Marker-Assisted Backcross Breeding for Improvement of Submergence Tolerance and Grain Yield in the Popular Rice Variety, ‘Maudamani’

Background: Submergence stress due to flash flood reduces rice yield drastically in sensitive varieties. Maudamani is a high yielding popular rice variety but highly susceptible to submergence stress. The Sub1 and yield component QTL, GW5 were transferred into the popular high yielding variety for enhancing submergence tolerance and yield. Methods: Sub1 and GW5 QTLs were transferred into Maudamani variety by adopting marker-assisted backcross breeding method. The target QTLs were selected by foreground selection in each backcross generation progenies to track the target QTLs. Recipient parent’s genome recovery was accelerated by adopting background selection in each backcross generation progenies carrying the target QTLs. Results: The two target QTLs were tracked in each backcross generation progenies by employing the tightly linked and direct markers. Background screening was performed in each backcross generation progenies carrying the target QTLs by using 57 background markers for enhancing the recovery of recipient parent’s genome cont ent. The selected progenies containing highest genome recovery of the parent was hybridized with recipient parent, Maudamani. Finally, the selected BC 3 F 1 plant containing highest recipient parent genome content and the two target QTLs was self-pollinated. In BC 3 F 2 generation, two QTLs, Sub1 and GW5 along with

recipient parent's yield component QTLs, OsSPL14, Gn1a and SCM2 were tracked for their homozygous state in the progenies. Amongst the developed pyramided lines, six lines showed tolerance to submergence for 14 days and also exhibited higher grain yield than both the parents.
The pyramided lines, MSS 607-116-541-117 and MSS 607-116-541-436 produced >9 t/ha grain yield showing an advantage of >5% over the popular recipient variety. Few pyramided lines were similar in appearance and quality traits with the recipient parent.

Conclusion:
The pyramided lines will be useful as potential donors for the QTLs Sub1+ OsSPL14+ Gn1a + GW5 + SCM2 and also as cultivars.

Background
Rice, the golden cereal is life for millions of people in the world. Importance of rice can be easily visualized from its use as worship material for important ceremonies and rituals in India since time immemorial. Rice provides important compounds namely carbohydrate, quality proteins, vitamins, specific oils, many minerals, dietary fibre and a few phyto-compounds which provide added health benefits. The crop is very unique in its adaptation and being cultivated from very high elevation to below sea level. World-wide, the crop covers around 160 million hectares of land. The crop is cultivated as a rainfed crop in about 45% of the total rice cultivated area. Rice crop provides livelihood to nearly 4 billion people which accounts for 55% of the global population. The global annual earnings from the crop are around $206 billion which is 17% of the total crop value [1]. In recent years, the crop has been highly affected by the bad effects of climate change. The higher production from rainfed rice cultivation is now challenged by the climate change related yield reducing factors in India. About 22 mha of rainfed rice area is 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60  61  62  63 64 65 cultivated in India of which 90% are confined to eastern region of the country. Submergence tolerance and high grain yield along with resistance to major diseases and insect pests should be transferred to a superior variety for ensuring stable production.
Although there has been impressive growth in rice production and productivity during recent years, still there continues to be a higher demand for rice, projecting upwards in future due to the population growth rate being higher than the rate of production of rice. Again, in the coming years, the production increase must be harvested under the adverse effects of climate change. In addition, the future production needs to be obtained from the least available land with less use of inputs, lower chemical usage and in a more environment friendly manner. The present breeding strategy focuses on simultaneous improvement of stress tolerance and yield in rice so as to fulfill the future requirement. Enhancement of yield potential in rice is possible by incorporating few reported yield component QTLs. Among the component traits, grain number enhancement is possible by QTLs, Gn1a [2], Ghd7 [3], Ghd8 [4,5]), APO1 [6]), DEP1 [7,8]), DEP2 [9]and DEP3 [10]. Grain weight and grain dimension increase by GW2 [11], GS3 [12], GS5 [13] and GW5 [14,15]. Tiller is controlled by the QTLs, MOC1 [16], LRK1 [17], EP3 [18] and IPA1 [19,20]. High yield through higher grain filling is by the QTLs, GIF1 [7] and FLO [21,22].
Absence of these yield component QTLs in those varieties need transfer from suitable donor parents using maker assisted breeding for enhancing the yield potential in rice.
Unpredictable and sudden floods are now a common occurrence in India, particularly in the eastern region of the country. This is a major cause of yield reduction in a susceptible variety affected by the submergence stress. Maudamani variety produces 7 to 9 ton/ha normally but under favorable conditions upto 11 t/ha grain yield can be achieved. However, total crop failure occurs if the crop is affected by flash flood causing submergence for more than a week. The submergence tolerance QTL, Sub1 confers tolerance to submergence for about two weeks [23]; [24]). Recently, gene based markers have become available for transfer of Sub1 QTL in markerassisted breeding. Submergence tolerance has been improved in many high yielding varieties using this QTL introgression [25][26][27][28][29]. Three high yielding varieties namely, CR Dhan 801, CR Dhan 802 and DRR Dhan 50 showing tolerance to both drought and submergence have been developed through marker-assisted breeding and released during 2019 for cultivation in India [29].

