Fine Mapping of the Affecting Tillering and Plant Height Gene CHA-1 in Rice

The plant architecture of rice is an important factor affecting yield. Strigolactones (SLs) are newly discovered carotenoid-derived plant hormones that play an important role in rice plant architecture. In this study, a high-tillering dwarf mutant, CHA-1, was identified by spatial mutagenesis. CHA-1 was located in the region of 31.52–31.55 MB on chromosome 1 by map-based cloning. Compared with the wild-type THZ, the CHA-1 mutant showed that ACCAC replaced TGGT in the coding region of the candidate gene LOC_Os01g54810, leading to premature termination of expression. Genetic complementation experiments proved that LOC_Os01g54810 was CHA-1, which encodes a putative member of Class III lipase. Expression analysis showed that CHA-1 was constitutively expressed in various organs of rice. Compared with those in THZ, the expression levels of the D17 and D10 genes were significantly downregulated in the CHA-1 mutant. In addition, the concentrations of ent-2′-epi-5-deoxystrigol (epi-5DS) in the root exudates of the CHA-1 mutant was significantly reduced compared with that of THZ, and exogenous application of GR24 inhibited the tillering of the CHA-1 mutant. These results suggest that CHA-1 influences rice architecture by affecting SL biosynthesis.


Introduction
Rice (Oryza sativa L.) is one of the most important food crops in the world and plays an important role in global food production and consumption [1]. More than half of the world's population depends on rice as a staple food [2]. Improving rice yield is still one of the main goals of rice breeding. Rice plant type is mainly determined by plant height, tillering and panicle structure, which directly affect the lodging resistance and yield potential of rice varieties [3].
The tillering development of rice is regulated by the initiation and outgrowth of axillary buds [4]. MOC1 is the first cloned key gene that positively regulates rice tillering and encodes plant-specific GRAS family proteins, which activate axillary buds and promote tiller bud outgrowth [5]. MOC3 encodes a protein with the highest homology to Arabidopsis WUS, which is a key factor in tiller bud initiation and regulates tiller bud outgrowth. FON1 positively regulates the outgrowth of tillering buds [6,7]. MOC3 binds to the promoter of FON1 and activates its expression. As a coactivator of MOC3, MOC1 enhances FON1 expression in the presence of MOC3 [8].
Plant hormones such as gibberellins (GAs), cytokinins (CKs), brassinosteroids (BRs), auxins, and strigolactones (SLs) are involved in the regulation of plant structure in rice. Among them, SLs are plant hormones, which have been discovered in recent years, that are produced by plant roots and transported upward to axillary buds, playing an important role in inhibiting the outgrowth of axillary buds [9,10].
The mutation in SL biosynthesis or signal led to plant dwarfing and increased tiller number [11]. D27, D17/HTD1 and D10 are key genes in the SL biosynthesis pathway   I, II, III, IV, and V represent  internode I, internode II, internode III, internode IV, and  The plant height of the CHA-1 mutant was only approximately half that of THZ at the mature stage ( Figure 1F). By comparing and analyzing the length of the upper five internodes of CHA-1 and THZ, it was found that the length of each internode of CHA-1 was shorter than that of THZ ( Figure 1C,E). Compared with THZ, the shortening proportion of CHA-1 increased gradually from I the internode to the V internode ( Figure 1D).
In addition, the seed setting rate of CHA-1 reached only 61.95%. KI-I2 staining was used to assess the pollen viability, and the results showed that the pollen viability of CHA-1 was only 62.9%, which was much lower than the 95.1% of THZ. We also found that the anthers of CHA-1 were smaller and thinner than those of THZ (Supplementary Figure  S1E-G). Therefore, the poor seed setting rate of mutant CHA-1 may be caused by low pollen fertility.

Microscopic Observation of the CHA-1 Mutant
To identify the factors responsible for the shortened internodes in CHA-1, longitudinal sections of the upper five internodes of CHA-1 and THZ were conducted to observe and compare their internode cell length under an optical microscope. Compared with THZ, the average longitudinal length of cells in internode I, internode II, and internode IV of CHA-1 was greater (Figure 2A,B), which indicated that the difference in internode length between THZ and mutant CHA-1 was not due to cell size. These results suggest that the shortening of internodes was due to the reduction in the number of longitudinal cells.

