Relative quantication of HLA-DRA1 and -DQA1 expression by real-time reverse transcriptasepolymerase chain reaction (RTPCR)

Blackwell Publishing Ltd.

S. Fernandez,* R. Wassmuth,* I. Knerr, C. Frank* and J. P. Haas*

Summary
Polymorphism in the upstream regulatory region (URR) of the MHC class II DQA1 gene denes 10 different alleles named QAP (DQA1 promoter). In vitro studies have suggested that allelic polymorphism in the HLA-DQA promoter region may result in differences in HLA-DQA1 gene expression. In the present study, we used real-time reverse transcriptasepolymerase chain reaction (RTPCR) to quantify differences in HLA-DQA1 gene expression. After the isolation of total mRNA, reverse transcription into cDNA was carried out using random hexamer priming and moloney murine leukaemia virus (MMLV) reverse transcriptase. Quantication of DQA1 mRNA species using a set of six group-specic primer pairs for the detection of HLA-DQA1*01, *02, *03, *04, *05 and *06 was carried out on an ABI PRISM GeneAmp 7700 Sequence Detection System (Perkin Elmer, Foster City, CA) with real-time detection and quantication taking advantage of the uorescence TaqMan technology (Perkin Elmer, Foster City, CA). Normalization of cDNA templates was achieved by glyceraldehyde-3-phosphate dehydrogenase (GAPDH) quantication. In addition, the total amount of mRNA produced by HLA-DQA1 and HLA-DRA1 expression was quantied for comparison. Subsequently, this approach was validated using Raji and HUT-78 cell lines and tested with peripheral mononuclear cells (PBMC) of 45 samples taken from healthy volunteers. The sensitivity was determined with 102 copies. Comparison of the allele-specic DQA1 expression with the total expression of DQA1 and DRA1 mRNA indicated that DQA1*04 expression was increased compared with the expression of other alleles of
* Institute for Clinical Immunology, Department of Medicine III, Friedrich Alexander University, Erlangen-Nuremberg, Germany, Institute for Transplantation Diagnostics and Cell Therapeutics, University of Dusseldorf Medical Center, Germany, Childrens Hospital, Friedrich Alexander University, Erlangen, Germany, and Department of Pediatrics, Division of Neonatology and Critical Care Medicine, University of Greifswald, Germany. Received 22 February 2002; revised 15 October 2002; accepted 6 December 2002 Correspondence: Johannes-Peter Haas, Department of Paediatrics, Division of Neonatology and Critical Care Medicine, University of Greifswald, Soldtmannstr. 15, D17487 Greifswald, Germany. Tel: 49 3834 866409; Fax: 49 3834 867377; E-mail: jphaas@uni-greifswald.de

the DQA1 gene. Thus, allele-specic quantication of DQA1 gene products could be achieved by real-time RTPCR suitable for the analysis of differential expression of DQA1 mRNAs in homozygote and heterozygote combinations.

Introduction
MHC class II molecules play a major role in shaping the antigen-specic immune response. Their expression is regulated in a cell-specic manner and controlled via cis-acting elementsw in the upstream regulatory regions (URRs) located in the 5-anking regions of HLA class II genes. These URRs consist of highly conserved sequences having promoter, repressor or enhancer functions, e.g. TATA, CCAAT, and X1-, X2-, Y- and W-box. Transacting factors such as RF-X, c-fos, c-jun and hXBP bind to cis-acting elements and either activate or repress transcription (Edwards et al., 1986; Auffray et al., 1987; Reith et al., 1995). Despite the highly conserved sequences of the cis-acting elements, polymorphism has been found in the URRs of HLA-D genes, namely in DRB1 (Perfetto et al., 1993), DQA1 (Del Pozzo et al., 1992) and DQB1 (Andersen et al., 1991). In the URR of DQA1, polymorphism is concentrated in the hypervariable region between 240 and 200 bp upstream of exon 1, dening 10 different allelic variants referred to as QAP (DQA1 promoter) alleles: 1.1, 1.2, 1.3, 1.4, 1.5, 2.1, 3.1, 3.2, 4.1 and 4.2 (Haas et al., 1994). These QAP alleles were found to be in linkage disequilibrium with DQA1 alleles dened by DQA1 exon 2 polymorphism (Haas et al., 1995). The URR of DQA1 differs from other MHC class II URRs in the absence of a TATA as well as a CCAAT sequence. Moreover, the Y-box in the URR of DQA1 (YC-box) differs from the consensus sequence (bp 123 A instead of G) found in all other Y-boxes of class II URRs (Auffray et al., 1987). This may be relevant to the decreased surface expression seen for HLA-DQ molecules, compared to HLA-DR and -DP (Edwards et al., 1986; Marley et al., 1987; Kimura & Sasazuki, 1992). In addition, the promoter variants QAP 4.1 (DQA1*0401 and *0601) and 4.2 (DQA1*0501) carry a second substitution in their Y-box (YM-box: bp 119 A to G base change). In vitro studies analysing cell lines have suggested a functional relevance of the allelic polymorphism in the DQA1 URR for DQA1 gene expression (Kimura & Sasazuki, 1992; Morzycka-Wroblewska et al., 1997).