Foreground and background selections
The target QTLs controlling the traits were validated in the donor and recipient parents before start of the hybridization and selection work. Presence of submergence tolerance QTL, Sub1 and the yield component QTLs Gn1a and GW5 were confirmed in parent, Swarna-Sub1 (Fig. 2). The recipient parent is a high yielding and popular variety. The genetic basis of high yield was checked by validating the presence of QTLs contributing higher yield namely Gn1a, OsSPL14 and SCM2 in the recipient parent, Maudamani. The common yield component QTLs namely Gn1a and SCM2 were observed in both the parents (Fig. 2). The gene based molecular markers for Sub1 QTL and yield QTLs were used for validation and tracking of target genes in the   1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60  61  62  63  64  65 parental and backcross derived lines (Table 1). Parental polymorphism survey was conducted by using 644 simple sequence repeats primers covering all the chromosomes. A total of 57 polymorphic markers were detected between the two parents and deployed for background screening ( Table 2). Maudamani was hybridized with 'Swarna-Sub1' and 870 F1 seeds were obtained. The hybridity in F1 plants were confirmed by genotyping the hybrid plants using the Sub1 specific marker. One true F1 plant was crossed with the recipient parent, Maudamani and a total of 132 BC1F1 seeds were generated. The backcross generation was grown for foreground screening in all the BC1F1 plants using the markers of two QTLs Sub1, and GW5 in the segregating progenies (Fig. 3).
The screening results in the BC1F1 progenies of the cross revealed the presence of Sub1 Additionally, the common yield components QTLs present in both the parents namely, Gn1a and SCM2 detected from both the parents are expected in homozygous state in these progenies.
Background screening was performed in all these 14 foreground positive BC1F1 progenies using 57 SSR markers. Out of these 14 plants, the progeny carrying maximum recipient genome content was selected for next backcross. The recipient parent's genome content in those 14 progenies varied from 64.58 -81.25% with an average of 76.26%. The backcross derivatives MSS128 showed highest recurrent genome content of 81.25%. The BC1F1 lines generated from MSS128 was crossed back to recipient parent, Maudamani to obtain BC2F1 derivatives. One hundred sixty nine BC2F1 plants were grown in the field for selection. Target QTLs were tracked by foreground selection using gene specific markers. Genotyping results of 169 BC2F1 progenies showed 93 positive progenies for Sub1 QTL. These 93 positive progenies were checked for the presence of GW5 QTL using the gene specific markers. Seventeen plants were detected to carry these two target QTLs (Fig. 4). In addition, the desired yield QTL from Genotyping results revealed the presence of 87 progenies positive to Sub1 QTL. These 87 carrier plants were genotyped for checking the presence of GW5 QTL. In the foreground analysis detected 12 plants carrying the QTLs (Fig. 5). Also, the desired yield QTL from Maudamani namely OsSPL14 would also be present in these 12 progenies. The common yield component QTLs present in both the parents will be in homozygous state in these progenies. The background analysis using 57 SSR markers in those 12 positive plants detected 92.7 to 96.875% recurrent parent's genome recovery with an average of 94.88% (Table 3). The highest recurrent genome containing plant MSS 128-102-97 was selfed and 31.5g seeds were produced for further 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60  61  62  63  64  65 evaluation in BC3F2 generation. Around one third of the selfed seeds were raised and 618 BC3F2 plants were subjected to foreground screening. Seven plants that were homozygous for Sub1 and GW5 QTLs combinations were selected in the foreground analysis. Also, the yield QTL being inherited from Maudamani parent was checked for its homozygous state. In addition, the inheritances of yield component QTLs Gn1a and SCM2 from both the parents to BC3F2 progenies were also confirmed through the foreground analysis using the gene specific markers.