Microscopic Observation of the CHA-1 Mutant
To identify the factors responsible for the shortened internodes in CHA-1, longitudinal sections of the upper five internodes of CHA-1 and THZ were conducted to observe and compare their internode cell length under an optical microscope. Compared with THZ, the average longitudinal length of cells in internode I, internode II, and internode IV of CHA-1 was greater (Figure 2A,B), which indicated that the difference in internode length between THZ and mutant CHA-1 was not due to cell size. These results suggest that the shortening of internodes was due to the reduction in the number of longitudinal cells.  Values are means ± SD. I, II, III, IV, and V represent internode I, internode II, internode III, internode IV, and internode V, respectively. *, p < 0.05; **, p < 0.01; ns, no significance (Student's t-test).

Fine-Mapping of CHA-1
The F 2 population derived from the cross between the CHA-1 mutant and japonica rice cultivar 02428 was used for fine mapping of the CHA-1 gene. The plant height in the F 2 population showed obvious high-dwarf separation and a continuous bimodal distribution, which indicated that there was a major gene in the CHA-1 mutant leading to dwarfing. The separation ratio of high-dwarf plants was 3.04:1, which showed that CHA-1 was a recessive gene (Table 1). According to the previously identified marker RM302 on the long arm of chromosome 1 [27], which was linked to CHA-1, we carried out further fine mapping of CHA-1 using dwarf plants with a homozygous genotype in the F 2 population by SSR markers and InDel markers. As a result, CHA-1 was flanked by markers DL1 and DL2, spanning the region from 31.51 Mb to 31.59 Mb of chromosome 1 ( Figure 3A). To further narrow the interval of CHA-1, 2393 recessive homozygous genotype plants were developed by expanding the F 2 population to perform fine mapping. Finally, CHA-1 was located in a region of 29.52 kb flanked by markers DL5 and DL8 ( Figure 3A), which covered three candidate genes ( Table 2).

Analysis of Candidate Genes of CHA-1
Sequencing analysis of three candidate genes within a 29.52 kb region on chromosome 1 showed that ACCAC replaced TGGT in 2382-2386 bp after the initial codon of LOC_Os01g54810, and the mutation site was Chr1.31524649-31524653 (-) ( Figure  3B). This mutation causes code shift mutation, termination codons appear in advance, and translation terminates in advance. There was no sequence difference among other candidate genes between the CHA-1 mutant and THZ. Therefore, LOC_Os01g54810 was the likely candidate gene of CHA-1. To confirm that LOC_Os01g54810 is CHA-1, we introduced a DNA fragment containing the entire genome sequence of LOC_Os01g54810 including its upstream 2 kb sequence into the CHA-1 background. The height, tillering,

LOC_Os01g54810
Class III lipase LOC_Os01g54850 Cyclin-like F-box domain containing protein LOC_Os01g54860 Enoyl-CoA hydratase/isomerase 2.4. Analysis of Candidate Genes of CHA-1 Sequencing analysis of three candidate genes within a 29.52 kb region on chromosome 1 showed that ACCAC replaced TGGT in 2382-2386 bp after the initial codon of LOC_Os01g54810, and the mutation site was Chr1.31524649-31524653 (-) ( Figure 3B). This mutation causes code shift mutation, termination codons appear in advance, and translation terminates in advance. There was no sequence difference among other candidate genes between the CHA-1 mutant and THZ. Therefore, LOC_Os01g54810 was the likely candidate gene of CHA-1. To confirm that LOC_Os01g54810 is CHA-1, we introduced a DNA fragment containing the entire genome sequence of LOC_Os01g54810 including its upstream 2 kb sequence into the CHA-1 background. The height, tillering, and pollen fertility of the T 1 transgenic plants were fully restored to those of the wild type ( Figure 3C,D). These results fully demonstrate that the mutation of LOC_Os01g54810 was responsible for the phenotypic variation of mutant CHA-1.

