Uncoordinated HLA-D Expression in a RFXANK-Defective Patient with MHC Class II Deficiency

This information is current as Ana-Maria Lennon-Duménil, Mohamed-Ridha Barbouche, of September 24, 2021. Jocelyn Vedrenne, Thomas Prod'Homme, Mohamed Béjaoui, Salma Ghariani, Dominique Charron, Marc Fellous, Koussay Dellagi and Catherine Alcaïde-Loridan J Immunol 2001; 166:5681-5687; ;

doi: 10.4049/jimmunol.166.9.5681 Downloaded from http://www.jimmunol.org/content/166/9/5681

References This article cites 48 articles, 19 of which you can access for free at: http://www.jimmunol.org/content/166/9/5681.full#ref-list-1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists by guest on September 24, 2021 • Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2001 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Uncoordinated HLA-D in a RFXANK-Defective Patient with MHC Class II Deficiency1

Ana-Maria Lennon-Dume´nil,2,3* Mohamed-Ridha Barbouche,2† Jocelyn Vedrenne,2‡ Thomas Prod’Homme,‡ Mohamed Be´jaoui,§ Salma Ghariani,† Dominique Charron,‡ Marc Fellous,* Koussay Dellagi,† and Catherine Alcaı¨de-Loridan4‡

We describe the analysis of a patient, JER, presenting classical immunological features of MHC class II deficiency. Unexpectedly, some HLA transcripts (HLA-DRA, HLA-DQA, and HLA-DMA) were found to be expressed in the JER cell line at nearly wild-type levels, while HLA-DPA and the HLA-D ␤-chain transcripts were not detected. Gene reporter experiments confirmed the differ- ential transcriptional activities driven by the HLA-D promoters in the JER cells. A defect in RFXANK was first suggested by genetic complementation analyses, then assessed with the demonstration of a homozygous mutation affecting a splice donor site downstream exon 4 of RFXANK. Because the severe deletion of the resulting cannot account for the expression of certain Downloaded from HLA-D , minor alternative transcripts of the RFXANK gene were analyzed. We thereby showed the existence of a transcript lacking exon 4, encoding a 28-aa-deleted protein that retains a transcriptional activity. Altogether, we characterize a new type of mutation in the RFXANK gene in a MHC class II-defective patient leading to an uncoordinated expression of the HLA-D genes, and propose that this phenotype is ensured by severely limited amounts of an active, although truncated RFXANK protein. The Journal of Immunology, 2001, 166: 5681–5687. http://www.jimmunol.org/

he MHC class II molecules (HLA-DR, HLA-DP, and The HLA-DR, HLA-DQ, and HLA-DP genes are coordinately HLA-DQ) present antigenic peptides to CD4ϩ T lympho- expressed in B lymphocytes, and this coordinated regulation is T cytes, and represent key elements in the onset of the cel- ensured at the transcriptional level by conserved promoter motifs, lular and humoral immune response (1). The crucial role of these the W, X1, X2, and Y boxes. Similar motives were found as well molecules is clearly illustrated by a severe immunodeficiency that in the promoters of Ii (invariant chain) and HLA-DM, whose prod- manifests upon their defective expression (2). MHC class II-defi- ucts are also involved in Ag presentation (6). The Y box of MHC cient patients suffer from recurrent bacterial and viral infections class II promoters binds NFY5 (nuclear factor binding to Y box),

appearing at an early age, and generally present reduced levels of a ubiquitous that includes three subunits, by guest on September 24, 2021 serum Igs and CD4ϩ T lymphocytes, in addition to the lack of in NFY-A, NFY-B, and NFY-C (7–9). The factor binding the X2 vitro T lymphocyte proliferation in response to Ags (3). This pa- box, X2BP, was recently characterized as the homodimeric cAMP thology is caused by autosomal recessive genetic defects, unlinked response element-binding protein (CREB) (10). MHC class II de- to the HLA genes themselves, but rather affecting regu- ficiency group B, C, and D cells are defective in the binding of the lating their transcription. Fusion experiments with cells from pa- RFX (regulatory factor binding to X box) complex to the X1 box tients suffering from this immunodeficiency have evidenced four (11, 12). The RFX complex is a heterotrimer, composed of RFX- complementation groups (A, B, C, and D), demonstrating the ge- ANK (13, 14), RFX5 (15), and RFXAP (RFX-associated protein) netic heterogeneity of the disease (4, 5). (16), respectively mutated in groups B, C, and D (13, 14, 17–19). Several publications demonstrate that the RFX, NFY, and X2BP

*Institut National de la Sante´et de la Recherche Me´dicale Unité 276, Institut Pasteur, complexes bind cooperatively to MHC class II promoters, and that Paris, France; †Laboratoire d’Immunologie, World Health Organization Collaborative their interaction increases the stability of the complexes to the Center for Research and Training in Immunology, Institut Pasteur de Tunis, Tunis, promoters (20, 21). More recently, RFX5 was shown to create a Tunisie, France; ‡Institut National de la Sante´et de la Recherche Me´dicale U396, Centre de Recherches Biome´dicales des Cordeliers, Paris, France; and §Centre de bridge in between NFY, RFXAP, and RFXANK (22). Functional Greffe de Moelle Osseuse, Tunis, Tunisie, France complementation of group A mutants has led to the identification Received for publication July 11, 2000. Accepted for publication February 22, 2001. of the CIITA gene (class II transactivator) (23). It encodes a non- The costs of publication of this article were defrayed in part by the payment of page DNA-binding protein, in agreement with data demonstrating that charges. This article must therefore be hereby marked advertisement in accordance MHC class II promoters are normally occupied in group A mutants with 18 U.S.C. Section 1734 solely to indicate this fact. (11, 24). The transactivating capacity of CIITA was established 1 This work was supported in part by grants from the Réseau International des Insti- through experiments showing its interaction with RFX5, RFX- tuts Pasteurs et Instituts Associés , Progre`s from the Institut National de la Sante´et de la Recherche Me´dicale, the Tunisian State Secretariat for Research and Technol- ANK, CREB, and subunits of the NFY complex (25). Most im- ogy, the Association pour la Recherche contre le Cancer, and the Fondation pour la portantly, CIITA is the key molecule controlling the tissue-specific Recherche Me´dicale. A.-M.L.-D. was supported by a fellowship from La Ligue Na- tionale contre le Cancer. J.V. was supported by a fellowship from the Ligue Nationale expression pattern of the MHC class II genes (26). contre le Cancer, then from the Fondation pour la Recherche Me´dicale. RFXANK, also named RFX-B, is coded by a 10-kbp gene com- 2 A.-M.L.-D., M.-R.B., and J.V. contributed equally to the work. posed of 10 exons (13, 14). This 33-kDa protein is characterized 3 Current address: Department of Pathology, Harvard Medical School, Boston, MA 02115. 4 Address correspondence and reprint requests to Dr. Cathrine Alcaı¨de-Loridan, Unite´ 5 Abbreviations used in this paper: NFY, nuclear factor binding to Y box; B-LCL, B d’Immunoge´ne´tique Humaine, Institut National de la Santé et de la Recherche Médicale lymphoblastoid cell line; CIITA, class II transactivator; CREB, cAMP response ele- Unité 396, Centre de Recherches Biome´dicales des Cordeliers, 15 rue de l’Ecole de ment-binding protein; Ii, invariant chain; PPD, purified protein derivative; RFX, reg- Me´decine, 75006 Paris, France. E-mail address: [email protected] ulatory factor binding to X box; RFXAP, RFX-associated protein; TT, tetanus toxoid.

