| Human Molecular Genetics | Pages |
Genetic and molecular definition of complementation group D in MHC class II deficiency
Introduction
Results
Classification of patients into complementation groups
Correction of HLA class II expression by transfection with the RFXAP cDNA
Mutations of the RFXAP gene in patients SS, AkO, ShA and ShG
MHC class II, DMB and Ii chain gene expression in complementation group D
Discussion
Materials And Methods
Patients
Cell culture
Immunofluorescence
Somatic complementation analysis
Transfections
PCR amplification and sequencing of RFXAP
RT-PCR detection of HLA class II, DMB, Ii and GADPH gene expression
Abbreviations
Acknowledgements
References
Genetic and molecular definition of complementation group D in MHC class II deficiency
Four complementation groups, A, B, C and D, have been described among cell lines defective in the coordinate expression of MHC class II genes. These include cell lines established from patients affected with MHC class II deficiency and experimentally generated mutant cell lines. Group D, in contrast to the other groups, was for a long time represented only by the 6.1.6 mutant cell line. The gene responsible for the defect in this group, RFXAP, recently was cloned and found to be mutated in the 6.1.6 cell line and in three patients. Here we report fusion experiments in several new HLA class II-deficient patients, completing the classification of the majority of known patients into the four complementation groups. Patients from five unrelated families were classified in complementation group D, while nine others fall into complementation groups A and B. None of the patients defined a new complementation group. Full correction of MHC class II expression was obtained in cells from patients belonging to group D by transfection with the RFXAP cDNA. The RFXAP coding region was found to be mutated in all patients. Mutations were found to be recurrent since only three different mutations have been found in the eight unrelated families reported to date.
INTRODUCTION
MHC class II antigen expression is tightly controlled in a cell-specific manner (1) and is essential for antigen-specific immune responses (2). Mechanisms which control transcription of MHC class II genes have been studied extensively, particularly by using cells deficient in HLA class II expression. These cells are either experimentally generated mutants, or cells from patients affected with MHC class II deficiency (3). This primary immunodeficiency resulting from defective expression of HLA class II antigens is an autosomal recessive disorder characterized by an extreme susceptibility to infections, and an absence of cellular and humoral T-cell responses upon antigen recognition (4). Cells from the patients do not transcribe [alpha] and [beta] MHC class II genes encoding the DR, DQ and DP HLA isotypes (3,5). Intrafamilial segregation analysis has shown that the genetic abnormality is localized outside the MHC region and involves trans-acting regulatory factor(s) (5). Both patient cell lines and experimental mutant cell lines were used, in somatic fusion experiments, to evaluate the number of transcription factors that were absolutely required to ensure normal MHC class II expression. Four complementation groups were identified, A, B, C and D (6,7).
The genes responsible for the defect in groups A and C were cloned by complementation cloning in the RJ 2.5.5 cell line (group A) and the SJO cell line (group C), respectively. The CIITA gene encodes a non-DNA-binding protein that controls both constitutive and inducible HLA class II expression and is mutated in group A patients (8). The RFX5 gene encodes a 75 kDa subunit of the RFX complex and is mutated in patients from complementation group C (9). A third gene, RFXAP (for RFX-associated protein), was cloned recently by Durand et al. (10). RFXAP encodes a small 36 kDa subunit of the RFX complex, and mutations of the gene are responsible for the defect in complementation group D. The RFX5 and RFXAP proteins form, together with a third protein of 41 kDa, a multimeric complex RFX which binds to the X box region of MHC class II promoters (11). It is possible that a gene encoding this third protein is mutated in complementation group B, in which the affected gene has not yet been identified. The binding of the RFX complex to DNA depends on cooperative binding with at least two pleiotropic factors, NFY and X2BP (12-14), but is not sufficient for class II transcription. CIITA must combine with RFX and the other transcription factors that bind to the X2 and Y boxes of the promoter to allow transcription of MHC class II genes.
