Genetic dissection of multifactorial disease models in the rat

Abstract

Type 2 diabetes (T2D) is characterized by two apparently distinct pathophysiologic defects - insulin resistance and a failure of the pancreatic Beta-cells to compensate for this resistance by appropriately increasing insulin secretion. There is strong evidence for an important role of genetics in both of these components. However identifying the genes responsible for the development of the disease has proven problematic because of complex background. Linkage analysis using the Goto-Kakizaki (GK) rat, a well-established genetic model for T2D, provides an ideal tool for understanding the pathogenesis of the disease by highlighting the chromosomal regions, which might harbor candidate diabetogenes. Furthermore by creating reciprocal crosses or changing environmental conditions, the effects of parent-of-origin or environmental factors on the manifestation of a quantitative trait locus (QTL) can be elucidated at the genome-wide level.

We have successively created three cohorts of F2 intercrosses of GK and F344 rats, founded by either GK grandmothers (F2GK) or grandfathers (F2F344). Parental, F 1, and F2 rats underwent a number of phenotypic assays i.e. intraperitoneal glucose tolerance test (IPGTT) with measurement of glucose, insulin and body weight. The first genome-wide scan with 253 markers revealed three glucose-related loci, Niddm I (LOD 11), Niddm2 (LOD 4.5) and Niddm3 (LOD 4.6), and one locus, Weight1 (LOD 6.2), for body weight on chromosomes 1, 2, 10, and 7 respectively, supporting the polygenic inheritance pattern of T2D in the GK rat. Furthermore clearly different models of inheritance for hyperglycemia at baseline vs. during the IPGTT were demonstrated, implicating different regulatory mechanisms at these two states. In this study paternal inheritance of diabetes of the GK rat was observed, in that GK fathers conferred significantly higher glucose and insulin levels in the IPGTT than GK mothers in the F I progeny.

By comparison of linkage analyses between two reciprocal crosses, four major parent-of-origin-specific QTLs were uncovered, implicating potential importance of mitochondrial and/or epigenetic effects in the GK rat. Niddm4, a major locus on chromosome 9 affected fasting insulin paternally (LOD 6.9 in F2F344 rats vs. 0.66 in F2GK rats). The locus also exerted a significant maternal-specific effect on glucose-induced insulin levels at 120 min, thus showing dual characteristics.

In this study we exhibited a distinct distribution pattern of QTLs in two reciprocal crosses, which indicated that genetic susceptibility genes to T2D might be fixed in the GK rat, but their effects were expressed in a manner that depends on the parent-of-origin. We also elucidated the additive and epistatic effects of QTLs. Mitochondrial genomes of the GK and F344 rats have been sequenced and the resulting sequences were checked against the published rat mtDNA sequence. In total 31 sequence variants were found in the GK rat as compared to the sequence of the F344 rat.

We uncovered seven gender-influenced QTLs giving strong evidence for linkage on chromosomes 1, 2, 5, 7, 10, and 18 in GK and F344 crosses. By analyzing the gene-gene interaction conditionally on gender effect, eleven locus-pairs showing epistatic interaction (p-value less than 0.0001) were identified to contribute to diabetic susceptibility in the F2 progeny.

Using the same approach in the study of experimental autoimmune neuritis (EAN) in rats, a model of GuillainBarré syndrome (GBS), a new region showing linkage to EAN phenotypes was found on chromosome 17. This locus showed strong evidence for linkage in affected rats vs. weak linkage in unaffected rats. Moreover the linkage to serum levels of anti-peripheral nerve myelin (anti-PNM) IgG2b and IgGI was much stronger in the cross with a Dark Agouti (DA) female founder and among male rats respectively.

A good linkage map with high accuracy and resolution is essential for a successful QTL mapping. Therefore we completed a linkage map of GK and F344 crosses based on a large number of meiosis (564 rats). The map consists of 530 markers and spans the entire rat genome in a total length of 1826.9 cM with an average interval of 3.4 cM. Subsequently based on the linkage map an integration of 7 linkage and 2 radiation hybrid maps of the rat was created. The integration provides reference to 8627 rat genetic markers. Among them there are more than 550 genes with homology information in human and mouse indicated.

In conclusion, by using the GK rat as a diabetic model we have demonstrated the complex genetics of T2D even in an inbred strain. In additional to the polygenic inheritance, effects of susceptibility genes were found to manifest conditionally, e.g. to depend on gender or parental origin, to respond to environment, or to interact with other genes. Our results provide an annotation on the genetics of TEND with detailed information and a platform to a further molecular and cellular analysis.