2003 Blackwell Publishing Ltd, European Journal of Immunogenetics 30, 141 148

141

142

S. Fernandez et al.

Association of HLA markers is a hallmark of some autoimmune diseases. Particularly, HLA-DQ molecules may play a signicant role in susceptibility to, for example, diabetes mellitus type 1, juvenile idiopathic arthritis ( JIA) and coeliac disease. Moreover, a positive association of those DQA1 URR alleles carrying the mutated YM-box with the oligoarticular subtype of JIA has been described, suggesting the relevance of regulatory effects on disease susceptibility in autoimmune disease (Haas et al., 1995). Thus, the aim of this study was to establish a polymerase chain reaction (PCR)-dependent assay to quantify HLA-DQA1 expression in an allele-specic manner in order to investigate the role of allele-specic expression in autoimmunity.

puried with high-performance liquid chromatography (HPLC). The uorogenic probes contained a reporter dye (FAM, 6-carboxy-uorescein) covalently linked at the 5 end and a quencher dye (TAMRA, 6-carboxytetramethyl-rhodamine) covalently attached at the 3 end. Extension from the 3 end was blocked by attachment of a 3-phosphate group.
cDNA standards

Materials and methods

Cell isolation and cell lines

Cell lines [Raji (ATCC CCL86) and HUT-78 (ATCC TIB161)] were cultured in R10, i.e. RPMI 1640 supplemented with 10% heat-inactivated foetal calf serum, 2 mm l-glutamine, 100 g ml1 penicillin and 100 g ml1 streptomycin (Invitrogen, Darmstadt, Germany). Peripheral blood mononuclear cells (PBMC) were isolated with Histoprep (BAG GmbH, Lich, Germany) from peripheral blood samples using standard methods.
DQA1 genotyping

As external controls for each target gene, plasmid recombinants containing the specic target sequence were generated for DQA1 alleles, as well as GAPDH and DRA1*01. For this purpose, total RNA from individuals positive for the allele of interest was extracted and reverse transcribed as described above. Following reverse transcription and allele-specic PCR, amplicons were cloned into pCR2.1 TOPO (Invitrogen Co., Carlsbad, CA). Recombinant plasmids were expressed in competent Escherichia coli (TOP 10F, Invitrogen). Plasmid DNA was isolated using silicea cartridges (QIAprep Spin Miniprep Kit Qiagen, Hilden, Germany). Sequences of the cloned amplicons were veried using an automated capillary sequencer (ABI PRISM 310, Perkin Elmer, Foster City, CA) with universal M13 primers. Concentrations of the recombinant plasmids were determined by optical density spectrometry (Eppendorf, Hamburg, Germany). Serial dilutions from the resulting clones were used for standardization, as described in detail in the manufacturers bulletin (Applied Biosystems, 1997).
PCR amplication and cDNA quantication

MHC class II polymorphism was investigated in 40 healthy unrelated volunteers, all of Caucasian origin. For DNA isolation, cell nuclei were prepared and DNA separated with guanidine isothiocyanate, followed by precipitation with isopropanol (Ciulla et al., 1988). PCR amplication of DQA1 alleles was carried out using exonspecic primers described elsewhere (Haas et al., 1994). DIG-11-ddUTP (digoxigenin-11-2-3-didesoxy-uridinetriphosphate) oligonucleotide labelling and detection were performed as described by Nevinny-Stickel & Albert (1993).
RNA extraction and cDNA synthesis

Total RNA was extracted from PBMC using TRIzol reagent (Invitrogen, Darmstadt, Germany). Total RNA was reverse transcribed using random hexamer priming and MMLV reverse transcriptase (RT) (ProSTAR FirstStrand RTPCR Kit, Stratagene, La Jolla, CA). One RNA sample of each preparation was processed without MMLV RT (RT reaction) to provide a negative control in subsequent PCR reactions.
Primers and probes

