Physiology, Genetics, and Breeding Strategies for Improving Anaerobic Germinability Under Flooding Stress in Rice

Abstract

Anaerobic germination (AG) is a pivotal trait for successful direct-seeded rice cultivation, encompassing rainfed and irrigated conditions. Elite rice cultivars are often vulnerable to flooding during germination, resulting in poor crop establishment. This drawback has led to the exploration of AG-tolerant rice landraces, which offer valuable insights into the genetic underpinnings of AG tolerance. Over the years, substantial progress has been made in identifying significant quantitative trait loci (QTLs) associated with AG tolerance, forming the basis for targeted breeding efforts. However, the intricate gene regulatory network governing AG tolerance remains enigmatic. This comprehensive review presents recent advances in understanding the physiological and genetic mechanisms underlying AG tolerance. It focuses on their practical implications in breeding elite rice cultivars tailored for direct-seeding systems.

1. Introduction

Rice (Oryza sativa L.) is one of the most important staple crops of more than half of the world’s population and is essential to food security. However, global climate change threatens the reliable production of this vital grain. The intensification of floods, alterations in rainfall patterns, increased salinity, and more frequent droughts, among other climatic shifts, have cast a shadow over rice cultivation. These difficulties are made worse by the fact that over one-third of rice production worldwide takes place in lowlands and flooded areas [1]. Failure to germinate under water is one of the main constraints of rice cultivation. However, the scarcity of water supply for irrigation and labor costs is increasing daily. These have driven a growing interest in adopting direct-seeded cultivation methods. Research has shown that direct-seeded rice (DSR) outpaces transplanted rice by 7–10 days in flowering, maturity, and overall life cycle [2,3]. Monsoon rainfall and flash floods are common in South and Southeast Asia [4], resulting in poor crop establishment in direct-seeded rice cultivation. Although most rice varieties withstand waterlogged conditions during the vegetative stage, they exhibit sensitivity to anaerobic conditions during germination and early embryo growth [5,6,7]. The ability of a plant to germinate and thrive under submerged conditions, devoid of oxygen, is referred to as anaerobic germination (AG) or AG tolerance [5,8]. Most modern rice varieties struggle to germinate beyond the early stages in prolonged oxygen-deprived environments. However, a wide genetic variability exists within rice varieties for anaerobic germination tolerance, making them prime candidates for studying AG tolerance mechanisms. Anaerobic germination tolerance is a cumulative mechanism that promotes quick germination and coleoptile elongation by switching from aerobic respiration to anaerobic respiration through enzymatic fermentation [9]. Mechanistically, AG-tolerant plants develop a hollow coleoptile, utilizing stored energy reserved in the seed. When the elongated coleoptile reaches the aerated water surface, it conveys oxygen to the roots and endosperm for establishing the plant. Recent studies have expanded our understanding of these responses at the molecular and physiological levels, highlighting hormonal signaling, starch metabolism, and gene regulatory networks as critical components of AG tolerance [10,11].

The interplay between anaerobic germination tolerance in rice and sustainable agriculture is of paramount importance in the face of climate change and resource constraints. By delving into the genetic and physiological intricacies of AG tolerance, this review aims to pave the way for developing rice cultivars that endure environmental challenges and promote sustainable farming practices.

2. The Mechanisms of Rice Response to Anaerobic Germination

The germination of rice in anaerobic conditions, characterized by a lack of oxygen, poses significant challenges to plant growth and nutrient uptake. Under these circumstances, a cascade of physiological changes occurs, including elevated levels of ethylene, carbon dioxide (CO₂) and the accumulation of phytotoxic substances like reduced iron (Fe²⁺), manganese (Mn²⁺), hydrogen sulfide (H₂S), and organic acids, ultimately leading to plant mortality [12]. However, tolerant genotypes have evolved strategies to escape this dire situation, exhibiting rapid coleoptile elongation and the potential to develop roots and shoots in shallow water [5,7]. Developing aerated tissues and aerenchyma formation under submerged conditions also facilitates the supply of vital oxygen to submerged plant parts. This intricate interplay of processes involves the maintenance of carbohydrate catabolism and anaerobic respiration, which provide the necessary energy for cellular expansion during embryo growth [8]. Tolerance to anaerobic germination in rice emerges as a complex trait governed by multiple genes that orchestrate essential biochemical and metabolic processes, including starch breakdown, glycolysis, and fermentation (Figure 1). Furthermore, hormonal and metabolic regulatory mechanisms modulate starch degradation and energy supply, contributing to the plant’s ability to endure oxygen-deprived conditions.