Copyright © 2001 by The American Association of Immunologists 0022-1767/01/$02.00 5682 UNCOORDINATED HLA-D EXPRESSION IN A RFXANK-DEFECTIVE PATIENT by three ankyrin domains located in its carboxyl terminus. The transferred to a polyvinylidene difluoride membrane (Millipore, Bedford, ankyrin motifs have been described as protein/protein interaction MA). Western blotting was performed with 1/5000 dilution of the anti- domains. In MHC class II-defective group B patients, two defects HLA-DR DA6.147 mAb (36) and 1/4000 dilution of peroxidase-labeled anti-mouse Ig Ab from Pierce (Rockford, IL), then revealed with the ECL (del26 and del58) were first described, resulting from deletions detection kit (Amersham). overlapping exon 6 affecting either the intron upstream (del26) or downstream exon 6 (del58) (13, 14). More recently, a founder Luciferase constructs and assays effect has been described for the del26 mutation in 17 patients of The pGL3 control, containing an SV40 promoter upstream the luciferase North African origin (27). Two additional patients with mutations gene, and the promoterless pGL3-basic plasmid constructs were purchased in exon 5 or 8 have been evidenced (19). from Promega (Madison, WI). The pGL3-DRA, pGL3-DQA, pGL3-DQB, and pGL3-DPA plasmids contain 250-bp promoter fragments including the We describe in this work the analysis of the JER cell line that conserved W, X1, X2, and Y boxes (37). Transient transfections were was established from a patient affected by a MHC class II defi- performed by electroporation with a mixture of the luciferase constructs ciency with classical immunological characteristics. Hereby, we and pSV␤gal vector (Promega) (10:1 molar ratio) using a Gene Pulser characterize a new type of genetic defect in RFXANK, with an apparatus (Bio-Rad, Richmond, CA) with a 960 ␮F/300 V/200 ⍀ pulse, unexpected uncoordinated expression of MHC class II genes. under conditions previously described (38). Luciferase activities were mea- sured with a Lumat luminometer (Berthold Instruments, Nashua, NH), and were further divided by the ␤-galactosidase activities (expressed as A420) Materials and Methods to correct for transfection efficiency. All the values correspond to an av- Flow cytometric analyses erage of three to four independent experiments. Immunofluorescence assays were done following classical procedures with Cell fusions a FACScan (Becton Dickinson, Franklin Lakes, NJ) using a CellQuest Downloaded from program. The phenotype of the JER PBLs, purified through Ficoll- Two days before the fusion, all cell lines were depleted of L112-positive Hypaque, was performed with CD3 (clone UCHT1), CD4 (clone 13B8.2), cells (caused either by reversion of the phenotype or nonspecific binding of CD8 (clone B9.11), PanB (clone DBB42), and HLA class II (clone 9-49) the Ab) through cell sorting using anti-mouse Ig-coupled magnetic beads mAbs from Immunotech (Luminy, France). Before labeling of B lympho- (Dynal, Great Neck, NY), as described previously (34). Cell fusions were blastoid cell lines (B-LCLs), FcRs were saturated by preincubation with performed with 107 cells of each fusion partner using PEG 4000 (Merck, heat-inactivated FCS. Indirect immunofluorescence was performed on the West Point, PA). As a control, homocaryons were prepared with each fusion partner to assess the complete lack of MHC class II-positive cells.

