when they were cultured for 24 h with medium alone (Fig. 5). In these patients, most T cells expressed CD95#x002F;Fas, resulting in enhanced apoptosis after treatment with 1 μg#x002F;ml of anti-Fas MoAb. These findings were also similar to those observed in Omenn syndrome [17]. Addition of IL-2 partially abrogated lymphocyte cell death in both unstimulated and Fas-stimulated cultures (data not shown).

Fig. 5.

Accelerated apoptosis of T cells from patients. Peripheral blood mononuclear cells (PBMC) were cultured for 24 h with (▪) or without (□) 1 μg#x002F;ml of anti-Fas MoAb. The percentage of apoptotic cells in T cells was analysed using FITC-conjugated Annexin V with a flow cytometer. The bars indicate the means ± s.d. of triplicate cultures.

DISCUSSION

The diagnosis of Omenn syndrome has been based on typical clinical and laboratory findings, such as early onset generalized erythroderma, lymphadenopathy, hepatosplenomegaly, eosinophilia, hypogammaglobulinaemia with elevated serum IgE, absence of circulating B cells and presence of activated, anergic T cells [2]. These characteristic features and the hereditary nature lead to the diagnosis without much difficulty. Similar findings caused by graft-versus-host illness can be ruled out by proving the presence of circulating allogeneic T cells. However, sporadic cases have been reported [21–23], and mild variant or related disorders without maternal T cell engraftment have also been described [3–6]. These cases sometimes pose diagnostic problems, especially when some of the cardinal features are missing. In particular, there exist a few atypical cases without elevated serum IgE, eosinophilia or decreased numbers of B cells, which were nevertheless diagnosed as Omenn syndrome [17,20,24,25]. Our cases 3, 4 and 5 also presented with extremely low serum IgE levels, and we regarded these cases as related disorders rather than Omenn syndrome. Relatively favourable responses to various therapeutic modalities and ultimately good prognosis may function as diagnostic aids to differentiate these cases from Omenn syndrome. We do not know, however, if these related disorders are a different clinical entity, based on distinct genetic abnormalities, or are only variations of a similar illness with a common pathophysiological background.

Recently Villa et al. reported on the mutational analysis of seven Caucasian patients with typical features of Omenn syndrome and found mutations in both alleles of RAG1 or RAG2 gene [7]. However, no data on immune functions and RAG gene mutations of a typical case of Omenn syndrome or related disorders are yet available. This prompted us to analyse the various abnormalities of RAG genes in the related disorders, if present, and evaluate their relevance to the clinical pictures. Two sibling cases, cases 1 and 2, had not only the typical clinical features of Omenn syndrome, but also missense mutations in each allele of RAG1 gene. One of the mutations, a R396C substitution, has been reported in two European cases, and it has been suggested that R396 residue is critical in DNA binding and recombination. Thus, this particular mutation may constitute a hot spot, regardless of the ethnic background. Another mutation, L885R, is a novel missense mutation, which is located within the active core of RAG1. Although we did not perform a functional analysis to see if this particular mutation is important, we regard this as significant because it was found in both of the sibling cases, and no other mutation was found in their RAG1 gene. A K820R substitution, found in both alleles of patient 3 and in an allele of patient 4, was also found in apparently normal individuals, and the frequency of the mutation was very high (up to 50%). Although no similar mutation has been described yet in the literature, it is most likely that this is a novel, functional polymorphism and will not lead to RAG1 enzyme dysfunction. Another mutation found in an allele of RAG1 gene in patient 4, R249H, has been previously reported as a polymorphism. Patient 5 did not have any mutation in RAG1 or RAG2 gene. Taken together, significant mutations were found only in typical cases of Omenn syndrome, but none of the cases with the related disorders has mutations within RAG1 or RAG2 gene.

Patient 3 had neither lymphadenopathy nor hepatosplenomegaly, and the disease onset was relatively late at 5 months. However, we included this case as one of the related disorders because some of the clinical features were reminiscent of Omenn syndrome. In addition to erythematous rash, low serum immunoglobulin and absent B cells, peripheral blood T cells had activated phenotype. They produced significant amounts of IL-4 but scarce IFN-γ upon stimulation, and were also prone to accelerated apoptosis. All of these immunological characteristics are also typical features of Omenn syndrome [18–20]. In marked contrast, a patient who had extreme atopy with eosinophilia and elevated serum IgE concentrations did not exhibit the abnormal cytokine profiles seen in the related disorders. Our data clearly show that Omenn syndrome, with RAG1 or RAG2 missense mutation, is not the sole cause of T cell immune dysfunction with extreme polarization toward Th2 function. However, these changes may reflect a common, but unique pathophysiology shared by Omenn syndrome and the related disorders, but not by atopic illnesses.

It has been recently reported that the majority of infiltrating T cells were expressing CD30 on the surface, supporting the hypothesis that predominant proliferation of Th2-type T cells plays a significant role in the pathogenesis of the characteristic symptoms [26]. Immunohistochemical analysis of lymph node specimens was available only for cases 1, 2, 4 and 5. They shared common histological features, including effacement of normal lymph node architecture, loss of B cells, infiltrations of CD45RO⁺ activated T cells and progressive depletion of lymphocytes. We could detect multiple small foci of CD30⁺ lymphocytes within a biopsy specimen of case 1, but it was not detectable within the second biopsy specimen performed 1 month later or within the lymph node of his younger sibling. The reason why we could not detect massive proliferation of CD30⁺ lymphocytes within the lymph nodes is not clear. Predominant infiltration of CD30⁺ lymphocytes may be apparent only at certain times in the disease course and may regress as the lymph node becomes progressively depleted of lymphocytes.

Extremely high levels of serum IgE in the absence of circulating B cells are characteristic of Omenn syndrome [2], although the reason for this apparent discrepancy is not known. Enhanced production of IL-4 and IL-5 is surmised to be responsible for the increased IgE synthesis by B cells. These B cells are thought to be located within sites other than peripheral circulation or lymph node, such as gastrointestinal tract. Serum IgE concentrations were constantly low in cases 3 and 4, although peripheral blood T cells produced IL-4 significantly. The marked differences in serum IgE concentration may be due to the critical differences in pathogenesis between Omenn syndrome with RAG1 or RAG2 gene mutation and the related disorders. Further analysis is required to draw any conclusion as to the significance of serum IgE levels in distinguishing Omenn syndrome among the constellation of similar illnesses.

RAG1 or RAG2 mutation may result in variable clinical manifestations, including disorders with only subtle immune dysfunction. At the same time, Omenn-related disorders or variations may have defects in some other proteins involved in recombination, or they may represent different types of immunodeficiency, totally independent of V(D)J recombination machinery. Further analysis of RAG1 and RAG2 missense mutations in Omenn syndrome and related disorders may lead to the elucidation of the cardinal clinical features of RAG1 or RAG2 mutations and reclassification of the related disorders with combined immunodeficiency.