Several factors affected the successes of the positional cloning of Ncfl. First of all, it is a gene with a large effect size, accounting for about 20% of the total phenotypic variation . Moreover, the mutation has a dominant effect on a well-characterized phenotype and also, the rest of the genome in the DA
background allows the functional expression of the allele. However, in many cases a QTL effect is the sum effect of several positionally linked genes and their interactions with other genes both within and outside the QTL. This certainly makes positional cloning more difficult.
One such example is two of the first QTLs reported to control an animal model of an autoimmune disease: the Eae2 and Eae3 loci associated with experimental autoimmune encephalomyelitis . Eae2 and Eae3 are located on chromosomes 3 and 15, respectively, and these loci have an interacting effect on the disease . We made congenic strains with these fragments from the resistant RIIIS/J strain on the susceptible B10.RIII background. The effect on disease in these mice was mild, far from being Mendelian and therefore not suitable for positional cloning of the underlying genes. Clearly both the environmental and genetic conditions needed to be better defined in order to increase the effect of the RIIIS/J alleles [37,38]. Firstly we found that both loci also controlled CIA. Moreover, studying environmental factors affecting the arthritis phenotype we found that inducing the disease without tubercle antigens in the adjuvant increased the penetrance of the congenic effect in both chromosome 15 and chromosome 3 congenics. We also utilized the observation that these loci interacted and developed a strategy to define the genetic interactions. The congenics were intercrossed for more than eight generations, resulting in a partial advanced intercross cohort of roughly 1,000 mice. This accumulated both the recombination density and statistical power to dissect the loci into several subloci and to identify the interactions between them. Within the previous QTL on chromosome 3 (Eae3), we were able to identify three separate loci (Cia5, Cia21, and Cia22) and within the previous QTL on chromosome 15 (Eae2) we found four loci (Cia26, Cia30, Cia31, and Cia32). Most of these loci interacted with each other (Fig. 5) and RIIIS/J alleles at the different loci could be either disease-promoting or disease-decreasing. For example; Cia5 affected the onset and early severity of arthritis in additive interaction with Cia30 on chromosome 15, whereas the Cia21 and Cia22 affected severity during the chronic phase of the disease in epistatic interaction with Cia31 on chromosome 15. The definition of the environmental conditions and the genetic interactions was a prerequisite to dissect the Eae2 and Eae3 QTLs. Human studies will have the same problem, but they are more readily seen and analyzed in animal model studies. In the case of the Eae2/Eae3 project, the interactions between the QTLs could be detected in the F2 intercross with no more than approximately 100 animals. However, the breakdown of the two QTLs into seven and detecting the specific interactions among them was only possible by collecting a larger number of recombinations in approximately 1,000 mice. What we learned was that by analyzing the genetic interactions in the F2 intercross we could get information
on how to design congenic strains with the optimal effect on the phenotype. The knowledge of the genetic interactions is then not only useful in the positional cloning process but also once the gene is identified and studied in humans.
The interactions within a QTL are another situation and they pose both new problems and possibilities. The problems are the difficulties in splitting up a tightly linked haplotype. This is an apparent problem in human studies, for example in addressing conserved haplotypes within the MHC region. In animal models, haplotypes can be intentionally broken up through the search of recombinations. Such recombinations will often lead to new and changed effects on the phenotype, possibly caused by a balancing interaction between closely positioned genes. It will be a challenging task to investigate in detail such clusters but it is also important to remember that the minimal congenic fragment, containing such conserved haplotypes, are likely to have been selected for a balanced control of a specific pathway. Such congenic strains will therefore be ofvalue for studies ofpathophysiologic pathways and thereby for understanding the fundamental laws determining the possibility of developing drugs with efficiency and with limited side-effects.
The possibility of genetically modifying mouse strains through transgene and embryonic stem cell-based technologies is a major advantage in the further dissection of the critical genes and haplotypes. Studies on the role of MHC genes are an excellent example of this. The establishment of H-2 (i.e., mouse MHC) congenic strains was the basis for the identification of the MHC regions and such strains are still useful for studies of the role of MHC. It has, however, been difficult to split this region in order to isolate each gene and thereby understand its function. Instead, transgenic experiments have been conducted with which it has been possible to insert both specific murine class II genes [39, 40] and their human counterparts [41-44], and thereby prove the role of these genes for the development of autoimmune disease such as RA and MS. These studies also show the risk in transgenic technology as overexpression of class II genes may also lead to an artifactual toxicity in B cells [45,46], which is the most likely explanation for a suppressive effect by transgenic class II genes, as has been reported in diabetes in the NOD mouse.
Was this article helpful?