cell lines with the following primary mAbs: L112 (anti-DR) (28), D112 http://www.jimmunol.org/ (anti-DR) (29), B7/21 (anti-DP) (30), L2 (anti-DQ) (31), and MHC class I After fusion, cells were cultured in RPMI 1640 supplemented with 10% (W6/32; Serotec, Oxford, U.K.). Direct immunofluorescence was done Myoclone serum for 48–72 h. Indirect immunofluorescence analysis was with the anti-DR L243 mAb FITC from Becton Dickinson. then performed on cells incubated with either anti-HLA-DR L112 or D112 or with the anti-HLA-DQ L2 mAbs. Cells were analyzed under a fluores- Immunological analysis of the patient JER cence microscope without any step of permeabilization of the cells. Serum Ig concentration was measured by radial immunodiffusion. The ex- RT-PCR analysis pression of HLA class II Ags was studied on patient and control PHA- cDNA preparation and PCR were performed with the reverse transcriptase stimulated blasts with the 9-49 mAb. The immune response was evaluated and Taq polymerase from Qiagen (Chatsworth, CA) following the instruc- by proliferation assays, with PBLs cultured in 96-well plates (106 cells/ml) tions of the manufacturer. Amplification was conducted as follows: 95°C, in RPMI 1640 supplemented with 10% human AB serum. PHA (10 ␮g/ml) 5 min, one cycle; 95°C, 1 min; 55°C, 1 min; 72°C, 2 min, 30 cycles; 72°C, by guest on September 24, 2021 or specific Ags (tuberculin (purified protein derivative, PPD, 10 U/ml) or 10 min, one cycle. Amplification of a 811-bp cDNA fragment of RFXANK, tetanus toxoid (TT, 4 ␮g/ml) were added, and the proliferation was as- corresponding to nt 416-1227 (exons 3–10) of the cDNA, was performed sessed 3 days (PHA) or 5 days (PPD and TT) later by the incorporation of with primers S-5Јp33 (5Ј-CCATGGAGCTTACCCAGCCTGCAGA-3Ј) [3H]thymidine. HLA-DR typing was performed on genomic DNA by PCR- and AS-3Јp33 (5Ј-AGTGTCTGAGTCCCCGGCA-3Ј), kindly provided sequence-specific oligonucleotide procedure (JER, DR13/DR7; JER’s fa- by K. Masternak (University of Geneva Medical School, Geneva, ther (TAO), DR3/DR7; JER’s mother (HAS), DR13/Ϫ). Switzerland). cDNA amounts were verified through 24-cycle PCR with Cell culture GAPDH-specific primers that amplify a 255-bp fragment (GAPDH sense, 5Ј-GTCGTATTGGGCGCCTGGTCAC-3Ј; GAPDH antisense, 5Ј-CAC The analyzed cell lines are B-LCLs grown in RPMI 1640 supplemented GACGTACTCAGCGCCAGCA-3Ј), used in a 1/15 concentration com- with antibiotics, 2 mM glutamine, and 10% of heat-inactivated FCS (for pared with the RFXANK primers. The analysis of the different RFXANK MHC class II-positive B-LCLs) or Myoclone (Life Technologies, Cergy- products in the JER cell line was performed with sets of primers (whose Pontoise, France) (for B-LCLs from MHC class II-defective patients). The sequences are available upon request) hybridizing on exons 3, 4, 5, 6, 7, JER, HAS, and TAO EBV-transformed B-LCLs were established using and 8, or on intron 4. HLA-DMA mRNA was amplified with primers DMA standard protocols. The BLS-2 and BCH cell lines belong to complemen- sense (5Ј-CTAAAGCTGGGTTGGTAGC-3Ј) and DMA antisense (5Ј- tation group A; the THF B-LCL to group C; and the ZAR cell line, a kind GCTGGCATCAAACTCTGGT-3Ј). DMB was amplified with primers gift from B. Grospierre-Lisowska, belongs to group D. The HAD B-LCL DMB sense (5Ј-ATCTTTACAGAGCAGAGCAT-3Ј) and DMB antisense belongs to group B (32), as well as the KHE cell line (unpublished data), (5Ј-CCTTCTCACTTGGAGTGGA-3Ј). To control the amplification effi- and the BLS-1 cell line (33). The HAD and BLS-1 are respectively del26 ciency, 15 ng of ␤-actin primers (␤-Act sense, 5Ј-CACCCTGTGCTGCT and del58 RFXANK mutants. The unrelated COM and GES B-LCLs were CACCGAGGCC-3Ј; ␤-Act-antisense, 5Ј-CCACACAGAGTACTTGCGC used as positive controls for MHC class II expression. TCAGG-3Ј) was introduced in each of the HLA-DM amplifications. Northern blot analysis DNA sequencing of the genomic RFXANK products Total RNA was isolated by the guanidinium thiocyanate procedure, as A DNA fragment encompassing exons 4–6 of the RFXANK gene described previously (34). RNA samples (15 ␮g/lane) were loaded on aga- was amplified from the patient and parent genomic DNA using the sense rose gels containing 2.2 M formaldehyde. The gels were transferred on ANK-Si3.4 primer that hybridizes to intron 3 (5Ј-GAACCCAGGAGGCA Nylon N membranes (Amersham, Arlington Heights, IL), and the filters GAGGTTG-3Ј) and the antisense primer ANK-ASe6 that recognizes exon were hybridized overnight with HLA-DRA, HLA-DQA, HLA-DPA, HLA- 6(5Ј-AGCAGGAAGCGAACGGTC-3Ј). The PCR products (831 bp) were DRB, HLA-DQB,orHLA-DPB isotype chain-specific cDNA probes (35) subcloned into the pCR2.1-TOPO plasmid (Invitrogen, San Diego, CA), labeled by random priming using the Multiprime Labeling Kit from Am- and nucleotide sequencing was performed with M13 and M13 reverse ersham. The mouse Ii, which cross-hybridizes with its human homologue, primers by the Big Dye Terminator cycle sequencing (PE Applied Biosys- and ␤-actin probes were previously described (34). tems, Foster City, CA) using an ABI 377 automatic sequencer (Perkin- Elmer, Norwalk, CT). Western blot analysis RFXANK constructs and transfections Cells were lysed in a solution containing 0.5% Triton X-100, 300 mM NaCl, 50 mM Tris-HCl (pH 7.5), and a cocktail of protease inhibitors. The RFXANK wild-type and ANK⌬4 cDNAs were obtained through RT- Proteins (10 ␮g) were loaded per lane on a 12% polyacrylamide gel, then PCR amplification of mRNA preparation from the control COM and JER The Journal of Immunology 5683

B-LCLs, using the S-5Јp33 and AS-3Јp33 sets of primers described above. These PCR products were cloned in the pCR2.1-TOPO plasmid, and fur- ther subcloned into the EcoRI site of the pIRES-Neo expression vector (Clontech Laboratories, Palo Alto, CA). Sequencing was performed on both inserts to assess the lack of mutations created during the process. A total of 20 ␮g of these constructs was next cotransfected with 5 ␮gofthe red fluorescent DsRed1 construct (Clontech) through electroporation with the ECM 630 electroporator from BTX (975 ␮F, 200 ⍀, 270 V). Cell surface HLA-DR expression was determined 2 days later by flow cytom- etry on Red-positive transfected cells stained with fluorescein-conjugated L243 mAb from Becton Dickinson. Southern blot Oligonucleotides were 3Ј-end labeled with the digoxigenin oligonucleotide 3Ј-end labeling kit from Roche following the instructions of the manufac- FIGURE 1. Analysis by indirect immunofluorescence of HLA-DR, turer. The gels containing the RT-PCR products were transferred on Nylon HLA-DP, HLA-DQ, and HLA-ABC cell surface expression, using the N membranes (Amersham), and the filters were hybridized overnight with L112, B721, L2, and W6/32 mAbs, respectively, on JER (dark histogram) the labeled probes. The membranes were further revealed with ready-to-use and HAS (light histogram) B-LCLs. CSPD from Roche (Meyan, France). The sequences of the exons 4- and 5-specific oligonucleotides were respectively: 5Ј-CCGGTTGGT- Ј Ј Ј GAGAGTGGTGGAGT-3 and 5 -CCATCCACCAGCTCGCAGCAC-3 . in MHC class II-deficient patients (3), MHC class I expression in Downloaded from Results the JER cell line was slightly decreased and appeared heterogeneous. Clinical features of the MHC class II-deficient patient JER The clinical features of the patient JER included a severe chronic The JER cell line displays normal expression of DRA and DQA diarrhea, recurrent pulmonary infections, and mucocutaneous can- mRNAs didiasis detected at the age of 5 mo. He died at the age of 7.5 mo We next investigated by Northern blot analysis the expression of following an overwhelming infection. Consanguinity was ob- the transcripts encoding the ␣- and ␤-chains of MHC class II iso- http://www.jimmunol.org/ served in the family with a second degree marriage. The familial types in the JER B-LCL, and we compared it with that of HAS, survey indicates the death at an early age of a first degree maternal TAO, and a MHC class II-expressing unrelated B-LCL. The re- female cousin. Consistent with a diagnosis of MHC class II defi- sults were quite striking showing a HLA-DRA and HLA-DQA tran- ciency (3), phenotyping of the JER patient revealed decreased script expression in the JER cell line quite similar to that of the ϩ numbers of CD4 T lymphocytes (Table I) and normal counts of wild-type controls (Fig. 2A). However, transcripts corresponding B lymphocytes. In addition, PBLs and PHA blasts from patient to the HLA-DRB, HLA-DQB, HLA-DPA, and HLA-DPB mRNAs JER did not stain with anti-pan MHC class II mAbs (data not were not detected in the JER B-LCL. Through semiquantitative shown). The PBLs from the JER did not proliferate in vitro in RT-PCR techniques (20-cycle amplifications), we observed that by guest on September 24, 2021 response to specific Ags (PPD and TT), although they retained the HLA-DRA mRNA of the JER cells is about one-half to one- normal proliferative capacities to PHA. Finally, evaluation of se- third of the levels observed in a B-LCL control cell line (data not rum immunoglobulins also revealed reduced levels compared with shown). The HLA-DRA transcript is functional, as the correspond- an age-matched normal control (Table I). ing chain was evidenced by Western blot analysis (Fig. 3). We then analyzed the expression of HLA class II-related molecules Lack of cell surface HLA-D expression in the JER B-LCL that are implicated in Ag presentation. Similar to many MHC class EBV-transformed cell lines were established from JER as well as II-deficient cell lines, the expression of Ii mRNA in the JER cell from his healthy parents, HAS (the mother) and TAO (the father), line was comparable with positive control cell lines (data not to characterize the molecular defect of the JER patient. Cell sur- face expression of the HLA-DR, HLA-DQ, and HLA-DP mole- cules in the three established cell lines was first analyzed by flow cytometry (Fig. 1 for the JER and HAS B-LCLs, not shown for the TAO cell line). As initially observed on fresh PBLs, the JER cell line is completely negative for the cell surface expression of the three MHC class II isotypes, in contrast to HAS and TAO B-LCLs, which express high levels of these molecules. As often evidenced

