ESHG Posters 1

ESHG - Posters: P 1 Analysis of Disorders and Traits with Complex Inheritance

An association study of schizophrenia and the human cadherin G-type receptor Celsr1

L. N. Georgieva¹, N. Poriazova², I. Nikolov¹, G. Jones³, D. Toncheva¹, M. J. Owen³, G. Kirov³;
¹Department of Medical Genetics, Medical University Sofia, Sofia, BULGARIA, ²Psychiatric Dispensary, Higher Medical Institute, Plovdiv, BULGARIA, ³Neuropsychiatric Genetics Unit, Department of Psychological Medicine, University of Wales College of Medicine, Cardiff, UNITED KINGDOM.

Cadherins play a critical role in morphogenesis and maintenance of neuronal connections in the adult brain. We examined a member of the non-classic type seven-pass transmembrane cadherins, the human homologue of Drosophila Flamingo gene - Celsr1. It maps to chromosome 22q13.31, a region of positive linkage results in schizophrenia and manic-depressive illness. The gene contains nine cadherin ectodomain repeats-seven EGF-like repeats, two laminin A G-type repeats coupled to a seven-pass transmembrane domain. Celsr1 is a neural-specific gene that plays a role in early embryogenesis, cell adhesion and signal transduction.
The first exon is 3,544nt in length and encodes for the signal peptide and all nine ectodomains in the protein. We screened this exon in 24 schizophrenic patients with dHPLC followed by sequencing. Three amino-acid changes were identified and submitted to HGBASE: CTG1665GTG (L555V, SNP001026397), TCG1989TGG (S663W, SNP001026398) and CGC3375TGC (R1125C, SNP001026399). An R119G SNP from public databases (TSC0242402) was not polymorphic in this population. The three SNPs were genotyped using primer extension on ABI373 sequencers on a sample of 243 Bulgarian schizophrenic parent-offspring trios from 243 nuclear families, as well as 179 schizophrenics and 163 matched controls from UK. The three SNPs were in complete LD. There was no preferential transmission of alleles from heterozygous parents to affected offspring when analyzed with TDT. In the UK population the rare alleles were even more common in controls, the difference almost reaching statistical significance. We conclude that variations in the nine ectodomains of Celsr1 do not increase susceptibility to schizophrenia.

Frequencies for the rare alleles (%)
	SZ patients, BG (N = 243)	Parents of SZ patients, BG (N = 466)	SZ patients, UK (N = 179)	Controls, UK (N = 163)
L555V	3.9	3.1	2.8	3.7
S663W	7.7	7.2	4.5	7.7
R1125C	7.9	7.5	4.5	8.0

Longitudinal Analysis of Heteroplasmy Levels in Families with Leber Hereditary Optic Neuropathy (LHON)

A. Puomila¹, T. Viitanen¹, M. Savontaus¹, E. Nikoskelainen², K. Huoponen¹;
¹University of Turku, Turku, FINLAND, ²Turku University Central Hospital, Turku, FINLAND.

Leber hereditary optic neuropathy (LHON) is a maternally inherited ocular disease characterised by acute or subacute bilateral loss of central vision. It is associated with point mutations in the mitochondrial DNA (mtDNA). 14 % of LHON families are heteroplasmic, with a mixture of wild-type and mutant mtDNA existing within the same individual. To study whether there is any preferential selection of either the wild-type or the mutant mtDNA within individuals over time, an extensive longitudinal analysis of the segregation of the primary LHON mutations ND1/3460 and ND4/11778 was accomplished. Blood samples from 9 heteroplasmic individuals from 4 LHON families were studied over a time period of 4 to 12 years. In addition to blood samples, hair follicle and urinary tract epithelium samples of one patient were also analysed. The quantification of heteroplasmy was performed using the solid-phase minisequencing method. In one individual, the proportion of the mutant mtDNA decreased from 47 % to 42 % (p=0,034*) in 12 years. In 8 individuals, only minor changes were observed but they did not reach statistical significance. The changes did not occur in a systematic manner, suggesting absence of any simple and general selection mechanism for or against the mutant mtDNA in LHON. The various outcomes of the segregation processes can be explained by random genetic drift and/or thus far unknown nuclear control ruling the segregation of the mtDNA in each individual.

No evidence for association between a TSSP polymorphism and coeliac disease in French population.

F. Clot¹, J. Hugot², M. C. Babron¹, F. Clerget-Darpoux¹;
¹INSERM U.535, Le Kremlin-Bicêtre, FRANCE, ²CEPH, Paris, FRANCE.

Celiac disease (CD) is a chronic inflammatory disease of the gut resulting from ingestion of gluten, occurring in genetically susceptible individuals. The strong genetic association of CD with HLA heterodimers is well known, but there is evidence for the involvement of additonal genetic risk factors in the HLA region. Lie et al (1999) suggested that a gene in the vicinity of D6S2223 could play a role in the pathogenesis of the disease. The TSSP gene (Thymus Specific Serine Protease), located in this region, is a good candidate because it is expected to play a role in antigen processing and presentation pathway in cortical thymic epithelial cells where it is specifically expressed. We sequenced the 12 exons of the gene and we identified 1 exonic polymorphism, a deletion of 15 bp in exon 12, at 8 bp before the stop codon (1520-1544delAGAGCCAGATTAAGG). We developed a genotyping method based on the electrophoresis on an agarose gel (2%) of PCR products and we analysed 130 French CD trios using the transmission disequilibrium test. Among 89 heterozygous parents, the deletion was transmitted in 55% of the cases (chi2=0.91, p=0.34). This result does not indicate a role of TSSP gene in the predisposition to CD. However, the identification of this deletion may be useful in the study of other auto-immune disorders.
This study was funded by the Commission of the European Communities (QLRT-1999-00037).

J. P. Hugot¹, F. Clot², S. Lesage¹, H. Zouali¹, M. C. Babron², F. Clerget-Darpoux²;
¹CEPH, Paris, FRANCE, ²INSERM U.535, Le Kremlin-Bicêtre, FRANCE.

Celiac Disease (CD) is a gluten-sensitive enteropathy characterized by malabsorption and mucosal injury of the small bowel. The disease is associated wih both genetic and environmental risk factors. Recently, the CARD15/NOD2 gene, encoding a member of the CED4/APAF1 family of apoptosis regulators, was reported to be involved in Crohn’s disease susceptibility, a chronic inflammatory disorder of the gastrointestinal tract which is associated with CD. We therefore tested the involvement of CARD15/NOD2 gene in the susceptibility to CD. Three main variants associated with Crohn’s disease were genotyped in twenty five simplex celiac families for the R702W and G908R variants and in 71 simplex celiac families for the 1007fs variant. The genotyping methods were an allele-specific PCR assay for R702W, a HhaI enzyme digestion of PCR products for G908R, and an electrophoresis on an acrylamide gel of labelled PCR products for 1007fs. Genotyping data were analysed using the transmission disequilibrium test. The frequencies of the R702W, G908R and 1007fs alleles were respectively 0.06, 0 and 0.01 in CD patients. These frequencies were not different from a control population. For R702W, only 5 parents were heterozygous and the allele was transmitted in 60% of the cases. For G908R, there were no heterozygous parents. For 1007fs, only 4 parents were heterozygous and the insertion was transmitted in 50% of the cases. Altogether, these observations do not indicate that the CARD15/NOD2 gene is a major risk factor for CD.
This study was funded by the Commission of the European Communities (QLRT-1999-00037).

E. Giardina¹, F. Capon¹^,2, A. Tacconelli¹, S. Chimenti¹, G. Zambruno³, G. Novelli¹;
¹Tor Vergata University, Rome, ITALY, ²University of Leicester, Leicester, UNITED KINGDOM, ³IDI-IRCCS Institute, Rome, ITALY.

Psoriasis, (PS, OMIM 177900) is an inflammatory skin disorder affecting approximately 2% of Caucasians. PS is widely regarded as a complex trait and genome-wide scans have mapped a number of loci (PSORS 1-7) contributing to disease susceptibility. We have assigned the PSORS4 locus to chromosome 1q21, and have recently refined the susceptibility interval to 100 kb. This minimal region contains the gene for loricrin (LOR), a keratinocyte structural protein. LOR is homologous to corneosdesmosin (CDSN), an extensively investigated positional candidate for the PSORS1 locus. Interestingly, both LOR and CDSN genes show an altered expression in psoriatic lesions.
We performed a genetic analysis of the LOR gene in a sample of Italian psoriatic trios. We first re-sequenced the LOR coding region and its 5'/3' UTRs in 8 patients and identified 2 novel SNPs and a 6bp in-frame duplication. We therefore analysed these variants, as well as a previously published 12 bp duplication, in 90 trios, each including an affected offspring and both parents. We tested for association using the Transmission Disequilibrium Test and by assessing deviation from Hardy-Weinberg equilibrium (HWE). This latter test identified a significant heterozygote excess (p = 1.8 x 10-4 ) for a coding SNP in exon 2. The analysis of 40 unrelated healthy controls confirmed that the SNP is in equilibrium among unaffected, indicating that the deviation from HWE observed in patients is likely to be disease-related.
Work funded by the Italian Ministry of Health

M. Shadrina¹, E. Kondratyeva¹, P. Slominsky¹, G. Levitsky², V. Skvortsova², S. Limborska¹;
¹Institute of Molecular Genetics, Moscow, RUSSIAN FEDERATION, ²Russian State Medical University, Moscow, RUSSIAN FEDERATION.

At present the cause of amyotrophic lateral sclerosis (ALS) remains uknown. The main known ALS-causing gene is the CuZn-superoxidedismutase gene. However, ALS is complex disease and other genetic systems may be involvement in the pathogenesis of this disease.Autopsy studies have revealed aggregation and abnormal assembly of neurofilaments (NF) in the perikarya and proximal axon of motor neurons in ALS. The potential importance of NF is underscored by the observation abnormal accumulation NF in ALS patients with various mutations in CuZn-SOD. There are the tail domains of neurofilament subunits of medium and heavy molecular weight (NFM and NFH), which contains a repeated amino asid motif. In human, there are two common variants of the NFH tail, one with 45 repeats and named long (L) allele, another with 44 repeats and named short (S) allele. Previous studies have been shown that NFH tail may be involved in the pathogenesis of ALS. To investigate whether L and S allele genotype associated with ALS, we study allelic frequency in 52 patients with SALS in Moscow and control unrelated population from Russia. We have found 17 patients with LL genotype, 18 patients with LS genotype and 17 patients SS genotype, compared with 14 SS, 28 LS and 3 SS in unrelated controls. Sufficient differences in SS genotype frequency between control population and patients were observed (X²=9.97, p< 0.005). We conclude that SS genotype of NFH gene, probably, is associated with the pathogenesis of ALS.

