Identification of QTLs responsible for culturability, and fine-mapping of QTL qCBT9 related to callus browning derived from Dongxiang common wild rice (Oryza rufipogon Griff.)

Abstract

Compared to japonica, the lower genetic transformation efficiency of indica is a technical bottleneck for rice molecular breeding. Specifically, callus browning frequently occurs during the culture of the elite indica variety 93-11, leading to poor culturability and lower genetic transformation efficiency. Here, 67 QTLs related to culturability were detected using 97 introgression lines (designated as 9DILs) derived from Dongxiang common wild rice (DXCWR, Oryza rufipogon Griff.) with 93-11 genetic background, explaining 4% ~12% of the phenotypic variations. The QTL qCBT9 on chromosome 9 was a primary QTL for reducing callus browning derived from DXCWR. Five 9DILs with light callus browning and high differentiation were screened. We evaluated the callus browning index (CBI) of 100 F₂ population crossed of 93-11 and 9DIL71 and the recombinant plants screened from 3270 individuals. The qCBT9 was delimited to a ~148kb region between the markers X16 and X23. RNA-seq analysis of DEGs between 9DIL71 and 93-11 showed three upregulated DEGs (Os09g0526500, Os09g0527900, Os09g0528200,) and three downregulated DEGs (Os09g0526700, Os09g0526800, Os09g0527700) were located in the candidate region of qCBT9. Furthermore, callus browning may be involved in cell senescence and death caused by oxidative stress. The differentiation of indica and japonica in this region suggested that qCBT9 was possibly a vital QTL contributed to better culturability of japonica. Our results laid a foundation for further cloning of the gene for reduced callus browning in O. rufipogon, and also provided a new genetic resource and material basis for improving the culturability and genetic transformation efficiency of cultivated rice.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11032-024-01470-z.

Introduction

Genetic transformation is an indispensable tool for functional studies of genes and the molecular breeding of crop plants. It is of vital importance to establish an efficient transformation system, especially for superior varieties that are difficult to transform. An efficient genetic transformation system has been established in rice (Oryza sativa L.) (Hiei et al. 1994; Nishimura et al. 2006; Nishimura 2020) based on the excellent culturability of japonica (Oryza sativa ssp. japonica), However, poor tissue characteristics, especially callus browning, usually result in few regenerated plants, which is a major obstacle to the genetic transformation of indica rice (Oryza sativa ssp. indica) (Lin and Zhang 2005; Swain et al. 2018). The stages of tissue culture include induction, proliferation and regeneration. Callus browning is a common phenomenon that occurs during callus proliferation, resulting in decreased regenerative ability, poor growth, and even death (He et al. 2009; Zhang et al. 2020). Although much work on the development of an efficient tissue culture system in indica rice has been done, a generally applicable culture system has not been established for most indica rice varieties because the varieties are highly variable. Instead, few relatively adaptive culture systems have been found for only a limited number of varieties, and large differences exist between them (Sahoo et al. 2011; Shri et al. 2013; Sundararajan et al. 2017).

There are many factors affecting callus browning, such as medium composition, explants type, antioxidants, adsorbents, salt concentration, exogenous hormones and culture environment (Tóth et al. 1994; Wang et al. 2000; Atanassov et al. 1995; Datta and Datta 2006; Krishna et al. 2008; Uchendu et al. 2011; Thomas 2008; Han et al. 2010; Dong et al. 2016; Huang et al. 2021). However, for different rice materials, improving these conditions could not fundamentally reduce the callus browning. The root cause of callus browning is genotypic (Abe and Futsuhara. 1986; Huang et al. 2021), therefore, genetic alteration might be employed to reduce the important problem of callus browning, which could lead to the improvement of gene functional studies, genome editing and molecular breeding (Jia et al. 2007).

Currently, many QTLs related to traits associated with tissue culture potential have been identified using mature embryos during callus induction, proliferation and regeneration in rice (Huang et al. 2021; Nishimura et al. 2005; Li et al. 2007; Li et al. 2013; Taguchi-Shiobara et al. 1997, 2006; Takeuchi et al. 2000; Ozawa and Kawahigashi 2006; Zhang et al. 2016; Zhao et al. 2009). Recently, Huang et al. (2021) detected a total of 8 QTLs associated with anther culturability in rice, and screened the candidate gene Os09g0551600 combined with an RNA-seq analysis of the parents (YZX and 02428), low- (L-Pool) and high-CIR RILs (H-Pool) after 16 and 26 days of culture (Huang et al. 2021). Additionally, map-based cloning has been successfully employed to clone genes related to rice culturability. The gene encoding the ferredoxin-nitrite reductase (NiR) gene, which controls the differentiation and regeneration potential of rice callus, has been identified (Nishimura et al. 2005). In a previous study, we isolated and characterized a gene named BROWNING OF CALLUS 1(BOC1) responsible for reducing callus browning and improving the efficiency of genetic transformation (Zhang et al. 2020). However, only a few genes related to callus browning have been fine-mapped, and the pathways of culturability traits also remain poorly understood.

