Minggu, 03 Juli 2016

Quantitative Trait Loci for Obesity and Type II Diabetes in Rodents

Source: http://www.cimedicalcenter.com/


Obesity and type II diabetes, which often coexist, are very cotnmon diseases in humans.
An estimated 65 or 31 % of U.S. adults are overweight or clinically obese, defined as a
body mass index greater than 25 or 30 kg/m2, respectively (1). In addition, about 90–95%
of diabetes patients suffer type II diabetes (2,3), and currently this accounts for
approximately 16 million patients in the United States.

The genetic contribution to human obesity and type II diabetes has been revealed
through twin, adoption, and family studies, demonstrating that an individual with obese
and/or diabetic relatives has a higher risk for being affected by these diseases [4–8].
There are some rare forms of obesity and type II diabetes that are caused by a single
gene mutation, and these include mutations of Ay (agouti), Cpe (carboxypeptidase E), Lep
(leptin), Lepr (leptin receptor), and Tub (tubby) and the maturity-onset diabetes of the
young (MODY) [9–11]. To date, six MODYs, MODY1, 2, 3, 4, 5, and 6 have been
reported in humans and the responsible genes are hepatocyte nuclear factor (HNF)-4α,
glucokinase, HNF-1α, insulin promoter factor-1/pancreas duodenum homeobox-1/islet
duodenum homeobox-1, HNF-1β, and NeuroD/BETA2, respectively (11). Mutation in
the insulin 2 gene (Ins2) has been also reported for one murine MODY in the Akita
mouse model (12).

Most common forms of obesity and type II diabetes in humans, however, follow
polygenic inheritances: i.e., multiple genes are involved in the development of these
diseases (13,14). These multiple genes are referred to as susceptibility genes, reflecting
the concept that these genes confer an increased susceptibility to a disease rather than the
certainty of developing the disease (14,15).

Animal models including rats and mice have long been an adjunct to human studies,
minimizing many difficulties encountered in carrying out genetic studies of obesity and
type II diabetes in human populations (16). For example, the capability of genetic and
environmental controls, the availability of inbred strains and the ability to generate large
experimental cohorts, and the short generation cycle can simplify and facilitate genetic
studies. Furthermore, as rodents and humans share basic biological and physiological
characteristics, and gene order over large distances has been conserved through
evolution, candidate genes or pathways found in rodents can readily be tested in humans
(17,18). Indeed, obesity genes known in humans, such as Lep and Lepr, were discovered
in mice first (19,20). This review will discuss common strategies for dissecting genetic
factors underlying obesity and type II diabetes using polygenic rodent models and the
related genetic studies.


2.1. Complex Traits and QTLs
When a one-to-one association between genotype and phenotype does not exist for
certain traits, these traits are called complex traits as opposed to simple mendelian traits
(21,22). This inconsistency between genotype and phenotype, despite an evidence of
strong heredity for the traits, results from the fact that complex traits, unlike single
mendelian traits, are determined by multiple factors including genes and environments.
Complexity is created by the presence of multiple genes contributing in different degrees
to the trait, possible interactions among those genes, and interactions between genes and
environments in the determination of the traits (15,23). The list of complex traits can
include natural traits such as skin color, wavy hair, height, and behavioral characteristics
as well as disease traits [called complex diseases) such as hypertension, obesity, diabetes,
alcoholism, and cancer (23–25).

Because of the involvement of multiple genes in controlling the traits, complex traits
are also referred as polygenic traits. These traits are usually quantitative or assessed
quantitatively, and thus the trait controlling genes (or loci) are called QTLs (21). A
review of terminology frequently used in genetics is available (26).
Quantitative trait loci for obesity and type II diabetes in rodents 15

2.2. Dissecting Genetic Factors for Complex Disease Using Polygenic Animal Models
2.2.1. Sources of Polygenic Models
Selected Strains. An often-used approach to create animal models harboring genetic
variation is long-term breeding with phenotypic selection (27,28). When individuals in a
population differentially express a trait that is heritable (often found in heterogeneous
outbred populations), breeding individuals ranking at phenotypic extremes (i.e., heaviest
or lightest) can produce offspring that also rank at the extremes (28). Repeating this
selective breeding (usually followed by inbreeding) over several generations can fix
genetic variants that contribute to the selected trait, creating new lines that possess the
extreme phenotypes (9,28).

Standard Inbred Strains. Inbred strains are defined to be homozygous for each gene
throughout the genome, and numerous inbred strains, especially for mice, are currently
available (25,29,30). Because standard inbred strains are generated via repeated brother-sister mating at random over at least 20 generations, these phenotypic variations among the strains result
from naturally occurring gene combinations (28). Indeed, many phenotypic variations
including adiposity and glucose metabolism have been reported among existing inbred
strains (27) (http://aretha.jax.org/pub-cgi/phenome/mpdcgi?rtn=docs/home). With the
currently growing database for standard inbred strains regarding genetic maps
(http://www-genome.wi.mit.edu/cgi-bin/rat/gmap_search; http://wwwgenome.
wi.mit.edu/cgi-bin/mouse/index) and phenotypic characterizations (31), these
can serve as very valuable resources for biomedical science including complex disease
field of obesity and diabetes.