Table I. Immunological features of the patient JER

CD4ϩ T cell counts 693/mm3 (n ϭ 1700–2800/mm3) Serum Ig levels IgG ϭ 1.75 mg/ml (n ϭ 2.69–9.13 mg/ml) Proliferative assays Unstimulated 1,544 cpm PHA 131,810 cpm FIGURE 2. A, Northern blot analysis of MHC class II mRNA expres- PPDb 1,302 cpm sion in GES (a control MHC class II-positive B-LCL), JER, HAS, and TT 1,215 cpm TAO cell lines. B, RT-PCR analysis of the expression of HLA-DMA and

ϩ HLA-DMB genes in a control MHC class II-expressing (Cont) and in JER a The N values correspond to the counts of CD4 cells and Igs for an age-matched normal control. B-LCLs. The JER cell line was cultured for 8 wk, and RNA preparations b PPD, tuberculin Ag; TT, tetanus toxoid Ag. were made on aliquots taken each 2 wk from the cell culture. 5684 UNCOORDINATED HLA-D EXPRESSION IN A RFXANK-DEFECTIVE PATIENT

Table II. Results obtained from the fusion experiments of JER cells and MHC class II deficient cell lines from complementation groups A–Da

Group A Group B Group C Group D FIGURE 3. Western blot analysis of the HLA-DR␣-chain expression BLS-2, BCH BLS-1, HAD, KHE THF ZAR detected with the DA6.147 mAb in cell extracts from the TAO, HAS, and HLA-DR ϩϩϩ ϩ ϩϩϩ ϩϩϩ JER B-LCLs, in addition to an unrelated MHC class II-expressing control HLA-DQ ϩϩ Ϫ ϩϩ ND cell line (GES) and a group B cell line presenting a del26 mutation in the RFXANK gene (HAD). a Cells were fused and heterokaryons analyzed by immunofluorescence for the membrane expression of HLA-DR (with L112 mAb) or -DQ (with L2 mAb) Ags. The symbols indicate: ϩϩϩ, restoration of MHC class II expression: Ϫ, no staining: ϩ, weak staining of the heterokaryons. shown). In contrast, the analysis of HLA-DM was puzzling because the expression of these mRNAs appeared to be unstable: we ini- tially observed that HLA-DMA mRNA was expressed in the JER The situation was more complex for heterokaryons between JER B-LCL at levels quite similar to that of control cell line, whereas and group B mutants. Three unrelated group B cell lines, BLS-1, HLA-DMB mRNA expression was residual. After several weeks in HAD, and KHE, were assayed to assess the complementation anal- culture, the expression of these mRNAs was progressively lost ysis. Strikingly, when using either the anti-DR L112 or D112 (Fig. 2B), whereas that of HLA-DRA and HLA-DQA remained sta- mAbs, a faint cell surface staining was detected with all types of ble under these conditions (data not shown). heterokaryons. This labeling was much weaker than that observed To establish the involvement of trans-acting factors in the un- in fusions involving JER and the other groups, but was perfectly Downloaded from usual HLA-D expression in the JER B-LCL, the transcriptional reproductive with all three group B cell lines. However, when activity of different HLA-D promoters was next analyzed through using the L2 anti-DQ Ab, staining was not detected for any of the luciferase reporter gene experiments. In agreement with the data JER ϫ group B heterokaryons. As controls, detection of HLA-DR presented above, the HLA-DRA and HLA-DQA promoters drive, expression was neither obtained with any homokaryon, nor with respectively, a luciferase expression in the JER cell line ϳtwo- heterokaryons obtained with the BLS-1, HAD, and KHE fusion thirds or one-half of the HAS-positive cell line, while the HLA- partners. Therefore, these data suggested a defect in the RFXANK http://www.jimmunol.org/ DQB and HLA-DPA promoters lead to weak transcriptional activ- gene in the JER cell line. ities close to that of the BLS-1 B-LCL (Fig. 4). These data indicate a transcriptional defect of HLA-D genes in the JER cells like clas- RFXANK is defective in the JER cell line sical MHC class II-deficient mutants, even though HLA-DRA and To test this possibility, we amplified by RT-PCR the whole coding HLA-DQA genes remain transcribed at nearly wild-type level. sequence of RFXANK mRNA using specific oligonucleotides hy- bridizing to exons 3 and 10 of the RFXANK gene. A major band Genetic complementation of the JER cell line was detected corresponding to the RFXANK mRNA form con- To further understand the particular defect in the JER B-LCL, taining exons 3–10 (wild type) in control MHC class II-positive by guest on September 24, 2021 genetic complementation analysis was undertaken. These experi- B-LCL, and in HAS and TAO cell lines. A minor band (⌬5), ments were performed by cell fusion between the JER cells and described by others as the outcome of an alternative splicing of cell lines belonging to the four MHC class II deficiency comple- exon 5 (14), was also observed in these cell lines. In BLS-1 group mentation group, followed by an analysis of cell surface expres- B cell line, the major and minor bands detected correspond re- sion of HLA-DR or HLA-DQ on the heterokaryons by spectively to mRNAs lacking both exons 5 and 6 (⌬5/6) or lacking immunofluorescence. exon 6 only (⌬6), as previously described (13). The heterokaryons obtained after fusion of JER with either BCH Interestingly, the JER cell line presented an unusual pattern with or BLS-2 cell lines from group A, with THF cells from group C, six main bands, in which none of these bands has the size of the or with ZAR cells from group D, displayed HLA-DR Ag expres- wild-type mRNA (Fig. 5). The more abundant form (band A) de- sion at their surface, with a slightly weaker expression of HLA-DQ tected in the JER cells was observed as well, in some experiments, (Table II). These results indicate that the JER cells are not defec- however very faintly, in both parents’ cell lines, but not in control tive for CIITA, RFX5, or RFXAP. B-LCLs. The size of band A was compatible with the presence of the intron (Int.4–5) located between exons 4 and 5. The presence

