healthmedicinet

Allelic Diversity of OsC1 Responsible for Variation in Leaf Sheath Colors

Plants accumulate anthocyanin in various tissues as an aid to survival and reproduction.
In rice, anthocyanin is deposited in the root, leaf sheath, internode, leaf blade,
lemma, palea, apiculus, stigma, and pericarp. Tissue-specific anthocyanin accumulation
is common in numerous genotypes. Purple pigmented traits in different tissues do not
always co-segregate (Sakamoto et al. 2001]). However, purple leaf sheath, apiculus, and stigma cosegregated in the F
₂
population of PSH TNG72 × GSH TCS17 (Fig. 1) and other populations (Fan et al. 2008]; Gao et al. 2011]). Anthocyanin accumulation in the apiculus had previously been related to OsC1 (Takahashi 1957]; Saitoh et al. 2004]) but in the leaf sheath was thought to be inherited by polygenes or a single gene
(Fig. 2; Fan et al. 2008]; Wang et al. 2009]; Gao et al. 2011]).

OsC1, conferring leaf sheath color, was isolated from the F
₂
population of TNG72 × TCS17 by positional cloning (Fig. 2), differing from PSH1 (t) but the same as a locus identified from the somaclonal line Z418 (Wang et al.
2009]; Gao et al. 2011]). The PSH allele was dominant to GSH in crosses between natural germplasm or mutant
lines. PSH TNG72 had a full length OsC1 allele encoding 272 amino acids while GSH TCS17 had a 10-bp deletion in the R3 Myb
domain in exon 3, resulting in truncated translations of 207 amino acids. GSH cultivar
Nipponbare had another mutated allele, a 3-bp deletion in the R3 Myb domain in exon
2 (Fig. 2). Among the 50 accessions analyzed herein, the 10-bp deletion was found in 17 indica landraces and improved cultivars from Taiwan, China, and the Philippines, and 2 japonica landraces from Taiwan (Fig. 4). The 10-bp deletion was conserved in 17 indigenous varieties in Northeast India
and only observed in indica varieties from Taiwan, China, India, and Indonesia (Saitoh et al. 2004]; Choudhury et al. 2014]).

Ten GSH accessions without the 10-bp deletion of OsC1 had other mutated alleles or genes, leading to absence of anthocyanin pigmentation.
Deletions of 3-bp in exon 2 and 2-bp in exon 3, both in the R3 Myb domain, were found
in japonica rice from Japan and China, respectively (Figs. 2 and 4; Saitoh et al. 2004]). Two GSH accessions, Midon and T65, had an amino acid substitution at position 918.
However, three other accessions with the same substitution, Asamurasaki, KSWSK, and
O. rufipogon, had blackish purple leaf sheath. Thus, mutations in genes other than OsC1 were suggested to also eliminate anthocyanin pigmentation, a hypothesis that was
supported by seven japonica accessions having the same OsC1 coding sequence as PSH TNG72 (Fig. 4). OsC1, containing an R2R3 Myb domain, is thought to function as a transcription factor,
regulating other genes involved in anthocyanin synthesis and enhancing another transcription
factor, bHLH, regulating anthocyanin structural gene DFR (Dooner et al. 1991]; Ithal and Reddy 2004]). Failure of anthocyanin synthesis could result from malfunction of OsC1 or any downstream proteins: for example, PSH1 (t) conferring purple leaf sheath was mapped on chromosome 1 (Wang et al. 2009]). One of two activator genes, OsB1 and OsB2 encoding basic helix-loop-helix (bHLH) transcription factors, incorporated with maize
C1 could induce anthocyanin synthesis in the aleurone layer; however, the lack of function
of these two genes in T65 resulted in green leaf blade and leaf sheath (Sakamoto et
al. 2001]). One japonica GSH accession, TNG67 which is a descendant of T65, had significantly reduced expression
of OsC1 despite having a full length amino acid encoding sequence (Fig. 3), suggesting feedback regulation of gene expression.