We describe here the somatic complementation analysis of 13 new HLA class II deficiency patients and two previously reported patients SS (15) and ABI (16), suspected to represent different complementation groups. One of the new patients was found to belong to group A and seven others to group B. However, four new patients fell into the same complementation group as patients SS and ABI. The latter group was found to represent group D by complementation analysis, transfection with the RFXAP cDNA and mutation analysis of the RFXAP gene.
RESULTS
Classification of patients into complementation groups
Somatic cell fusion experiments were performed using either B cells or fibroblasts representing complementation groups A, B, C and D. Both B cells and fibroblasts were available only for patient ZM. Figure 1. shows an example of HLA expression obtained in heterokaryons of both B cells and fibroblasts. In fibroblast heterokaryons, 3-10% of HLA class II-positive cells were detected 48 h after interferon-[gamma] (IFN-[gamma]) treatment. In B-cell heterokaryons, no more than 5% HLA class II-positive cells could be detected. Table 1 summarizes the data obtained by somatic cell fusions, and Table 2 gives the classification of the newly studied patients with respect to previously published data (6,7). One patient of Pakistani origin was classified into group A, seven patients were placed in group B and five were classified in group D. None of the patients belonged to group C.
Figure 1. HLA-DR expression at the surface of heterokaryons obtained between cells from groups D and B (a and b) or D and A (c and d), 48 h after fusion. (a) and (b) show B-cell heterokaryons, ZM×ABL; (c) and (d) show fibroblast heterokaryons, ZM×HVM, induced (d) or not (c) by IFN-[gamma]. [Igr]n (b), (c) and (d), cells were stained with anti-class II monoclonal antibody Bu-27. Table 1
Group
Cell line
Cell type
HVJ
AABL
BSJO
C6.1.6
DSS
ZM
AkO
ShA
ShG
ABI
A
HVJ
B, F
-
B
ABL
B, F
+
-
C
SJO
B
+
+
-
D
6.1.6
B
+
+
-
nd
SSa
F
+
+
nd
nd
-
ZM
B, F
+
+
+
-
-
-
AkO
F
+
+
nd
nd
-
-
-
ShA
F
+
+
nd
nd
-
-
nd
-
ShG
F
+
+
nd
nd
-
-
nd
-
-
ABIb
F
+
+
nd
nd
-
-
-
nd
-
-
Classified in group:
D
D
D
D
D
D
D
Table 2
| A | B | C | D | |
| `Founder' cell line | Patient BLS2RJ 2.5.5 exp mutant | Patient BLS1 | Patient SJO | 6.1.6 exp. mutant |
| Affected gene | CIITA | RFX5 | RFXAP | |
| Patients | 6a | 22b | 3 | 8c |
| families | 5 | 20 | 3 | 6 |
| ethnic origin | Spanish, 4Pakistani, 1 | North African, 15Italian, 2 Turkish, 2Saudi Arabian, 1 | North American, 1Dutch, 1 | Druze 2North African 2Turkish 2 |
| consanguinity | 2/4 | 15/19 | 1/3 | 4/5 |
As Table 1 shows, fibroblasts from four patients, ZM, AkO, ShA and ShG, complemented cells representing complementation groups A and B but did not complement the SS fibroblast cell line, which had not been classified (15). ZM B cells complemented B cells from groups A, B and C but did not complement the HLA class II-negative 6.1.6 B cell line defining group D. No complementation was obtained after fusion between ZM, AkO, Sha and ShG fibroblasts. In addition, none of these fibroblast cell lines, incuding ZM, complemented cell line ABI (16). Thus, cells from patients SS, AkO, ShA, ShG ABI and ZM all belong to complementation group D.
Correction of HLA class II expression by transfection with the RFXAP cDNA
As previously reported, in transient transfection experiments with the RFXAP cDNA, a correction of MHC class II expression was obtained in ZM and ABI fibroblasts (17). Figure 2. depicts full correction of constitutive HLA class II gene expression in cells from patient ZM, the only patient for whom a B-cell line was available, following transfection with the RFXAP cDNA. In contrast, transfection with CIITA or RFX5 cDNAs did not restore HLA class II expression; neither did transfection with the empty vector. Similar results were obtained for IFN-[gamma]-induced MHC class II expression in fibroblasts from patients AkO, SS and ShA (data not shown).