Broad phenotype of child speech disorder shows strong evidence of linkage at candidate gene region 7q31

J. H. Schick¹, A. M. Kundtz¹, H. K. Tiwari¹^,2, H. G. Taylor³, L. A. Freebairn³, L. D. Shriberg⁴, B. A. Lewis³, S. K. Iyengar¹;
¹Department of Epidemiology & Biostatistics, Case Western Reserve University, Cleveland, OH, ²Department of Biostatistics, University of Alabama Birmingham, Birmingham, AL, ³Department of Pediatrics, Rainbow Babies & Childrens Hospital, Case Western Reserve University, Cleveland, OH, ⁴Waisman Center on Mental retardation & Human Development, University of Wisconsin-Madison, Madison, WI.

Although the etiologies of many child speech-sound disorders are largely unknown, family and twin studies have indicated a significant genetic component for one of the more rare subtypes. A locus (SPCH1) with an autosomal dominant mode of inheritance for developmental apraxia of speech in combination with a grammatical deficit was previously localized to chromosome 7q31. We tested the hypothesis that a candidate locus on 7q31 also segregates for more common subtypes of speech disorder in 89 families (n = 196 sib pairs) ascertained through pre-school probands with moderate-to-severe speech-sound disorders of unknown origin. Our analysis included 18 markers spanning a 25.66 cM region on 7q31 positioned by using a map from the Weizmann Institute (http://bioinformatics.weizmann.ac.il). We conducted an affected sib-pair analysis using SIBPAL2 (S.A.G.E.© 4.0). Speech disorder was treated as a binary trait and adjusted for age and socio-economic status. Multi-point linkage results suggest speech disorder locus located near a 5 kb region containing markers D7S1812 (p < 0.006), D7S821 (p < 0.002), and D7S1796 (p < 0.003). This study provides support for the earlier hypothesis that a putative gene for speech disorders localizes to 7q31. Our studies are now being expanded to include continuous phenotypes based on a suite of metrics obtained from conversational speech samples. The next research phase will be to perform SNP analyses and radiation hybrid mapping to isolate the first putative gene for the most common type of child speech-sound disorder. Supported by NIH grants NIDCD-00528 and NIDCD-04005-01.

Chronic recurrent multifocal osteomyelitis (CRMO): evidence for a susceptibility gene located on chromosome 18q21.3-18q22

A. Golla¹, A. Jansson², J. Ramser³, H. Hellebrand³, T. Meitinger⁴, B. H. Belohradsky², A. Meindl⁵;
¹IMBIE, University of Bonn, Bonn, GERMANY, ²Dr. von Hauner´sches Kinderspital, University of Munich, München, GERMANY, ³Abteilung Medizinische Genetik, University of Munich, München, GERMANY, ⁴Institut für Humangenetik, TU Munich and GSF, Neuherberg, München, GERMANY, ⁵Abteilung Mediznische Genetik, University of Munich, München, GERMANY.

Chronic recurrent multifocal osteomyelitis (CRMO) is characterized by recurrent inflammatory lesions in the metaphyses of long bones and usually affects children and adolescents. Similarity with an autosomal recessive mouse disorder (cmo, chronic multifocal osteomyelitis) prompted us to perform a family based association study with two markers on chromosome 18q in the region homologous to the cmo localisation of the mouse. We found a significant association of CRMO with a rare allele of marker D18S60, resulting in a haplotype relative risk (HRR) of 18. This suggests the existence of a gene in this region contributing in a significant manner to the etiology of CRMO and concomitantly demonstrates evidence for a genetic basis of CRMO for the first time. This gene is not identical with RANK, which is mutated in familial expansile osteolysis (FEO), because no mutations were detected in RANK. Mutation search in RANK and the genes PIGN and KIAA1468 lead to detection of two variants (one in RANK and one in PIGN), which are in linkage disequilibrium with the rare D18S60 allele, but not independently associated with CRMO.

LDL receptor-related protein (LRP) expression pattern in male coronary patients: Inverse regulation on transcriptional and translational level

S. Schulz¹, G. Birkenmeier², P. Greiser¹, T. Süß¹, D. Rehfeld¹, A. Nordwig¹, U. Schagdarsurengin¹, K. Werdan³, C. Gläser¹;
¹MLU, Inst. of Human Genetics, Halle, GERMANY, ²Inst. of Biochemistry, Univ. Leipzig, Leipzig, GERMANY, ³MLU, Dep. of Internal Med., Halle, GERMANY.

LRP is a multifunctional cell receptor which internalizes a variety of important ligands and is therefore considered to be a candidate gene for complex diseases like atherosclerosis. Materials and methods: We investigated the individual LRP-mRNA and protein expression in native monocytes from 72 male probands, 36 patients with angiographically proven severe coronary atherosclerosis (age 51.7 years) and 36 healthy long-standing blood donors (age 47.3 years). The investigations on transcriptional level were carried out using a competitive RT-PCR. For specific detection of LRP protein expression we applied a macro array analysis. As a reference we used a commercially available LRP standard protein (Biomac). Results: We measured a significantly 1.82 fold higher LRP-mRNA expression in coronary patients in comparison to healthy controls (223ag/cell vs. 122.3ag/cell, p<0.001). However the investigation of LRP-protein expression revealed an inverse pattern: Whereas the expression of the coronary patients amounted to 1.6pg/cell the controls showed a significant higher protein expression of 6pg/cell (p<0.001). Obviously a high LRP-mRNA expression was associated with a low protein expression, and vice versa. These results suggest a complex regulatory mechanism of the LRP at transcriptional and translational level. The detected lower protein expression in coronary patients may be due to a severe unbalanced metabolism in atherosclerosis. This could lead to a diminished receptor-mediated endocytotic pathway in coronary patients which may then be compensated by increasing the mRNA expression. These findings supply evidence for the importance of the expression pattern of the receptor in the assessment of atherosclerosis development.

Power of genomic variants of TNFa and TNFß as major risk factors for coronary macroangiopathies

T. Süß¹, S. Schulz¹, U. Schagdarsurengin¹, P. Greiser¹, D. Rehfeld¹, A. Nordwig¹, A. Kabisch², U. Müller-Werdan³, K. Werdan³, C. Gläser¹;
¹MLU, Inst. of Human Genetics, Halle, GERMANY, ²MLU, Bloodbank, Halle, GERMANY, ³MLU, Dep. of Internal Med., Halle, GERMANY.

The cytokines TNFa and TNFß are important cytokines in the complex signaling pathways involved in the development of atherosclerosis. Materials and methods: We studied three polymorphisms located in functional important regions (TNFa-promoter: G-308A, TNFß: exon2, T492C, CysàArg; exon3, C720A, ThràAsp) in 198 patients with angiographically confirmed coronary diagnosis (49.7y, SD 8.5; 136 males). The patient group consists of 99 CAD-patients and 99 patients without any coronary afflictions as a control group (age- and gender-matched random probands). Results: The analysis of the TNFa-polymorphism (G-308A) showed an increased number of the homozygous mutation-carriers AA among CAD-patients (6 vs. 1, n.s.). Investigating the genotype frequencies of the exon2-polymorphism of TNFß we detected more mutation-carriers in the control group (0.56 vs. 0.48, n.s.). However, the examination of the TNFb-C720A-polymorphism resulted in significant differences in the two subgroups for the A-recessive model: the homozygous mutation-carriers were more often found among CAD-patients (0.18 vs. 0.07, p<0.004). Regression analysis including 5 major risk factors of CAD (smoking, hypertension, hypercholesterolemia, low LDL, Diabetes mellitus) showed a significant 3.6 times higher coronary risk for AA-carriers (95%-CI: 1.131-11.633; p<0.035). Conclusions: For the investigated polymorphisms in the promoter region of TNFa as well as in exon2 of TNFß no significant relation to the occurrence of CAD could be determined. However the polymorphism in exon3 of TNFß was shown to be significant associated with the risk of the development of a coronary macroangiopathy and increased the power of the investigated major clinical coronary risk factors.

I. Nikolov¹, D. Toncheva¹, P. Blagoeva², V. Sekoulov³, R. Vladimirova⁴, G. Kirov⁵;
¹Dept. Medical Genetics, Medical University Sofia, Sofia, BULGARIA, ²Plovdiv Psychiatric Dispensary, Plovdiv, BULGARIA, ³Sofia Psychiatric Dispensary, Sofia, BULGARIA, ⁴Dept. Psychiatry, Medical University Sofia, Sofia, BULGARIA, ⁵Neuropsychiatric Genetics Unit, UWCM, Cardiff, UNITED KINGDOM.

Anticipation refers to an earlier age at onset and increased severity of illness in offspring, compared to parents. It is usually caused by dynamic mutations. Anticipation has been observed in psychiatric disorders. We have collected 608 parent-offspring trios where the proband had a diagnosis of Schizophrenia (SZ), Bipolar Affective Disorder (BP) or Schizoaffective Disorder (SA). Diagnoses were made on the basis of clinical records and structured clinical interviews of probands. In 40 families (6.5%) the proband had a parent who had one of the above diagnoses. We used age at onset (AO) as the main variable for assessment of anticipation. A well-known bias operates in such studies, because psychiatric patients are less likely to have children after they get ill, so that affected parents of probands are likely to have a later AO. An additional bias operates in our sample, as all patients were quite young (all their parents were still alive). In order to reduce the bias, we looked at two subsamples: a) families where parents became ill before their child was born, and b) families where the affected offspring has children. The results were analysed with paired-samples T-test. The offspring had an earlier AO not only in the general sample, but also in a) and b) subsamples (Table 1).
This study adds to the evidence that anticipation might be a true phenomenon in major psychiatric disorders, as a difference was still present after these corrections.

Table 1. Results
	BP	SZ	SA	AO [SD] (years)	Mean paired difference [SD] (years)	Significance
General sample, N=40
Parents	20.0%	65.0%	15.0%	33.7 [11.2]	12.4 [12.6]	t = 6.21, p=0.000
Offspring	25.0%	55.0%	20.0%	21.4 [5.0]	12.4 [12.6]	t = 6.21, p=0.000
Subsample a), N=14
Parents	-	78.6%	21.4%	22.6 [3.9]	2.8 [6.3]	t = 1.65, p=0.123
Offspring	14.3%	42.9%	42.9%	19.9 [4.8]	2.8 [6.3]	t = 1.65, p=0.123
Subsample b), N=12
Parents	33.4%	50.0%	16.7%	33.9 [12.2]	12.2 [12.9]	t = 3.28, p=0.007
Offspring	33.3%	41.7%	25.0%	21.8 [5.3]	12.2 [12.9]	t = 3.28, p=0.007

High complexity in the genetic basis underling Factor VII deficiency in two Spanish families.