Previous studies related to the QTL analysis of rice culturability mostly used japonica rice as the donor, because of its excellent tissue culture potential, however, no research on fine-mapping of the culturability of japonica has been done. Also, only a few studies have identified QTLs associated with culturability in wild rice. In this study, we detected QTLs related to culturability traits using O. rufipogon introgression lines derived from a cross between a Dongxiang common wild rice (DXCWR) donor and the elite indica variety 93-11 (O. sativa) as the recipient. Additionally, a major QTL qCBT9 associated with the reduction of callus browning was identified on chromosome 9. We subsequently constructed a segregation population from a cross between 93-11 and 9DIL71, an introgression line with low callus browning, and fine-mapped the qCBT9 locus. We also used the transcriptome data from 93-11 and 9DIL71 for a cluster analysis. Furthermore, we performed an evolution analysis using publicly available genome resequencing data. These results provided the foundation for further positional mapping and cloning of the genes associated with tissue culture potential in rice, providing a gene target for improving culturability and genetic transformation.

Materials and methods

Plant materials

A population consisting of 97 Oryza rufipogon Griff. introgression lines (BC₃F₃, named 9DILs) was developed from a cross between a Dongxiang common wild rice and the elite indica variety 93-11. The 9DIL71 line that had a low level of callus browning was selected from the 9DILs, and backcrossed with 93-11 to construct an F₂ population from which the exchange individual plants were selected for fine mapping.

Media component

The NB1 medium for QTL mapping consisted of NB medium (containing N6 macronutrient components (Chu et al. 1975), B5 micronutrient components (Gamborg et al. 1968), and organic components), supplemented with 2 mg/L of 2,4-D, 30 g/L of sucrose, and phytagel of 3g/L. The pH of the medium was adjusted to 5.8~5.9 with 1 N KOH.

The NB differential medium consisted of NB medium (containing N6 macronutrient components, B5 micronutrient components, and organic components), 30 g/L of sucrose, casein hydrolysate of 2 g/L, sorbitol of 30 g/L, 2-morpholine ethyl sulfonic acid of 1.1 g/L, NAA of 2 mg/L, KT of 1 mg/L, phytagel of 4 g/L. The pH of the medium was adjusted to 5.8~5.9 with 1 N KOH.

Tissue culture procedure

Proliferation procedure

Mature, healthy, dehusked seeds of 9DILs and the F₂ population were sterilized by immersion in 75% ethanol for 30 s, then in 15% sodium hypochlorite solution for 15 min with shaking, then rinsing 5 times with sterile water. The seeds were inoculated on the appropriate medium for induction and proliferation. Approximately 75 mature seeds per line, evenly distributed among three petri dishes, were incubated for 30 days at 28 °C in the dark. The traits of the calli in each petri dish were recorded.

Differentiation procedure

After 15 days on callus induction medium, the scutellum- derived calli were transferred to the differentiation medium. Three plant tissue culture dishes, containing nine calli each, were used for the regeneration culture for each line. The regeneration culture was performed in the light (16 h light/8 h dark) at 28 °C 18000 Lx light intensity for 30 days.

Phenotypic evaluation

The degree of callus browning was categorized into five levels as follows: 0) browning of less than 1/10 of the callus tissue area (recorded as no browning); 1) browning between 1/10 and 1/3 of the callus tissue area (recorded as light browning); 2) browning between 1/3 and 2/3 of the callus tissue area (recorded as medium browning); 3) browning between 2/3 and 1 of the callus tissue area (recorded as deep browning); and 4) browning of all the callus tissue area (recorded as total browning). The degree of fragmentation was divided into four levels as follows: 1) callus with no fragmentation; 2) callus with light fragmentation; 3) callus with medium fragmentation; 4) callus with deep fragmentation. Callus culturability was assessed based on seven parameters i.e., CIR (Callus Induction Rate), CBR (Callus Browning Rate), CBI (Callus Browning Index), ICW (Induced-Callus Weight), CFI (Callus Fragmentation Index), CGR (Callus Greening Rate), and RR (Regeneration Rate). The calculation formulae were:

• CIR = (number of calli / number of seeds each dish) × 100%

• CBR = (number of browning calli / number of transferred calli) ×100%

• CBI = (Σ number of calli at each browning level × corresponding browning level)/ (number of transferred calli × highest browning level), the browning level is measured on a scale of 0-4, with level 4 being the highest degree of browning;

• ICW = weight of total calli per dish / number of total calli per dish;

• CFI = (Σ number of calli at each fragmentation level × corresponding fragmentation level) / (number of total calli × highest fragmentation level), the fragmentation level is measured on a scale of 0-4, with level 4 being the highest degree of fragmentation;

• CGR = (number of greening calli / number of differentiated calli ) ×100%;

• RR = (number of calli producing green buds/number of transferred calli) ×100 %.

Scanning electron microscopy observation

At 30 days after inoculation, callus samples were placed in a vacuum in 2.5% FAA fixative for 2 h. The samples were rinsed with phosphoric acid buffer (0.2 mol/L pH=7.2) 3-5 times and fixed with OsO₄, then dehydrated in an ethanol gradient for 20 min each. The samples were placed into a critical point dryer sample chamber (HCP-2, HITAVHI, Japan) and then glued on the sample table and coated using an ion sputtering apparatus. A scanning electron microscope (HITAVHI S-3400N) was used to observe and photograph the samples.