Analysis of recipient genome recovery on the carrier chromosomes in pyramided lines
The background analysis for recipient genome recovery and genetic drag of donor segment were assessed using 57 polymorphic SSR markers. All the chromosomes were carefully covered with markers in background screening. The were selected to ( Table 2) (Fig. 11).
Gene pyramiding works for transfer of various traits in rice have been published earlier [29,30-40, 46,47]. By using this precision breeding for transfer of trait, pyramids containing Sub1+ OsSPL14+ Gn1a + GW5 (MM allele)+ SCM2 QTLs along with recipient parents' genome of > 95% in the pyramided lines was possible. The undesirable drag expected from the donor genome may come during selection of additional unlinked loci in backcross generations [36]. In our investigation, such effects were detected in the elite pyramided lines while transferring the Sub1 QTL into Maudamani background. The graphical representation of genotyping data as seen in the diagram constructed for the pyramided lines showed the linkage drag in the chromosome carrying carrying the target QTLs (Fig. 11). However, no linkage drag was observed in the chromosome 8 carrying OsSPL14 as the QTL was inherited from the recipient parent and was not from the donor parent. Less linkage drag from donor parents are also reported by earlier researchers in marker-assisted breeding in rice using more background markers [29,[36][37][38][39]. Here the donor parent was a popular variety and hence the drag may not show any undesirable effects in the developed pyramided lines (Fig. 11). Similar findings in other publications suggest the use of improved variety as donor results less or no undesirable drag than the wild and landraces source [29,38,39,42,43] . .   1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60  61  62  63  64 Few elite pyramided lines were as like the recipient parent's important features though variation was seen among the pyramided lines. All the pyramided lines and recipient parent were observed in one quadrant in the biplot diagram drawn based on 16 morpho-quality traits indicate minor variations among the lines (Fig. 10). In addition, the dendrogram drawn based on 15 traits also indicate grouping into main three clusters with similarity within the clusters (Fig. 7A).  Table 4). The transfer of traits and achieving similar or more yield in pyramided lines were also reported earlier in few gene pyramiding publications [29,[30][31][32][33][34][35][36][37][38][39][40].
Biplot diagram places all the pyramided lines in one quadrant along with the recipient parent while the donor parent is quite away and in a separate quadrant. This shows the resemblance among the pyramided and recipient lines and also no undesirable drag from donor parent in the transfer of target genes into the pyramided lines. Also, the placement pattern of few pyramided lines near to origin indicates the stability of the lines for the studied traits over the years. Performance of few pyramided lines was better than the recipient parent in yield, quality and morphologic traits ( Fig. 7; Fig. 8; Table 4). The analysis of background genotyping results showed a higher recovery of recipient parent's genome in few pyramided lines than the expected value in various backcross generations. Again, it revealed that transfer of Sub1 and the yield component QTLsinto one genetic background may not show antagonistic effects for yield and other traits [29,[36][37][38][39]. 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60  61  62  63

Plant materials and breeding program
Rice variety Swarna-Sub1 having Sub1 QTL for submergence tolerance and yield component QTLs, SCM2 and GW5 was used in the hybridization program as donor male parent. The recipient parent, Maudamani is a high yielding variety of eastern India containing the yield component QTLs OsSPL14, Gn1a and SCM2 but shows susceptibility to submergence stress.
One true F1 plant was crossed with recipient parent during rainy season, 2014 to generate BC1F1 generation seeds. True hybridity was checked using the direct Sub1 marker Sub1-A203 as well as a co-dominant marker RM8300. All the BC1F1 seeds were grown for foreground selection of the yield component and submergence tolerance QTLs by using the established molecular markers ( Table 1)

Genomic DNA isolation, Polymerase Chain reaction and Marker analysis
Genomic DNA was isolated following standard protocol [48]. The PCR reaction was performed following the procedure used in the previous publication [29]. The information on the markers used in the polymerase chain reaction are presented in Table 1. Agarose gel electrophoresis was used to separate the amplification products obtained from PCR reactions. The images were 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60  61  62  63  64  65 recorded in a gel documentation system (SynGene, Germany). Eight gene specific and tightly linked markers for the two target QTLs and four recipient's QTL were used for tracking the QTLs in foreground selection (Table 1). A total of 644 SSR markers publicly available were used for the study of polymorphism between the two parents. The polymorphic markers detected were used for background selection (Table 2). Data analysis and dendrogram construction were performed following the standard publications [49][50][51]. Graphical Geno Types (GGT) Version 2.0 software was used to construct the genome recovery graph of recipient parent in the pyramided lines based on the SSR marker data [52].

Consent for publication: Not applicable
Availability of data and material: The data generated or analyzed in this study are included in this article.