Figure 2. Correction of membrane HLA class II expression by transfection with RFXAP cDNA. Seven (left) or twenty two (right) days after transfection with pREP4-RFXAP (p36) or pREP4 (vv) plasmids, ZM (group D) and Kh (group A) cells were treated with IFN-[gamma], stained 48 h later for HLA class II expression and analysed by flow cytometry (FACScan).
Mutations of the RFXAP gene in patients SS, AkO, ShA and ShG
Full-length RFXAP cDNAs were amplified by PCR from patients SS, AkO, ShA and ShG, as described (10), subcloned in a pGEM vector and sequenced. In all four patients, mutations in the RFXAP cDNA were identified. These mutations were then confirmed by direct sequencing of genomic PCR products from the patients and their parents.
In family SS, with two affected children, a homozygous deletion of a G at nucleotide 484 was found (484delG) (Fig. 3). The same mutation was detected in family ZM (17). The resulting frameshift leads to an out-of-frame stop codon at nucleotide 525. The parents were found to be heterozygous for this mutation.
Figure 3. Mutation analysis of RFXAP in patient SS. PCR products from the patient and the parents were sequenced. In family AkO, a substitution of a glutamine codon by a stop codon (CAG->TAG) at position 279 of the RFXAP gene was detected (C279X) (Fig. 4). This mutation is predicted to lead to a truncated protein of 52 amino acids. The same mutation has been detected previously in patient ABI (17). The parents of AkO are heterozygous for this mutation. Figure 4. Mutation analysis of RFXAP in patient AkO. PCR products from the patient and the parents were sequenced. In family Sh, with two affected first degree cousins ShA and ShG, an insertion of seven nucleotides `GCGGGCG' at position 151 of the RFXAP cDNA was found. This mutation is designated 151ins7. It represents a duplication of the wild-type sequence spanning nucleotides 144-150. It leads to a frameshift resulting in a stop codon at nucleotide 329. Both patients from this family inherited this homozygous mutation (Fig. 5). The parents of both patients were heterozygous (data not shown). Figure 5. Mutation analysis of RFXAP in patients ShA and ShG. PCR products from both patients were sequenced.
MHC class II, DMB and Ii chain gene expression in complementation group D
Since residual expression of DRA and DPB genes had been reported in cells from patient ABI (16), MHC class II gene expression was examined in group D patients by RT-PCR. A variable level of transcription was observed for DPA and/or DPB genes in four of the patients, but not in a control patient from group A (Fig. 6). However, even in patient ShG cells, in which both DPA and DPB transcripts were detected, there was no detectable membrane expression of the DP[alpha][beta] complex (Fig. 7). In none of the patient cell lines were HLA DR, DQ, DMB or Ii chain gene mRNA detected by RT-PCR (Fig. 6).
Figure 6. Expression of MHC class II, DMA, Ii and GADPH genes. Radioactive RT-PCR was performed using RNA from IFN-[gamma]-induced patient and control fibroblasts. In the last lane (KhA), the RNA used was prepared from cells of a patient classified in complementation group A. Figure 7. HLA-DR and HLA-DP expression at the surface of fibroblasts from control and patients ShA and ShG treated for 48 h with IFN-[gamma]. Cells were stained with anti-DR 112 and anti-DP 227 mAbs, respectively.
DISCUSSION
Despite the growing number of patients examined and the pleiotropic factors shown to be involved in the control of MHC class II expression, the number of complementation groups, and therefore factors mutated in MHC class II immunodeficiency patients, remains at four (Table 2). With the 13 new HLA class II-deficient patients studied in this work, the majority of reported patients have now been classified into complementation groups and none defines a fifth complementation group. Altogether, of the 39 patients from 34 families, there are 6 in complementation group A, 22 in group B, 3 in group C and 8 in group D. Two brothers with a partial defect of MHC class II expression and with a mild form of immunodeficiency recently were classified in an additional complementation group (18). These atypical patients probably represent another syndrome, since they are able to respond to vaccination (19).