FIGURE 4. Luciferase assay performed with pGL3-DRA, pGL3-DQA, pGL3-DQB, and pGL3-DPA promoter constructs in the JER, HAS, and BLS-1 B-LCLs, in addition to a MHC class II-expressing cell line (GES). FIGURE 5. RT-PCR analysis of RFXANK expression on RNA ex- Background luciferase expression determined with the promoterless pGL3- tracted from the patient JER, its parent HAS and TAO B-LCL, BLS-1 basic vector was deduced from these values. Luciferase relative counts (group B), and MHC class II-expressing B-LCLs. GAPDH amplification were standardized through the measurement of the cotransfected pSV␤gal was also performed to assess that equal amounts of cDNA were used for plasmid. these samples. The Journal of Immunology 5685 of this intron was further confirmed through RT-PCRs and South- ern blotting of the RT-PCR products with Int.4–5-specific oligo- nucleotides (data not shown). Sequencing of the cDNA obtained through RT-PCR amplification with primers hybridizing with ex- ons 3 and 6 showed the presence of Int.4–5 in the mRNA and the existence ofaGtoCpoint mutation in the canonical donor splice site downstream exon 4 (not shown). Accordingly, genomic PCR of the intron 3 to exon 6 region from RFXANK was amplified in both parents’ and patient JER DNA, and further subcloned. Sequencing of these DNA fragments confirmed the presence of the same G to C mutation in all six sequenced clones from the patient JER (Fig. 6). Concerning the parents’ DNA, the mutation was found in two of four clones for the mother and in three of four clones in the father, indicating the heterozygocity of both parents. Therefore, these data show that the JER cell line presents a defect in a splice donor site downstream exon 4 of the RFXANK gene. The lack of splicing of Int.4–5 generates a premature stop codon in the RFXANK cDNA, thereby producing an RFXANK mutant protein lacking Downloaded from all three ankyrin domains. FIGURE 7. Immunofluorescence analysis of cell surface expression of We next performed a functional complementation analysis HLA-DR in the JER and BLS-1 B-LCLs transiently cotransfected with the through transient transfection of the wild-type cDNA encoding red fluorescent DsRed1 vector and the indicated RFXANK cDNA construct RFXANK. As seen in Fig. 7, the construct was able to restore the or empty vector. HLA-DR expression was detected with the L243 mAb cell surface expression of HLA-DR in both the JER and BLS-1 cell stained with fluorescein (FL1-H). Values in the upper right of the quad- rants represent the number of events. The mean fluorescence intensity ap- lines, further assessing the belonging of this cell line to group B. http://www.jimmunol.org/ pears in between parentheses. The JER cell line expresses a mutant form of RFXANK

We next investigated the uncoordinated pattern of expression of tive splicing forms of the RFXANK mRNA. Four of these bands, the HLA-D genes in the JER B-LCL and the peculiar genetic including band A, were either containing Int.4–5, or were splice complementation data obtained when the JER cells were fused variants in which the exon junctions were not in a proper reading with group B B-LCLs. Bandshift experiments (data not shown) frame. The fifth band presented an alternative splicing of both suggested that this mutant protein would be in low amounts or that exons 4 and 5, with the junction of exons 3 and 6 remaining in a it would render the RFX complex unstable: these experiments did proper open reading frame (Fig. 8). This RT-PCR product was by guest on September 24, 2021 not show the binding of RFX-containing complexes with nuclear dismissed as a possible candidate for the HLA-DRA transcription, extracts from the JER cells using the HLA-DRA promoter, thereby as it has been previously demonstrated that a cDNA lacking exon resembling classical group B cell lines. Given the detection of six 5(⌬5) restored HLA-DRA expression 15 times less efficiently than RFXANK RT-PCR products in the JER B-LCL, we hypothesized the wild-type construct (19). that a mutant form of RFXANK might retain a residual activity and Therefore, we focused on band B (Fig. 8), which is lacking exon a selective affinity for certain HLA-D promoters. This was inves- 4 and whose alternative splicing maintains a proper open reading tigated through size analysis of the PCR products and Southern frame. We hypothesized that the ANK⌬ex4 mutant protein, which blotting with oligonucleotides specific for different introns and ex- presents a 28-aa deletion, was responsible for the selective expres- ons of the RFXANK gene. We thereby identified different alterna- sion of HLA-DRA and HLA-DQA mRNAs in the JER cell line. The corresponding cDNA, the ANK⌬ex4 form, was subcloned, and the lack of exon 4 was confirmed by nucleotide sequencing. Considering the complementation analysis, where the fusion of the JER and BLS-1 cell lines led to a residual expression of HLA-DR at the cell surface of the heterokaryons, we thus investigated the possibility of the rescue of HLA-DR expression in the BLS-1