PCR reactions contained at a nal concentration: 300 nm forward and reverse primers, 200 nm TaqMan probe (Table 1), 200 m dATP, dCTP and dGTP, 400 m dUTP, 0.025 U l1 AmpliTaq Gold, 0.01 U l1 uracil-Nglycosylase and 2.5 l cDNA in a total volume of 25 l. Each PCR amplication was performed in triplicate wells using the following temperature and cycling prole: 50 C for 2 min and 95 C for 10 min, followed by 40 cycles of 95 C for 15 s and 60 C for 1 min (DQA1*04 primers: 66 C for 1 min, DQA1*05 primers: 62 C for 1 min). Reagents were obtained from Applied Biosystems (Weiterstadt, Germany). For quantication, an ABI PRISM 7700 Sequence Detection System (Perkin Elmer) was used (Gibson et al., 1996; Heid et al., 1996; Applied Biosystems, 1997; Knerr et al., 1999).
Detection of genomic DNA contamination

PCR primers and uorogenic probes were designed for all target genes according to the published sequences (Olerup et al., 1993; Marsh & Bodmer, 1995) using Primer Express software (Applied Biosystems, Foster City, CA) They were obtained from Eurogentec (Seraing, Belgium)

Exclusion from the PCR amplication of contaminating genomic DNA was accomplished by coamplication with a pair of primers located in the rst DQA1 intron (Table 1). The copy number of this intron PCR product corresponds to the number of genomic DNA molecules and it was thus used to estimate the genomic DNA content of the samples and compared with the copy number of the allele-specic products.

2003 Blackwell Publishing Ltd, European Journal of Immunogenetics 30, 141148

Quantication of HLA-DRA1 and -DQA1 expression by real-time RTPCR

143

Table 1. Primers and probes Amplicon length (bp) 105

Name DQA1*01 DQA01F DQA01R DQA01FATA DQA02F DQA02R DQA0203FATA DQA03F DQA03R DQA0203FATA DQA04F DQA04R DQA0405FATA DQA05F DQA05R DQA0405FATA DQA06F DQA06R DQA06FATA DQAF DQAR DQAFATA DQAintrF DQAintrR DQAintrFATA DRAF DRAR DRAFATA GAPDH3 GAPDH5 GAPDHFATA

Sequence (5 3) GAAGGAGACTGCCTGGCG ATGATGTTCAAGTTGTGTTTTGC* CAAATTTGGAGGTTTTGACCCGCAGG ACGGTCCCTCTGGCCAGTT* TTGCGGGTCAAATCTAAGTCTGT ATGAATTTGATGGAGACGAGGAGTTCTATGTGG GGTCCCTCTGGGCAGTACAG CAAATTGCGGGTCAAATCTTCT* ATGAATTTGATGGAGACGAGGAGTTCTATGTGG GAGCAGTTCTACGTGGACCTGG GGAACCTCATTGGTAGCAGCA* ACTGTCTGGTGTTTGCCTGTTCTCAGACAA AGATGAGCAGTTCTACGTGGACC AGAGTTGGAGCGTTTAATCAGAC* ACTGTCTGGTGTTTGCCTGTTCTCAGACAA ACGGTCCCTCTGGCCAGTT* CGGGTCAAATCTAAATTGTCTGAGA* AATTTGATGGAGACGAGCAGTTCTACGTGGA CACAGCTCAGAGCAGCAACTG AGCCACAATGTCTTCACCTCCA CCTTGGGAAGAGGATGATCCTAAACAAA GTTGCCCGTTTCTTTCTCTCA TGGACTCCTTTACCCACTCCC ATTTCCACATGGGAACTGGCACAGGT GGACAAAGCCAACCTGGAAA AGGACGTTGGGCTCTCTCAG TACTCCGATCACCAATGTACCTCCAGAG GCCATCAATGACCCCTTCATT TTGACGGTGCCATGGAATTT CCTCAACTACATGGTTTACATGTTCCAATATGATTCCAC

Length (mer) 18 23 26 19 23 33 20 22 33 22 21 30 23 23 30 19 25 31 21 22 33 21 21 26 20 20 34 21 20 39

Tm (C)
58 56 69 60 59 69 58 59 69 60 59 68 58 56 68 60 60 70 59 60 70 58 59 68 59 58 69 60 60 70

DQA1*02

122

DQA1*03

127

DQA1*04

170

DQA1*05

153

DQA1*06

119

DQA1 (total DQA1 primer) DQA1 (intron primer) DRA1

127

80

120

GAPDH

89

* Primers modied according to the sequences published by Olerup et al. (1993). Forward (F) and reverse (R) primers for DQA1*01, *02, *03, *05 and *06 and DRA1*01 are located in a single exon; primers for DQA1*04 and GAPDH are located in two exons, thus spanning one intron. Fluorogenic probes (FATA) are FAM-labelled at the 5-end and TAMRA-labelled at the 3-end.

