Biology Reference
In-Depth Information
plant genes, represent genuine pleiotropy. However, this may not be the
case—one of the main issues with QTL analysis is the large confidence
intervals when an F2 or backcross (BC) format are used, due to the relatively
low number of recombinations that may have occurred between closely linked
genes ( Lander and Botstein, 1989; Lynch and Walsh, 1998 ). In this case, such
large confidence intervals will be unable to distinguish close linkage from plei-
otropy without very large sample sizes.
As an illustration, in the chicken study by Wright et al.( Wright et al.,
2010, 2012 ), which analyzed a wide variety of behavioral (anxiety and
fear
avoidance), morphological (growth and bone density) and life-history
(onset of sexual maturity and fecundity) traits, a very large sample size and
multivariate statistical analysis were used to attempt to dissect pleiotropy
from linkage. Whereas the central loci in the modules were shown to be
indistinguishable from pleiotropy, those on the periphery were linked, rather
than pleiotropic. A follow-up study using an advanced inter-cross (to increase
the number of recombinations present in the cross) examined one module,
which affected comb mass and bone allocation, and this was genuinely
pleiotropic ( Johnson et al., 2012 ). Pooling these different results from domesti-
cated species, it appears that these modules may contain central “cores” of
pleiotropic loci, surrounded by more loosely linked loci.
This idea of loose-linkage was first alluded to by Grant (1981) .He
noticed that even with a strong domestication phenotype, multiple correla-
tions between different aspects of the domestication phenotype would rapidly
break down when reverse selected or crossbred. Although he noticed this
with several plant-based examples, the same is true with the correlations
between different domestication characteristics breaking down rapidly during
inter-crossing ( Albert et al., 2008, 2009; Wright et al., 2010 ). The uses
of this in domestication would be to enable new phenotypes to be rapidly
bred into a certain background (far harder if traits are generally all pleiotro-
pic), and the pre-existing domestic phenotype regained. Though this may be
an advantage, as it allows desirable characteristics to be rapidly bred into the
population and then the previous domesticated phenotype to be rapidly
regained, it is unclear whether this ability would be specific to domesticated
species (potentially explaining the rather limited numbers of domesticated
species) or is a general feature of strong directional selection on phenotypic
evolution, which is then exploited in the domestication process.
In addition to modularity being observed in the organization of loci
affecting different aspects of the domestic phenotype, modularity can also be
seen in genome organization and gene expression in general. Using the
whole genome expression variation, it appears that the differences between
wild and domestic populations are due to hundreds or thousands of gene
expression changes ( Lai et al., 2008; Natt et al., 2012; Rubin et al., 2007 ).
Once again, these differentially expressed genes tend to be grouped together
into modules that are once again close to one another. In effect, what this
Search WWH ::




Custom Search