Bc19 Alters Mechanical Strength and Cell Wall Composition but Has No Distinct Influence on Growth and Grain Yield
Researches on brittle mutants have showed several genes underlying cellulose synthesis and cell wall formation, while mutation of these genes not only affected mechanical strength of plant bodies, but also resulted in other pleiotropic abnormalities in rice. The bc3 mutant showed easily snapped culms, dwarf plant bodies, and short roots (Hirano et al. 2010). Besides brittle culms and leaves, bc10 also exhibited yellowish seedlings, fewer and shorter roots, decreased plant height, and fewer tillers (Zhou et al. 2009). The other brittle mutant, bc12, also showed abnormal growth and morphogenesis, including dwarfism and reduced root length (Zhang et al. 2010). In addition, alterations in cell wall components of these mutants were varied. Briefly, cellulose contents in bc3 and bc10 were decreased but not affected in bc12; lignin content was increased in both bc10 and bc12 (Hirano et al. 2010; Zhou et al. 2009; Zhang et al. 2010).
In case of CesA mutants in rice, different mutation types and sites also resulted in varied phenotypes. NE1031, NC0259, ND8759, and ND2395 were four brittle mutants generated by Tos17 insertion in OsCesA genes (NE1031 relative to OsCesA4, NC0259 and ND8759 relative to OsCesA7, and ND2395 relative to OsCesA9), and growth and development of them were all aberrant, mainly in plant height, leaf size, culm thickness, and fertility. Moreover, cellulose contents of the four mutants were sharply decreased to 8.9% − 25.5% of that of the wild type plant (Tanaka et al. 2003). bc7 was another mutant allele of OsCesA4 obtained through 60Co-γ radiation, resulting in the premature transcription due to the 7-bases deletion in the junction of exon 10 and intron 10 (Yan et al. 2007). Besides weak mechanical strength, plant height of bc7 was slightly shorter than wild type, and cellulose content and cell number of the parenchymatous tissues were all reduced. Other 7 mutants of OsCesA genes were point mutations as reported, including 2 of OsCesA4 (bc11 and fc17), 1 of OsCesA7 (S1-24), and 4 of OsCesA9 (Bc6, S1-60, bc13, and bc-s1) (Zhang et al. 2009; Li et al. 2018; Wang et al. 2016a; Kotake et al. 2011; Wang et al. 2012; Song et al. 2013; Jin et al. 2016). In addition to prominent brittle phenotype, 3 mutants (bc11, S1-24, and S1-60) among them also showed shorter plant bodies, and/or other abnormal morphologies, such as poor fertility, fewer tillers and shorter roots. In most of those mutants, there were reductions in cellulose contents and compensatory increases in hemicellulose and lignin contents, which was different from the brittle mutant in this study (Yan et al. 2007; Wang et al. 2016a; Kotake et al. 2011; Wang et al. 2012; Li et al. 2018). Here, we reported another brittle mutant Bc19, in which cellulose, hemicellulose and lignin contents were reduced by 12.5% − 28.5%, but apparent morphology, growth and grain yield were not affected (Figs. 1, 3). Additionally, different from most recessive brittle mutants such as fc17, we verified Bc19 as a new semi-dominant mutant allele of gene OsCesA4, and the mutant gene influenced mainly the SCW synthesis (Figs. 1, 2, 3, 4).
In recent years, constantly raising grain yield attracts breeders' increasing concern of lodging resistance in rice cultivars. Enhancing culm strength can increase the lodging resistance in rice and confrontation capacity against natural disasters such as strong wind and rain (Li et al. 2014a). On the other hand, reduced proportion of components in cell walls could make rice straw easily degraded after harvest, improving its utilization efficiency either as animal feed or biofuels (Wang et al. 2005; Johnson et al. 2006). In addition, engineering improvement of rice straw as biofuel resources will be a sustainable solution to solve environmental problems owing to straw burning (Himmel et al. 2007; Ragauskas et al. 2014). In this study, cellulose content was reduced while growth and yield were not affected in Bc19 mutant, thus the mutant gene could be a potential genetic resource for rice straw recycling, which could maintain grain yield as well (Figs. 1, 3). What's more, the specific semi-dominant characteristic of brittle gene Bc19 makes it more convenient to implement of high grain yield, heavy biomass and their efficient utilization in breeding of grain-straw dual-purpose hybrid rice (Peng et al. 2010; Li et al. 2019).