Relative expression of different DQA1 alleles

The relative expression of DQA1 alleles was determined with reference to the total amount of HLA-DQA1 mRNA after normalization against GAPDH as implemented in the ABI PRISM 7700 Sequence Detection System software (Applied Biosystems, 1997). Results were considered only if the analysis for allele 1 (e.g. DRA1) showed all reactions to have the same amount of amplication as allele 2 (e.g. GAPDH). This procedure allowed comparison of group-specic HLA-DQA1 expression as well as the total expression levels of HLA-DQA1 and -DRA1.

Results
Design of primers and probes

For the allele-specic quantication of DQA1 gene products, a real-time RTPCR set of primer pairs (n = 6) and probes (n = 6) was designed and optimized (Table 1). HLA-DQA1 specic primers and probes were designed to encompass allele group-specic detection of HLA-DQA1

gene expression. Primers were designed based on the DQA1 genotyping system described by Olerup et al. (1993). The primers for DQA1*01, *02, *03, *05 and *06 were located in the hypervariable region within exon 2 (see Fig. 1). For DQA1*04, exon-spanning primer pairs were employed. For normalization, GAPDH and HLADRA1 gene expression were used as endogenous reference. Primers for DRA1 were also located within the second exon, while for GAPDH exon-spanning primer pairs were employed. Melting temperatures (Tm) were chosen to range from 56 to 60 C (see Table 1). Probes were constructed to have an annealing temperature at least 10 C higher than the primer Tm, thus approximately 6870 C. The amplicon lengths were kept between 71 and 173 bp. The reproducibility of the assay was determined by quantication of different cDNA target samples carrying the same allele. Quantications were obtained by performing individual experiments with repeated runs of the same preparation. As shown in Table 2, the standard error of the mean (SEM) was within 310%. The specicity of DQA1 allele-specic amplication in different homozygous and heterozygous combinations was conrmed by the analysis of a panel of 51 individuals.

2003 Blackwell Publishing Ltd, European Journal of Immunogenetics 30, 141 148

144

S. Fernandez et al.

Figure 1. Sequences of HLA-DQA1 alleles, showing the sequence alignment of published (Marsh & Bodmer, 1995) nucleotide sequences containing the rst 100 amino acids (exons 13) of HLA-DQA1 alleles *0101 to *0601. Numbers above the sequences refer to the amino acid position. Nucleotide sequences were aligned to the DQA1*0101 allele. Dashes indicate identity of nucleotides. The reverse 3-primer DQA04R extends into part of the third exon. Primers are shaded dark grey, and allele-specic probes light grey.

Table 2. Reproducibility of the assay, showing quantication of DQA1*05 expression by real-time PCR in three different cDNA target samples Sample 1 Assay 1A 1B 1C 1D Mean SEM 2A 2B 2C Mean SEM Mean SEM (total) Ct 24.98 25.03 25.12 24.88 22.65 22.81 22.76 Copy number 72250 69516 64963 77398 71032 2600 74863 67323 69647 70611 2229 70851 1628 Ct 25.98 26.06 26.04 26.03 23.71 23.83 23.75 Sample 2 Copy number 34019 32091 32604 32692 32852 411 37799 35032 36689 36507 803 34418 828 Ct 27.05 27.15 27.13 27.12 25.07 24.43 24.64 Sample 3 Copy number 15241 14208 14332 14429 14552 234 15668 23684 20707 20020 2339 16896 1420

All quantications were obtained by performing individual experiments with quadruplicate (1A1D) and triplicate (2A2C) runs of the same preparation.

Criteria for inclusion in the analysis were: (i) HLA-DQA1 genotyping, (ii) positive result for GAPDH and DRA, and (iii) positive result for DQA1 expression for at least one DQA1 allele. Six individuals did not fulll these criteria and were thus excluded. The other 45 individuals covered a broad range of DQA1 alleles and allelic combinations (see Table 3). There were ve individuals each homozygous for DQA1*01 and DQA1*05. In the remaining 35 individuals and both cell lines, gene expression of two different DQA1 alleles was seen. There were no discrepancies between expression data and the DQA1 genotype. The rare allele DQA1*0601 was not represented in our non-selected study population.