M. Sabater-Lleal, J. M. Soria, E. Martinez-Marchan, E. Martinez-Sanchez, I. Coll, C. Vallvé, J. Mateo, J. C. Souto, J. Fontcuberta;
Unitat d'Hemostasia i Trombosi. Hospital de la Santa Creu i Sant Pau, Barcelona, SPAIN.

Factor VII (FVII) is a vitamin k-dependent serine protease enzyme essential for initiating the coagulation cascade via the extrinsic pathway. Reduced FVII levels cause bleeding disease, whereas increased levels are associated with thrombosis. Some mutations have been described which modulate levels of this protein.
We have studied two different assymptomatic patients with FVII levels lower than 3% of normal values. We have sequenced the entire FVII gene (promotor, exons, introns and 3’-UTR: 15kb) to identify mutations of this gene in families.
A Gln100Arg mutation located in exon 5 was detected in one patient in the homozygote state. In this patient, we also identified a novel G to A substitution at nucleotide 3294. In another patient, we identified two different mutations: Met298Ile and Gly331Ser in exon 8. This patient was heterozygous for both of these mutations. Moreover, other FVII genetic variants that influence FVII levels, are co-segregating in these families.
Although some of these mutations have been described as closely related with bleeding disease, the presence of several putative functional genetic variants could be responsible for the high variability in phenotype observed in the family members with the same mutation. These results clearly show the complexity of FVII deficiency and the importance of a global effect of multiple quantitative trait locus (QTLs) in determining FVII levels. Further investigation should help to identify other QTLs that influence variation in FVII levels and may help to quantify the relative risk for thrombosis or bleeding disease.

P. Levy, J. Ferreira, L. Breitenfeld, M. Bicho;
Genetic Laboratory, Medical School, University of Lisbon, Lisbon, PORTUGAL.

The serotonin transporter gene is a likely candidate in autistic disorder, based on efficacy of potent serotonin transporter inhibitors in reducing rituals and routines and of elevated serotonin levels been consistently found in 30%±50% of autistic patients. The aim of this study is to determine a possible association between the polymorphism in serotonin transporter gene and autism.
Materials and Methods
Blood samples were obtained from forty-eight individuals with autism (DSM-IV) with no associated disorders, thirty-two mothers, twenty-seven fathers (twenty-three nuclear families) and 49 normal controls.
Each family was seen for a developmental assessment, and also by a clinical geneticist for identification of eventually associated aetiology.
Ages ranged between 3.6 and 41 years.
Genotyping were performed by PCR methods. Chi-square analysis was applied to the results.
Results
Within the three main possible alleles (9, 10, 12), we only observed 10 and 12. There were no differences between the frequencies distribution of autistic individuals and control group.
In autistic individuals there was no prevalence of any specific genotype. There was no difference in allele frequency between the autistic group versus mothers, fathers, or mothers and fathers.
We observed a preferential 12/12 genotype in patients with mothers carrying the same genotype. This characteristic was not observed regardind fathers genotype.
Conclusions
In this study we did no found any association of autism with the polymorphism in serotonin transporter gene .
It seems to exist a preferential 12/12 genotype in autistic individuals with mothers carrying the same genotype.

Examination of KIAA1327 on chromosome 4p15.33 for association with bipolar affective disorder

G. Kirov, M. J. Owen;
University of Wales College of Medicine, Cardiff, UNITED KINGDOM.

The region on chromosome 4p15-p16 has been implicated in the aetiology of bipolar affective disorder (BP) through several linkage findings. One family collected at our department showed lod=1.96. We saturated a 9cM region on 4p with 32 microsatellites and genotyped them on sets of pooled DNA from 110 BP patients and their parents, as well as 178 patients and 184 controls. One marker was significant in the trios pool and the association was confirmed by individual genotyping. The closest gene to this marker is 170kb away: hypothetical protein KIAA1327. We screened the 17 exons, as well as 3,000bp of 5’ flanking sequence which showed high homology with mouse. We used DHPLC on 15 individuals, including three key members from the linked family. We sequenced the fragments with shifts and every fragment from the key individuals.
We found five SNPs, but none of them was unique to the disease chromosome in the linked family. Four of the SNPs, including the only amino-acid change (Pro>Ile in exon 1) were genotyped in the 110 trios. None of the results reached statistical significance when analysed with TDT. The Ile allele of the Pro>Ile was transmitted 34 times and not transmitted 21 times, p=0.08. The result could not be replicated in a set of 110 trios collected in Bulgaria: 38 parents transmitted and 37 did not transmit that allele.
An interesting finding is that this SNP was in strong LD (D’=0.5) with the microsatellite that produced the original association, despite a distance of 170kb.

J. Tzenova¹, B. J. Kaplan²^,3, T. L. Petryshen⁴, L. L. Field¹;
¹University of British Columbia, Vancouver, BC, CANADA, ²University of Calgary, Calgary, AB, CANADA, ³Alberta Children's Hospital, Calgary, AB, CANADA, ⁴Whitehead Institute, Cambridge, MA.

Dyslexia is a common and complex genetic trait that manifests as a specific reading disability independent of intelligence and educational opportunity. Rabin et al. (1993, Lancet 342:178-79) found suggestive evidence for linkage of dyslexia to Rh on chromosome 1p34-36. More recently, Grigorenko et al. (2001, AJMG 105:120-29) reported significant linkage to the same region, but using a quantitative definition of dyslexia. We have tested for the presence of a dyslexia gene in this region on chromosome 1p in a sample of 100 Canadian families using both qualitative and quantitative definitions of the phenotype. With the qualitative phenotype, parametric linkage analysis produced a maximum lod score of 1.7 (q = 0.3) at D1S1597 under a recessive model with incomplete penetrance. Multipoint analyses using GENEHUNTER generated a maximum HLOD of 2.6 between markers D1S1597 and D1S3669 under the same model. Non-parametric analysis of a sub-sample of 351 sib-pairs indicated strongest linkage to D1S3669 (SIBPAL p = 0.00015). The multipoint NPL score for all the families maximized over a 15cM region spanning D1S1597 – D1S3669 – D1S199 – D1S552 (lowest p = 0.017). We also used quantitative measures of the phenotype (spelling and phonological coding) to test for linkage by the variance components method (GENEHUNTER). Using a model with QTL additive and dominance variance and polygenic additive variance, the multipoint lod score maximized over the 15cM region with a peak of 3.11 near D1S3669 for spelling. In conclusion, our study confirms and strengthens evidence for a dyslexia susceptibility gene on chromosome 1p.

J. H. Edwards¹, H. Eiberg²;
¹Oxford University, Oxford, UNITED KINGDOM, ²Institute of Medical Biochemistry and Genetics, Copenhagen, DENMARK.

Penrose introduced sib-pair analysis in 1935: an essential feature was ascertainment for one or more affected sibs providing controls data against unrelated sib-similarity. (www.gene.ucl.ac.uk/anhumgen/).
Most recent sib-pair studies are restricted to affected sib pairs and assume the parental gametes present at ascertainment are representative of those present before fertilisation, each allele having an equal chance of both achieving fertilisation and surviving birth.
Such equality is not to be expected in view of the strong preference for gametic differences in plants, with evolutionary advantages likely to be exploited in other species, and losses between conception and birth from embryonic lethals. Both would lead to excess sib-similarity at neighbouring loci. There are few data relating to sib-similarity in normal sibs.
The Danish set of normal families, occasionally supplemented by families with Mendelian disorders, was started in 1972 and has provided key information on several assignments, including CF and Batten's disease. It includes over 6000 typed individuals in 850 families with about a million genotypes.
We present estimates of the same:different ratio of paternal and maternal alleles passed from parents who were heterozygous and differed in genotype from the other parent.
Similar displays from affected sib-pairs with may be seen on www.bioch.ox.ac.uk/~jhe or www.angis.org.au/medvet

A. Czibula¹, M. Morocz¹, A. Csiszar¹, C. Bachrati¹, E. Olah², F. Szeszak³, E. Morava⁴, L. Szappanos⁵, I. Rasko¹;
¹Institute of Genetics,Biological Research Center of Hungarian Academy, Szeged, HUNGARY, ²Department of Pediatrics, University Medical School of Debrecen, Debrecen, HUNGARY, ³Department of Human Genetics, University Medical School of Debrecen, Debrecen, HUNGARY, ⁴University School of Pecs, Department of Medical Genetics and Child Development, Pecs, HUNGARY, ⁵County Hospital of Mezotur, Mezotur, HUNGARY.

Idiopathic scoliosis (IS) is a complex disease, with a strong genetic influence. This is supported by the familiar accumulation of the disease. It is the most frequent spine deformity of adolescence. The etiology is unknown, however difference in sex incidence was observed, the girls being affected 10 times more likely than boys. The genetic susceptibility loci of IS have not been identified so far. We accidentally were able to reveal a possible susceptibility allele for IS, in the bromodomain PHD finger transcription factor (BPTF). Two alleles with different size could be obtained with a simple PCR reaction. A 8.3 kb in homo- and heterozygotes and a 3.5 kb long in heterozygotes. No homozygotes were detected for the shorter allele. The 3.5 kb allele possibly resulted via a deletion in the last intron of BPTF and it appears more frequently in IS (p=0.29) than in control groups (p=0.16). According to a recent publication the BPTF is the human ortholog of Drosophila Nurf301, which is the largest subunit of a chromatin remodeling NURF complex. Malfunction of BPTF in early ontogenesis, together with intense growth rate during adolescence and other environmental factors could influence the development of IS.

F. Gurrieri, L. Russo, T. Giordano, E. De Vincenzi, G. Neri;
Catholic University, Rome, ITALY.

Pervasive Developmental Disorders (PDD) represent an heterogeneous group of behavioural deficits, including autism and atypical autism, characterized by impaired communication and social interaction, restricted interests and stereotyped behaviours. PDD affect around 1:2500 individuals within the first three years of life with a sex ratio M:F = 4:1.
A strong genetic basis has been recognized and several genomic studies, as well as cytogenetic observations, have identified at least 12 candidate loci for genes predisposing to PDD. Among these regions, 15q11-q13, 7q31 and, recently, 22q13, have been most frequently reported in association with PDD.
In order to identify the prevalence of 22q13 rearrangements (deletions) in PDD we have genotyped, by microsatellite markers analysis, 110 patients selected through a collaboration with various italian Child Neuropsychiatry centers. We detected a deletion at the D22S1169 locus in two patients (one paternally and one maternally derived) and a maternally derived deletion at the D22S1170 locus in a third patient. These patients presented with an Angelman-like phenotype, characterized by hypotonia, developmental delay and absent speech. However, they did not show epilepsy. Although the pathogenicity of these rearrangements needs to be confirmed by further studies, our results suggest that deletions within the 22q13 region might play a role in determining PDD in 2.7% of patients. Considering the marked genetic heterogeneity of PDD, this percentage appears to be relatively high.