2.1. Phenotypic Adaptation

Identifying target traits for phenotyping is important in assessing anaerobic germination tolerance in rice. Rapid coleoptile elongation is a critical indicator of anaerobic germination. This elongated, hollow coleoptile serves as a conduit for air when it breaches the water surface, facilitated by the energy stored in the seed. Coleoptile elongation is the result of cell elongation. In the first 48 hrs of submergence, cell division remains active, where oxygen is essential. But the plant modifies the physiological process itself in O₂-deprived conditions. As cell elongation requires less O₂ than cell division, cell elongation proceeds first [13,14]. Rapid coleoptile growth during submergence is primarily driven by auxin-mediated cell elongation [15]. The microRNA miR393 targets and degrades the mRNA of TRANSPORT INHIBITOR RESPONSE 1 (TIR1)/AUXIN SIGNALING F-BOX 2 (AFB2), key auxin receptors involved in regulating AUXIN RESPONSE FACTOR (ARF) expression [16]. When submerged, the expression of miR393 is suppressed, allowing the auxin signaling pathway to become active, which enhances ARF expression. This process promotes stomatal development and coleoptile elongation in rice [17]. Certain ‘Expasine’ genes also correlates with cell wall loosening and elongation [5,14]. EXPA7 and EXPB12 expression were observed in rice coleoptiles under anaerobic conditions [18].

Another indicator of anaerobic germination tolerance is the survival of seedlings at 21 days after sowing, a trait distinguishing tolerant and non-tolerant genotypes [5,7]. In some rice genotypes, germination and coleoptile emergence may occur under stress; however, subsequent developmental arrest is frequently observed. In contrast, tolerant varieties demonstrate sustained growth and resilience, facilitating continued development beyond the coleoptile stage.

2.2. Hormonal Regulation

In adverse environmental conditions, plants deploy phytohormonal signals as a survival strategy. Among these, ethylene is a pivotal player in the submerged rice seeds. Studies have demonstrated ethylene’s regulatory role in coleoptile elongation under anoxic conditions [13,19,20,21]. However, the mechanism of ethylene in coleoptile elongation under complete oxygen deprivation is not clear yet. A study reported that ethylene and its precursor, 1-aminocyclopropane 1-carboxylic acid (ACC), require oxygen for shoot elongation [22]. Conversely, some reports proposed ethylene-independent mechanisms in anoxia signaling [23,24]. Despite these, recent research has shown that rice seed germination rate increases under complete submergence when pretreated with ACC [25]. This suggests a potential priming or post-anoxia recovery role for ethylene. This implies that ethylene may function in a temporal and oxygen-sensitive manner, playing a limited role during early germination under strict anoxia, but becoming more influential during later stages when oxygen becomes available, facilitating coleoptile elongation and growth [26]. In addition to ethylene, other phytohormones, such as abscisic acid (ABA) and jasmonic acid (JA), also play significant roles in anaerobic germination (AG) tolerance. In silico analysis, as reported by Mohanty et al. [27], unveiled a distinctive co-expression pattern of genes under submerged conditions, revealing ABA as a positive regulator. A recent study showed that a glucosyltransferase-encoding gene OsUGT75A regulates seed germination in submerged conditions through the glycosylation of ABA and JA [10]. These intricate interactions of phytohormones underscore their multifaceted contributions to the remarkable adaptation of rice seeds to germinate in anaerobic conditions.

2.3. Changes in Carbohydrate Metabolism

AG-tolerant rice varieties exhibit significant alterations in carbohydrate metabolism. The key mechanism for tolerance is the breakdown of starch, where ATP production occurs in anaerobic conditions, so that the stored carbohydrates can be utilized to support seed germination [28]. Expression studies showed that α-amylase (αAmy) is necessary for starch breakdown, which is highly correlated with germination under water. The correlation of αAmy with AG, coleoptile length, and anaerobic respiration was revealed by enzyme assays, which suggested its emergent role in AG tolerance [5]. The transcription factor MYB SUCROSE 1 (MYBS1) activates the αAmy promoter by interacting with the cis-acting TATCCA element during sugar starvation [29,30]. SUCROSE NONFERMENTING 1-RELATED PROTEIN KINASE 1A (SnRK1A), a conserved sugar and energy sensor in eukaryotic cells, acts as an essential upstream kinase of MYBS1. It facilitates seed germination and seedling growth under normal and anaerobic conditions in rice when the levels of sugar or oxygen are low [31,32]. Two hypoxia-inducible negative regulators, SKIN1 and SKIN2, interact with SnRK1A to suppress MYBS1 and αAmy3 expression, hindering starch breakdown and slowing germination and seedling growth during submergence [33]. MYB SUCROSE 2 (MYBS2), another transcription factor, competes with MYBS1 to inhibit αAmy activity under submergence conditions [34]. CIPK15, a CBL-INTERACTING Ser/Thr PROTEIN KINASE, connects low oxygen signals to the SnRK1A sugar starvation pathway, activating αAmy to regulate sugar and energy production during anaerobic germination and seedling development (Figure 1) [32].