Quantication

Standard curves for the direct quantication of the cDNA levels of HLA-DQA1 and -DRA1 were established using serial dilutions of the corresponding recombinant plasmid clones. For each standard curve, a linear range of concentrations covering 6 log units (102108) was employed (Fig. 2). The precision and reproducibility of amplication were conrmed for input amount of copies ranging from 108 down to 50 copies. Standard errors of the mean (SEM) for assay variation (Table 2) were calculated with maximums of 8.4% (interassay) and 11.7% (intra-assay). At 102 copies, the minimum input amount of cDNA was

2003 Blackwell Publishing Ltd, European Journal of Immunogenetics 30, 141148

Quantication of HLA-DRA1 and -DQA1 expression by real-time RTPCR

145

Figure 2. Efciency of PCR amplication in Raji and plasmids. Five-fold serial dilutions of target (cDNA prepared from Raji cell lines) and control cDNA (cloned DQA1 cDNA prepared from plasmids) were carried out. Triplet amplication was performed with HLA-DQA1 specic primers using the ABI Prism 7700 Detection System. Obtained Ct values are plotted against the relative amount of input copies.

Figure 3. Ratio of expression of different DQA1 and DRA1 alleles. Ratios (given in numbers above the bars) of expression were calculated from the concentrations of the different DQA1 alleles, and the total amounts of DQA1 (t-DQA1), DRA1 (t-DRA1) and GAPDH mRNA. The numbers of samples tested are given in parenthesis.

determined to be above 102 copies with a markedly increased intra-assay variation (data not shown).
Genomic DNA contamination

Comparison of copy numbers of amplied DQA1 intron sequence with DQA1 exon 2 allele-specic amplications indicated contamination of the cDNA with genomic DNA. The contamination rate was found to range from 0.01 to 4.2% (mean 0.8%, SEM 0.7).
Relative expression of different DQA1 alleles

rates of expression of particular DQA1 alleles in different heterozygous combinations, DQA1 interallelic ratios were determined in heterozygous individuals. The results are summarized in Table 3. Taking DQA1*01 as an example, the relative expression ranged from 0.12 with DQA1*03 expressed on the second haplotype up to 0.2 with DQA1*02 on the second haplotype. Another example is the combination of HLA-DQA1*01 and *04, with eight individuals found to be positive. While DQA1*04 showed a relative expression of 0.37 (SEM 0.13), DQA1*01 showed a relative expression of 0.13 (SEM 0.02) compared with GAPDH.

HLA-DQA1 was found to have 4.15-fold lower average expression than HLA-DRA1 (Fig. 3). The analysis of the expression levels of individual HLA-DQA1 alleles showed that DQA1*04 expression was signicantly increased (ratio 0.37) compared to the expression of GAPDH. All other HLA-DQA1 alleles, namely DQA1*01, *02, *03 and *05, showed very similar ratios, from 0.09 to 0.14 relative to GAPDH expression. Thus, the relative expression of HLA-DQA1*04 was observed to be 2.6 4fold higher than that of any other HLA-DQA1 allele. Relative to HLA-DRA1, the expression of HLA-DQA1*04 was observed to be 2.2-fold higher. In homozygous individuals (DQA1*01/*01 and *05/ 05), expressions levels for individual alleles were not higher than their corresponding levels in heterozygous individuals (DQA1*01/*05). In order to examine relative

Discussion
Recent in vitro studies have suggested that variability of cis-acting elements may contribute to the differential regulation of HLA-DQA1 expression (Kimura & Sasazuki, 1992; Morzycka-Wroblewska et al., 1997). Nevertheless, the relevance of DQA1 promoter polymorphism for allele-specic differences in DQA1 expression in vivo remains to be determined (Cesari et al., 1999). For this purpose, we have developed a real-time RTPCRbased system to allow quantication of HLA-DQA1 gene expression in an allele group-specic manner in heterozygous combinations without any in vitro manipulation. Studies on DQA1 gene expression in homozygous cells have shown differences in DQA1 expression to correlate with nucleotide substitutions in conserved