Single Nucleotide Polymorphism Detection and Association Results - Exclusion of ITGb7 and VDR (Chromosome 12q) as Candidate Genes for Asthma

C. Vollmert¹, T. Illig¹, J. Altmüller¹, S. Klugbauer¹^,2, S. Loesgen¹, M. Wjst¹;
¹GSF Research Center for Environment and Health, Institute of Epidemiology, Neuherberg, GERMANY, ²current address: Biomax Informatics AG, Martinsried, GERMANY.

The human genes coding for the integrin b7 (ITGb7) and the vitamin D receptor (VDR) are two of several candidate genes for asthma and related phenotypes in a promising candidate region on chromosome 12q pointed out in several genome-wide screens and candidate gene approaches. Therefore we screened the promoter region as well as all exons of the ITGb7 gene, including parts of the neighbouring introns for common polymorphisms in 32 German probands. Moreover we then tested single nucleotide polymorphisms (SNPs) for linkage/association with asthma and related traits (total serum IgE level, eosinophil cell count, peak flow and SLOPE of the dose-response curve after bronchial challenge) in a Caucasian sib-pair study (187 families with at least two affected children). We also analysed one already described SNP in the human VDR gene. We could identify three new single nucleotide polymorphisms in the ITGb7 gene. Two of them are non-coding (intron 2 and intron 6) while the SNP in exon 3 causes an substitution of the amino acid GLU to VAL.
None of the SNPs neither of the ITGb7 nor the VDR gene showed significant linkage/association with asthma or related phenotypes in this family study. From these findings we conclude that both the human ITGb7 and the VDR gene seem not to influence the pathogenesis of asthma or the expression of atopic asthmatic phenotypes as eosinophilia and changes in total IgE levels.

Significant linkage of phonological coding dyslexia to dopamine receptor type 4 (DRD4) on chromosome 11p15.5

G. R. Hsiung¹, B. Kaplan², T. Petryshen³, S. Lu¹, L. Field¹;
¹University of British Columbia, Vancouver, BC, CANADA, ²University of Calgary, Calgary, AB, CANADA, ³Whitehead Institute, Cambridge, MA.

Phonological coding dyslexia (PCD) is a specific language disability that is independent of general intelligence and educational opportunity and is highly heritable. Because the 7-repeat allele of the dopamine D4 receptor (DRD4) exon III has been implicated in attention deficit hyperactivity disorder (ADHD), and there is known comorbidity between ADHD and dyslexia, we investigated DRD4 as a candidate gene for PCD. In our 2-point screen (FASTLINK) of the 11p15.5 region in 100 families with at least two siblings affected with PCD, we found highly suggestive evidence of linkage to markers D11S1363 (LOD=2.31, theta=0.2), DRD4-exon III repeat (LOD=2.75, theta=0.2), and HRAS (LOD 2.53, theta=0.2) . Allowing for heterogeneity (HOMOG), the HLOD for DRD4 was 3.10 with alpha=0.85. Using non-parametric affected sib-pair analysis (SIBPAL) on 254 nuclear families derived from our dataset, we also found significant linkage to DRD4 (p=0.0007) and HRAS (p=0.0011) . With multipoint linkage analysis (GENEHUNTER), we obtained a maximum between D11S1363 and DRD4 (HLOD=2.14, alpha=0.75, NPL 2.37, p=0.009). QTL analyses (SOLAR) also demonstrated a multipoint maximum near the DRD4 locus with spelling (MLOD=3.61) and phonological coding traits (MLOD=2.51). However, preliminary analysis using family-based association studies (AFBAC and ETDT) did not show significant association of PCD with the DRD4-7 repeat allele. It is possible that the DRD4-7 repeat allele is not pathogenic in dyslexia, but other mutations in the DRD4 gene not in linkage disequilibrium with the repeat are involved. Alternatively, it is possible that another gene closely linked to DRD4 in the 11p15.5 region influences susceptibility to dyslexia.

N. Elleuch-Bougacha¹, A. Maalej², H. Makni³, M. Bellasouad³, J. Jouida⁴, F. Ayadi⁵, A. Rebaï⁶, M. Abid⁷, H. Ayadi⁸;
¹Faculty of Medicine, Sfax, TUNISIA, Sfax, TUNISIA, ²Faculty of Medicine Sfax, TUNISIA, Sfax, TUNISIA, ³Hédi Chaker Hospital, Sfax, TUNISIA, ⁴Bir El Hfay Hospital, Sidi Bouzid, TUNISIA, ⁵Habib Bourguiba Hospital, Sfax, TUNISIA, ⁶Center of Biotechnology, Sfax, TUNISIA, ⁷Hedi Chker hospital, Sfax, TUNISIA, ⁸Faculty of Medicine, Sfax, TUNISIA.

The autoimmune thyroid diseases (AITDs) include two related disorders, Graves’ disease (GD) and Hashimoto’s thyroiditis (HT). The pathogenesis of the AITDs involves a complex interaction between genetic and environmental factors. Recently, five potential susceptibility loci for AITD have been mapped to chromosomes 14q31, Xq21.33, 20q11.2, 2q33 and 18q21 in different populations. In this study, we have investigated a large consanguineous Tunisian family affected with AITDs. The search for susceptibility genes in this family was undertaken by means of both linkage and association analyses. To perform linkage study, a genome screening was done using microsatellite markers. A single marker located on chromosome 2 (D2S171) showed evidence of linkage with a MMLS-c score of 3.03. However, no evidence of linkage was found for some candidate regions covering MHC, Ig VH, Cb TCR and as well as the five reported loci (Maalej et al. Genes and Immunity. 2001 Apr;2(2): 71-5). Since association analysis is more powerful than linkage analysis to detect minor susceptibility genes of complex diseases, such studies were performed on HLA loci and the CTLA-4 gene polymorphisms. Using the TDT, we have reported genetic association of GD with HLA-B*3701 (chi2= 6.12; p=0.0134), (Bougacha et al. Clinical Endocrinology 2001, 55: 1-3) and lack of association for the intragenic CTLA-4 (AT)n and (A/G) dimorphism using the FBAT approach (p=0.406 and p= 0.466 respectively) (Maalej et al. Human Immunology 2001, 62:1245-1250). In conclusion, our data showed linkage of AITDs with the D2S171 microsatellite marker, and genetic association between GD and HLA-B*3701 allele.

Y. Weigl¹, L. Peleg², A. Dotan³, I. E. Ashkenazi³;
¹The Sackler Faculty of Medicine, Tel-Aviv, ISRAEL, ²The Danek Gertner Genetic Institute, Ramat-Gan, ISRAEL, ³The Sackler Faculty of Medicine, Tel Aviv, ISRAEL.

Three month old BALB/c, c57BL/6J mice and their F1 offspring were exposed, for three weeks, to 12:12 light:dark illumination. Then, over a period of 30 hours at nine equidistant times, three male and female mice of each group were sacrificed and WBC count, kidney creatine phosphokinase (CPK) and alkaline phosphatase (AP) activities and kidney urea (U) were determined. Rhythms significance and parameters like Period, acrophase (Peak time) and amplitudes were analyzed (by Curvefit). Results: Comparison among rhythms revealed Gender dependent differences which in turn were also strain dependent. For example, WBC-count rhythm acrophases differed between c57BL/6J genders but not in BALB/c. The periods of AP activity rhythms differed only among c57BL/6J genders while acrophases of AP activity rhythms differed only among BALB/c gender. The amplitude of CPK activity rhythm was significantly higher in females of BALB/c then in males while the reverse was observed among c57BL/6J Genders. Gender differences were recorded also in the F1 groups. Depending on the examined variable, the rhythm differences between F1 genders may resemble those shown by one of the strains or exhibited a new difference range which didn’t follow any of the gender differences (or similarity) exhibited in the parental stains. Conclusions: 1. Gender dependence differences exist in variable rhythms even under normal identical conditions. 2. The range of the differences among genders is strain dependent. 3. The inheritance mode suggests that each rhythm parameter is individually controlled (inherited).

I. Greenbaum¹, Y. Weigl², V. Glauber, Dr.³, A. Pearl, Prof.³, G. Barkai², E. Gak²;
¹Danek Gertner Institute of Human Genetics, Ramat-Gan, ISRAEL, ²Danek Gertner Genetic Institute of Human Genetics, Ramat-Gan, ISRAEL, ³Department of Reanimation and Anesthesiology, Ramat-Gan, ISRAEL.

Malignant Hyperthermia (MH) is clinically and genetically heterogeneous disorder. At least six distinct chromosomal loci have been associated with this condition, among them RYR1 gene encoding calcium-channel accounts for about 50% of MH cases.
In this study we focused on a unique Israeli family, where MH segregated along three living generations and 8 out of 10 family members were tested in a skeletal muscle contractility in-vitro assay (IVCT). A comparative genetic analysis of MH susceptible (MHS and MHE) versus normal (MHN) family members was based on polymorphic markers from two major loci linked to MH, RYR1 on chromosome 19 and DHPR on chromosome 1. Four markers within 2Mb interval spanning RYR1 gene, D19S191, D19S224, D19S228 and D19S897, were conclusive as to the linkage of MH to chromosome 19 RYR1 gene localization. Sequence analysis of RYR1 exons 11, 17, 39, 40, 45 and 46, where mutations have been previously reported in Caucasian MH pedigrees, did not reveal any variation from wild type. A potential functional polymorphism comprising non-perfect trinucleotide repeat within exon 35 was detected by sequencing and fragment analyses. Further studies on a population scale are needed to assess the population frequency and phenotypic significance of this variation. Moreover, two markers from chromosome 1, D1S2853 and D1S2683, were consistent with MH susceptible (MHS) versus MH equivocal (MHE) family members, suggesting possible involvement of DHPR gene on a refined MH phenotype.
This family might represent a unique case of complex inheritance of two genetic loci involved in predisposition to MH.

A single nucleotide polymorphism in exon 24 of the MYH7 gene may be associated with left ventricular hypertrophy in essential hypertension and with left ventricular outflow obstruction in hypertrophic cardiomyopathy

K. V. Puzyrev¹, M. V. Golubenko²^,3, E. N. Pavlukova¹, V. P. Puzyrev²^,4, H. P. Vosberg³, O. V. Makeeva², C. Selignow³, R. S. Karpov¹;
¹Research Institute of Cardiology, Tomsk, RUSSIAN FEDERATION, ²Research Institute of Medical Genetics, Tomsk, RUSSIAN FEDERATION, ³Max-Plank Institute for Physiological and Clinical Research, Bad Nauheim, GERMANY, ⁴Siberian State Medical University, Tomsk, RUSSIAN FEDERATION.