PCR analysis

Genomic DNA was isolated from the leaves of parents and populations using the CTAB method (Rogers et al. 1989). SSR markers and InDel markers were used to identify the polymorphism of the parents and the population. The 10 µL PCR system contained 2 µL of DNA, 2 μL of primer, 1 µL of distilled water and 5 µL of PCR master mix (10×Taq buffer, MgCl₂, 0.2 mM dNTPs, 1 U Taq polymerase). PCR amplification was performed using a PTC-100 PCR instrument, and the procedure was as follows: pre-denaturation at 95°C for 5 min, 35 cycles of denaturation at 95°C for 30 s, annealing at 58°C for 30 s, extension at 72°C for 40 s, and final extension at 72°C for 10 min. The amplification products were denatured by adding 5 µL of denaturant (3×Loading buffer), then electrophoresed on 8% denaturing polyacrylamide gels or 5% agarose gel electrophoresis to analyze the genotype.

Statistical analysis

The culturability traits data from 97 introgression lines were combined and sorted in Excel, and a frequency distribution histogram was made. An analysis of variance (ANOVA) and a correlation analysis of the culturability traits were performed using the software SPSS V 25.0.

QTL analysis

We selected 139 pairs of SSR markers combined with 97 introgression lines for a QTL analysis using the Map Manager QTXb20 (Manly et al. 2001) software. A probability value less than 0.05 was taken as the threshold to confirm the existence of a QTL, and a QTL analysis was conducted by a single marker regression method. A total of 100 F₂ populations constructed by the hybridization of 93-11 and 9DIL71 were used for the phenotypic identification of callus browning. Combined with a genotype analysis, the Map Manager QTXb20 and QTL IciMapping software (Meng et al. 2015) were used for a single point analysis and an interval mapping analysis, respectively.

RNA extraction and RNA-seq analysis

RNA from 93-11 and 9DIL71 was extracted from callus cultivated for 30 days using three biological replicates. Paired-end libraries were constructed and sequenced on the Illumina HiSeq 2500 platform at the Novogene Company (China). FPKM (fragments per kilobase per million mapped reads) of each gene and the differentially expressed genes (DEGs; |log₂fold change| >1, P value<0.05) in the test group 9DIL71 compared to the control group 93-11 were calculated (Trapnell et al. 2010). GO and KEGG enrichment were analyzed using the Cluster profiler software.

RT-qPCR analysis

Total RNA (~2 μg) was used for cDNA synthesis using FastKing gDNA Dispelling RT Supermix according to the manufacturer’s instructions. RT-qPCR was performed using a CFX96 Real-Time System (BIO-RAD, US). The rice Actin (Os11g0163100) gene was used as an internal control to normalize the gene expression data using the relative quantification method (2^–ΔΔCT). Each reaction contained 10 ng of first-strand cDNA, 10 μM of gene-specific primers and 10 μL of real-time PCR SYBR MIX (TB Green® Premix Ex Taq^TM, Takara). The sequences of all primers used are listed in Supplementary Table S5.

Results

Phenotypic variations for the culturability of 9DIL introgression lines

To identify genes associated with culturability traits, we constructed a set of introgression lines (BC₃F₃, designated as 9DIL) derived from a cross between an O. rufipogon accession Dongxiang common wild rice (DXCWR) as the donor and the elite indica variety 93-11 (O. sativa) as the recipient. In a previous study, we investigated the callus browning trait of DXCWR after 30 days of cultivation on NB unimproved medium (designated as NB1, Wang et al. 2023). The results showed that DXCWR was relatively resistant to browning, and the callus browning was light, however, in 93-11, it was serious (Wang et al. 2023). We used NB1 medium to incubate 97 lines in dark culture for 30 days. The callus induction rate (CIR), the callus browning rate (CBR), the callus browning index (CBI), the induced-callus weight (ICW), and the callus fragmentation index (CFI) ranged from 24%~100%, 40%~ 100%, 0.10~0.80, 0.04~0.20, 0.26~0.85, respectively (Fig. 1; Supplementary Fig. S1-2). The ANOVA showed that the culturability traits were highly significantly different among the introgression lines (Supplementary Table S1), indicating that genotype had a major effect on all of the traits. The four culturability traits of the introgression lines showed a continuous distribution, and the CIR, CBR, CBI, ICW and CFI were mainly distributed in 60%~70%, 84%~98%, 0.30~0.40, 0.06 ~ 0.08g and 0.30~0.40, respectively (Fig. 1; Supplementary Fig. S2), which basically conformed to a normal distribution and thus could be analyzed for QTL.

Fig. 1.