2003 Blackwell Publishing Ltd, European Journal of Immunogenetics 30, 141 148

146

S. Fernandez et al.

Table 3. Relative expression of DQA1 alleles in 45 individuals. Mean values obtained from analyses of relative HLA-DQA1 gene expression comparing different haplotypic combinations are shown. Gene expressions are given relative to GAPDH DQA1 *01 *02 *03 *04 *05 *06 *01 0.17 (n = 5) 0.20 (n = 3) 0.12 (n = 7) 0.13 (n = 8) 0.13 (n = 13) SEM 0.08 0.03 0.03 0.02 0.07 *02 0.10 (n = 3) 0.06 (n = 2) 0.01 SEM 0.02 *03 0.09 (n = 7) 0.07 (n = 2) 0.02 SEM 0.03 *04 0.37 (n = 8) SEM 0.23 *05 0.10 (n = 13) 0.07 (n = 2) 0.05 (n = 2) 0.09 (n = 5) SEM 0.05 0.02 0.02 *06 0.08

SEM = standard error of the mean.

regulatory boxes of the DQA1 promoter region (Morzycka-Wroblewska et al., 1997). Since polymorphic sites within the W-, X- andY-boxes of the DQA1 promoter have been shown to be in strong linkage disequilibrium with DQA1 allele groups including DQA1*01, DQA1*02, DQA1*03, DQA1*04, DQA1*05 and DQA1*06, the level of qualitative resolution for quantication was chosen accordingly (Haas et al., 1994) Thus, a set of six primers and probes was constructed allowing the quantication of DQA1*01 (including *0101, *0106), DQA1*02 (including *0201), DQA1*03 (including *0301, *0302), DQA1*04 (including *0401), DQA1*05 (including *0501) and DQA1*06 (including *0601) in homozygous and heterozygous combinations. Because exon-specic primers were used, genomic DNA contamination was monitored by amplication of DQA1 intron sequences in parallel. To allow reproducible standardization and quantication, a panel of recombinant plasmid clones containing the DQA1 exon 2 target sequences was established and used for the determination of sensitivity and specicity of amplication. When this panel was applied, we were able to show linearity of amplication efciency in a range above 6 log units. Moreover, the sensitivity and specicity were found to correspond to those obtained using real-time RTPCR-based quantication systems targeting other genes (Heid et al., 1996; Knerr et al., 1999). Since HLA-DRA1 shows only limited polymorphism in exon 2 and in the promoter region and is constitutively expressed in MHC class II positive cells (Abdulkadir et al., 1995), HLA-DRA1 expression was used as a marker for HLA class II expression. Comparison of the total amounts of HLA-DRA1 and -DQA1 expression in PBMC showed HLA-DRA1 expression to be on average 4.15-fold higher in all individuals tested. This relationship corresponds to the relative proportion of cell surface expression of HLA-DR and -DQ (Brooks & Moore, 1988). The difference in the expression of HLA-DR and -DQ genes may result from structural differences in cis-acting elements which in turn lead to differential binding and gene activation via trans-acting elements, as has

recently been observed for the activation of HLA-DR vs. -DQ by class II transactivator (CIITA) in hematopoietic cells (Liu et al., 1999). Structurally, the lower expression of the DQA1 gene has been linked to the difference in the Y-box sequence (YC-box) unique to the HLA-DQA1 gene (Auffray et al., 1987; Kimura & Sasazuki, 1992; MorzyckaWroblewska et al., 1997). The interallelic comparison of HLA-DQA1 alleles indicated rather homogeneous levels of expression in PBMC from healthy donors. Surprisingly, expression in homozygous individuals was not observed to be signicantly increased as compared to heterozygous combinations, thus excluding signicant gene dosage effects at the mRNA level. Although only a limited number of heterozygous combinations could be analysed for the purpose of establishing the real-time RTPCR detection system, no DQA1 allele hierarchy for the expression of particular HLA-DQA1 allelic groups was seen when different heterozygous combinations were compared. However, a consistency in expression patterns was seen for the DQA1*04 and DQA1*01 alleles, as in all DQA1*04/*01 heterozygotes DQA1*04 was expressed at a higher level than DQA1*01. Interestingly, DQA1*04 carries a promoter (QAP 4.1) which is very similar to that found with DQA1*05 and *06 (QAP 4.2). QAP 4.1 and QAP 4.2 carry the YM-box characterized by a second nucleotide exchange compared to the HLA class II consensus sequence (Haas et al., 1994). However, QAP 4.2 as compared with the sequence of QAP 4.1 carries some remarkable nucleotide substitutions apart from those in the Y-box. One is a T to G mutation at position 134, and the other is a G to T mutation at position 209. Both regions have been determined to be located in the cisacting sequence for a transcription factor regulated by tumour necrosis factor (TNF)- (Kimura & Sasazuki, 1992). In vitro studies with cell lines carrying the QAP 4.1 or the QAP 4.2 promoter showed decreased levels of expression in the presence of the mutated YM-box found in both promoter alleles (Morzycka-Wroblewska et al., 1997; Indovina et al., 1998). This suggests other regions located within the DQA1 promoter region might be