In addition to αAmy, alcohol dehydrogenase (ADH), aldehyde dehydrogenase (ALDH), and pyruvate decarboxylase enzyme (PDC) are also associated with maintaining energy production during germination in submerged conditions [9]. It has been reported that the expression of ALDH2 is upregulated in submerged conditions [35]. In addition, lower ADH activities were observed in adh1-deficient rice mutants with a reduced germination rate and coleoptile growth, suggesting its importance in AG tolerance [36]. Different types of enzymes, proteins, and peroxidases act differently under submerged conditions. Higher activities of peroxidases resulted in a negative correlation with coleoptile growth under submerged conditions [5]. On the other hand, Tubulin a-1 chain (TUBA 1) and Actin Depolymerizing Factor 4 (ADF 4) are positively associated with seed germination and faster coleoptile elongation under water [37].

All of the mechanisms discussed above are interconnected. Seed germination and rapid coleoptile elongation are key phenotypic traits for AG tolerance. These traits are regulated by different hormones such as auxin, ethylene, and abscisic acid, which influence downstream gene expression. These hormones also modulate metabolic pathways, including starch breakdown and anaerobic respiration, that provide the energy required for germination and seedling survival. Together, the coordinated responses form an integrated physiological network enables rice seeds to germinate and establish under submerged conditions.

3. Genetic Mapping for AG Tolerance

A number of genetic variants have been identified for AG tolerance in rice, which provide excellent resources for genetic studies. A series of linkage mapping and association mapping identified several QTLs and candidate genes with major and minor effects on AG tolerance, which could be used as bases for future studies.

3.1. Bi-Parental Linkage Mapping

Identification of QTLs and candidate genes facilitate the selection of target traits with low heritability. Phenotypic screening is costly and associated with a fluctuating environment. QTL mapping is efficient for selecting the target region of a trait. Several QTLs were reported from different genetic populations that impart AG tolerance in rice. QTLs that have been identified in different studies are summarized in Figure 2 and Table 1.

QTLs associated with AG tolerance were identified in chromosomes 1, 2, 5, and 7 (qAG-1, qAG-2, qAG-5a, qAG-5b, qAG-7-2) using restriction fragment length polymorphism (RFLP) markers from the recombinant inbreed line (RIL) population of japonica (Kinnaze) and indica (DV85) rice [38]. A similar region was later identified with two QTLs named qAG-1 and qAG-7-2 [7]. A Myanmar variety, ‘Khao Hlan On’ (KHO), was backcrossed with IR 64 where IR64 was the recurrent parent. In the backcross population, five putative QTLs were found on chromosome 1 (qAG-1-2), 3 (qAG-3-1), 7 (qAG-7-2), and 9 (qAG-9-1 and qAG-9-2). Among the identified QTLs, qAG-9-2 was the most significant, showing superfluous phenotypic variations [7]. A study was conducted to determine the candidate region within qAG-9-2 QTL. This region was fine mapped to about 55 Kb, and the OsTPP7 gene was identified. The identified gene is related to trehalose-6-phosphate (T6P) metabolism and confers anaerobic germination tolerance. Another major QTL, qAG7-1, was identified in the AG-tolerant rice cultivar Ma-Zhan Red. QTL mapping was performed using an F_2:3 population derived from a cross between Ma-Zhan Red and the susceptible cultivar IR42. In addition to qAG7-1, five other QTLs were identified on chromosomes 2, 5, 6, and 8 [39]. Chromosome 7 is considered one of the most important regions for AG tolerance. A major QTL, qAG7, has been reported on chromosome 7, derived from the F₂ population of Nanhi (an aus landrace) and IR64. QTL qAG11 was also identified from Nanhi with a small effect [40]). Four QTLs have been reported from an IR64 and Kharsu 80A F_2:3 mapping population, among which three QTLs are located on chromosome 7 (qAG7-1, qAG7-2, and qAG7-3) and one on chromosome 3 (qAG3) [41]. Composite interval mapping from the F_2:3 population of TN (tolerant, a Vietnamese variety) and Anda (sensitive) has revealed three QTLs: qAG1a and qAG1b on chromosome 1 and qAG8 on chromosome 8 [42]. There are other QTLs identified that are related to seedling survival and seedling height under anaerobic conditions. The identified QTLs derived from the BC1F_2:3 mapping families are from the cross of Kalarata (indica landrace) and two recurrent parents, NSIC Rc222 and NSIC Rc238. Five QTLs were identified on chromosomes 3, 5, 6, 7, and 8 for survival named qSUR3-1, qSUR5-1, qSUR6-1, qSUR7-1, and qSUR8-1. Kalarata contributes to all the identified QTLs associated with AG tolerance except the QTL on chromosomes 5 and 8 [43]. Photo-blastic rice (PBR) is a Korean wild rice and it has been reported as an AG-tolerant germplasm [44]. PBR crossed with another Korean cultivar, ‘Nampyeong’, and the RIL population was developed in the eighth filial generation for QTL mapping [45]. The results revealed three AG tolerance-related QTLs on chromosomes 1, 3, and 11 which were not found in previous studies. Another QTL was identified from the ‘Koshihikari’ rice variety, which is related to seed germination and coleoptile elongation in submerged conditions through fermentative metabolism [46]. In a further study, high-density bin mapping has been conducted using the RIL population from YZX and 02428. This study initially identified twenty-five loci, which were further narrowed down to thirteen stable loci and pyramiding of these thirteen loci improved AG (Table 1) [47].

