Molecular Biology of Lignification in Grasses - Botanical Research - page 95

Colocalizations between lignin or cell wall degradability QTL and candi-
date genes involved in cell wall biosynthesis and assembly ( Barri`re et al.,
2009a,b ) were investigated, based on the physical positions (in the B73
sequenced line, Schnable et al., 2009 ) of genes and putative QTL estimated
physical positions. Colocalizations between QTL and genes involved in
monolignol biosynthesis were quite rarely observed, except for ZmCCR1 in
bin 1.07. The most frequent associations were shown between cell wall
related traits and transcription factors or genes involved in tissue patterning.
Such colocalizations with ZmMYB were observed in bins 1.07 and 6.06, and
NAC in bins 4.05, 4.08, 9.02 and 10.03/04. As a tentative conclusion, natural
variation for cell degradability appeared more likely related to regulation
mechanisms than to efficiency variation of monolignol enzymes.

VI. NETWORK AND CO-EXPRESSION DATA
OF LIGNIN GENES

Sequence and gene expression data have greatly enhanced our understanding
of how biological systems function. As of January 2012, more than 3000
genomes from prokaryotic, eukaryotic and archae organisms have been fully
sequenced, with more than 7000 sequencing projects currently in progress
( Liolios et al., 2010 ). Microarrays and RNA deep sequencing studies are now
generating a growing repository of gene expression data for various organ-
isms, tissues and perturbations, with more than 25,000 microarray samples
available for Arabidopsis thaliana alone (Gene Expression Omnibus data-
base, Edgar et al., 2002 ). A major finding, arising from analysis of available
transcriptome data, is that functionally related genes tend to be transcrip-
tionally coordinated (i.e., co-expressed; reviewed in Usadel et al., 2009 ).
Using guilt-by-association approaches, co-expression analyses have proved
invaluable for rapid inference of gene function, subcellular localization of
gene products and biological pathway discovery ( Persson et al., 2005; Wei
et al., 2006; Yonekura-Sakakibara et al., 2008 ).
An example of co-expressed genes can be seen in Fig. 2 , where genes PAL
(blue line) and 4CL (red line) both potentially involved in lignin biosynthesis
show visible similarities of expression profiles across microarray data repre-
senting different tissues. Ribosomal subunit S15A (green line), which is not
involved in lignin biosynthesis, is included as a control. While many different
metrics can score similarities of gene expression profiles, the most widely used
is metric Pearson correlation coefficient (r-value), which can range from 1
(perfect correlation), 0 (no correlation) to 1 (perfect anti-correlation).

Using the genes in Fig. 2 , for example, r(PAL,4CL)

0.88 (strong positive

¼