2003 Blackwell Publishing Ltd, European Journal of Immunogenetics 30, 141148

Quantication of HLA-DRA1 and -DQA1 expression by real-time RTPCR

147

relevant to the observed differences in gene expression. Indovina et al. (1998) recently reported expression data from a heterozygous lymphoblastoid cell line positive for HLA-DQA1*04 and 02. In view of the strong linkage disequilibrium between DQA1 alleles within the promoter and the second exon, this cell line is likely to be positive for QAP 4.1 and QAP 2.1. The authors demonstrated decreased expression of DQA1*04 in this homozygous lymphoblastoid cell line using competitive RTPCR and northern blotting (Indovina et al., 1998). The expression of DQA1 alleles was previously analysed by transfection assays or semiquantitative RTPCR in mostly lymphoblastoid cell lines. In this study, we have used bulk PBMCs to explore the differential expression patterns seen in heterozygous individuals. With respect to previous reports, differences may relate to the fact that, in transformed cell lines, HLA class II genes may be regulated differently, and that PBMCs constitute a heterogeneous group of cell types. Cell type-specic differences may therefore exist and the overall expression may consequently be altered compared with that found for isolated cell types. This, however, remains to be determined. Regulation of HLA-DQ gene transcription is a complex phenomenon. Allelic polymorphism is present in the DQA1 (encoding the -chain) as well as in the DQB1 (encoding the -chain) genes. Moreover, both genes have been found to have allelic polymorphism within their URRs. Strong linkage disequilibria have been observed between QAP and DQA1 alleles (Haas et al., 1994) as well as between QBP and DQB1 alleles (Reichstetter et al., 1996). As both genes may be subject to allelic regulation, complex patterns of interaction at the level of gene expression, pairing and surface expression may be present, inuencing the level of allelic mRNA expression. HLA-DQ molecules are associated with susceptibility to autoimmune diseases such as diabetes mellitus type 1, JIA and coeliac disease. Moreover, particular haplotypic combinations have been shown to be of central importance in disease susceptibility and protection. Because of the very close proximity of the HLA-DR and -DQ genes there is still a debate over whether HLA-DR or HLA-DQ is the more relevant association in JIA (Smerdel et al., 2002). A hierarchy of MHC class II associations has been observed in JIA, where DQA1 alleles (*0401, *0501 and *0601) are responsible for disease association while DQB1 alleles (susceptibility: *0301 and *0402; protection: *0201) are crucial for the gene effect. Under the hypothesis that quantitative differences in allelic expression may lead to alterations in the composition of the functionally active heterodimer, which in turn may give rise to qualitative differences in the ability of the - heterodimer to bind and present peptides, the study of allelic expression of DQA1 alleles may contribute to our understanding of the role of HLA-DQ in autoimmunity.

and by the BMFT-funded Center for Interdisciplinary Clinical Research at the Friedrich Alexander University Erlangen-Nrnberg [IZKF Erlangen (01 KS 9601)/project B17, B31 and C8]. We would like to thank the healthy volunteers who participated in the study. The generous provision of cell samples with rare DQA1 alleles by A. McNickolas and E. D. Albert, Immunogenetic Laboratory, Ludwig Maximillians University, Munich is gratefully acknowledged.