The Point Mutation in Bc19 Could Partially Abolish OsCesA4 Function
OsCesA4, together with OsCesA7 and OsCesA9, encoding CesAs, are indispensable and irredundant during biosynthesis of cell wall in rice (Tanaka et al. 2003). The three OsCesA subunits share strong similarity in amino acid sequences, and they possess similar function domains including a Zinc Finger domain, a D,D,D,QXXRW motif, a Plant-conserved Region (P-CR), a Class-specific Region (C-SR) and totally 8 transmembrane domains (TMDs) (Somerville 2006; Li et al. 2014b) (Fig. 8). Recently it was visually identified that the PttCesA8 subunit in poplar (Populus tremula × tremuloides) possessed 7 TMDs, but none of such discoveries was found in rice or other species (Purushotham et al. 2020). Nevertheless, the functional cellulose synthase complex (CSC) has been identified as a hexamer, the so-called rosette structure which is anchored to the plasma membrane (Atanassov et al. 2009). Through interactions induced by the Zinc Finger domain with each other, single CesA subunit first forms homodimers and then six homodimers are polymerized into the rosette structure (Kurek et al. 2002; Timmers et al. 2009; Hill et al. 2014; Nixon et al. 2016). Although formation of the rosette structure relies on what motif of CesA subunits remains unclear, it is predicted that the D,D,D,QXXRW motif as well as the P-CR in the second cytoplasmic region between TMD2 and TMD3 are essential for the catalytic activity during cellulose synthesis (Taylor et al. 2000; Doblin et al. 2002; Arioli et al. 1998; Rushton et al. 2017). In another word, a functional CSC must possess three elements, correct Zinc Finger motif to form the rosette, correct second cytoplasmic region for the catalytic activity, and correct TMDs for locating to the plasma membrane.
NE1031, bc11 and Bc19 are allelic mutants of OsCesA4 with different mutation sites, while they exhibit different morphological abnormalities. NE1031 was found the Tos17 insertion in the sixth exon, which resulted in the premature transcription of OsCesA4 (Tanaka et al. 2003). The point mutation in bc11 was located in the end of TMD5, resulting in decreased abundance of OsCESA4 in the plasma membrane (Zhang et al. 2009). Despite weak mechanical strength, these two mutants also displayed dwarfed plant bodies; and NE1031 even exhibited small leaves, thinner culms and low fertility. Moreover, cellulose contents were all sharply reduced in NE1031 and bc11, by 79.6% and 57%, respectively. In this study, the P507S mutation within the second cytoplasmic region of OsCesA4 only influenced the mechanical strength of Bc19, accompanied by 28.5% reduction in cellulose content and thinner SCW (Figs. 1, 2, 3). Meanwhile, transcript levels of Bc19 were not apparently affected in the mutant (Figs. 6d, 7b). Considering normal Zinc Finger domain and TMDs of OsCesA4, the rosette probably could still be secreted and trafficked to the plasma membrane in Bc19. Evidence was that in the irx1 (AtCesA8) mutant in Arabidopsis, antibodies specifically combining to IRX3 (AtCesA7) and IRX5 (AtCesA7) could still coprecipitate IRX1, suggesting that the point mutation in the third Asp residue of the D,D,D,QXXRW motif in mutant irx1 didn't affect interactions of the three kinds of CesA subunits (Taylor et al. 2000; Taylor et al. 2003). Therefore, different from complete loss of function of OsCesA4 in NE1031 due to its premature transcription termination, catalytic efficiency of the mutated OsCesA4 in Bc19 might just be reduced to some extent, instead of thoroughly abolished. After all, cellulose content in Bc19 was not sharply reduced as it was in NE1031 (Fig. 3) (Tanaka et al. 2003).