FIGURE 8. Southern blot of the RT-PCR products (nt. 416-1227, exons 3–10) obtained by the amplification of RFXANK transcripts from JER (ethidium bromide staining, C) and control (COM) B-LCLs using the FIGURE 6. Sequence of the exon 4/intron 4–5 junction of the RFXANK S-5Јp33 and AS-3Јp33 primers. Blots were next hybridized with oligonu- gene from the patient JER and HAS cell lines. cleotides corresponding to exon 4 (A) and exon 5 (B). 5686 UNCOORDINATED HLA-D EXPRESSION IN A RFXANK-DEFECTIVE PATIENT transfectants by the ANK⌬ex4 cDNA. The ANK⌬ex4 construct (33). Indeed, to our knowledge, the genetic defect in the KEN and KER was transiently transfected in the JER and BLS-1 B-LCLs, and cell cell lines has not been identified. surface HLA-DR expression was analyzed by cytofluorometry. In With the absence of a clear indication concerning noncoordinate agreement with the above hypothesis concerning the BLS-1 trans- HLA-D expression, we hypothesized that a truncated RFXANK fectants, HLA-DR expression was detected in these cells. There- protein might be responsible for the peculiar phenotype displayed fore, these data show that the ANK⌬ex4 cDNA can rescue both by the JER cells. One of the transcripts of RFXANK could be HLA-DRA and HLA-DRB expression. In the JER-transfected encoding such a protein, ANK⌬ex4, as the skipping of exon 4 cells, the ANK⌬ex4 cDNA restored HLA-DR expression with val- maintains a proper open reading frame. We show in this work that ues quite similar to those of the BLS-1 cell line. Therefore, these this mutant protein, although lacking 28 aa, is retaining the capac- data establish that the ANK⌬ex4 protein, although deleted of 28 ity to rescue cell surface expression of the HLA-DR molecules aa, retains, when overexpressed in transfected cells, a transcrip- (Fig. 7). This 28-aa region is a Thr- and Ser-rich region (four Thr tional activity quite similar to that of the wild-type protein. and four Ser residues) that is highly conserved in the mouse pro- tein (27 of 28 aa) (13). Its localization between the acidic amino- Discussion terminal region and the ankyrin domains of the RFXANK protein In most group B cell lines from patients of North African origin may indicate that this region is a hinge between these two domains. with MHC class II deficiency, a similar 26-bp deletion upstream However, our data exclude an essential role for this region in MHC exon 6 of the RFXANK gene was described (13, 14, 27). It was class II gene transcription. recently shown that this phenomenon is not due to a hot-spot site Based on the RT-PCR analysis of RFXANK mRNA, the ANK⌬ex4 of mutagenesis, but is linked to the genetic transmission of the protein is most likely expressed in very limited amounts in the JER cell Downloaded from mutation in these patients (27). In the JER patient, who is of North line. This is in agreement with our bandshift experiments in which RFX- African origin, we have characterized a new type of defect in RFX- containing complexes were not detected with the JER B-LCL nuclear ANK. A homozygous single nucleotide substitution abolishing the extracts. It is also possible that the ANK⌬ex4-containing RFX complex, splice site downstream of exon 4 was assessed by gene sequencing although highly likely formed in vivo, might be unstable in our experi- and by functional complementation with the wild-type RFXANK mental conditions. Our data suggest that the few ANK⌬ex4 molecules