Table 1. Marker and position of identified QTLs from several studies.

QTL Name	Marker	Chromosome	Position (cM)	References
qAG-1	XR2635–R1485	1	78	[38]
qAG-2	R418–C560	2	110
qAG-5a	G260–X105	5	80
qAG-5b	X105–C43	5	97
qAG-7	X379–C213	7	108
qAG3	id3002377–id3004190	3	29.8–37.9	[41]
qAG7.1	id7000519–id7002260	7	24.8–53.4
qAG7.2	id7002427–id7003359	7	61.6–90.3
qAG7.3	id7003853–id7004429	7	96–97.1
qAG-1-1	RM582–RM10713	1	49–60.9	[7]
qAG-1-2	RM11125–RM104	1	83.7–159
qAG-2-1	RM327–RM6318	2	78.8–97.5
qAG-3-1	RM7097–RM520	3	115.6–138.7
qAG-7-1	RID12i–RM5606	7	43.8–60.8
qAG-7-2	RM21868–RM172	7	82.6–118.6
qAG-8-1	RM210–RM149	8	86.7–100.6
qAG-9-1	RM8303–RM5526	9	0.8–15.3
qAG-9-2	RM3769–RM105	9	36–40.7
qAG2.1	id2001831–id2003094	2	13.8–39.5	[40]
qAG2.2	id2006621–id2007526	2	75–83.2
qAG3	id3007932–id3010875	3	97.7–115.8
qAG7	id7000465–id7002784	7	29.4–153.5
qAG11	id11009201–id11010245	11	109.8–122.1
qAG1a	43,902–48,214	1	10	[42]
qAG1b	id1006871–id327392	1	78
qAG8	id8001299–id8107849	8	37
qAG11	id11003544–id11194923	11	58
qAG2	RM263–RM5378	2	145.4–154.8	[39]
qAG5	RM5361	5	0–16.2
qAG6	RM204–RM402	6	0–29
qAG7.1	RM3583–RM21427	7	73.5–79.9
qAG7.2	RM7338–RM346	7	86.1–114.2
qAG7.3	RM21803–RM234	7	121.5–143.4
qAG9	RM553–RM3808	9	95.4–109.3
qAG12	RM313–RM28766	12	67.6–133.1

Table 1. Marker and position of identified QTLs from several studies.

QTL Name	Marker	Chromosome	Position (cM)	References
qAG-1	XR2635–R1485	1	78	[38]
qAG-2	R418–C560	2	110
qAG-5a	G260–X105	5	80
qAG-5b	X105–C43	5	97
qAG-7	X379–C213	7	108
qAG3	id3002377–id3004190	3	29.8–37.9	[41]
qAG7.1	id7000519–id7002260	7	24.8–53.4
qAG7.2	id7002427–id7003359	7	61.6–90.3
qAG7.3	id7003853–id7004429	7	96–97.1
qAG-1-1	RM582–RM10713	1	49–60.9	[7]
qAG-1-2	RM11125–RM104	1	83.7–159
qAG-2-1	RM327–RM6318	2	78.8–97.5
qAG-3-1	RM7097–RM520	3	115.6–138.7
qAG-7-1	RID12i–RM5606	7	43.8–60.8
qAG-7-2	RM21868–RM172	7	82.6–118.6
qAG-8-1	RM210–RM149	8	86.7–100.6
qAG-9-1	RM8303–RM5526	9	0.8–15.3
qAG-9-2	RM3769–RM105	9	36–40.7
qAG2.1	id2001831–id2003094	2	13.8–39.5	[40]
qAG2.2	id2006621–id2007526	2	75–83.2
qAG3	id3007932–id3010875	3	97.7–115.8
qAG7	id7000465–id7002784	7	29.4–153.5
qAG11	id11009201–id11010245	11	109.8–122.1
qAG1a	43,902–48,214	1	10	[42]
qAG1b	id1006871–id327392	1	78
qAG8	id8001299–id8107849	8	37
qAG11	id11003544–id11194923	11	58
qAG2	RM263–RM5378	2	145.4–154.8	[39]
qAG5	RM5361	5	0–16.2
qAG6	RM204–RM402	6	0–29
qAG7.1	RM3583–RM21427	7	73.5–79.9
qAG7.2	RM7338–RM346	7	86.1–114.2
qAG7.3	RM21803–RM234	7	121.5–143.4
qAG9	RM553–RM3808	9	95.4–109.3
qAG12	RM313–RM28766	12	67.6–133.1