The phenotype of callus browning in introgression lines of DXCWR and 93-11 (A-F) The phenotype of callus browning of 93-11and 9DILs ranging from level 0 to level 4. Bars=1 cm. (G-H) Frequency distribution of CBR and CBI in rice introgression lines populations. CBR indicates the callus browning rate; CBI indicates the callus browning index

Correlation analysis for the culturability of 9DIL introgression lines

A correlation analysis of the culturability traits showed a high negative correlation between CBR, CBI and ICW (Table 1). In other words, the callus weight of the introgression line with lower callus browning was heavier, indicating the callus of these lines grows well. In particular, the correlation between CBR and CBI was the highest, with a positive correlation coefficient of 0.876 (Table 1), which illustrated that CBR and CBI described the degree of callus browning in a highly consistent manner. Our study did not find a significant correlation between CBR or CBI and CFI. Furthermore, we randomly selected 9 introgression lines with high, medium and low fragmentation degree for differentiation phenotype analysis, respectively. The result showed that inducing the appropriate level of callus fragmentation could enhance differentiation in tissue culture (Supplementary Fig. S5F-I; Fig. 3; Dabul et al. 2009), however, excessive or insufficient callus fragmentation led to a decrease in callus differentiation and emergence (Supplementary Figure S5A-E, J-M, N-T), which indicating that the impact of CFI on callus regeneration may be confined to a specific threshold range.

Table 1.

Analysis of correlation between the 5 indices for tissue culture ability traits

	CIR	CBR	CBI	ICW	CFI
CIR	1
CBR	-0.407**	1
CBI	-0.332**	0.876**	1
ICW	-0.078	-0.416**	-0.445**	1
CFI	-0.299**	-0.058	-0.021	0.215*	1

CIR indicated callus induction rate; CBR indicated callus browning rate;

CBI indicated callus browning index; ICW indicated induced-callus weight;

CFI indicated callus fragmentation index

^* Correlation is significant at the 0.05 level (two-tailed)

^** Correlation is significant at the 0.01 level (two-tailed)

Fig. 3.

Phenotype of tissue differentiation for 5 introgression lines with light callus browning (A-F) Investigation of differential traits in 93-11 and 5 9DILs with light browning. Bars=2cm. (G-H) Statistics of differentiated characters. Data are mean ± SD (n = 3). The double asterisks represent a significant difference determined by Student’s t-test at P < 0.01 and one asterisk represents a significant difference determined by Student’s t-test at P < 0.05

QTL analysis for culturability of 9DIL introgression lines

A total of 14, 16, and 10 QTLs related to CIR, ICW and CFI were respectively detected, which could explain 4%-15% of the phenotypic variation (Fig. 2; Supplementary Table S2). Among the QTLs, QTL qI1-2, qI8-3, and qI9-1 explained 10%, 9%, and 8% of the phenotypic variation, respectively, and the additive effects were all negative, indicating that alleles from DXCWR reduced the callus induction rate (Fig. 2; Supplementary Table S2). The QTLs qW1-1, qW1-2, qW4-1, qW11-3, and qW11-4 explained 12%, 10%, 12%, 12%, and 12% of the phenotypic variation, respectively, and the additive effects were all positive, indicating that alleles from DXCWR could increase the callus weight (Fig. 2; Supplementary Table S2). The QTLs qF2-1, qF3-2 and qF8-2 contributed 15%, 8% and 8%, respectively, and the additive effect was positive, which showed that alleles from DXCWR increased the degree of CFI (Fig. 2; Supplementary Table S2).

Fig. 2.

Linkage map of QTLs detected in the 9DILs population responsible for the tissue culture potential

A total of 27 QTLs related to CBR and CBI, which reflected the frequency and degree of callus browning were detected, and could respectively explain 4%-10% of the phenotypic variation (Fig. 2; Table 2). A total of 8 QTLs were co-localized in the CBR and CBI, and qB2-1, qB4-1, qB9-4 and qB9-5 respectively explained 7%, 10%, 7% and 8% of the CBI variation (Fig 2; Table 2). The additive effect was negative (Fig. 2; Table 2), indicating that alleles from DXCWR reduced callus browning. Furthermore, the loci qB9-4 and qB9-5 that regulated CBI (designated as qCBT9, the QTL regulating the callus browning trait) were newly identified, and had not been reported in previous studies (Fig. 2; Table 2).

Table 2.

QTLs analysis of callus browning trait in rice introgression lines derived from DXCWR

Trait	Chr.a	Locus	Marker	PV(%)b	Pc	Addd
CBR	Chr2	qCBR2-1	RM341	10	0.00114	-10
	Chr3	qCBR3-1	RM1352	5	0.02534	8
	Chr4	qCBR4-1	RM335	8	0.00414	-8
	Chr4	qCBR4-2	RM518	5	0.03446	-4
	Chr5	qCBR5-1	RM459	4	0.0415	-29
	Chr7	qCBR7-1	RM248	5	0.02199	-11
	Chr9	qCBR9-1	RM201	5	0.03294	-6
	Chr9	qCBR9-2	RM107	5	0.02874	-6
	Chr9	qCBR9-3	RM5384	6	0.01618	-6
	Chr9	qCBR9-4	RM6294	6	0.01733	-6
	Chr11	qCBR11-1	RM27181	4	0.03842	-6
CBI	Chr2	qCBI2-1	RM341	7	0.00617	-0.09
	Chr3	qCBI3-1	RM1352	9	0.00199	0.11
	Chr3	qCBI3-2	RM85	5	0.02205	0.09
	Chr4	qCBI4-1	RM335	10	0.00184	-0.08
	Chr4	qCBI4-2	RM518	6	0.01266	-0.05
	Chr7	qCBI7-1	RM180	4	0.04339	-0.07
	Chr7	qCBI7-2	RM214	4	0.04082	-0.07
	Chr7	qCBI7-3	RM320	4	0.03739	-0.07
	Chr7	qCBI7-4	RM505	4	0.03699	-0.07
	Chr7	qCBI7-5	RM248	6	0.01599	-0.12
	Chr9	qCBI9-1	RM7390	5	0.02657	0.09
	Chr9	qCBI9-2	RM201	6	0.01097	-0.08
	Chr9	qCBI9-3	RM107	5	0.03366	-0.06
	Chr9	qCBI9-4	RM5384	7	0.00829	-0.07
	Chr9	qCBI9-5	RM6294	8	0.00474	-0.08
	Chr10	qCBI10-1	RM228	5	0.03133	0.11
	Chr10	qCBI10-2	RM333	7	0.01016	0.14
	Chr11	qCBI11-1	RM202	4	0.04824	-0.08