Fourteen PSH accessions exhibited variation in intensity of purple coloring of the
leaf sheath, classified as tyran rose, pansy purple, red purple, and blackish purple.
The anthocyanin contents of these 14 PSH accessions differed significantly (Fig. 3). However, OsC1 gene expression among three accessions with diverse anthocyanin contents, TNG72,
SCTT, and KSWSK, showed only slight and non-significant differences. These three accessions
represented three different haplotypes, H3, H7, and H10, respectively (Fig. 5). Allelic variation of OsC1 accounted for anthocyanin accumulation and pigmentation in the leaf sheath, for which
nonsynonymous mutations at the C-terminal domain were found herein. The C terminus
of R2R3 Myb transcription factors is an activation and repression domain (Dubos et
al. 2010]). Allelic variation of OsC1 and Purple leaf (Pl) coding regions also caused diverse intensities of purple apiculus pigmentation (Sakamoto
et al. 2001]; Saitoh et al. 2004]). Allelic variation resulting in diversified phenotypes is very common, as exemplified
by the great impact of different alleles of Wx and starch synthesis related genes on rice grain appearance, cooking and eating quality
(Tian et al. 2009]; Zhang et al. 2012]; Wu et al. 2015]). However, other genes participating in anthocyanin synthesis and environmental effects
cannot be neglected (Sakamoto et al. 2001]). Anthocyanin accumulation is frequently responsive to abiotic stresses, such as
UVB radiation, temperature, and soil acidity (Reddy et al. 2004]). A cis-element related to light regulation (?10PEHVPSBD, Thum et al. 2001]) was found in the promoter region of OsC1 in TNG72 by using the PLACE database (https://sogo.dna.affrc.go.jp/, Higo et al 1999]). We noted that the leaf sheath accumulated more anthocyanin when rice plants were
grown in soil under natural light than in hypotonic solution in the growth chamber.

The Selection and Genealogy of OsC1

Traits considered part of domestication syndromes are favored by artificial selection
during domestication, and consequently distinguish cultivated plants from their wild
progenitors. During domestication, genetic diversity of whole genomes can be dramatically
reduced because of the ‘genetic bottleneck’ effect of selecting a few individuals
as a founder population. The genetic diversities of genes contributing to domestication
syndromes and improvement traits are often accompanied by artificial selection for
production and culture, and the geographic distribution of alleles might be altered
by human migration (Kovach et al., 2007]; Kovach and McCouch 2008]; Olsen and Wendel 2013]). In rice, OsC1 conferring purple leaf sheath, apiculus, and stigma has been suggested as a domestication
gene (Choudhury et al. 2014]). Wild types are generally purple while cultivars tend to be green, although exceptions
exist as noted above. The nucleotide diversity of OsC1 was higher in O. rufipogon than in O. sativa, and OsC1 of the haplotype of Asian cultivated accessions showed evidence of selection (Saitoh
et al. 2004]). In addition to perennial O. rufipogon and annual O. rufipogon (also called O. nivara), OsC1 sequences of 4 other wild species, O. glaberrima, O. barthii, O. glumaepatula, and O. meridionalis which had not been previously reported were subjected to allelic variation analysis.
Sequence variation existed both in noncoding and coding regions. While these species
had PSH, a few nonsynonymous amino acid substitutions were detected (Fig. 4). O. meridionalis (from Australia) had extremely divergent OsC1 DNA sequence, forming a distinct haplotype by itself, and all parameters regarding
nucleotide divergence of OsC1 decreased when it was not included in analysis (Table 2). African cultivated species, O. glaberrima, its progenitor O. barthii, and South American species, O. glumaepatula, also had their own specific sequences and formed 2 different haplotypes (Figs. 4 and 5). On the other hand, OsC1 in Asian rice shared sequence similarity and formed 13 interconnected haplotypes.
Allelic variation of OsC1 in Asian, African, South American, and Australian rice reflected independent mutation
without gene flow because of their geographic distribution. Two accessions of O. nivara and one accession of O. rufipogon together with 24 accessions of O. sativa were in Group B. The genealogy of OsC1 in the collected Asian rice accessions herein and other studies showed no distinct
correlation with geographic distribution. The 10-bp deletion conferring GSH was prevalent
in indica landraces and improved cultivars and rare in japonica accessions from many countries (Saitoh et al. 2004]; Choudhury et al. 2014]). The accessions in haplotypes H3 and H10 were collected from several countries (Fig. 5). Gene flow due to human activities might be an important factor in the geographic
distribution of OsC1 alleles.

Purple leaf sheath, stigma, and apiculus are widespread in wild forms and often found
in landraces, while green leaf sheath as well as colorless stigma and apiculus are
common in modern cultivars. OsC1 nucleotide divergence was up to 90 % lower in GSH than PSH although twice as many
GSH accessions were studied (Table 2). However, no values from three neutral tests, Tajima’s D, Fu and Li’s D F, were
significantly differently from neutral expectations, indicating that OsC1 had not been subjected to selection–a conclusion that is supported by findings regarding
OsC1 in indigenous rice varieties in Northeast India (Choudhury et al. 2014]). Genetic diversity was higher in indica than japonica accessions both in 14 PSH and 20 GSH accessions, revealed by parameters reflecting
nucleotide segregation at total polymorphic sites and silent sites, ?
_T
, ?
_T
, ?
_sil
, and ?
_sil
(Table 2, Fig. 4). This phenomenon is congruent with evidence that genetic diversity is larger in
subspecies indica than japonica (Garris et al. 2005]).