cDNA. Interestingly, mutations affecting exon 5 or 8 of the RFX- expressed in the JER cell line would only bind to promoters for which http://www.jimmunol.org/ ANK gene have been recently described in two other patients from they present the highest affinity. However, comparison of the HLA-D group B (19). In contrast to CIITA mutations altering the carboxyl- promoter sequences, mainly in the 3Ј region of the X1 motif in which terminal part of the protein (26), or RFXAP defects, in which all RFXANK binding sites have been identified (45), did not reveal any the mutations have been found in exon 2 (17), it appears that in the sequence that would correlate with the selective HLA-DRA, HLA-DQA, RFXANK-defective patients different regions of the gene can be and HLA-DMA transcription in the JER cell line. Affinities of the RFX affected. complex for the different HLA-D promoters were shown as DRA ϭ The presence of intron 4 in the most abundant RFXANK tran- DPA Ͼ DQB ϾϾ DQA ϭ DPB Ͼ DRB (46). Therefore, our data suggest script form from the JER cells creates a premature stop codon that that the truncation of the RFXANK protein modifies the affinities of the leads to a protein lacking all three ankyrin domains. The resulting mutant RFX complex for the different HLA-D promoters in the JER by guest on September 24, 2021 mutant protein, if expressed, cannot be active, as it has been shown cell line. that the lack of the three domains suppresses the function of this Genetic complementation experiments (Table II) led to highly inter- protein in ϳ20 group B BLS cell lines (13, 14, 19, 27). However, esting data: the JER cell line was complemented with cells from groups two characteristics of the JER cell line therefore remained unex- A, C, and D, but displayed residual expression of HLA-DR Ags when plained: the nearly wild-type expression of HLA-DRA and HLA- fused with three different group B cell lines. One of these latter cell lines DQA mRNAs, in addition to HLA-DMA and Ii, and the residual (HAD) has the classical 26-nt deletion upstream exon 6 of RFXANK (27), expression of HLA-DR in the heterokaryons generated by the fu- while BLS-1 is displaying a 58-nt deletion downstream this exon (14). sion of the JER B-LCL with three group B cell lines. We propose that the faint staining detected on the JER/group B B-LCLs Differential expression of the HLA-D genes has been described mainly heterocaryons is due to ␣-complementation phenomenons between the in tumor cell lines (reviewed in Refs. 39 and 40). CIITA-independent two RFXANK mutant proteins provided by each fusion partner. In agree- selective expression of HLA-DQ in clone 13 (41) and the expression of ment with our data, Lin et al. have demonstrated the oligomerization of HLA-DP in a neuroblastoma cell line (42), for instance, have been re- Tvl-1, the mouse homologue of the RFXANK protein (47). Accordingly, ported. In these cells, compensation mechanisms mediated by the dys- it was recently demonstrated that a del26 cell line, Ramia, was comple- regulation of expression or activity of certain oncogenes (ras), antionco- mented by a fragment of RFXANK encompassing residues 69–260 (48). genes (Rb), or cellular transcriptional regulators (CREB) (40) may Therefore, our data suggest that the ANK⌬ex4 mutant protein from JER account for the unusual phenotype. In cell lines originating from MHC and the 122 amino-terminal part of the del26 or del58 RFXANK proteins class II-deficient patients, uncoordinated HLA-D expression has been oc- from the other group B cell lines present a capacity to dimerize and to casionally described. However, this expression is residual in these cells weakly activate HLA-DR expression. It was shown that a mutant and is most likely due to the mutant CIITA or RFX proteins they contain. Tvl-1 protein lacking the 130 carboxyl-terminal amino acids lost As an example, HLA-DP transcripts are detected at low levels in certain the capacity to oligomerize (47). Taken as a whole, these data RFXAP-defective cells (17). The only case of a clearly uncoordinated suggest that the individual monomers interact through different pattern of expression of HLA-D are B-LCLs established from two broth- domains, and that the amino-terminal domain is required as well ers, KEN and KER, which express HLA-DRA, HLA-DQA, HLA-DPB, for the dimerization of the RFXANK protein. and Ii (43), but not HLA-DM (33) mRNAs. This phenotype does not In summary, we report that a 28-aa truncated RFXANK protein entirely match the JER phenotype, in which HLA-DPB is not expressed, retains a transcriptional activity. Therefore, we propose that the although HLA-DMA is. The observation that both KEN and KER broth- deleted region is not essential for the activity of the protein, but the ers displayed very mild immunodeficiency symptoms compared with pa- low amount of this protein triggers the transcription of the HLA-D tient JER (44) further differentiates these patients. In addition, genetic promoters most likely presenting the highest affinity for this pro- complementation analysis showed that KER cells do not belong to groups tein or for the protein complex containing RFXANK. This report A, B, or C, and are therefore not defective for CIITA, RFXANK, or RFX5 further emphasizes the necessity to study mutant RFXANK cDNA The Journal of Immunology 5687 constructs in cell lines presenting a complete lack of the endoge- 23. Steimle, V., L. A. Otten, M. Zufferey, and B. Mach. 1993. Complementation nous protein, to avoid misleading interpretations linked to cloning of an MHC class II transactivator mutated in hereditary MHC class II deficiency (or ). Cell 75:135. ␣-complementation phenomenons occurring between the endoge- 24. Riley, J. L., and J. M. Boss. 1993. Class II MHC transcriptional mutants are nous and exogenous mutant proteins. defective in higher order complex formation. J. Immunol. 151:6942. 25. Zhu, X. S., M. W. Linhoff, G. Li, K. C. Chin, S. N. Maity, and J. P. Ting. 2000. Transcriptional scaffold: CIITA interacts with NF-Y, RFX, and CREB to cause Acknowledgments stereospecific regulation of the class II major histocompatibility complex pro- We are grateful to Drs. R. Al-Daccak. N. Setterblad, N. Mooney, and G. moter. Mol. Cell. Biol. 20:6051. Barbieri for critical reading of the manuscript. We thank Dr. B. Lisowska- 26. Mach, B., V. Steimle, E. Martinez-Soria, and W. Reith. 1996. Regulation of MHC class II genes: lessons from a disease. Annu. Rev. Immunol. 14:301. Grospierre for highly helpful guidance. 27. Wiszniewski, W., M. C. Fondaneche, N. Lambert, K. Masternak, C. Picard, L. Notarangelo, K. Schwartz, J. Bal, W. Reith, C. Alcaı¨de, et al. 2000. Founder References effect for a 26-bp deletion in the RFXANK gene in North African MHC class II deficient patients belonging to complementation group B. Immunogenetics 51: 1. Germain, R. N. 1994. MHC-dependent antigen processing and peptide presenta- 261. tion: providing ligands for T lymphocyte activation. Cell 76:287. 28. Bono, R., F. Hyafil, J. Kalil, V. Koblar, J. Wiels, E. Wollman, and M. Fellous. 2. Lisowska-Grospierre, B., D. J. Charron, C. de Pre´val, A. Durandy, C. Griscelli, 1979. Monoclonal antibodies against HLA-DRw antigens. Transplant. Clin. Im- and B. Mach. 1985. A defect in the regulation of major histocompatibility com- munol. 11:109. plex class II gene expression in human HLA-DR negative lymphocytes from 29. Carrel, S., R. Tosi, N. Gross, N. Tanigaki, A. L. Carmagnola, and R. S. Accolla. patients with a combined immunodeficiency syndrome. J. Clin. Invest. 76:381. 1981. Subsets of human Ia-like molecules defined by monoclonal antibodies. 3. Klein, C., B. Lisowska-Grospierre, F. LeDeist, A. Fischer, and C. Griscelli. 1993. Mol. Immunol. 18:403. Major histocompatibility complex class II deficiency: clinical manifestations, im- 30. Watson, A. J., R. DeMars, I. S. Trowbridge, and F. H. Bach. 1983. Detection of munologic features, and outcome. J. Pediatr. 123:921. a novel human class II HLA antigen. Nature 304:358. 4. Be´nichou, B., and J. L. Strominger. 1991. Class II-antigen-negative patient and mutant B-cell lines represent at least three, and probably four, distinct genetic 31. Fermand, J. P., C. Schmitt, and J. C. Brouet. 1985. Distribution of class II DQ Downloaded from defects defined by complementation analysis. Proc. Natl. Acad. Sci. USA 88: antigens on normal and leukemic lymphoid cells. Eur. J. Immunol. 15:1183. 4285. 32. Lisowska-Grospierre, B., M.-C. Fondaneche, M.-P. Rois, C. Griscelli, and 5. Hume, C. R., and J. S. Lee. 1989. Congenital immunodeficiencies associated with A. Fischer. 1994. Two complementation groups account for most cases in inher- absence of HLA class II antigens on lymphocytes result from distinct mutations ited MHC class II deficiency. Hum. Mol. Genet. 3:953. in trans-acting factors. Hum. Immunol. 26:288. 33. Douhan, J., I. Hauber, M. M. Eibl, and L. H. Glimcher. 1996. Genetic evidence 6. Benoist, C., and D. Mathis. 1990. The regulation of major histocompatibility for a new type of major histocompatibility complex class II combined immuno- complex class-II genes: X, Y and other letters of the alphabet. Annu. Rev. Im- deficiency characterized by a dyscoordinate regulation of HLA-D ␣ and ␤ chains. J. Exp. Med. 183:1063.

munol. 8:681. http://www.jimmunol.org/ 7. Dorn, A., J. Bollekens, A. Staub, C. Benoist, and D. Mathis. 1987. A multiplicity 34. Lennon, A., C. Ottone, A. Peijnenburg, C. Hamon-Benais, F. Colland, S. Gobin, of CCAAT box-binding proteins. Cell 50:863. P. van den Elsen, M. Fellous, R. Bono, and C. Alcaide-Loridan. 1996. The RAG 8. Van Huijsduijnen, H. R., X. Y. Li, D. Black, H. Matthes, C. Benoist, and cell line defines a new complementation group of MHC class II deficiency. Im- D. Mathis. 1990. Co-evolution from yeast to mouse: cDNA cloning of the two munogenetics 43:352. NFY (CP1/CBF) subunits. EMBO J. 9:3119. 35. Piatier-Tonneau, D., P. Turmel, C. Auffray, and D. Charron. 1986. Construction 9. Sinha, S., S. N. Maity, J. Lu, and C. B. De. 1995. Recombinant rat CBF-C, the of chain- and locus-specific HLA class II DNA probes: study of HLA-class II third subunit of CBF/NFY, allows formation of a protein-DNA complex with transcripts in leukemias: construction of chain- and locus-specific HLA class II CBF-A and CBF-B and with yeast HAP2 and HAP3. Proc. Natl. Acad. Sci. USA DNA probes. J. Immunol. 137:2050. 92:1624. 36. Gruneberg, U., S. M. Van Ham, G. Malcherek, I. Broker, A. Melms, and 10. Moreno, C. S., G. W. Beresford, P. Louis-Plence, A. C. Morris, and J. M. Boss. J. Trowsdale. 1997. The structure and function of the novel MHC class II mol- 1999. CREB regulates MHC class II expression in a CIITA-dependent manner. ecule, HLA-DM. Biochem. Soc. Trans. 25:208S.