Expression and Evolutionary Analysis of CHA-1
LOC_Os01g54810 encodes a putative member of Class III lipase, which hydrolyzes the ester bond of triglycerides. Using the RiceXPro database [29] (http://ricexpro.dna.affrc. go.jp/, accessed on 20 February, 2022) to analyze its expression profile, it was revealed that LOC_Os01g54810 was expressed in multiple tissues of rice ( Figure 4A). qRT-PCR was performed in several rice tissues, and the results were consistent with those in the database. LOC_Os01g54810 was expressed in multiple tissues including the leaf lamina, leaf sheath, root, flower, culm, and seed ( Figure 4B).
Phylogenetic tree analysis and amino acid sequence alignment revealed that CHA-1 homologs were conserved in several monocots and dicots, respectively (Supplementary Figure S3). Furthermore, the CHA-1 homology showed obvious separation between the dicot and monocot species groups in the phylogenetic tree analysis (Supplementary Figure  S2), indicating that it underwent a different evolution after the differentiation of the dicot and monocot species groups.
Six CHA-1 homologs were identified in rice. Sequence alignment showed that the Nterminal amino acid sequence of CHA-1 was quite different from that of other homologs and that the amino acid sequence of CHA-1 was longer (Supplementary Figure S5). This implied that CHA-1 may have a unique biological function. LOC_Os01g14080 and LOC_Os01g47610 were similar to CHA-1 in gene structure, but the CDD-Blast results showed that only CHA-1 had the III lipase domain (Supplementary Figure S4).
In order to detect whether CHA-1 has lipase activity, the lipase activity of the fusion protein GST-CHA-1 was determined using p-nitrophenyl butyrate as a substrate. Compared with the control and GST alone, GST-CHA-1 hydrolyzed the lipid substrate effectively ( Figure 4C), indicating that CHA-1 has lipase activity and is a functional lipase.

CHA-1 Affects the SL Biosynthesis
In rice, mutation of the biosynthetic and signaling pathways of SLs leads to dwarfing and high-tillering phenotypes [30]. The phenotype of the CHA-1 mutant was similar to the phenotype of mutants related to the biosynthetic and signaling pathways of SLs. Therefore, we speculated that CHA-1 may influence the rice architecture by affecting the SL pathway. To determine whether CHA-1 affected the SL pathway, we measured the epi-5DS content in the root exudates of THZ and CHA-1 by HPLC-MS/MS. Compared with THZ, CHA-1 had significantly lower amounts of epi-5DS ( Figure 5C). The results of qRT-PCR analysis showed that the expression levels of the D17, D10, D14, and D3 in the CHA-1 mutant were significantly downregulated ( Figure 5B). After 1 µmol synthetic strigolactone (GR24) treatment, the CHA-1 and THZ tillers were inhibited; that is, the CHA-1 phenotype with multiple tillers was rescued ( Figure 5A). These results demonstrate that the mutation of CHA-1 led to the downregulation of SL biosynthesis genes D17 and D10, which resulted in the decrease in endogenous SL content.
( Figure 3C,D). These results fully demonstrate that the mutation of LOC_Os01g54810 was responsible for the phenotypic variation of mutant CHA-1.

Expression and Evolutionary Analysis of CHA-1
LOC_Os01g54810 encodes a putative member of Class III lipase, which hydrolyzes the ester bond of triglycerides. Using the RiceXPro database [29] (http://ricexpro.dna.affrc.go.jp/, accessed on 20 February, 2022) to analyze its expression profile, it was revealed that LOC_Os01g54810 was expressed in multiple tissues of rice ( Figure 4A). qRT-PCR was performed in several rice tissues, and the results were consistent with those in the database. LOC_Os01g54810 was expressed in multiple tissues including the leaf lamina, leaf sheath, root, flower, culm, and seed ( Figure 4B).

Plant Materials and Phenotypic Characterization
An F2 population was constructed using 02428 (Oryza sativa L. ssp. japonica) an CHA-1 mutant (Oryza sativa L. ssp. indica) as the male and female parents, respect All rice populations were planted in the experimental field of the South C Agricultural University Campus Teaching & Research Base (Guangzhou, 23. 113.36°E), where there was no disease stress. There were six rows per plot, six plan row, and a planting interval of 20 × 20 cm. Mature pollen grains were stained with 1 KI solution and photographed with an Olympus CX31 light microscope (Oly Corporation, Tokyo, Japan). Mature anthers were photographed with an Optec D Digital Camera system (Optec, Chongqing, China).

Histological Analysis
At the mature stage of rice, five internode tissues (approximately 2-3 mm) fro stems of the main panicle were fixed with FAA (5% formaldehyde, 5% acetic acid Plants were treated without GR24 as mock-treated control. (B) Relative expression levels of SL pathway genes between THZ and CHA-1. D53, D14, and D3 were SL signaling pathway genes. D27, D17, and D10 were SL biosynthesis pathway genes. (C) Comparison of epi-5DS contents in the THZ and CHA-1 root exudates. Values are the means ± SD. *, p < 0.05; **, p < 0.01; ns, no significance (Student's t-test).