Left ventricular hypertrophy (LVH) is known to occur as a cardiac complication of essential hypertension (EH) or as a typical structural change in dominantly inherited hypertrophic cardiomyopathy (HCM). In minority of HCM patients hypertrophy isfor unknown reasons associated with obstruction of the left ventricular outflow tract. We have investigated the association of an exon 24-SNP in the ß-myosin heavy chain gene (MYH7) with the LVH phenotype. Pyrosequencing was used for the analysis. The SNP is a silent T/C transition in codon 989. Allele frequencies in EH (n=43) and HCM patients (n=23) did not differ significantly (EH: C=0.32, T=0.68; HCM: C=0.26, T=0.74). The left ventricular mass index (in g/m²)was calculated based on echocardiographic assessments. The following index values were obtained for CC and CT: 140 ±9.5 and 166±51 g/m², resp.; and for TT: 253±58 g/m² (significant difference by ANOVA: F=4.02, P=0.02). Outflow obstruction in HCM patients as documented by Doppler echocardiography was virtually absent in carriers of the C allele (CC or CT; n=10). However, in carriers homozygous for T (n=13) both obstruction (n=8) and non-obstruction (n=5) were observed. The difference between CC/CT carriers and TT carriers was significant as was deduced from crosstabulation tables (Pearson c²=8.06, P=0.0045). Conclusion: we take these results to suggest that TT homozygosity is in two different clinical conditions a marker associated with LVH in EH patients, and with obstruction in HCM patients, respectively. Conversely, the presence of the C allele may protect against LVH or left ventricular outflow obstruction.

Lack of association between a G-protein beta3-gene C825T polymorphism and attention deficit hyperactivity disorder

B. Zagradisnik¹, N. Potocnik Dajcman², M. Gajsek², N. Kokalj Vokac¹;
¹General Hospital Maribor, Maribor, SLOVENIA, ²Primary Health Centre Maribor, Maribor, SLOVENIA.

G-proteins are important elements in the regulation of cellular responses where they conduct signals between receptors and effector proteins. A recently identified polymorphism C825T of a G-protein beta3 subunit (GNB3) has been shown to be associated with increased signal transduction and ion transport activity. In patients with attention deficit hyperactivity disorder (ADHD) signal transduction through several neurotransmitter systems is known to be altered. The C826T polymorphism in the GNB3 gene might contribute to the change in neurotransmitter activity. The aim of this study was to establish an association between C825T polymorphism of the GNB3 gene and ADHD. A group of 52 Slovenian children affected with ADHD was genotyped using PCR-RFLP method. The distribution of genotypes was not significantly different in ADHD patients and in healthy controls (0.403, 0.462, 0.135 vs. 0.468. 0.468, 0.064, chi square = 4.51, p = 0.1047). Also, haplotype-based haplotype relative risk analysis of 46 ADHD families (father, mother and affected child) showed no significant association between T allele and ADHD (chi square = 2.0588, p = 0.1513). Therefore, the C825T polymorphism of the GNB3 gene is unlikely to be an important genetic susceptibility factor for ADHD.

Lack of association of myo-inositol monophosphatase 2 (IMPA2) polymophisms and bipolar affective disorder

A. Z. Dimitrova¹, I. Nikolov¹, C. Kostov², D. Toncheva¹, M. J. Owen³, G. Kirov³;
¹Medical University, Sofia, BULGARIA, ²First Psychiatric Clinic, Sofia, BULGARIA, ³University of Wales, College of Medecine, Cardiff, UNITED KINGDOM.

Lithium is used as a mood-stabilizing drug in the therapy of bipolar affective disorder (BP). The IMPA2 gene codes for an enzyme in the phosphatidylinositol signalling system, which is inhibited by lithium. The IMPA2 gene is located on 18p11.2, a region implicated as a BP susceptibility locus by several linkage studies. We examined SNPs identified within this gene in 123 Bulgarian patients affected with BP and their parents. We tested 3 SNPs implicated as a disease haplotype in schizophrenia in a study by Yshikawa et al., (Mol Psych 2001, 6:202-210) and an amino-acid change (Arg148Glu ) identified by Sjoholt et al., (Mol Psych 2000, 5:172-180). All four SNPs were genotyped in a multeplex assay using primer extension.
The 558C>T and 490+13-14insA (reported in the study on patients from Japan) were found to be much rarer in our population, (f of the rarer alleles = 2.4% and 0.8%, respectively). This made them unsuitable for LD mapping and we stopped genotyping after the first 30 trios. The remaining two SNPs were genotyped in 123 BP trios from Bulgaria. 58 parents were heterozygous for the 97-15G>A of whom exaclty 50% transmitted allele A to their affected offspring.
The Glu allele of Arg148Glu was present in seven parents (f=1.4%). Four of them transmitted and three did not transmit it. We genotyped a further 120 BP trios recruited in the UK. There were 8 heterozygous parents (f=1.7%) but only 2 transmitted the mutation. We cannot find support for the involvement of IMPA2 in bipolar disorder.

Autistic Spectrum Disorder: A Case Study Using High-Resolution Microarray-Based CGH

S. Shah¹, M. Mohammed¹, S. Damani¹, R. Locker¹, W. Yu¹, J. Gregg²;
¹Spectral Genomics, Inc., Houston, TX, ²University of California at Davis, Davis, CA.

Autism is a severe developmental disorder characterized by impairment in social interactions and in language and communication skills, and by restricted repetitive behaviors and activities. The etiology of this disorder is poorly understood, although a variety of etiologic mechanisms have been suggested, including genetic, immunologic, infectious, neurologic and gastrointestinal abnormalities. Family and genetic studies support a strong genetic susceptibility to autistic spectrum disorders and a high incidence in family members of nonautistic PDD variants. These studies have been completed using karyotypic analysis which is limited to detecting chromosomal aberrations greater than ~ 5 Mb (<650 band resolution). With this low-resolution scan of the genome, there is a high likelihood that smaller disease associated deletion and duplication events are being missed. A modified CGH to BACs immobilized on a glass slide (instead of metaphase chromosomes) can provide a much higher resolution, potentially resolving to the size of individual BACs. We developed a human genomic array in which BACs are spaced at ~ 1 Mb intervals, on average, throughout the genome. Clinical specimens and established cell lines with a broad spectrum of known chromosomal abnormalities were tested. Test and reference genomic DNAs were differentially labeled with fluorochromes. After co-hybridization of labeled test and reference genomic DNAs, the BAC arrays were scanned and analyzed with software developed specifically for this purpose. Control hybridizations with normal male-to-male, male-to-female and female-to-female reference samples were performed and showed the expected results. This study details the results identified in the screening of over 25 autistic patients.

Variations in the vitamin-D binding protein (Gc locus) and parathyroid cell function in dialysis patients with end-stage renal disease

J. Fibla¹, L. Piedrafita¹, G. Nieto¹, A. Velasco¹, E. Fernandez²;
¹Department of Basic Medical Sciences, University of Lleida, Lleida, SPAIN, ²Nephrology Service. Universitary Hospital Arnau de Vilanova, Lleida, SPAIN.

Vitamin-D inhibits both PTH expression and proliferation of parathyroid cells. Patients with end-stage renal disease (ESRD) show low levels of vitamin-D due to renal dysfunction. Limiting amounts of vitamin-D do not allow a normal regulation of parathyroid function and predisposes patients to either adynamic-bone disease (ABD) or secondary hyperparathyroidism (2HP). Vitamin-D binding-protein (DBP) (Gc locus) is the major carrier protein for vitamin-D metabolites. Two variants at codons 416(Asp-Glu) and 420(Thr-Lys) of the exon-11 of the Gc gene define the three common electrophoretic variants of the DBP protein (Gc1F, Gc1S and Gc2). We have studied exon-11 alellic variants as risk factors for ABD and 2HP predisposition and parathyroid function in dialysis patients. A population of 155 patients with more than 1 year in haemodialysis were recruited and genotyped for exon-11 polymorphisms. Alelle frequencies at both codons do not differ between patients and controls. Distribution of the genotypes at both codons and the genotypes defined by the combination of the two codons were also similar in patients and controls (tables 1&amp;amp;2). After applying excluding criteria for treatment and time on dialysis, genotype and alelle frequencies of patients grouped as “high” and “low” PTH do not differ from those observed in control subjects. Differences for PTH levels were observed between homozygous 420LysLys and homozygous 420ThrThr patients (48.8 [95%CI:35-62] vs. 18.8 [95%CI:2.6-34.9] pmol/L, respectively, P=0,045). Variability at exon11 of the Gc locus seems to be not related to predisposition to ABD and 2HP. In contrast, variability at codon 420 seems to affect PTH levels.

Table 1.- Distribution of Gc genotypes at codons 416 and 420 in control and patients
Groups considered	codon 416 Genotypes N (%)			codon 420 genotypes N (%)
Groups considered	AspAsp	AspGlu	GluGlu	ThrThr	ThrLys	LysLys
Controls (N=139)	23(16)	83 (60)	33 (24)	65 (47)	61 (44)	13 (9)
Dialysis Patients (N=155)	31(20)	72 (47)	52 (33)	80 (51)	57 (37)	18 (12)
High PTH Group (N=35)#	4(11)	18 (52)	13 (37)	21 (60)	12 (34)	2 (6)
Low PTH Group (N=39)#	11 (28)	19 (49)	9 (23)	18 (46)	14 (36)	7 (18)
#Serum PTH levels in “high” PTH group were higher than 60 pmol/L and in “low” PTH group were lower than 12 pmol/L

Table 2.- Distribution of Gc genotypes and alelles defined by polymorphisms at the two codons
Groups considered	Gc Genotypes, N (%)						Gc Alelles , N (%)*
Groups considered	FF	SS	22	FS	F2	S2	F	S	2
Controls (N= 139)	3 (2)	33 (24)	13 (9)	29 (21))	7 (5)	54 (39)	42 (15)	149 (54)	87 (31)
Dialysis Patients (N= 155)	1 (1)	52 (34)	18 (12)	27 (17)	12 (8)	45 (29)	41 (13)	176 (57)	93 (30)
High PTH group (N= 35)#	0 (0)	13 (37)	2 (6)	8 (23)	2 (6)	10 (29)	10 (14)	44 (63)	16 (23)
Low PTH group (N= 39)#	1 (3)	9 (23)	7 (18)	8 (21)	3 (8)	11 (28)	13 (17)	37 (47)	28 (36)
* F = haplotype 416Asp-420Thr, S = haplotype 416Glu-420Thr and 2 = haplotype 416Asp-420Lys #Serum PTH levels in “high” PTH group were higher than 60 pmol/L and in “low” PTH group were lower than 12 pmol/L

Lack of evidence for association of the endothelial nitric oxide synthase gene polymorphism in intron 4 and progression to end stage renal disease in Greek population

P. Zirogiannis¹, H. Psara², G. Staikos², C. Deltas³, K. Lamnissou²;
¹Dept of Nephrology, "G. Gennimatas" Hospital, Athens, GREECE, ²Dept of Biology, University of Athens, Athens, GREECE, ³Dept of Molecular Genetics, The Cyprus Institute of Neurology & Genetics, Nicosia, CYPRUS.