a: Chromosome; b: Phenotypic variance of the callus browning; c: The probability that the marker genotype had no effect on the trait; d: additive effect of allele from DXCWR (Oryza rufipogon)

Furthermore, we screened 18 lines with CBI values less than 0.30 to differentiation culture by using NB differentiation medium. The result showed that there was a positive correlation between CGR (the callus greening rate) and RR (the regeneration rate) (Supplementary Table S3). Compared with 93-11, we selected 5 introgression lines with better differentiation potential, including 9DIL156, 9DIL272, 9DIL280, 9DIL300 and 9DIL343 (Fig. 3). For example, the CGR of 9DIL280 was as high as 94.44%, and the RR of DIL156 and DIL272 were 25.93% and 29.63%, respectively (Fig. 3), which would be used as elite indica rice genetic transformation receptors and parent materials for the fine mapping of culturability traits.

Callus browning trait analysis of the introgression line 9DIL71

To identify genes related to callus browning, we screened an introgression line 9DIL71 including a chromosomal segment at the qCBT9 locus on the long arm of chromosome 9. We inoculated 93-11 and 9DIL71 in NB1 medium and incubated the cultures in the dark for 30 days to investigate the callus browning rate (CBR) and the callus browning index (CBI). The CBR and CBI of 9DIL71 were lower than those of 93-11 (Fig. 4). Therefore, we constructed a separate population by crossing 93-11 and 9DIL71. Compared with 93-11, the callus browning phenotype of the F₁ was significantly reduced, and similar to 9DIL71 (Fig. 4), which indicated that the reduced-callus browning gene was dominant. Scanning electron microscopy (SEM) showed that the surface of 9DIL71 was rough and uneven, and the cells were close together, while the surface structure of 93-11 seemed smooth, and the cells lost their regular roundness and clustered poorly, showing an inactive flat phenotype with small and few gaps between cells (Fig. 4).

Fig. 4.

The callus browning traits of 93-11, 9DIL71, and F₁ crossed by 93-11 and 9DIL71 (A-C) The phenotypes of 93-11, 9DIL71, and F₁ crossed between 93-11 and 9DIL71 inoculation for 30 days. Bars=1 cm. (D-F) Scanning electron microscopy of 93-11, 9DIL71, and F₁. Bars=400 μm. (G-H) Comparison of the CBR and CBI of 93-11, 9DIL71, and F₁. Values in (G-H) are means ± SD (n = 3). Two-tailed Student’s t-tests were performed to determine significant differences

Fine-mapping the gene responsible for callus browning trait

To verify the existence of qCBT9, we constructed an F₂ segregating population derived from a cross between 93-11 and 9DIL71. Firstly, we randomly selected 100 individual plants to determine their callus browning phenotypes. An ANOVA analysis showed that the CBI of the 100 plants was different between individuals (Supplementary Table S4), indicating that callus browning was mainly affected by genotype. As shown in Supplementary Fig. S3, CBI showed a bimodal distribution, which indicated that the callus browning trait of this population was similar to a quality trait, providing a good foundation for the fine-mapping of qCBT9. We then performed a QTL analysis (i.e., a single point analysis and interval analysis) by using the Map manager QTX20 and Icimapping software packages, respectively. The results suggested that the callus browning trait QTL qCBT9 mapped between LX11-2 and LX13-3, explaining 13% of the phenotypic variances related to reduced callus browning (Table 3).

Table 3.

QTL analysis of callus browning trait in rice 100 F₂ segregation population crossed by 93-11 and 9DIL71

	Single point analysis				Interval analysis
Chr	Locus	PV(%)	P	Add	Locus	LOD	PV(%)	Add
Chr9	LX11-2	10	0.00436	-0.08	LX11-2~LX13-3	2.54	13	-0.089
Chr9	LX13-3	9	0.00857	-0.06

For fine mapping the callus browning trait QTL qCBT9, we selected recombinant plants from the 3270 individuals such that the genotypes were heterozygous at LX11-2 and homozygous at LX13-3, which was consistent with 93-11, and heterozygous at LX13-3 and homozygous at LX11-2, which was also consistent with 93-11, to evaluate the CBI (Fig. 5A-B). Additionally, we developed 8 InDel markers and identified the genotypes of recombinant plants. The possible candidate region for qCBT9 mapped between markers X16 and X23, a ~148kb region (Fig. 5B-C).