3.2. Genome-Wide Association Study (GWAS) for AG Tolerance in Rice

Genome-wide association studies (GWASs) opened a new era to identify candidate regions through single-nucleotide polymorphisms (SNPs). They are advantageous over linkage mapping because they uncover allele variations in a wide range of natural populations. A GWAS has been conducted using two types of mapping populations, i.e., 153 diverse accessions of indica and japonica, and a 144 RIL population derived from Nipponbare and IR64. The result indicated that several potential candidate genes are associated with AG tolerance. Important candidate genes are summarized in Table 2. LOC_Os01g53930 (encodes for hexokinase, HXK 6) was reported as a potential candidate gene for seed germination and coleoptile elongation under submerged conditions, which functions as a glucose sensor [48]. Moreover, eleven significant SNPs for flooded coleoptile growth and nine for flood tolerance index have been reported [49]. Eventually, they found a gene in chromosome 6 named LOC_Os06g03520. This gene encodes for a domain of unknown function (DUF) and could be a potential candidate for AG tolerance [50]. The detailed mechanisms for AG tolerance are still unclear; research is being carried out to elucidate the mechanism on a genetic and physiological basis. OsTPP7 has already been reported to control coleoptile elongation under water [51]. There are other varieties with the gene TPP7 but they do not show the same tolerance level as KHO. In order to identify the new region associated with coleoptile growth, a GWAS was conducted using 273 japonica rice genotypes that harbor TPP7. The genotypes with longer coleoptiles at four days after sowing (DAS) did not elongate further, or elongation was not significant. The study also revealed, there was no significant variation in the expression of starch catabolism genes between the genotypes with longer and shorter coleoptile grown in anaerobic conditions. However, the study found eleven significant marker–trait associations in chromosomes 1, 2, and 6 that were not significantly associated with the SNPs of the TPP7 gene and the observed traits. This suggests that there are additional mechanisms involved in the final coleoptile length under submerged conditions [52].

3.3. Omics Studies for AG Tolerance in Rice

There are a significant number of studies that explore the adaptation mechanisms of seed germinability and seedling survival in submerged conditions, but understanding how oxygen deficiency is sensed and how regulatory cascades govern transcriptional and translational modification are not clear. A gene expression study found that sucrose metabolism was increased in oxygen-deprived conditions. There were thirty-seven reactions regulated at the transcriptional level under both aerobic and anaerobic conditions. The study identified different transcription factors (TFs), i.e., MYB, bZIP, ERF, and ZnF, that were engaged with sucrose metabolism and fermentation during anaerobic germination [53]. A whole-genome transcriptome profile of six varieties, including one susceptible variety (IR64) and five tolerant varieties, showed that forty-three genes were significantly upregulated in five tolerant varieties but not in susceptible IR64. The study detected significant induction of genes encoding α amylase (LOC_Os08g36910), pyruvate decarboxylase (LOC_Os05g39310), and alcohol dehydrogenase (LOC_Os11g10480 and LOC_Os11g10510) in the five tolerant genotypes, suggesting the function of the identified genes in AG tolerance. Among the five tolerant genotypes, four genotypes showed high tolerance. An additional twenty-six genes were specifically upregulated in the four highly tolerant genotypes. Among these upregulated genes, several genes were related to cell walls and membranes including LOC_Os07g35480, LOC_Os10g40430, LOC_Os10g40440, and LOC_Os10g40520 [54]. A comparative transcriptome analysis was conducted in other tolerant (AC41620) and susceptible cultivars (Naveen), which identified a total of 906 and 823 upregulated genes in AC41620 and Naveen, respectively. The identified upregulated genes are involved in starch degradation, carbohydrate metabolism, nitrogen metabolism, redox handling, and ethylene regulation, suggesting their association with AG tolerance. Moreover, α-Amy transcripts were upregulated in the tolerant cultivar AC41620, indicating that enhanced starch degradation contributes to seed germinability in anaerobic conditions [55]. Yang et al. (2019) identified 13 stable loci for AG tolerance using high-density bin mapping. An RNA-seq analysis was further conducted across 13 stable loci. The gene expression profile combined with gene ontology (GO) resulted in three novel loci. Subsequent analysis revealed several candidate genes (Table 2) which are linked to AG tolerance in rice [47].