Plant Materials and Phenotypic Characterization
An

Histological Analysis
At the mature stage of rice, five internode tissues (approximately 2-3 mm) from the stems of the main panicle were fixed with FAA (5% formaldehyde, 5% acetic acid, and 90% ethanol), dehydrated with a gradient alcohol, and then embedded into paraffin. The sections were stained with toluidine blue and observed under an Olympus IX70 light microscope (Olympus, Tokyo, Japan).

Fine Mapping of CHA-1
DNA samples of the F 2 population were extracted from leaves by the cetyltrimethylammonium bromide (CTAB) method [31]. SSR marker primer sequences were obtained by using the Grame database (https://www.gramene.org, accessed on 20 February 2022) and developing new Indel marker sequences for the fine-scale localization of CHA-1 (Table 3). Linkage analysis was performed on 3205 F 2 plants with recessive homozygous genotypes using polymorphic markers. Table 3. Primers used for the fine mapping of CHA-1 and quantitative real-time PCR.

Phylogenetic Tree Construction and Analysis
CHA-1 and homologous sequences were downloaded from the NCBI database. Phylogenetic trees were constructed using MEGA7.0 [32]. Sequence relationships were inferred using the maximum likelihood (ML) method. The conserved motifs were predicted using the online multiple expectation maximization for motif elicitation (MEME) program [33]. Images were generated using TBtools software [34].

Lipase Activity Assay
The CDS sequence of CHA-1 was amplified by Phanta Max Super-Fidelity DNA Polymerase (Vazyme, Nanjing, Jiangsu, China) and cloned into the pGEX-4T-1 vector. The recombinant vector was transformed into Escherichia coli strain BL21 (DE3) and the expression of the recombinant protein was induced by isopropylthio-β-galactoside. Purification of CHA-1 and the lipase assay were performed as described previously [35]. The recombinant protein was incubated with 1 mM p-nitrophenyl butyrate substrate in enzyme reaction buffer (0.5 M HEPES, pH 6.5) at 30 • C for 1 h. The absorbance was measured every 10 min at 405 nm for 120 min.

RNA Extraction and Quantitative Real-Time PCR (qPCR)
Total RNA was extracted from various organs using a TRIzol Kit (Invitrogen) following the manufacturer's method. cDNA was reverse transcribed using 5× HiScript III qRT Super-Mix (Vazyme). qPCR experiments were performed using AceQ qPCR SYBR Green Master Mix (Vazyme) and the ABI Step One Plus Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). The relative expression was estimated by the 2 −∆∆CT method [36].

Vector Construction and Transformation
The full-length DNA and upstream promoter (approximately 2 kb from the start codon) of CHA-1 were cloned into the pCAMBIA1300 vector, which was introduced into Agrobacterium tumefaciens strain EHA105 and then transformed into rice callus, as previously reported [37].

GR24 Treatment
SL analysis in the rice root exudates was quantified by Webiolotech (Nanjing, China) as previously reported [16] with minor modifications. One-week-old seedlings were grown on hydroponic culture medium without phosphate for 14 days. The hydroponic culture medium was collected, and the fresh root weight was recorded. A total of 50 mL hydroponic culture medium was loaded into a pre-balanced Oasis HLB 3cc cartridge (Waters, Milford, MA, USA) and the column was washed with deionized water. Fractions containing SLs were eluted with acetone, collected and dried under nitrogen, then reconstituted in acetonitrile and subjected to HPLC-tandem mass spectrometry (HPLC-MS/MS) analysis. SLs analysis was performed on a quadruple linear ion trap hybrid MS (QTRAP 6500, AB SCIEX) equipped with an electrospray ionization source coupled with a HPLC (Aglient1290, Aglient, Santa Clara, CA, USA).
Hydroponic culture medium preparation and GR24 (Coolaber, Beijing, China) treatment were performed as previously described [9]. The 1-week-old seedlings were grown in a climatic cabinet under 16 h light at 28 • C and 8 h dark at 25 • C for another 28 days, and the hydroponic culture medium with or without 1 µmol GR24 was renewed every week.