Fig. 5.

Map-based cloning of QTL qCBT9 related to callus browning (A) The QTL qCBT9 was preliminarily located between LX11-2 and LX13-3 markers by analyzing 100 F₂ plants. (B-C) The QTL qCBT9 was delimited to a ~148 kb region between the X16 and X23 markers by evaluating the phenotypes of heterozygous recombinants (R1 to R6). In graphical genotypes of recombinants, the black and white regions represent homozygosity for the DXCWR genome and homozygosity for 93-11 genome respectively. Data are means ± SD (n = 3). Two-tailed Student’s t- tests were performed between 93-11 and transgenic recombinant plants (**P < 0.01)

RNA-seq analysis between 9DIL71 and 93-11

To elucidate the molecular mechanism of callus browning, we performed an RNA-seq experiment between 9DIL71 and 93-11 cultured for 30 days, and analyzed the DEGs. A total of 3,348 upregulated genes and 3,116 downregulated genes were identified in 9DIL71 versus 93-11 callus (Fig. 6A). GO and KEGG analyses indicated that callus browning was associated with transcription regulator activity, DNA binding, oxidoreductase activity, heme binding, biosynthesis of amino acids, plant-pathogen interaction, carbon metabolism, plant hormone signal transduction, MAPK signaling pathway, and cysteine and methionine metabolism (Fig. 6B-C). Additionally, the DEGs are categorized based on gene annotation, primarily including the types illustrated in Figure 6D. A total of 162 DEGs were upregulated, which were associated with stress response and cell death (Fig. 6D; Supplementary Fig. S4A-G). Significantly, several signal molecules, such as calciumion, mitogen-activated protein kinase (MAPK) and calmodulin-dependent protein kinase (CAMK), have been reported to regulate antioxidant defense systems (Supplementary Fig. S4A; Jiang and Zhang 2003; Ma et al. 2012; Zhang et al. 2007). Compared to 93-11, the higher expression of OsCATA and OsCATC encoding catalase in 9DIL enhances the clearance of hydrogen peroxide (H₂O₂), maintains reactive oxygen species (ROS) homeostasis, and prevents cell death (Supplementary Fig. S4C, Ye et al. 2014; Lin et al. 2012). A total of 5 up-regulated DEGs encoding glutathione peroxidases (GPXs) were identified, which modulated the levels of H₂O₂, GSH and thioredoxins (Supplementary Fig. S4C, Passaia and Margis-Pinheiro. 2015). The NIR gene encoding ferredoxin-nitrite reductase, which was first cloned as associated with regeneration in rice by genetic transformation, was upregulated (Supplementary Fig. S4G) and used as a selection marker (Nishimura et al. 2005). In addition, DEGs related to auxin and cytokinin responses (Supplementary Fig. S4B), negatively regulating cell death (Supplementary Fig. S4D), dehydration response (Supplementary Fig. S4E), Dof transcription factor (Supplementary Fig. S4F) might maintain the ROS balance during a stress response. We also found some downregulated DEGs encoding xyloglucan galactosyltransferase and glycosyl hydrolase involved in cell wall degradation (Fig. 6D, Supplementary Fig. S4H), and 80 DEGs encoding receptor-like kinase and disease resistance proteins associated with the immune response (Fig. 6D, Supplementary Fig. S4I-J). Clearly, a total of 16 PCD-related DEGs were down-regulated and two DEGs were up-regulated (Fig. 6D), in which there were 13 DEGs associated with cysteine protease with down-regulation and two with up-regulation (Supplementary Fig. S4K). Previous study showed that the mRNA level of cysteine protease was also increased during specific developmental stages of plant programmed cell death (PCD), leading to cell death (Shi. 2002). A total of 13 oxidase-related DEGs are down-regulated and two were up-regulated (Fig.6D; Supplementary Fig. S4L). The oxidase genes related to monocopper, reticuline, and cytochrome c have oxygenase activity as a source of ROS (Zangar et al. 2004). Taken together, these observations indicate that a decrease in cell senescence and PCD caused by ROS could alleviate callus browning.

Fig. 6.

RNA-seq analysis between 93-11 and 9DIL71 (A) DEGs between 93-11 and 9DIL71 at 30 days after inoculation; (B-C) Gene ontology (GO) and Kyoto encyclopedia of genes and genomes (KEGG) analysis of DEGs. (D) Hierarchical clustering of DEGs associated with stress response and cell death. The numbers in the boxes represent the number of genes per cluster