2.3. Mapping of a QTL
Mapping, i.e., identifying the chromosomal location, is the first step to dissect the genetic
factors contributing to a trait. Genome-wide QTL linkage analysis (or genome-wide scan)
has been a powerful way for comprehensively mapping genetic factors underlying
complex diseases (23). Because this approach does not require any molecular knowledge
of the traits of interest, it has the potential of discovering new genes or pathways not
previously known.

The approach consists of studying whether there is an association (cosegregation)
between the genotype at a marker locus and the trait values, and if there is, then this
indicates that the marker locus is close by (or linked) to the putative disease QTL (14,26).
2.3.1. Genetic Crosses to Create Segregating Populations

Commonly, QTL mapping is initiated by crossing two different inbred strains that show
contrasting phenotype and genotypic variation (9,27). The resultant F1 mice are then
intercrossed (sib mating) or backcrossed to one of the parental strains. This will generate
F2 or backcross (BC) progeny in which both phenotypes and genotypes segregate unlike in the grandparental and parental generations. The origin of this segregation is the
recombination of the DNA segments between the homologous chromosome pairs
occurring during meiosis in the production of germ cells from the F1 parent. The
recombination events are random and, consequently, individual F2 or BC mice possess a
unique combination of progenitor alleles, which gives rise to segregation of genotypes
and concurrently phenotypes. Phenotypes can, however, be influenced by environmental
factors, such as high and low fat diets (25).

The choice of methods for genetic crosses is reviewed in detail in elsewhere (21,32).
In addition to F2 or BC mice, recombinant inbred (RI) strains, which have a fixed
genotype, have been used for QTL mapping, and the usage of RI strains is thoroughly
reviewed elsewhere (33,34). A segregating population as described in the preceding paragraphs is then genotyped throughout the entire genome using a series of genetic markers that are polymorphic, i.e., they differ between the two parental strains. Using inbred strains provides a great
advantage in that only two alleles of genetic factors originating from each of the
crossbred strains segregate in the population, and this makes all polymorphic markers
informative for the genotype at all loci (32).

The most commonly used genetic markers are microsatellite repeats, also known as
simple sequence repeats, that are present throughout all mammalian genomes examined
and found to be highly polymorphic (there is variation in the number of repeats rather
than in the sequence) (35,36). Microsatellite loci do not appear to have any functionality
(36). The most frequently found microsatellites contain a (CA)n multimer, often referred
to as a CA repeat. Microsatellite loci can be easily typed by polymerase chain reaction
(PCR) amplification with primers designed from unique flanking sequences on each side
of the repeats (36,37). Variations in the length of the PCR products can be detected by
separation on agarose gels or poly-acrylamide gels or by automated system using
fluorescent-labeled primers (38,39) (Fig.1).

Single nucleotide polymorphisms (SNPs) can also be employed as genetic markers
(40). SNPs are single-base variations in DNA sequences which are present throughout the
mammalian genomes examined so far and which occur with an average frequency of 1
per 1000 base pairs (in humans), thus is overly superior to the microsatellite frequency
(41). Currently, high-density SNP maps are produced in the human genome as well as in
mouse genome (42,43), and SNPs will be more commonly used for linkage analysis
when more cost-efficient high-throughput technology of SNP genotyping is available.
Individual F2 or BC animals are also scored for the trait of interest, such as body
weight, plasma glucose and insulin levels, or core temperature. Subsequently, using
statistical methods, the individual phenotypic scores are examined for correlation with the
genotypes of the polymorphic markers (44). Regardless of the statistical methods
applied, the basic tenet is that markers that are significantly associated with the trait lie
close to (are linked to) QTLs that are responsible for variation in the trait of interest. This
is based on the assumption that the recombination during meiosis less likely occurs
between closely linked loci on the same chromosome, resulting in cotransmission of
Genomics and proteomics in nutrition 18 these closely linked loci (15). With known map positions of genetic markers, the genomic locations of QTLs on the chromosomes can be estimated statistically.
Further details about statistical analysis for QTL mapping including methods, number
of markers and animals required, and software available are very well reviewed
elsewhere [44–48]. One statistical model for data analysis derived from a cross between
outbred strains has been discussed by Nagamine and Haley (49).

Author: Jung Han Kim
Nutrition Department, University of Tennessee, Knoxville, Tennessee,

Tidak ada komentar:

Posting Komentar