Nitric oxide (NO) is thought to be an important factor in the deterioration of renal function. A variable number tandem 27-bp repeat in intron 4 of the endothelial nitric oxide synthase (ecNOS) gene has been found to affect the plasma levels of NO metabolites. Two alleles are of varied frequencies in different populations (a and b). The shorter allele, a, has been associated in Japanese populations with the progression of renal disease. We studied the association of this polymorphism in a Greek population of patients with end stage renal disease (ESRD) by studying the genotypes of 108 ESRD patients and 105 healthy subjects. The frequencies of aa, ab, bb were 0.69, 0.28, 0.03 in the control group and 0.70, 0.26, 0.04 in the patient group. The dada between the two groups were analyzed by chi-square test. Our results from controls show that the frequencies of these three genotypes in the Greek population are similar to those observed in some other Caucasian populations. But the results from the patient group showed that the frequency of aa genotype in the patient population was not significantly different than in the control group. This work indicate that ecNOS4 polymorphism do not show any association with the development of end stage renal disease in the studied European population.

M. Werner, J. Altmüller, N. Herbon, M. Wjst;
GSF National Research Center for Environment and Health, München-Neuherberg, GERMANY.

SNP Scanning in Pooled DNA - A Chance for Rapid Case-Control-Studies
Chromosome 6p21-22 is a commonly discussed region for candidate genes concerning autoimmune diseases. One possibility to find the responsible genes is to compare SNP frequencies in case-control-studies.
In a "proof of principle’ approach we tested 550 SNPs in a region of 20 Mbp on chromosome 6p. All SNPs were selected from public available databases (NCBI dbSNP, TSC Consortium). SNP allele frequencies were determined by a previously validated MALDI-TOF MS based pooling approach.
Our first sample comprised 288 individuals from the Southern German population. Only SNPs with a frequency between 5% and 50% were tested. The estimated frequencies f were distributed as follows: 7,8% of all SNPs (f<10%), 21,4% (10%<f<20%), 22,4% (20%<f<30%), 26,3% (30%<f<40%) and 22% (40%<f<50%).
In a second sample of 122 German/Swedish DNAs the same 550 SNPs were tested and differences between SNP frequencies were calculated. The mean difference was 5,4%. Only 5% of all tested SNPs differed more than 15% in both samples.
The distribution of the mean allele difference will allow the assignment of possible disease associated loci. Based on this experience, we would suggest that SNP frequency differences between two samples of a case-control-study should not exceed 5-7%. This may be a strict threshold but seems practicable as large study sizes are common practice in studies of complex diseases.

Immune response to Hepatitis B and A vaccination - estimating heritability and assessing different sources of variation in twins

C. M. Freitag¹, E. Reuss², N. Evers², N. Dietrich², J. Vollmar³, P. M. Schneider⁴, C. Rittner⁴, R. Fimmers¹, T. Höhler²;
¹Institute for Biometry, Informatics and Epidemiology, Friedrich-Wilhelms University, Bonn, GERMANY, ²Medizinische Klinik und Poliklinik, Johannes Gutenberg-University, Mainz, GERMANY, ³GlaxoSmithKline, Munich, GERMANY, ⁴Institute for Legal Medicine, Johannes Gutenberg University, Mainz, GERMANY.

The immune system is highly influenced by genetic and environmental factors. Prototypic genetic markers are the different alleles and loci of the HLA system. We estimated heritability and different sources of variation (genetic, environmental) of the immune response to Hepatitis B and A vaccination and assessed its association with several HLA-DRB1*alleles.
Methods: 96 monozygotic (MZ) and 95 dizygotic (DZ) antibody negative twin pairs were vaccinated with a combined HAV/HBsAg vaccine. Anti-HAV and anti-HBs were measured 4 weeks after the last vaccination. All twins were typed for HLA-DRB1*alleles using sequence specific PCR amplification.
Heritability was estimated based on intrapair variances. Generalized estimating equations were calculated to assess the impact of BMI, age, gender, smoking , alcohol, and HLA-DRB1*alleles. A sex limitation structural equation model was fitted to evaluate different sources of variation.
Results: HBsAg vaccine response was weaker with increasing age, body mass index and among male subjects. Several HLA-DRB1*alleles showed a positive correlation with the response. Heritability was 0.61. A model of additive genetic and environmental components best explained the observed variances. Anti-HAV response was weaker among male subjects and heritability was lower (0.35).
Conclusion: Our data suggest a major genetic contribution to the immune response to the Hepatitis B, and a minor one to the Hepatitis A vaccination. The appropriate statistical assessment of the contribution of the HLA-DRB1*alleles will be discussed.

Family based association and linkage analysis of the CD14 gene with allergic asthma in Italian familes

G. Malerba¹, E. Trabetti¹, M. Lauciello¹, R. Galavotti¹, C. Migliaccio¹, L. Pescollderungg², A. L. Boner³, P. F. Pignatti¹;
¹DMIBG, Biology and Genetics, University of Verona, Verona, ITALY, ²Institute of Paediatrics, Hospital of Bolzano, Bolzano, ITALY, ³DMIBG, Paediatrics, University of Verona, Verona, ITALY.

The receptor for bacterial lipopolysaccharides (CD14) appears to be involved in APC-mediated Th1/Th2 cell differentiation. A polymorphism in the flanking region of the gene has been recently described to be associated with circulating soluble CD14 levels and with total serum IgE. The CD14 gene maps on chromosome 5q31, a region that has been linked to asthma and atopic responses. We investigated whether 1 polymorphism located in the promoter of the CD14 gene (-159C/T) was associated to allergic asthma or intermediate phenotypes such as skin prick test positivity to common allergens (SPT), bronchial hyperresponsiveness to methacholine challenge (BHR), and total serum elevated IgE (IgE), in a sample of 182 asthmatic families.
Non parametric linkage analysis in affected sib-pairs did not reveal any significant result.
The transmission disequilibrium test revealed a positive association of the polymorphism with asthma (-159C p:0.0005 ), SPT (-159C p:0.013 ), BHR (-159C p:0.0001), IgE (-159C p:0.0.009 ).
A multivariate analysis performed on family founder members only, did not show any significant association between the gene polymorphisms and any of the phenotypes investigated, suggesting that the CD14 gene might be an allergy susceptibility factor during childhood.
In conclusion, we confirmed the association of polymorphism -159C/T in CD14 gene with allergic asthma phenotypes in an Italian population.

IL4R and NOD2 constitute independent genetic risk factors for Crohn’s disease in the IBD1 region

F. A. Hol, T. H. J. Naber, H. G. Brunner, D. J. de Jong;
University Medical Center, Nijmegen, NETHERLANDS.

In Crohn's disease consistent linkage has been found with the IBD1 locus on chromosome 16 and IBD2 locus on chromosome 12. An insertion polymorphism within the NOD2 gene constitutes the only identified genetic risk factor to date. However, NOD2 can only partially explain the genetic risk contributed by the IBD1 locus. The identity of further susceptibility genes remains unknown.
We tested candidate genes for allelic association using Transmission Disequilibrium Tests in patient-parent pairs. Polymorphisms in the IL-4R, CD11b and NOD2 genes on chromosome 16 and the STAT6 gene on chromosome 12 were tested for association. DNA was available from 91 patients and 147 parents. Significant association was observed with the Q576R polymorphism in IL-4R (Chi2 5.82, p < 0.05) and a polymorphism in the STAT6 gene (Chi2 5.23, p < 0.05). No association was observed with the CD11b gene, and a tendency towards significance with the NOD2 gene. The IL-4R results remain significant after excluding families carrying the C-insertion allele of the NOD2 gene that was previously shown to be a risk factor for Crohn’s disease.
Our results indicate that IL4R and NOD2 constitute independent genetic risk factors for Crohn’s disease in the IBD1 region. Our finding that the IL-4R/STAT6 pathway of T lymphocyte differentiation is involved in Crohn's disease supports and extends current theories of the pathogenetic mechanisms of this disease.

CTLA-4 gene polymorphism in coeliac patients stratified by the presence of the HLA-DQ2 heterodimer

B. Mora¹, M. Bonamico², P. Indovina¹, F. Megiorni¹, M. Ferri², M. Mazzilli¹;
¹Dept. of Experimental Medicine and Pathology - University 'La Sapienza', Rome, ITALY, ²Dept. of Pediatrics - University 'La Sapienza', Rome, ITALY.

Coeliac disease (CD) is a gluten sensitive enteropathy with multifactorial aetiology. Susceptibility to CD is strongly associated with particular HLA class II alleles, while genetic factors other than HLA remain to be determined. The cytotoxic T-lymphocyte antigen 4 (CTLA-4), a downregulator of T cell activation, has been reported both in linkage and association with CD in French and Scandinavian populations. We performed case-control and family-based association studies to investigate if the polymorphism at position 49 of the CTLA-4 exon 1 was associated with the development of CD in the Italian population. The +49 A/G dimorphism was analysed in 195 CD patients, 318 relatives and 144 ethnically matched controls by PCR-RFLP method. The A allele frequency resulted increased in patients compared with healthy controls (75.6% vs. 65.6%; p=2.9x10^-3), mostly in the homozygous form (57% vs. 45.8%; p=2.8x10^-2). The segregation analysis showed a preferential transmission of the A allele to the probands (61%; c²_TDT=4.15). In order to test for an interaction between CTLA-4 and HLA, the patients were stratified according to the presence of the high risk HLA-DQ2 heterodimer. The CTLA-4 AA genotype raised to 81.5% in the DQ2 negative group showing a difference statistically significant versus both controls (p=5.2x10^-4) and DQ2 positive patients (p=4x10^-3). In the families where the affected child was DQ2 negative, the A allele resulted transmitted in 8 out of 9 informative cases (c²_TDT=5.4). In conclusion our study confirms CTLA-4 as predisposing gene for CD, with a prominent role in patients without the high risk HLA-DQ2 molecules.

The potassium-chloride cotransporter 3 gene (KCC3) is excluded from a newly defined 10.9 centiMorgan candidate region for chromosome 15 related schizophrenia (SCZD10)

J. Meyer¹, K. Schraut¹, G. Ortega¹, F. Rüschendorf², G. Nürnberg², G. Stöber¹, R. Mössner¹, K. Saar³, T. F. Wienker⁴, A. Reis², K. P. Lesch¹;
¹University of Wuerzburg, Department of Psychiatry and Psychotherapy, Wuerzburg, GERMANY, ²Max-Delbrück-Centrum for Molecular Medicine, Gene Mapping Center, Berlin, GERMANY, ³Max-Planck-Institute for Molecular Genetics, Berlin, GERMANY, ⁴University of Bonn, Institute of Medical Biometry, Informatics and Epidemiology, Bonn, GERMANY.