Candidate gene analysis of qCBT9 and evolutionary analysis of candidate segments

To further screen the candidate gene qCBT9, we analyzed the RNA-seq data and selected DEGs in the region between markers X16 and X23. A total of three upregulated DEGs (Os09g0526500, Os09g0527900, and Os09g0528200) and three downregulated DEGs (Os09g0526700, Os09g0526800, and Os09g0527700) were identified between the markers markers X16 and X23 (Fig. 7A-F; Table 4). The RT-qPCR analysis of six DEGs in the ~148kb region showed the same results as RNA-seq (Fig. 7G-I). But only one DEG was cloned: OsIAA26 with downregulation encodes atypical Aux/indole acetic acid (IAA) protein (Chen et al. 2018). Os09g0526500 with up-regulation encodes a Bon-association protein (BAP) and BAP genes serve as general negative regulators of PCD induced by biotic and abiotic stimuli including reactive oxygen species (Yang et al. 2007). Another two upregulated DEGs (two transcription factors) were involved in the stress response and cell death (Table 4). Os09g0527900 with upregulation encodes a zinc finger family protein, and in previous study, zinc finger proteins (ZFPs) have been shown to play important roles in the responses of plants to oxidative and abiotic stress (Zhang et al. 2014). Os09g0528200 encoding homeobox associated leucine zipper, after being activated by phosphorylation, bZIP transcription factor participates in ABA-mediated transcriptional regulation by binding to ABREs in gene promoters, thereby regulating rice abiotic stress response and senescence (Han et al. 2023). In addition, Os09g0526700, encoding UDP-galactose/glucose epimerase 3 with downregulation, mediated the redox reaction by formatting temporary ketone intermediates and OsUGE3 positively regulates cellulose and hemicellulose biosynthesis, enhancing cell wall by increasing polysaccharide deposition to improve biomass, mechanical strength and salt tolerance (Tang et al. 2022). Os09g0526800 encoding lecithin retinol acyltransferase, participated in phospholipid metabolism with catalytic activity (Anantharaman and Aravind 2003).

Fig. 7.

Candidate genes and domestication analysis of the candidate interval (A-F) The FPKM values of three upregulation and three downregulation. (G-L) The RT-qPCR results of six genes in the ~148 kb region (n = 3). (M)The values of were individually calculated. The blue line shows the F_ST between indica and japonica, and the orange line indicates the F_ST between O. sativa and O.rufipogon. The values were calculated for each sliding window of 10 kb with an increment of 1 kb. Two-tailed Student’s t-tests were performed to determine significant differences, **P < 0.01: extremely significant difference

Table 4.

Candidate genes in the ~148 kb genomic region between markers X16 and X23 of qCBT9

gene_id	log2FoldChange	pvalue	padj	Description
Os09g0528200	1.89	1.74E-16	4.11E-15	homeobox associated leucine zipper, putative, expressed
Os09g0527900	1.20	5.97E-12	9.12E-11	B-box zinc finger family protein, putative, expressed
Os09g0526500	2.29	3.46E-08	3.31E-07	BAP2, putative, expressed
Os09g0527700	-1.14	3.61E-04	1.66E-03	OsIAA26, Aux/IAA protein; Indoleacetic acid-induced protein 26
Os09g0526800	-1.20	5.73E-08	5.35E-07	lecithin retinol acyltransferase, putative, expressed
Os09g0526700	-1.71	1.79E-12	2.89E-11	UDP-galactose/glucose epimerase 3, putative, expressed

Finally, we analyzed the fixation index (F_ST, the level of population differentiation) in the candidate interval on chromosome 9 using publicly available genome resequencing data containing 446 accessions of O. rufipogon and 1,083 O. sativa on a sliding window (Huang et al. 2012). The mean F_ST level in this ~148kb candidate interval was 0.43 between indica and japonica, and higher than that between O. sativa and O. rufipogon (Fig. 7G), suggesting that the genes responsible for the differentiation of indica and japonica existed in this candidate interval and further indicating that DXCWR may carry the allele from japonica.

Discussion

Evaluation of phenotype in tissue culture ability

From the viewpoint of the rice tissue culture potential, the callus induction rate (CIR) reflects the capacity for callus induction. The induced callus weight (ICW) indicating the accumulation of dry matter in the process of proliferation. Similarly, the callus browning rate (CBR) and callus browning index (CBI) provide insights into the metabolic state and proliferation capacity. While the callus fragment index (CFI) is important for differentiation and regeneration potential, emphasizing the callus's differentiation capacity (Li et al. 2013). In our study, the low correlation of CIR and ICW, CBR and CFI, CFI and CBI (Table 2) implied that the tissue culture stages of induction, proliferation and differentiation ability might be independent. Furthermore, the CBI not only be associated with the browning rate but also the degree of browning, reducing potential errors from human observation thus minimizing the error caused by human investigation (Zhang et al. 2020). Using CBI as a phenotypic index, we examined callus browning in an F₂ segregation population derived from 93-11 and 9DIL71 (Fig. 3), successfully identifying the QTL qCBT9 (Table 3).