Selection might not be a driving force for reducing genetic diversity in GSH. There
is no significant evidence that OsC1 deviated from neutral expectations in indica PSH or japonica PSH and GSH accessions. Nevertheless, indica GSH accessions might be experiencing relaxed purifying selection, indicated by neutrality
tests at significance levels of 0.10??P??0.05 (Table 3). Unlike other domestication syndromes directly related to productivity and other
desirable traits that were selected for particular purposes over several thousand
years, GSH might have been selected unintentionally.

The Asian cultivated species, O. sativa, evolved from Asian wild rice progenitors O. nivara (annual) and O. rufipogon (perennial). O. sativa was domesticated from divergent wild populations about 10,000 years ago and diversified
into two major subspecies, indica and japonica, subsequently being subjected to a long period of natural and artificial diversifying
selection (Gross and Zhao 2014]). Indica and japonica subspecies are distinguishable in morphology and physiology, already recognized as
Hsien (long grain) and Keng (short grain) in the Han dynasty, China, over 2,000 years
ago (Oka 1988]; Callaway 2014]). Numerous genes related to differentiation between these two subspecies experienced
mutation and diversifying selection, e.g., Phr1 responsible for phenol reaction; and GS3, qSW5 and GS5 responsible for grain shape (Yu et al. 2008]; Lu et al. 2013]). The 10-bp deletion conferring GSH was prevalent in indica landraces and improved cultivars and rare in japonica, but not in its progenitor species. The OsC1 allele with 10-bp deletion was suggested to have originated and been an early target
of domestication in subspecies indica (Figs. 4 and 5; Saitoh et al. 2004]; Choudhury et al. 2014]). In addition, 3-bp and 2-bp deletions in exon 2 and exon 3 of the R3 Myb domain
were found in japonica rice from Japan and China, respectively. These two alleles were independent from
the gene lineage of indica, which suggested mutation after subspecies divergence (Figs. 2 and 4; Saitoh et al. 2004]).

Although reproductive barriers such as hybrid sterility and hybrid breakdown impede
gene flow between cultivated rice and its wild progenitors, numerous interspecific
crosses and successful introgressive hybridizations have been performed to unravel
useful alleles and genes of wild species. Gene flow confounded with selection has
been revealed at the genome level (Zhao et al. 2010]; He et al. 2011]; Yang et al. 2011]) and in domestication-related genes including Wx, GS3, SD1, and qSH1 (Yamanaka et al. 2004]; Konishi et al. 2006]; Takano-Kai et al. 2009]; Asano et al. 2011]).

The genealogy of OsC1 suggests some gene flow events. O. rufipogon (Taiwan type 1) had an OsC1 sequence identical to that of Taiwan landrace KSWSK. One SNP aligned to japonica Asamursaki and Taichung 65 and indica Midon, accessions that were clustered in Group III and haplotype B3 (Fig. 4 and 5). Two O. nivara accessions also shared similar sequences to most japonica accessions and were classified as haplotype B1. Thus, OsC1 in three Asian wild accessions closely resembled that of japonica but not indica, a finding which might support the hypothesis that japonica and indica were domesticated independently from O. rufipogon (Yang et al. 2011]; Wei et al. 2014]). Gene flow between subspecies was not rare, as revealed by genealogy of OsC1 in 23 indica and 20 japonica accessions. In haplotype H1, three japonica landraces from Taiwan also possessed the 10-bp deletion specific to indica; in haplotype H10, one japonica landrace (Shang Chi Tsao Tao) and three indica accessions (Taiwan landrace Jinya-149, India landrace G124 and improved line IR1535)
had the same allele (Figs. 4 and 5).

The genealogy of OsC1 might not be in agreement with rice phylogeography because of human behavior. In
Taiwan, O. rufipogon and O. nivara were once found in several swamp sites (Chang 1976]) but, unfortunately, all habitats were destroyed several decades ago. Archaeological
evidence shows that tropical japonica or indica had been cultivated over 5,000 years by ancient indigenous peoples in Taiwan (Hsieh
et al. 2011]). In the early 17th century, numerous Chinese migrated and carried many landraces
(mostly indica) from coastal regions of Fujian and Guangdong Provinces of China to Taiwan. By the
early 20th century, 1,197 indica accessions were identified officially in Taiwan, and 1,256 japonica accessions were introduced from Japan (Iso 1964]). More than 1,000 of these accessions were deposited in The T.T. Chang Genetic Resources
Center at the International Rice Research Institute (IRRI), the Philippines. Taiwanese
rice germplasm was thus an admixture of indigenous wild species, landraces, and introduced
germplasm from China and Japan; and was subsequently spread over Southeast Asia via
the germplasm deposited in IRRI. As a result, introgression of OsC1 may have occurred by hybridization between subspecies and both artificial and natural
selection, clouding the true genealogy of OsC1.