Despite the fact that schizophrenia is commonly regarded as a complex disorder, a seperate entry in the Online Mendelian Inheritance in Man Database (OMIM) has been created for chromosome 15 related, hereditary, catatonic schizophrenia (SCZD10, OMIM 605419), which is inherited in an autosomal dominant manner. A possible role of the chromosome 15q14-15 region in the pathogenesis of schizophrenia and manic depressive disorder has recently been reported by several scientific groups. Our group is currently investigating two large families with catatonic schizophrenia, both families support strongly the chromosome 15 locus. Genotyping of family members did lead to the definement of a 10.9 centiMorgan region between genetic markers D15S1042 and D15S659, respectively. A number of genes expressed predominantly in the brain, and localized adjacent to this chromosomal region, are excluded from the candidate gene panel by mutational analysis and the new fine mapping data. These include genes encoding the nicotinic alpha-7 receptor subunit, the potassium-chloride cotransporter 3, the ryanodine receptor 3, and connexin 36. All these proteins are important factors for brain development and function. Other genes, like the gene encoding the NOTCH ligand DLL4, remain potential candidates. We are currently narrowing down the region of interest by investigating more families and defining the gene(s) responsible for the disease by mutational analysis.

Association between TNFR2 and familial but not sporadic rheumatoid arthritis provides evidence for genetic heterogeneity.

P. Dieude¹, E. Petit-Teixeira¹, S. Cailleau-Moindrault¹, J. Osorio y Fortea¹, C. Pierlot¹, M. Martinez², O. Alibert³, S. Lasbleiz⁴, S. Faure⁵, T. Bardin⁴, B. Prum⁶, F. Cornelis¹;
¹GenHotel/Laboratoire de Recherche pour la Polyarthrite Rhumatoide, 91000, Evry, FRANCE, ²2 Institut National pour la Santé et la Recherche Médicale U358, Hôpital Saint-Louis, Paris, FRANCE, ³Commissariat à l’Energie Atomique, 2, rue Gaston Crémieux CP 5722 91057, Evry, FRANCE, ⁴Centre Viggo-Pertersen, Hôpital Lariboisière, Assistance Publique des Hôpitaux de Paris 75010, Paris, FRANCE, ⁵Généthon, 1, rue de l’internationale 91002, Evry, FRANCE, ⁶Laboratoire Statistique et Génôme, Génopole, tour Evry2 , 91000, Evry, FRANCE.

Background. Tumor necrosis factor alpha (TNFa), involved in rheumatoid arthritis (RA) binds TNFR1 and TNFR2 receptors. Genome scans have suggested the TNFR2 locus as a candidate RA locus. A case-control study in a UK Caucasian population has shown an association between a TNFR2 genotype (196R/R in exon 6) and familial, but not sporadic RA. Objective. To test this association in the French Caucasian population.
Methods. To test for an association in sporadic RA, 100 families were genotyped for the 196M/R polymorphism and analysed using the transmission disequilibrium test and the haplotype relative risk (HRR). To test for an association in familial RA, RA index cases from affected sib pair (ASP) families (n = 100) were genotyped for 196M/R. Linkage analysis was performed with 3 TNFR2 microsatellite markers.
Results. The TNFR2 196R/R genotype was not associated with sporadic RA (P= 0.72), but with familial RA (P= 0.026). The association was most marked in the context of TNFR2 “twin-like” RA sibs (affected sibs sharing both TNFR2 haplotype) (P= 0.0017). Linkage analysis was consistent with the association, the subgroup of families with 196R/R ASP index cases provided most of the TNFR2 linkage evidence.
Conclusion. This study represents the first replication of the involvement of TNFR2 in RA genetic heterogeneity. Our data refine the initial hypothesis, to suggest that a TNFR2 recessive factor, in linkage disequilibrium with the 196R allele, plays a major role in a subset of families with multiple RA cases.

D. Valverde¹, S. Bernal², I. Lorda³, A. Giménez³, M. Baiget², C. Ayuso³;
¹Universidad de Vigo, Vigo, SPAIN, ²Hospital de Sant Pau, Barcelona, SPAIN, ³Fundación Jiménez Díaz, Madrid, SPAIN.

Bardet-biedl syndrome (BBS) is a genetically and clinical heterogeneous disorder/>Until now six BBS loci have described; BBS1 on 11q13, BBS2 on 16q21, BBS3 on 3p112, BBS4 on 15q22.2-q233, BBS5 on 2q31, and BBS6 on 20p1122 with evidence for at least one more locus. For locus BBS6, the gene has been described and characterized as a chaperonin protein, with a suggested role for protein processing in limb, cardiac, and reproductive system development. Mutations in this gene are responsible for Mc Kusick Kauffman disease and Bardet Biedl syndrome.
Recently a new hypothesis has been proposed, a model of triallelic inheritance, in wich three mutant alleles segregate with the disorder.
We found three families out of 20 (15%) with anormal electrophoretic pattern in the SSCP. For family M-176 we found several variations, in exon 3 two polymorphism, silent mutations P39P and I178I, and an insertion of a C in position 938 of the cDNA that leads to a stop codon three codons ahead in position 20 of the protein. In exon 6 we found the two polymorphism described R517C and G532V and an insertion of a G in position 2557 of the coding region that leads to a stop codon 547nt.
For the family M523, we found the same allele of the later family in exon 6 with R517C, G532V and 547stop codon. In exon 4 we found another change in the intronic region of exon 3 a transition of a adenine to guanine .

Chromosome 22 investigation in rheumatoid arthritis using GenScore, a tool for candidate gene ranking in complex diseases.

J. Osorio y Fortea¹, E. Petit-Teixeira¹, G. Bana¹, S. Cailleau-Moindrault¹, P. Dieude¹, C. Pierlot¹, C. Sardou¹, S. Lasbleiz², L. Michou², S. Faure³, O. Alibert⁴, M. Martinez⁵, T. Bardin², F. Clerget⁶, B. Prum⁷, F. Cornelis¹;
¹GenHotel/Laboratoire de Recherche pour la Polyarthrite Rhumatoide, 91000, Evry, FRANCE, ²Centre Viggo-Pertersen, Hôpital Lariboisière, Assistance Publique des Hôpitaux de Paris 75010, Paris, FRANCE, ³Généthon, 1, rue de l’internationale 91002, Evry, FRANCE, ⁴Commissariat à l’Energie Atomique, 2, rue Gaston Crémieux CP 5722 91057, Evry, FRANCE, ⁵Institut National pour la Santé et la Recherche Médicale U358, Hôpital Saint-Louis, 75010, Paris, FRANCE, ⁶Hôpital Kremlin Bicêtre 78-80 rue du Général Leclerc, 94276, Le Kremlin Bicêtre, FRANCE, ⁷Laboratoire Statistique et Génôme, Génopole, tour Evry2 , 91000, Evry, FRANCE.

.
Multifactorial diseases such as rheumatoid arthritis (RA), the most frequent autoimmune disorder, involve several susceptibility genes. Typically, genome scans only provide suggestive loci, resulting in a large number of candidate genes. The aim of this work was to develop GenScore, a new tool for candidate genes ranking in complex disease studies, integrating functional and linkage data. Equal weights were given to both components, producing a GenScore ranging from 0 through 16. RA GenScore was determined for the 247 chromosome 22 genes which function was known, using our 3 cM resolution linkage data (22 microsatellites typed in 88 ASP families). The 5 prioritized genes were IL17R, MIF, IL2RB, CG12-1 cytokine and IGL gene-segment (GenScores ranking from 6 through 7). We propose GenScore as a new tool for candidate genes ranking in multifactorial diseases. Results obtained on chromosome 22 for RA will be posted at www.GenHotel.com.

Linkage disequilibrium mapping and sequence analysis of a psoriasis locus, PSORS5, on chromosome 3q21.

Lena Samuelsson, Camilla Lawer, Tommy Martinsson, Annica Inerot1 and Jan Wahlstrom
Dept Clinical Genetics, Sahlgrenska University Hospital /East and 1) Dept of Dermatology and Venereaology, Sahlgrenska University Hospital, Gothenburg, Sweden
Psoriasis is a chronic skin disorder affecting 2% of the population in northern Europe. The disease is characterised by hyperproliferation of keratinocytes and inflammatory infiltration. Psoriasis is today regarded to be a multifactorial disease with a complex genetic background.
In order to identify genetic alterations rendering predisposition to psoriasis several genome scans have been performed. This has led to the identification of several candidate loci but as of today, no single gene have been identified as disease-causing.
The psoriasis susceptibility locus PSORS5 on chromosome 3q21 was identified by our group in a genome-wide screen using a Swedish nuclear family set of 134 affected sib-pairs. Linkage was mainly found in families originating from south-west Sweden and the disease locus is likely to be caused by a founder mutation.
In order to fine-map the PSORS5 locus an SNP map spanning 900-1200 kb was created. A total of 26 SNP markers were genotyped for a large number of Swedish families. The transmission/disequilibrium test (TDT) was used to assess linkage disequilibrium between marker and disease. Five of the 26 SNPs showed significant association (p > 0.05). All five markers are located within a 160 kb region.
This region have been screened for disease-involved polymorphisms by direct sequencing of coding regions, association analysis and expression studies of candidate genes.

H. Zouali¹, S. Lesage¹, F. Merlin¹, M. Chamaillard¹, J. Cézard², J. Belaiche³, S. Almer⁴, C. Tysk⁵, C. O’Morain⁶, M. Gassull⁷, V. Binder⁸, Y. Finkel⁹, R. Modigliani¹⁰, C. Gower-Rousseau¹¹, J. Colombel¹², G. Thomas¹, J. Hugot²;
¹Fondation Jean Dausset-CEPH, Paris, FRANCE, ²Hôpital Robert Debré, Paris, FRANCE, ³CHU, Liège, BELGIUM, ⁴Hospital, Linkôping, SWEDEN, ⁵Hospital, Orebro, SWEDEN, ⁶Hospital, Dublin, IRELAND, ⁷Hospital, Badalona, SPAIN, ⁸Hospital, Herlev, DENMARK, ⁹Hospital, Stockholm, SWEDEN, ¹⁰Hôpital Saint Louis, Paris, FRANCE, ¹¹Hôpital Calmette, Lille, FRANCE, ¹²Hôpital Claude Huriez, Lille, FRANCE.