Among the immunodeficient patients classified in this report, only family Kh, with two affected siblings, belongs to group A. While previously described group A patients have been of Spanish descent (15), these new cases were of Pakistani origin, showing that CIITA mutations are not restricted to one ethnic origin. Seven of the new patients fall into group B and are of North African, Italian, Turkish and Saudi Arabian descent. Altogether, including the BLS1 founder patient, there are 22 families in group B.
Four new patients in three families, together with patient SS (15), fall into group D. They are of North African, Turkish and Druze descent. Their clinical and biological features did not differ from other HLA-deficient patients. However, in all but one, transcription of DPA and/or DPB genes could be detected. Residual DPA and DPB transcription was also reported in patient ABI (16). In the patients studied in this work, despite the presence of DP transcripts, mature DP[alpha]-[beta] heterodimers were not detected at the cell surface, perhaps because the Ii and DM genes are not expressed. Thus, no functional consequences are to be expected from this leakiness in the defect, as is indeed observed in vivo.
Three different lines of evidence demonstrate that these patients belong to complementation group D, i.e. somatic cell fusion analysis, transfection of the patients' cells with the RFXAP cDNA and mutational analysis. Unambiguous cell fusion results were obtained by using both fibroblasts and B-cell lines, as demonstrated for patient ZM. The absence of complementation between ZM cells and the 6.1.6 B cell line, on one hand, and between ZM and SS fibroblasts, on the other, validated all other fusion data, including those of the ABI fibroblasts. Previous misclassification of this patient into group `E' (16) was probably due to the leakiness of the 6.1.6 cell line used and/or to fusion experiments performed across cell lineages (B-fibroblast cell fusions). A full correction of HLA class II expression in cells from group D, and not from the other groups, by transfection of the RFXAP cDNA confirmed previously reported data from transient transfection experiments (10,17). The RFXAP cDNA was found to be mutated in patients SS, AKO, Sh and ShG, as also previously shown for ZM and ABI (17).
Three types of recurrent mutations of RFXAP were found, i.e. substitution, deletion and insertion (Fig. 8). A very high CG content in the RFXAP gene leading to polymerase errors could account for the generation of these mutations (20). The C279X transition detected in patient AKO is identical to the mutation previously detected in patient ABI (17) and the 484delG mutation in patient SS is identical to that detected in patients DA and ZM (10,17). In both cases, the mutations are predicted to give rise to a frameshift and to the synthesis of severely truncated proteins of 52 and 136 amino acids, respectively. A previously undiscovered mutation was found in the Sh cousins, i.e. an insertion of duplicated nucleotides 144-150: `GCGGGGC'. This mutation also results in a frameshift and a truncated product. All mutations were detected within a region spanning nucleotides 116-540 of the RFXAP cDNA, which correspond to a single exon.
Figure 8. Schematic representation of the RFXAP gene indicating the three mutations found in patients from six unrelated group D families. Therefore, only three different mutations of the RFXAP gene have been observed so far in six different families, including patient DA (10). The ethnic origin of the families could indicate an ancestral origin of these mutations since patients SS, RA and ZM are of North African origin, while patients AKO and ABI are of Turkish descent. However, remarkably, only in group D of the MHC class II deficiency have such recurrent mutations been detected thus far. In group A, for example, despite a high prevalence of patients of Spanish descent (15), no recurrent allele has been detected among the different patients (8, and B. Lisowska-Grospierre et al., unpublished). In all complementation groups, most of patients come from the Mediterrenean area, where consanguineous marriages are frequent. Why RFXAP gene mutations, probably resulting in non-functional proteins, are associated with residual DPA/B gene transcription is not understood. This observation provides evidence for a limited but significant dyscoordinate regulation of MHC class II gene transcription regulation. These results require further study to determine a putative specific role of RFXAP in DRA/B, DQA/B versus DPA/B transcription.