Table 2. Summary of identified candidate genes for AG tolerance.

Gene ID	Description	Chromosome	References
LOC_Os06g03520	DUF domain-containing protein	6	[49]
LOC_Os01g53090	Putative infection-related protein	1	[48]
LOC_Os01g53930	Glucose sensor	1
LOC_Os05g51390	Cytokinin-activating enzyme	5
LOC_Os05g48990	Related to cytokinin pathway	5
LOC_Os06g35140	Acts as MYB transcription factor	6
LOC_Os06g35160	Related to CBL-interacting protein kinase	6
LOC_Os10g18480	Encodes for an indole-4-acetate β ghucosyltransferase	10
LOC_Os10g18530	Encodes for cytoplasm O-glucosyltransferase	10
LOC_Os06g04510	Involved in substrate level phosphorylation to build ATP	6	[47]
LOC_Os02g0271900	MYB family transcription factor	2
LOC_Os06g0109600	Adenylate kinase, putative, expressed	6
LOC_Os06g0110000	Cytochrome P450, putative, expressed	6
LOC_Os06g0110200	Late embryogenesis abundant group 1, putative, expressed	6
LOC_Os07g0638300	Peroxiredoxin, putative, expressed	7
LOC_Os07g0638400	Peroxiredoxin, putative, expressed	7
LOC_Os07g0639400	Peroxidase precursor, putative, expressed	7
LOC_Os09g0531701	Glycosyl transferase family 8 protein, expressed	9
LOC_Os09g0532900	MYB family transcription factor, putative, expressed	9
LOC_Os10g0566800	Peroxidase precursor, putative, expressed	10
LOC_Os12g0539751	Expressed protein	12
LOC_Os12g0626500	Late embryogenesis abundant protein D-34, putative, expressed	12
LOC_Os10g0390500	Alanine aminotransferase	10	[56]

Table 2. Summary of identified candidate genes for AG tolerance.

Gene ID	Description	Chromosome	References
LOC_Os06g03520	DUF domain-containing protein	6	[49]
LOC_Os01g53090	Putative infection-related protein	1	[48]
LOC_Os01g53930	Glucose sensor	1
LOC_Os05g51390	Cytokinin-activating enzyme	5
LOC_Os05g48990	Related to cytokinin pathway	5
LOC_Os06g35140	Acts as MYB transcription factor	6
LOC_Os06g35160	Related to CBL-interacting protein kinase	6
LOC_Os10g18480	Encodes for an indole-4-acetate β ghucosyltransferase	10
LOC_Os10g18530	Encodes for cytoplasm O-glucosyltransferase	10
LOC_Os06g04510	Involved in substrate level phosphorylation to build ATP	6	[47]
LOC_Os02g0271900	MYB family transcription factor	2
LOC_Os06g0109600	Adenylate kinase, putative, expressed	6
LOC_Os06g0110000	Cytochrome P450, putative, expressed	6
LOC_Os06g0110200	Late embryogenesis abundant group 1, putative, expressed	6
LOC_Os07g0638300	Peroxiredoxin, putative, expressed	7
LOC_Os07g0638400	Peroxiredoxin, putative, expressed	7
LOC_Os07g0639400	Peroxidase precursor, putative, expressed	7
LOC_Os09g0531701	Glycosyl transferase family 8 protein, expressed	9
LOC_Os09g0532900	MYB family transcription factor, putative, expressed	9
LOC_Os10g0566800	Peroxidase precursor, putative, expressed	10
LOC_Os12g0539751	Expressed protein	12
LOC_Os12g0626500	Late embryogenesis abundant protein D-34, putative, expressed	12
LOC_Os10g0390500	Alanine aminotransferase	10	[56]