Identification of QTLs for culturability traits

QTLs responsible for culturability derived from DXCWR were found to co-locate with previously identified QTLs by utilizing a population created from indica and japonica rice. A total of 67 QTLs linked to tissue culture potential were identified, including qI11-2 near RM21 on chromosome 11, which aligns with a QTL regulating callus induction frequency detected in a population of CSSL from Zhenshan 97B and Nipponbare (Zhao et al. 2009). Another notable QTL, qB4-1 near RM335 on chromosome 4, showed similarities to qSc4 from a introgression line constructed by Koshihikari and Kasalath (Taguchi-Shiobara et al. 2006). QTLs regulating CBI near RM341 on chromosome 2, RM335 on chromosome 4, RM505 on chromosome 7, and RM202 on chromosome 11 were consistent with qBI2-2, qBI4-1, qBI7-1, qBI10-1, and qBI11-1 loci detected in a BC₄F₄ population from Yuanjiang common wild rice and the indica rice variety 93-11 (Zhang et al. 2016). A novel QTL, qCIR9.1, on chromosome 9 was identified for callus induction rate (Huang et al. 2021), while qIc9 and qIw9 were found to regulate callus color and induced-callus weight in lines derived from Koshihikari and Kasalath (Taguchi-Shiobara et al. 2006). In our study, we identified the QTL qCBT9 using 97 9DILs derived from DXCWR (Fig. 2; Table 2), and narrowed the range down to ~148kb on the long arm of chromosome 9 by using 3270 individuals genotypes and CBI measurement of recombinant plants (Fig. 5). These results showed that the end of the long arm on chromosome 9 was a hot spot responsible for tissue culture potential.

Callus browning may be involved in cell senescence and death caused by oxidative stress

Callus is subject to oxidative stress during cultivation, leading to ROS production and accumulation (especially of H₂O₂), which also promotes PCD. A recent study suggested BOC1(BROWNING OF CALLUS 1) may decrease the cell senescence and death caused by oxidative stress, resulting in callus browning (Zhang et al. 2020). In our study, GO and KEGG analyses of DEGs in the transcriptome data indicated some genes involved in ROS accumulation and scavenging, stress response, immune response, defense response and cell death were enriched during the regulation of callus browning (Fig. 6). Furthermore, Dongxiang common wild rice (DXCWR), the northernmost common wild rice known, has a wealth of new gene/allele sources for tolerance to abiotic stress (Zhao et al. 2016). In a recent study, we identified the tissue culture potential of DXCWR, especially callus browning, and the results showed that the callus browning of DXCWR was significantly lighter (Wang et al. 2023). These results indicated that DXCWR may harbor favorable alleles for reducing callus browning which have been lost in cultivated rice. Possibly, DXCWR favorable QTLs/alleles related to culturability introgressed into elite indica-type cultivars using backcross or genetic manipulation showed lower callus browning, higher regeneration efficiency and then improved genetic transgenetic efficiency.

qCBT9 has potential applications for rice biotechnology

Hiei et al. (1994) was the first researcher to report an efficient Agrobacterium-mediated transformation system for japonica rice with a transformation frequency of 10–30%, which indicated that, in contrast to indica rice, japonica rice had better culture potential. We proposed in a previous study that the effect of BOC1 on callus browning showed no clear differences between indica and japonica (Zhang et al. 2020). However, in our study, significant differences between indica and japonica were found in the ~148kb interval (including qCBT9), and the DXCWR was closer to that of japonica rice, which facilitates the identification of excellent genes responsible for culturability in japonica rice for potential applications in rice biotechnology.

In the future, we intend to isolate and modify the major QTLs for culturability, especially the QTL qCBT9. Firstly, we added the molecular markers between markers X16 and X23 and screened the recombinant plants by genotyping the 3,270 individuals, and then evaluated the CBI of these recombinant plants. The candidate gene would be identified combined with six DEGs in the ~148 kb region and genetic manipulation, which would not only facilitate greater understanding of the molecular mechanisms of callus browning, but might also be employed as selection markers for genetic transformation and allow the creation of an optimized transformation system in rice. Therefore, the development of molecular markers related to these QTLs may potentially improve applications in indica rice.

Conclusion

In conclusion, we detected 5 indices related to tissue culture ability, and combined with SSR markers, 67 QTLs was identified by 97 introgression lines derived from Dongxiang common wild rice. Five 9DILs with elite culturability were selected which could be used as genetic transformation receptors. Furthermore, we screened 9DIL71 with light callus browning and constructed F₂ segregation population crossed by 93-11 and 9DIL71. Combined with genotypic identification and phenotypic investigation of callus browning of recombinant plants, we narrowed the QTL qCBT9 to ~148kb region between markers X16 and X23. The candidate interval existed the differentiation of indica and japonica. Moreover, the transcriptome showed callus browning may be involved in cell senescence and death caused by oxidative stress.

Supplementary information

Below is the link to the electronic supplementary material.

Author contribution

X.L. performed the experiments, analyzed the data and wrote the manuscript. J.S. helped to assess the phenotype. Y.X. helped to identify the genotype of F₂ population. M.C., Q.Z. and Y.L. helped to perform the experiments. K.Z. conceived and designed the experiments, analyzed the data, wrote and modified the manuscript. C.S. provided a good experiment platform and guided the experiment. Y.F. conceived and designed the experiments, supervised the research and modified the manuscript. All authors read and approved the final manuscript.

Funding

National Natural Science Foundation of China, 32172016, Kun Zhang , Sanya Yazhou Bay Science and Technology City, SYND-2021-4,Yongcai Fu

Data availability

The RNA-seq data of 93-11 and 9DIL71 have been deposited in the Sequence Read Archive (SRA) under accession code PRJNA1062274.

Declarations

Conflict of interests

The authors declare no conflicts of interest.