Background : IBD are complex genetic disorders of unknown etiology. We have recently identified CARD15/NOD2 as being a susceptibility gene for Crohn’s disease (CD).
Aim : Because CARD4/NOD1 shares many structural and functional similarities with CARD15/NOD2, we tested its putative role in Inflammatory Bowel Disease (IBD).
Patients and methods : The IBD families were recruited through a large European consortium. The 11 exons and intron-exon boundaries of CARD4/NOD1 were screened for the presence of variants in 77 unrelated IBD patients (62 CD and 15 UC patients) using direct sequencing. The genotyping of identified variants in IBD families was carried out using a PCR-RFLP procedure and the Transmission Disequilibrium Test (TDT) was computed by GENEHUNTER 2.0 program.
Results : Nine sequence variations were identified in the coding sequence of the CARD4 gene. Five of them (E266K, D372N, R705Q, T787M and T787K) were non conservative variants but only one (E266K) was present in more than one IBD patient. This variant was genotyped in 373 IBD families including 235 CD, 57 Ulcerative Colitis (UC) and 81 mixed families. TDT failed to demonstrate any association between the E266K variant and any of the three phenotypes : IBD, CD and UC (p>0.05). The analysis of the phenotype-genotype relationship do not show any specific characteristics neither in patients who are homozygous for E266K, nor in those carrying the 4 other non synonymous variations (D372N, R705Q, T787M and T787K).
Conclusion: These results suggest that CARD4/NOD1 do not play a major role, if any, in IBD genetic susceptibility.

Interactive effect of DQA1*0101 and the HB*14 AChRa subunit polymorphism on anti-AChR autoantibodies in autoimmune Myasthenia Gravis.

M. Giraud, B. Eymard, C. Tranchant, P. Gajdos, H. J. Garchon;
INSERM U25, Paris, FRANCE.

The HB*14 microsatellite allele located within the CHRNA gene, that encodes the alpha subunit of the muscle acetylcholine receptor (AChR), the HLA classe II DQA1*0101 allele and the DR3 haplotype were previously associated with an increased risk of acquired autoimmune generalized Myasthenia Gravis (MG) using a case-control design. We looked for an influence of these three markers on anti-AChR antibody titers of 480 seropositive MG patients by ANOVA. We took thymus histopathology into account since it markedly influences anti-AChR antibody titers. Their distribution was normalized by logarithmic transformation. A synergistic effect of HB*14 and DQA1*0101 on autoantibody titers was observed (P<0.03). In DQA1*0101+ patients, the increase in autoantibody titers associated with the presence of HB*14 was 10-fold in average in the subgroup of thymectomized patients with thymus hyperplasia or with a normal thymus (P=0.0006). Consistent with these findings, HB*14 was preferentially transmitted to DQA1*0101+ MG patients of this subgroup (P=0.031). This effect of HB*14 was enhanced in haploDR3+ patients, with a 18-fold increase of their autoantibody titers (P=0.0006). Moreover, patients with high levels of autoantibody titers (>100nM) had an increased frequency of these three markers (OR=15.71 ; P=0.0032). Our data strengthen the hypothesis of an immunological role of the CHRNA gene product in MG predisposition. HB*14 would influence the immunogenicity of the alpha subunit of the AChR, perhaps through presentation of a novel epitopic variant by the DQA1*0101 gene product. This effect would be enhanced in the context of the DR3 haplotype, that would provide non-antigen specific immune dysregulation.

S. Lesage¹, H. Zouali¹, F. Merlin¹, M. Chamaillard¹, J. Cézard², J. Belaiche³, S. Almer⁴, C. Tysk⁵, C. O'Morain⁶, M. Gassull⁷, V. Binder⁸, R. Modigliani⁹, C. Gower-Rousseau¹⁰, J. Colombel¹⁰, G. Thomas¹¹, J. Hugot²;
¹Fondation Jean Dausset/CEPH, Paris, FRANCE, ²Hôpital Robert Debré, Paris, FRANCE, ³CHU, Liège, BELGIUM, ⁴IHM, Linköping Universitet, Linköping, SWEDEN, ⁵örebro Medical Center Hospital, Örebro, SWEDEN, ⁶Adelaide&Meath Hospital, Dublin, IRELAND, ⁷Hospital Universitari Germans Trias i Pujol, Baladona, SPAIN, ⁸Herlev Hospital, Herlev, DENMARK, ⁹Hôpital Saint-Louis, Paris, FRANCE, ¹⁰Hôpital Calmette, Lille, FRANCE, ¹¹Dondation Jean Dausset/CEPH, Paris, FRANCE.

Introduction: Recently, the IBD1 locus on chromosome 16 has been identified as CARD15, which encodes an intracellular protein involved in NFkB activation and apoptosis. It has been shown that mutations in the CARD15 gene have been associated with Crohn’s disease (CD) but not with Ulcerative Colitis (UC) or in controls.
Aim: To assess the relative importance of CARD15 mutations in « mixed families » where CD and UC co-existed.
Patients and methods: The entire coding sequence of CARD15 gene was screened either by sequencing or by dHPLC in an European cohort of 167 patients (85 CD and 82 UC) from 78 mixed families and 103 healthy controls (HC). Allele frequencies in CD and UC groups were compared with those in HC using the chi-square test.
Results: Twenty-six variants were identified, including the 3 main CD-associated variants (R702W, G908R and 1007fs) in the CD and UC groups. Ten rare mutations identified in both groups were considered as potential disease causing mutations. The allele prevalence of the three main variants R702W, G908R and 1007fs was 4.1%, 4.7% and 8.8% for CD patients; 1.8%, 4.3% and 3.0% for UC patients and 4.4%, 1.0% and 1.9% for controls, respectively.
Conclusion: In the « mixed » families, the G908R variant was associated to both CD and UC (p=0.03 and p=0.04, respectively), whereas the variant 1007fs was only associated to CD (p=0.002). In contrast, the allele frequency of R702W which was found associated to CD did not differ significantly in the CD group and the controls.

Haplotypic and cladistic analyses revealed a marked dispersion of CD-predisposing nucleotide variations in CARD15/NOD2

M. Chamaillard¹^,2, H. Zouali², S. Lesage², J. Cézard³, J. Belaiche⁴, S. Almer⁵, C. Tysk⁶, S. Montague⁷, M. Gassull⁸, V. Binder⁹, Y. Finkel¹⁰, P. Puig¹¹, C. Gower-Rousseau¹², J. Macry¹³, J. Colombel¹², M. Sahbatou², G. Thomas¹⁴^,2^,11, J. Hugot³^,2^,11;
¹INSERM, Paris, FRANCE, ²Fondation Jean Dausset, Paris, FRANCE, ³PEWG-IBD, Department of Paediatric Gastroenterology, Hopital Robert Debré, Paris, FRANCE, ⁴GETAID, Department of Gastroenterology, CHU de Liège, BELGIUM, ⁵Division of Gastroenterology and Hepatology, IHM, Linköpings Universitet, Linköping, SWEDEN, ⁶Department of Gastroenterology, Örebro Medical Center Hospital, Örebro, SWEDEN, ⁷Department of Gastroenterology, Adelaide & Meath Hospital, Dublin, IRELAND, ⁸Department of Gastroenterology, Hospital Universitari Germans Trias i Pujo, Barcelona, SPAIN, ⁹Department of Gastroenterology Herlev Hospital, Herlev, DENMARK, ¹⁰Department of Gastroenterology, Karolinska Children’s Hospita, Stockholm, SWEDEN, ¹¹INSERM U434, Paris, FRANCE, ¹²Registre EPIMAD, Hopital Calmette, Lille, FRANCE, ¹³INSERM U458, Hopital Robert Debré, Paris, FRANCE, ¹⁴Department of Gastroenterology, Hopital Saint Louis, Paris, FRANCE.

Crohn's disease (CD) is a worldwide early-onset complex condition of uncontrolled inflammation of the gastrointestinal mucosa. We have recently identified three dominating CD‘s predisposing variants (Arg702Trp, Gly908Arg, Leu1007fs) within CARD15/NOD2 gene (Nature, 411 : 599-603, 2001). CARD15 is a new member of the CED4/Apaf1 superfamily. The more drastic mutation, Leu1007fs, has been established as a causative variant with a loss of LPS-induced NFkappaB activation. The role in pathogenesis of 32 others non-conservative amino acid changes are currently investigated.
In order to explore the origin and dispersion of CD-predisposing alleles, a total of 232 CD families were recruited through a large European consortium : France (n=128), Belgium (n=23), Scandinavia (Sweden and Denmark, n=17), Mediterranean area (Italy, Spain, Portugal and North of Africa, n=37), Ireland (n=2), Poland (n=2), originating from France and another country (n=20), and from India (n=1), Sri-Lanka (n=1) and Iran (n=1). All these families were genotyped for 15 markers in and nearby NOD2/CARD15 including the three non-synonimous sites described above. Haplotypes were built using Haplodump implemented in the GeneHunter package. We report herein a distinct origin of Arg702Trp, Gly908Arg, Leu1007fsinC. The allele frequency of Arg702Trp, and Leu1007insC is partitioned within Europe (p=0,05 and p=0,02 respectively), which is consistent with recent and possibly multiple recurrent mutations events. Susceptibility haplotype distribution in Scandinavian CD families is statistically marked (p<0.05), which could reflect by genetic heterogeneity. In addition, we make optimal use of flanking SNPs to perform cladistic analyses illustrating the importance of phylogenetic analyses and present-day haplotypes complexity for gene-mapping strategy.

GenHotel, a new approach for candidate gene studies in multifactorial diseases applied to rheumatoid arthritis.

L. Michou¹, J. Osorio², E. Petit², C. Pierlot², T. Bardin¹, S. Lasbleiz¹, F. Cornélis for ECRAF²;
¹Unité de Génétique Clinique, Hôpital Lariboisière, Paris, FRANCE, ²GenHotel/Laboratoire de Recherche Européen pour la Polyarthrite Rhumatoïde, ECRAF-Université Paris VII, Evry-Genopole, FRANCE.

Introduction : Genome scans in multifactorial diseases result in a large number of candidate genes. To facilitate candidate gene investigation for rheumatoid arthritis (RA), the most common autoimmune disease, we set up GenHotel : the invitation to come and test hypotheses on a common resource (www.GenHotel.com).
Aim of the study : to illustrate the GenHotel approach with the test of RA associated HLA-DRB1 alleles.
Patients and methods: DNA of 100 caucasian French families with one RA patient and both parents were
genotyped for HLA-DRB1. Analysis was performed with the haplotype relative risk (HRR) and the transmission disequilibrium test (TDT). P < 0.05 was considered suggestive and < 10-6 demonstrative.
Results:
HRR : the allele frequency of RA-associated alleles (A) was 57% in transmitted chromosomes versus 28% in non-transmitted, chi 2 = 35,6 (P<10-7)
TDT : out of 97 heterozygous parents A/X, the A allele transmission was 79%, versus 50% under Mendel’s lawx, chi 2 = 33,5 (P<10-7)
Conclusion: The results demonstrated the HLA-DRB1 contribution to RA. The GenHotel appoach advantages include genotype quality control, increased robustness from familial based analysis and ability to test complex hypotheses involving haplotypes, inprinting and gene interactions.