Immunity 10:143. 37. Peijnenburg, A., R. Van den Berg, M. J. Van Eggermond, O. Sanal, J. M. Vossen, by guest on September 24, 2021 11. Reith, W., S. Satola, C. Herrero-Sanchez, I. Amaldi, B. Lisowska-Grospierre, A. M. Lennon, C. Alcaide-Loridan, and P. J. Van den Elsen. 2000. Defective C. Griscelli, M. R. Hadam, and B. Mach. 1988. Congenital immunodeficiency MHC class II expression in an MHC class II deficiency patient is caused by a with a regulatory defect in MHC class II gene expression lacks a specific novel deletion of a splice donor site in the MHC class II transactivator gene. HLA-DR promoter binding protein, RF-X. Cell 53:897. Immunogenetics 51:42. 12. Stimac, E., S. S. Urieli, S. Kempin, and D. Pious. 1991. Defective HLA DRA X 38. Lennon, A. M., C. Ottone, G. Rigaud, L. L. Deaven, J. Longmire, M. Fellous, box binding in the class II transactive transcription factor mutant 6.1.6 and in R. Bono, and C. Alcaide-Loridan. 1997. Isolation of a B-cell-specific promoter cells from class II immunodeficient patients. J. Immunol. 146:4398. for the human class II transactivator. Immunogenetics 45:266. 13. Masternak, K., E. Barras, M. Zufferey, B. Conrad, G. Corthals, R. Aebersold, 39. Alcaide-Loridan, C., A. M. Lennon, M. R. Bono, R. Barbouche, K. Dellagi, and J. C. Sanchez, D. F. Hochstrasser, B. Mach, and W. Reith. 1998. A gene encoding M. Fellous. 1999. Differential expression of MHC class II isotype chains. Mi- a novel RFX-associated transactivator is mutated in the majority of MHC class II crobes Infect. 1:929. deficiency patients. Nat. Genet. 20:273. 40. Blanck, G. 1999. HLA class II expression in human tumor lines. Microbes Infect. 14. Nagarajan, U. M., P. Louis-Plence, A. DeSandro, R. Nilsen, A. Bushey, and 1:913. J. M. Boss. 1999. RFX-B is the gene responsible for the most common cause of 41. Ono, S. J., and Z. Song. 1995. Mapping of the interaction site of the defective the bare lymphocyte syndrome, an MHC class II deficiency. Immunity 10:153. transcription factor in the class II major histocompatibility complex mutant cell 15. Steimle, V., B. Durand, E. Barras, M. Zufferey, M. R. Hadam, B. Mach, and line clone-13 to the divergent X2-box. J. Biol. Chem. 270:6396. W. Reith. 1995. A novel DNA-binding regulatory factor is mutated in primary 42. Ponzoni, M., F. Guarnaccia, M. V. Corrias, and P. Cornaglia-Ferraris. 1993. MHC class II deficiency (bare lymphocyte syndrome). Genes Dev. 9:1021. Uncoordinate induction and differential regulation of the HLA class I and class II 16. Durand, B., P. Sperisen, P. Emery, E. Barras, M. Zufferey, B. Mach, and expression by ␥-interferon in differentiating human neuroblastoma cells. Int. J. W. Reith. 1997. RFXAP, a novel subunit of the RFX DNA binding complex is Cancer 55:817. mutated in MHC class II deficiency. EMBO J. 16:1045. 17. Fondaneche, M. C., J. Villard, W. Wiszniewski, E. Jouanguy, A. Etzioni, 43. Hauber, I., H. Gulle, H. M. Wolf, M. Maris, H. Eggenbauer, and M. M. Eibl. F. Le Deist, A. Peijnenburg, J. L. Casanova, W. Reith, B. Mach, et al. 1998. 1995. Molecular characterization of major histocompatibility complex class II Genetic and molecular definition of complementation group D in MHC class II gene expression and demonstration of antigen-specific T cell response indicate a deficiency. Hum. Mol. Genet. 7:879. new phenotype in class II-deficient patients. J. Exp. Med. 181:1411. 18. Villard, J., B. Lisowska-Grospierre, P. van den Elsen, A. Fischer, W. Reith, and 44. Wolf, H. M., I. Hauber, H. Gulle, V. Thon, H. Eggenbauer, M. B. Fischer, B. Mach. 1997. Mutation of RFXAP, a regulator of MHC class II genes, in S. Fiala, and M. M. Eibl. 1995. Twin boys with major histocompatibility complex primary MHC class II deficiency. N. Engl. J. Med. 337:748. class II deficiency but inducible immune responses. N. Engl. J. Med. 332:86. 19. Nagarajan, U. M., A. Peijnenburg, S. J. Gobin, J. M. Boss, and 45. Westerheide, S. D., and J. M. Boss. 1999. Orientation and positional mapping of P. J. van Den Elsen. 2000. Novel mutations within the RFX-B gene and partial the subunits of the multicomponent transcription factors RFX and X2BP to the rescue of MHC and related genes through exogenous class II transactivator in major histocompatibility complex class II transcriptional enhancer. Nucleic Acids RFX-B-deficient cells. J. Immunol. 164:3666. Res. 27:1635. 20. Reith, W., C. A. Siegrist, B. Durand, E. Barras, and B. Mach. 1994. Function of 46. Louis-Plence, P., C. S. Moreno, and J. M. Boss. 1997. Formation of a RFX- major histocompatibility complex class II promoters requires cooperative binding X2BP-NF-Y multiprotein complex on the conserved regulatory regions of HLA between factors RFX and NF-Y. Proc. Natl. Acad. Sci. USA 91:554. class II genes. J. Immunol. 159:3899. 21. Reith, W., M. Kobr, P. Emery, B. Durand, C.-A. Siegrist, and B. Mach. 1994. 47. Lin, J.-H., A. Makris, C. McMahon, S. E. Bear, C. Patriotis, V. R. Prasad, Cooperative binding between factors RFX and X2bp to the X and X2 boxes of R. Brent, E. A. Golemis, and P. N. Tsichlis. 1999. The -containing MHC class II promoters. J. Biol. Chem. 269:20020. adaptator protein Tvl-1 is a novel substrate and regulator of Raf-1. J. Biol. Chem. 22. Villard, J., M. Peretti, K. Masternak, E. Barras, G. Caretti, R. Mantovani, and 274:14706. W. Reith. 2000. A functionally essential domain of RFX5 mediates activation of 48. DeSandro, A. M., U. M. Nagarajan, and J. M. Boss. 2000. Associations and major histocompatibility complex class II promoters by promoting cooperative interactions between bare lymphocyte syndrome factors. Mol. Cell. Biol. binding between RFX and NF-Y. Mol. Cell. Biol. 20:3364. 20:6587.