References
Abdulkadir, S.A., Krishna, S., Thanos, D., Maniatis, T., Strominger, J.L. & Ono, S.J. (1995) Functional roles of the transcription factor Oct-2A and the high mobility group protein I/Y in HLA-DRA gene expression. Journal of Experimental Medicine, 182, 487. Applied Biosystems (1997) User Bulletin #2: ABI PRISM 7700 Sequence Detection System. Applied Biosystems, Weiterstadt, Germany. Andersen, L.C., Beaty, J.S., et al. (1991) Allelic polymorphism in transcriptional regulatory regions of HLA-DQB genes. Journal of Experimental Medicine, 173, 181. Auffray, C., Lillie, J.W., et al. (1987) Structure and expression of HLA-DQ alpha and -DX alpha genes: interallelic alternate splicing of the HLA-DQ alpha gene and functional splicing of the HLA-DQ alpha gene using a retroviral vector. Immunogenetics, 26, 63. Brooks, C.F. & Moore, M. (1988) Differential MHC class II expression on human peripheral blood monocytes and dendritic cells. Immunology, 63, 303. Cesari, M., Caillens, H., et al. (1999) In vivo analysis of HLA-DQ gene expression in heterozygous cell lines. Immunogenetics, 50, 309. Ciulla, T.A., Sklar, R.M., et al. (1988) A simple method for DNA purication from peripheral blood. Analytical Biochemistry, 174, 485. Del Pozzo, G., Perfetto, C., et al. (1992) DNA polymorphisms in the 5-anking region of the HLA-DQA1 gene. Immunogenetics, 35, 176. Edwards, J.A., Durant, B.M., et al. (1986) Differential expression of HLA class II antigens in fetal human spleen: relationship of HLA-DP, DQ, and DR to immunoglobulin expression. Journal of Immunology, 137, 490. Gibson, U.E., Heid, C.A., et al. (1996) A novel method for real time quantitative RTPCR. Genome Research, 6, 995. Haas, J.P., Kimura, A., et al. (1994) Polymorphism in the upstream regulatory region of DQA1 genes and DRB1, QAP, DQA1, and DQB1 haplotypes in the German population. Human Immunology, 39, 31. Haas, J.P., Kimura, A., et al. (1995) Early-onset pauciarticular juvenile chronic arthritis is associated with a mutation in the Y-box of the HLA-DQA1 promoter. Tissue Antigens, 45, 317. Heid, C.A., Stevens, J., et al. (1996) Real time quantitative PCR. Genome Research, 6, 986. Indovina, P., Megiorni, F., et al. (1998) Different binding of NF-Y transcriptional factor to DQA1 promoter variants. Human Immunology, 59, 758. Kimura, A. & Sasazuki, T. (1992) HLA-DQA gene is differently regulated from other HLA class II genes. In: Molecular Approaches to the Study and Treatment of Human Diseases (ed. by T. O. Yoshida and J. M. Wilson), p. 97. Elsevier Science Publishers, Cambridge. Knerr, I., Repp, R., Dotsch, J., Gratzki, N., Hanze, J., Kapellen, T. & Rascher, W. (1999) Quantitation of gene expression by real-time PCR disproves a retroviral hypothesis for childhood-onset diabetes mellitus. Pediatrics Research, 46, 57.

Acknowledgements
This work was supported by the Deutsche Forschungsgemeinschaft (grant nos DFG-HA 2306 and SFB263/C8)

2003 Blackwell Publishing Ltd, European Journal of Immunogenetics 30, 141 148

148

S. Fernandez et al.

Liu, A., Takahashi, M., et al. (1999) Regulation of the expression of MHC class I and II by class II transactivator (CIITA) in hematopoietic cells. Hematological Oncology, 17, 149. Marley, N.J., Macartney, J.C., et al. (1987) HLA-DR, DP and DQ expression in the small intestine of patients with coeliac disease. Clinical and Experimental Immunology, 70, 386. Marsh, S.G. & Bodmer, J.G. (1995) HLA class II region nucleotide sequences, 1995. Tissue Antigens, 46, 258. Morzycka-Wroblewska, E., Munshi, A., et al. (1997) Differential expression of HLA-DQA1 alleles associated with promoter polymorphism. Immunogenetics, 45, 163. Nevinny-Stickel, C. & Albert, E.D. (1993) HLA class II typing in a microtitre plate format using digoxigenin-labelled amplied DNA and biotin-labelled oligonucleotide probes. European Journal of Immunogenetics, 20, 419.

Olerup, O., Aldener, A., et al. (1993) HLA-DQB1 and -DQA1 typing by PCR amplication with sequence-specic primers (PCR-SSP) in 2 hours. Tissue Antigens, 41, 119. Perfetto, C., Zacheis, M., et al. (1993) Polymorphism in the promoter region of HLA-DRB genes. Human Immunology, 36, 27. Reichstetter, S., Brunnler, G., et al. (1996) DQB1 promoter sequence variability and linkage in caucasoids. Human Immunology, 51, 73. Reith, W., Steimle, V., et al. (1995) Regulation of MHC class II gene expression. Immunobiology, 193, 248. Smerdel, A., Ploski, R., et al. (2002) Juvenile idiopathic arthritis (JIA) is primarily associated with HLA-DR8 but not DQ4 on the DR8-DQ4 haplotype. Annals of Rheumatic Disease, 61, 354.

2003 Blackwell Publishing Ltd, European Journal of Immunogenetics 30, 141148