4. Breeding for AG Tolerance in Rice

4.1. Phenotypic Screening Method for AG Tolerance

Precise phenotyping plays a crucial role in advancing plant breeding programs, as it reflects both the plant’s genetic makeup and its interaction with environmental factors. Researchers explore different growing conditions to screen AG-tolerant genotypes, such as different growing media, temperatures, sowing depths, light/dark conditions, etc. For screening, seeds are kept in an oven to break dormancy. Phenotypic screening is carried out mainly on two growing media, i.e., soil and water. The optimum temperature ranges from 25 °C to 30 °C. Temperature plays a vital role in anaerobic germination and coleoptile elongation. Moreover, 30 °C is the most acceptable temperature in most of the studies. For the soil medium, the experimental period is greater (21 days) than for water (6–10 days). Seeds are submerged at different depths from the surface, ranging from 3.1 cm to 15 cm. Following a defined period, traits associated with AG tolerance are evaluated. Since coleoptile length has been widely recognized as an indicator of AG tolerance, it is commonly measured across various studies. Germination and survival rates are also essential and associated with AG tolerance (

Table 3. Screening methods for AG tolerance in different studies.

Medium	Depth (cm)	Temp. (°C)	Period	Measured Traits	References
Soil	10	26–31	21	Survival rate (21st DAS *)	[40,41]
Soil	10		21	Seedling survival (21st DAS)	[42]
soil	10	28	10	Germination rate, coleoptile length (10th DAS)	[58]
Soil	8		21	Germination percentage, seedling height (14th and 21st DAS)	[43]
Soil	10	26	21	Survival rate, shoot and root length (21st DAS)	[5]
Soil	10	-	7	% germination, shoot length, root length, seedling length (7th DAS)	[59]
Soil	40	-	15	Germination rate, seedling length, vigor index (15th DAS)	[60]
Soil	10	26.2–30.9	7	-	[55]
Soil	10	30	21	Germination rate (7th DAS), survival (21st DAS)	[39]
Soil	10	30.1	21	Germination rate and survival (21st DAS)	[7]
Water	10	30	10	Coleoptile length	[49]
Water	5	25	7	Coleoptile length	[54]
Water	3.2	30	8	Coleoptile length, shoot length (4th and 8th DAS) length, and weight of intact hulled seed	[52]
Water	10	30	6	Coleoptile length, coleoptile surface area, coleoptile volume, and coleoptile diameter (6th DAS)	[47]

* DAS—days after sowing

). However, it has been reported that coleoptile length and survival do not correlate. A study has been conducted using two screening methods (protray and beaker) with both soil and water medium, and it found consistent results across both screening methods [57].

4.2. AG-Tolerant Rice Germplasms

AG tolerance is a relatively rare trait in rice germplasm, with only 0.23% identified as tolerant from more than 800 accessions screened (

Table 4. List of tolerant and susceptible parents used for AG tolerance studies.

Tolerant Cultivar	Population	Origin	References
FR 13A	Parental line	-	[62]
AC41620	Parental line	-	[55]
F291, F274-2a	RIL	-	[54]
8391	Parental line	Laos	[53]
8753	Parental line	Indonesia
Khao Hlan On	BC2F2	Myanmar	[7]
Dholamon 64-3	Parental line	Bangladesh
Liu-Tiao-Nuo	Parental line	China
Sossoka	Parental line	Guinea
Kaolack	Parental line	Guinea
Khao Hlan On	BC2F2, BC3F2	Myanmar	[63]
Kinmaze	RIL	-	[38]
Ma-Zhan Red	F2:3	China	[39]
Tai Nguyen	F2:3	India	[42]
Kalarata	BC1F2:3	India	[43]
Nanhi	F2	-	[40]
Kharsu 80A	F2:3	Pakistan	[41]
PBR	RIL	Republic of Korea	[45]
Khaiyan	Parental line	Bangladesh	[5]
Kalonchi	Parental line	Bangladesh
Cody	Parental line	USA
Lamone, Arborio	Parental line	-	[14]
Anaikomban, Muthuvellai, Rajamannar	Landraces	-	[57]
CR1009	Parental line	-
498-2A BR8, Barkhe tauli, Improved blue rose, Para nellu, Jangli boro, Subo, Hwanggeumnodeul	Parental line	-	[58]
E775, E1810, E596, E1786, E 753, E773, E1846, E1195, E1049, E1772, E1723, E1701 and E1777	Parental line	-	[60]
Solpuna	Parental line	India	[59]

) [7]. Among these, six accessions showed high tolerance: Khaiyan, Kalongchi, Nanhi, Khao Hlan On, Ma-Zhan Red, and Cody [61]. Notably, Khao Hlan On, a traditional variety from Myanmar, is the donor of the OsTPP7 gene [51]. Other than these six accessions, Photo-blastic Rice (PBR), a Korean wild rice, has been reported as an AG-tolerant germplasm [44]. RIL populations from F291 and F274-2a are also considered tolerant to AG. They exhibit tolerance primarily through seed germination and coleoptile elongation under water. The underlying genetic mechanisms contributing to AG tolerance involve genes associated with carbohydrate metabolism, phosphate-dependent energy production, and ethylene signaling pathways [54].