Discussion
Compared with other physical mutations, space mutation has the characteristics of non-replication. Each space mutation is subject to different spatial radiation, and the obtained mutants showed various types with different frequencies [25]. Space mutation breeding has a unique advantage and role in creating excellent new germplasms, inducing new gene resource mutations and cultivating new varieties of crops, which is an effective method for crop genetic improvement [24,26,38,39].
In a previous study, the this1 mutant showed a phenotype with dwarfing, high-tillering, and poor pollen fertility, which was similar to that of the CHA-1 mutant, and THIS1 was found to be LOC_Os01g54810 [40]. In this study, the CHA-1 mutation site was not consistent with the this1 mutation site. LOC_Os01g54810 in the this1 mutant had a 356-bp deletion covering the intron and the second exon, while LOC_Os01g54810 in the CHA-1 mutant had a mutation of ACCAC replaced TGGT on the second exon. There was no genotype similar to the CHA-1 mutant in the locus of LOC_Os01g54810 in approximately 3000 natural varieties of rice using the Rice Functional Genomics and Breeding (RFGB) database [41] (http://www.rmbreeding.cn/, accessed on 20 February 2022). Therefore, CHA-1 was identified as a new allele of THI1S1 obtained by space mutation in this study.
At present, several genes related to the SL pathway have been identified, mainly including the SL signaling genes D53, D14, and D3 and SL biosynthesis genes D27, D17, and D10 [11]. In the mutant of SL signaling genes, SL biosynthesis genes are upregulated by negative feedback regulation [17,42]. In d53, d14, and d27 mutants, the expression of the D10 gene was significantly upregulated by feedback regulation [42]. In contrast, the expression of D10 was significantly downregulated in the d17 mutant [30]. The CHA-1 mutant showed the phenotype of dwarfing and high tillering ( Figure 1A,B), which was similar to the phenotype of the mutant associated with the SL pathway. Histological analysis of the d53 mutant showed that the shortening and thinning of the stem of the d53 mutant were mainly caused by the decrease in cell number [17]. Histological analysis of the CHA-1 mutant and THZ also proved that the difference in stem node length was associated with the number of cells (Figure 2A,B). The expression of SL biosynthesis genes was downregulated in the CHA-1 mutant ( Figure 5B). The CHA-1 mutant had significantly lower concentrations of epi-5DS in the root exudates, and the tillering of the CHA-1 mutant was inhibited after GR24 treatment ( Figure 5A,C). Therefore, CHA-1 influences rice architecture by affecting SL biosynthesis.
Lipase regulates plant disease resistance and floral organ development by affecting the auxin pathway [35,43,44]. Auxin positively regulates the expression of SL biosynthesis genes D17 and D10 in Arabidopsis and rice [45]. The mutation of CHA-1 led to the downregulation of D17 and D10 expression without affecting D27 expression ( Figure 5B). The CHA-1 mutant has some mutant phenotypes that SL pathway mutants do not have. The expression pattern of CHA-1 regulated SL biosynthesis genes is similar to that of auxin regulated SL biosynthesis genes. Therefore, we speculated that CHA-1 indirectly regulates SL biosynthesis by affecting the auxin pathway, and its molecular mechanism remains to be further studied.
Several lipase genes have been identified to affect pollen fertility in rice [35,43,46]. The anthers of the rms2 mutant with abnormal pollen atrophy resulted in complete male sterility. Cytological and genetic analysis showed that RMS2 had lipase activity and was necessary for anther development and pollen fertility [35]. OsGELP34 encodes a putative GDSL lipase/esterase. The osgelp34 mutant exhibits an abnormal outer wall, leading to pollen sterility [46]. EG1 encodes a class III lipase family protein involved in the regulation of lipid metabolism in floral organs [23]. CHA-1 encodes a III lipase protein and has lipase activity ( Figure 4C). The CHA-1 mutant showed anther atrophy, poor pollen fertility, and a decreased seed setting rate. Accordingly, CHA-1, as a lipase, affects anther development and pollen fertility.

Conclusions
The CHA-1 mutant showed dwarfing, multi-tillering, and poor pollen fertility. Based on map-based cloning and candidate gene sequencing, LOC_Os01g54810 was proven to be CHA-1, which regulated rice architecture and pollen fertility. CHA-1 regulates the rice architecture by affecting strigolactone biosynthesis. The study on CHA-1 will contribute to further research of the strigolactone pathway in rice.

Data Availability Statement:
No new data were created or analyzed in this study. Data sharing is not applicable to this article.

Conflicts of Interest:
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as potential conflicts of interest.