Possible Explanation for the Dominant Negative Effect of the Point Mutation in Bc19
Although more than 20 brittle mutants have been reported in rice, almost all of them were recessive, and the only exception was Bc6, a semi-dominant mutant allele of OsCesA9 in rice (Kotake et al. 2011). In this study, we characterized another semi-dominant brittle mutant Bc19, which was allelic to OsCesA4. Till now, 8 single-base substitution mutant alleles of CesA proteins have been located on different functional domains of them (Fig. 8). S1-24, mutated on the Zinc Finger domain of protein OsCesA7, was presumed to affect interactions of OsCesA7 with other CesA subunits, and the mutant phenotype was recessive (Wang et al. 2016a). Similar events occurred in mutant bc13 and bc-s1, of which the missense mutations were mapped closely before and behind the Zinc Finger domain, respectively. Recessive mutants bc11 and S1-60, possessing point mutations at the beginning and the end of TMD5, respectively, were speculated to impede enrichment of the corresponding CesA subunits (OsCesA4 and OsCesA9) on the plasma membrane (Fig. 8) (Zhang et al. 2009; Wang et al. 2012). Furthermore, over-expression of the mutated cDNA of bc11 in the bc11 mutant background unexpectedly rescued the brittle phenotype, suggesting that excessive amount of the mutant version of OsCesA4 could make up the abundance of OsCesA4 in the plasma membrane, which implied that this point mutation did not change its normal catalytic activity, and might only affect secretion of the complex from the endomembrane to the plasma membrane (Zhang et al. 2009). As for the two semi-dominant mutants, Bc6 reported before and Bc19 in this study, the former one carrying a R588G mutation in OsCesA9 and the latter one with a P507S mutation in OsCesA4, both of them mutated in the second cytoplasmic domain between TMD2 and TMD3 (Kotake et al. 2011). Multiple alignment of CesA subunits from rice and other species showed that the interval between the P507 of OsCesA4 and the R588 of OsCesA9 was only one amino acid (Fig. 5). It's worth noting that the P507 residue is highly conserved among CesA family members, and it exactly corresponds to P557 of AtCesA7, which was changed to a threonine residue in the semi-dominant mutant fra5 in Arabidopsis (Figs. 5, 8) (Zhong et al. 2003).
Controlling mechanism of whether a particular mutated CesA subunit acts in a dominant or recessive way might be associated with different mutation sites in different function domains of them. In the study of fra5, it was hypothesized that the second cytoplasmic domain might participate in interactions of CesAs with other components in the rosette complex, and the missense mutation in fra5 perturbed this kind of interaction and then affected the normal function of CSC, which resulted in the dominant negative effect (Zhong et al. 2003). However, Kotake et al. (2011) speculated that the P-CR of the second cytoplasmic domain might be involved in interactions between CesA subunits during the formation of higher order CesA oligomerization, so that the mutation in Bc6 might interfere the formation of functional CSC. In this study, overexpression of Bc19 in the wild type caused more severe brittle phenotype than the Bc19 mutant, while the mutant phenotype of F1 plants was milder than Bc19 (Figs. 6b-d, 3, 1a, b, j, k). It seems that the severity of the mutant phenotype is tightly related to the level of Bc19 expression, implying that despite the dominant-negative effect, Bc19 might also act in a dosage way. Considering these characteristics of mutant gene Bc19, we raise the possibility here that in the hybrid F1 plants, the encoded product of the mutant gene Bc19 can randomly bind to the CSC and can be secreted to the plasma membrane together with the normal CSCs, but the abnormal CSCs cannot catalyze cellulose synthesis as efficiently as the normal CSC, resulting in milder cellulose deficiency compared to the homozygous mutant. Similarly, in the Bc19 overexpression plants, even more abnormal CSCs carrying the mutated OsCesA4 subunit gather onto the plasma membrane, thus severely impacting efficiency of cellulose synthesis, which fits the dominant-negative effect of a dose-dependent fashion. In cases of hybrid plants of point mutations in Zinc Finger domain or TMDs, abnormal subunits encoded by mutated genes either cannot form the CSC or cannot be secreted to the plasma membrane, thus CSCs transferred onto the plasma membrane are all functional CSCs compromising of normal subunits encoded by the correct genes, which can efficiently catalyze cellulose synthesis, and that’s why these mutated genes act in recessive ways. Intensive exploration on action mechanism of P-CR of CesA subunits and on activities between components within the CSC complex will help to uncover how these P-CRs participate in cellulose biosynthesis.