6th Annual BMRP Investigator Meeting - Abstract

 

R Thiebaut¹, Sophie Esmiol-Welterlin², Françoise Merlin¹, Jean-Frederic Colombel³, J Belaiche and the GETAID³, Vibeke Binder³, Sven Almer³, M Finkel³, Curt Tysk³, C O’Moraim³, Miguel Gassull³, Mark Lathrop⁴, Patrick Lécine Lécine², Leigh Pascoe⁵, Vincent Ollendorff², Jean-Pierre Hugot¹

 

¹INSERM and Paris Diderot University (France); ²INRA and Université Montpellier 2 (France); ³European consortium on the genetics of IBD; ⁴Centre National de génotypage; ⁵Fondation Jean Dausset (Paris, France).

 

Card15/Nod2 is a well established gene for Crohn’s Disease (CD) susceptibility. The working hypothesis is that genes coding for proteins able to physically interact with Nod2 could be important for IBD susceptibility.

 

Using an upgraded yeast two hybrid (Y2H) system with three reporter genes and NOD2 as bait, we have identified 15 double positive genes coding for physical partners of Nod2, including RICK/RIP2 and 14 new candidate genes. Y2H experiments were done on lung and colon DNA libraries (for a total of 30 10⁶ clones). Confirmation of the physical interactions by coiimunoprecipitation experiments is in progress.

 

For the 15 genes, we defined 144 single nucleotide polymorphisms (SNPs) after sequencing 30 independent IBD cases and based on the hapmap database. These tested SNPs define the vast majority of haplotypes with frequencies higher than 0.05. 300 IBD European families (corresponding to 1500 patients and their relatives) were genotyped for these 144 markers. Data analyses are in progress.

 

Among the 80 SNPs yet analysed, it was possible to define a specific haplotype associated with CD for the candidate gene CG10 (TDT test: transmited/untransmited haplotype 60/33 P<0.005). A . The frequency of the associated haplotype was estimated to 0.1 in these Caucasian families. Considering the large number of tests done, this result was only indicative. We thus performed a replication study on an independent set of 500 (less informative) IBD families. We were able to replicate the association with the same haplotype and the CD phenotype (Transmitted/Untransmitted : 40/20; P<0.01). Pooling the two datasets, the P value was lower than 2 10^-4. The transmission disequilibrium was significantly higher in non mutated patients than in the group of patients carrying CARD15/NOD2 mutation (P<0.03) providing an independant argument to link CG10 to CD. No association was found with the UC phenotype.

 

We are currently exploring CG10 by resequencing a large number of IBD patients and by performing functional analyses.

 

^aPrincipal Investigator