MATERIALS AND METHODS
Patients
Thirteen hitherto undescribed patients from eleven unrelated families affected with MHC class II immunodeficiency were studied. Eleven families were consanguineous. They were of North African (three), Italian (two), Turkish (three), Druze (two), Pakistani (two) and Saudi Arabian (one) origin. In all patients, HLA class II antigens were undetectable on the cell surface of B cells, monocytes and activated T cells. Clinical presentation in all was characterized by recurrent infections, severe diarrhoea and failure to thrive. The absence of antigen-specific humoral and cellular responses was observed in all, and hypogammaglobulinaemia in several of the patients. In this report, detailed studies are reported on the cells of patients, AkO, ShA, ShG , ZM. For ZM, partial results were presented previously (17). Patient SS has already been described (15). Characteristics of patient ABI, included in somatic complementation experiments, have already been reported (16).
Cell culture
B-cell lines were established from patients' B cells by Epstein-Barr virus (EBV) infection. Patients' skin fibroblasts were obtained by skin biopsy, and transformed with SV40, as described previously (15). All cell lines were grown at 37°C in 5% CO2 in RPMI-1640 (Gibco BRL Lifesciences) supplemented with 20% heat-inactivated fetal calf serum. Fibroblasts or their heterokaryons were treated by IFN-[gamma] (200 IU/ml) for 40 h before analysis of MHC class II expression.
Immunofluorescence
Anti-HLA-DR antibody 20.6 (21), anti-class II Bu27 (The Binding Site, Birmingham, UK) and anti-class I antibody W6/32 (ServaLab) were used. B cells were stained in suspension, while fibroblasts were fixed with ethanol prior to incubation with antibodies, as described (15). Cell suspensions were analysed with a Beckton Dickinson cytofluorograph, and fixed cells with a Leitz Ortoplan microscope.
Somatic complementation analysis
Fibroblast and B-cell lines from the newly established patients as well as experimental cell lines (HVJ, ABL, SJO and 6.1.6.) from patients previously classified to complementation groups A, B, C and D, respectively, were used in these experiments. The 6.1.6 cell line was a geneticin-selected fully negative MHC class II variant, kindly provided by C. Alcaide (Institut Pasteur). In addition, fusion experiments included: patient SS cell line, representing a putative complementation group X (15), and patient ABI cell line, reported to represent group E (16), kindly provided by P. van den Elsen. Transient heterokaryons of both cell types, B cells and fibroblasts were obtained by the electrofusion method described in detail previously (15). Phenotypic reversion was tested by immunofluorescence or by northern blot 48-72 h after cell fusion. In the case of fibroblasts, the cells were cultured in the presence or absence of 200 IU/ml of recombinant IFN-[gamma] (Genex).
Transfections
The plasmids used for transfection were pREP4 (Invitrogen), pREP4-RFXAP and EBO-Sfi (empty vector) or EBO-CIITA, kindly provided by V. Steimle (8). Transfections of B-cell lines and fibroblasts were carried out as described (22) with 1-10 µg of each plasmid, using a JOUAN GTH 128/A electropulser, as described (15). Stable transfectants were generated by selection with hygromycine B (250 µg/ml; Calbiochem). Transfected fibroblasts were treated with 200 IU/ml of IFN-[gamma] before testing for MHC class II expression.
PCR amplification and sequencing of RFXAP
Full-length RFXAP cDNA clones from patients were isolated by RT-PCR and subcloned into pGEM-T vector (Promega), according to standard procedures, and analysed for mutations. Mutations were then confirmed in the DNA of patients and their parents by studying PCR-amplified genomic DNA fragments spanning nucleotides 116-540, as described (17). PCR was performed with the Expand High Fidelity PCR System (Boehringer Mannheim), as described (10). Sequencing of recombinant plasmids was performed using the ABI PRISM dye terminator cycle sequencing kit (ABI) and an Applied Biosystems DNA sequencer. Genomic PCR products were cycle sequenced directly by using radiolabelled sequencing primers and a ThermoSequenase kit (Amersham), and analysed on a sequencing gel.