4.3. Marker-Assisted Breeding

Marker-assisted selection enables more accurate and efficient breeding compared to traditional approaches by utilizing DNA markers to track desirable traits [64]. The identified AG-tolerant varieties containing the AG1 QTL are mostly landraces with poor agronomic features, such as poor yield [65]. Incorporating AG-tolerant QTLs into elite backgrounds can facilitate the development of high-performing, AG-tolerant rice cultivars. QTL AG1 from Khao Hlan On has been successfully introgressed into elite rice cultivars IR64 and Ciherang-Sub1. Additionally, a previous study identified a closely linked molecular marker to AG1, which can be utilized for efficient breeding programs [65]. Four japonica backcrossed lines have been developed by introducing qAG9-2 and qAG7-2 from Khao Hlan On as a donor parent. The advanced backcrossed lines exhibit better performance against water stress during seed germination and seedling establishment [63]. Transgenic plants Ciherang-Sub1-AG2 and Ciherang-Sub1-AG1-AG2 performed better than other AG-tolerant lines in terms of anaerobic germination and growth [66]. Another study showed that transgenic plants overexpressing OsARD1 (OsARD1-OE) germinate under water and have faster coleoptile elongation to escape submergence stress. OsARD1-OE also showed increased shoot growth and inhibition of root growth in submerged conditions, highlighting its potential role in anaerobic germination [67].

5. Perspectives

Anaerobic germination intolerance is a significant challenge for direct-seeded rice (DSR) cultivation. Various strategies have been employed to identify the underlying mechanism and enhance anaerobic germination (AG) tolerance in rice. The selection of tolerant germplasms through large-scale screening can expedite breeding programs. However, most breeding programs have focused on limited germplasm pools. Diversifying germplasm sources by exploring landraces and wild rice varieties is essential. Identifying tolerant genotypes from underutilized rice germplasms would broaden the genetic base for breeding and enhance resilience against flooding during germination.

Next-generation sequencing (NGS), combined with bioinformatics tools, enhances the precision of gene identification and molecular breeding by identifying novel molecular markers and decoding favorable allele combinations in elite rice cultivars. Expression quantitative trait loci (eQTL) analysis and transcriptomics can reveal the genetic regulatory networks and transcription factors driving AG tolerance, highlighting candidate genes that can be targeted for breeding programs (Figure 2).

Significant progress has been made in identification of QTLs associated with AG tolerance. Several GWAS pinpointed key chromosomal regions contributing to AG tolerance. QTL qAG9-2 led to identifying the OsTPP7 gene, a breakthrough in AG tolerance research. However, several potential QTLs remain unexplored or partially characterized, particularly on chromosome 7. Future research should focus on fine mapping the potential QTLs to uncover novel candidate genes. Different studies showed the involvement of enzymes related to carbohydrates, fatty acids, and hormones in AG tolerance. Notably, ADH, ALDH, and PDC have been strongly associated with AG tolerance, warranting further investigations to elucidate their roles.

Hybridization with AG-tolerant landraces and the application of genetic engineering will accelerate the breeding of resilient varieties. Expanding germplasm screening, integrating multi-omics approaches, and conducting field trials will ensure the development of AG-tolerant varieties capable of thriving under real-world conditions. AG-tolerant rice varieties can be screened by germinating the seeds under flooded conditions. The selected seedling needs to be collected for genomic DNA extraction, followed by next-generation sequencing to obtain genome-wide data. These genomic data could be used to perform a GWAS to identify the QTLs and candidate genes related to AG tolerance. RNA-Seq could be performed for transcriptome profiling to detect differentially expressed genes. RNA-Seq data could be used for eQTL studies to identify the genetic regulatory network. The identified gene can be utilized as a genetic marker for marker-assisted selection (MAS) or genomic selection (GS) to enhance breeding efficiency. We can also use the conventional breeding approach, where the validated allele could be incorporated into elite cultivars through backcrossing. However, introgression of AG-tolerant QTLs into elite cultivars is challenging, including linkage drag, where undesirable traits are transferred with target loci and genotype × environment (G × E) interactions affect QTL expression stability. Emerging genomic tools such as genetic engineering and genome editing technologies (e.g., CRISPR/Cas9) could be applied to directly introduce desirable alleles into high-yielding varieties, expediting the development of AG-tolerant rice adapted for direct-seeded and flood-prone environments. Three distinct approaches can be employed to develop rice varieties with enhanced tolerance to anaerobic germination (Figure 3).