RT-PCR detection of HLA class II, DMB, Ii and GADPH gene expression
Total cytoplasmic RNA was extracted from IFN-[gamma] induced fibroblasts by the guanidium thiocyanate method (23). First strand cDNA was synthesized with avian myeloblastosis virus (AMV) reverse transcriptase (Boehringer Manneheim, SA) using 10 ng of oligo(dT) as a primer and 5 µg of cytoplasmic RNA. PCR amplification was performed in a final volume of 50 µl using 1/30 of the cDNA, 1 µM of each primer, 200 µM of each d NTP, 1.25 IU of AmpliTaq polymerase (Perkin-Elmer, Forster City, CA) and 0.1 µCi [[alpha]-32P]dCTP (Amersham) in each sample. The DNA was denatured at 94°C for 1 min, annealed for 1 min at 55°C to detect all genes except GADPH and [alpha]DR, for which a 58°C annealing temperature was used. Extension at 72°C was for 1 min. This cycle was repeated 25 times for GADPH and [alpha]DR and 27 times for the other genes, followed by 10 min elongation at 72°C. Samples were electrophoresed on 5% acrylamide gels at 150 V for 2 h. Gels were fixed and dried before exposure in a PhosphorImager box (Molecular Dynamics) or to X-Omat film (Kodak). The sequences of oligonucleotide primers used in PCR amplification were: for HLA-[alpha]DQ, -[beta]DQ and DMB as described by Peijnenburg (16), for GADPH, -[alpha]DR and Ii chain as described by Vedrenne et al. (24), and for -[beta]DR, -[alpha]DP and -[beta]DP as described by Hauber et al. (25).
ABBREVIATIONS
B-cell line, B-lymphoblastoid cell line; IFN-[gamma], interferon [gamma]; MHC, major histocompatibility complex; RT-PCR, reverse transcription-polymerase chain reaction.
ACKNOWLEDGEMENTS
We specially thank Drs Peter van den Elsen for allowing us to include the ABI cell line in fusion experiments. We are grateful to Drs Andrew Cant and Mario Abinun for blood samples and skin biopsies from patients treated in their Unit for BMT, and to Dr Klaus Schwarz for sending us the blood samples from two patients. We also thank Françoise Selz for her expert collaboration and E. Barras for excellent technical assistance.
REFERENCES
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 4 Apr 1998
Copyright© Oxford University Press, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M. Peretti, J. Villard, E. Barras, M. Zufferey, and W. Reith Expression of the Three Human Major Histocompatibility Complex Class II Isotypes Exhibits a Differential Dependence on the Transcription Factor RFXAP Mol. Cell. Biol., September 1, 2001; 21(17): 5699 - 5709. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Wiszniewski, M.-C. Fondaneche, F. Le Deist, M. Kanariou, F. Selz, N. Brousse, V. Steimle, G. Barbieri, C. Alcaide-Loridan, D. Charron, et al. Mutation in the Class II trans-Activator Leading to a Mild Immunodeficiency J. Immunol., August 1, 2001; 167(3): 1787 - 1794. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A.-M. Lennon-Dumenil, M.-R. Barbouche, J. Vedrenne, T. Prod'Homme, M. Bejaoui, S. Ghariani, D. Charron, M. Fellous, K. Dellagi, and C. Alcaide-Loridan Uncoordinated HLA-D Gene Expression in a RFXANK-Defective Patient with MHC Class II Deficiency J. Immunol., May 1, 2001; 166(9): 5681 - 5687. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Nekrep, N. Jabrane-Ferrat, and B. M. Peterlin Mutations in the Bare Lymphocyte Syndrome Define Critical Steps in the Assembly of the Regulatory Factor X Complex Mol. Cell. Biol., June 15, 2000; 20(12): 4455 - 4461. [Abstract] [Full Text]
|
|
|
|
|
|
J. Villard, M. Peretti, K. Masternak, E. Barras, G. Caretti, R. Mantovani, and W. Reith A Functionally Essential Domain of RFX5 Mediates Activation of Major Histocompatibility Complex Class II Promoters by Promoting Cooperative Binding between RFX and NF-Y Mol. Cell. Biol., May 15, 2000; 20(10): 3364 - 3376. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||