Regardless of the variation of different starch components, both soft and normal rice are estimated with almost 90% TS which is similar to the rest of rice (Kim et al., 2021). Low and high AC content in studied rice lines (soft and normal lines ) also showed low and high RS content, respectively, which indicates that AC has a great role in determination of RS content in rice (Shen et al., 2022). The lower amount of AC with a higher amount of AP in soft rice is in contrast to normal rice having high AC and low AP which reconfirm their japonica and indica nature, respectively (Panja et al., 2022, Vyas, 2020, Inukai, 2017). Due to low GT, soft rice required short time to cook in both CTBT and CTRT in comparison to normal rice but only soft rice showed RTE trait because of its low AC and also for its special types of physicochemical properties of starch as reported in our earlier work (Panja et al., 2022). Additionally, soft and normal rice showed soft and medium gel length, respectively due to their different amount of AC which has strong role in determining these quality traits (Yu and Wang, 2007, Juliano, 1979). With respect to rice reference genome (IRGSP 1.0, Oryza sativa Japonica Group), soft rice showed less number of overall variation in contrast to normal rice, which also indirectly supports the closeness of soft rice with japonica rice, as also represented in Kinship analysis. But variant analysis with 65 SSRGs associated with target traits showed that concerning rice reference genome number of variants was higher in soft rice than normal rice, which revealed that soft rice has unique type of sequence diversity for SSRGs, which is different from both indica and japonica rice and possibly due to long course of evolution these type of diversity in SSRGs occurred to acquire softness in rice. But in comparison to normal rice, soft rice possesses fewer variants in WX1 and ALK loci encoding GBSSI and SSIIa, which are exclusively responsible for AC and AP synthesis in developing endosperm (Kumar et al., 2018, Fujita et al., 2022) indicating the presence of unique sequence diversity in soft rice that results in low AC, RS and high AP content. It can be hypothesized that this soft glutinous rice, possibly originated from Thailand and Burma, migrated to Majuli Island of the upper Brahmaputra basin, Assam (Rajak and Bhuyan, 2013, Panja et al., 2023b) and became endemic to that region. Under these restricted agro-climatic conditions (particularly moisture, temperature and day length period during grain filling) they might have evolved the RTE trait. Transcriptome analysis of developing grains of soft and normal rice also indicated that with respect to normal rice, soft rice has special starch synthesis machinery functioning during grain filling to achieve all its unique type of starch components, which resulted in its unique RTE and GR through down-regulating most of the SSRG loci. With respect to normal rice, significant downregulation of GBSSI loci in soft rice proved its possible implication for low AC content, which affects not only the requirement of less cooking time but also for lower RS content. As the maximum number of significantly regulated DEGs are involved in carbohydrate metabolism related pathways, it confirmed that carbohydrate metabolism has a prime role in grain development (Hong et al., 2021) and compared to normal rice, regulation of these metabolism related genes is very unique in soft rice. This uniqueness in regulation may be responsible for its special RTE trait, which differs much from normal rice with no RTE trait. Despite two different types of rice transcriptome, both showed five major types AS events after fertilization and A3SS and IR events were the most common AS events during grain filling in rice, a finding quite similar to that of Li et al., (2021). During grain development, differentially expressed transcripts of soft rice showed less number of variants about starch synthesis machinery of japonica rice, which again supports the japonica origin of soft rice. Large number of identified SSRs with 150bp flanking sequences can be utilized for development of SSRs-based functional molecular markers essential for molecular breeding program of soft rice. AD in basal leaf sheath and stigma of inbreed lines are supposed to be inherited from P2 plant as it possesses AD in different parts of the body and P1 is devoid of AD. In comparison to presence of awn in grains, AD is very dominant trait in soft rice and due to its dominancy, all inbreed lines (F1) showed AD whereas awn was only inherited when P2 is the recipient plant (♀). Segregation analysis of RILs also suggests the same inheritance pattern by showing ¾ F2 populations with AD in immature spikelets. Both the SSRs (RM190 and RM332) may be used in soft rice breeding due to their co-segregation with the targeted traits. EF and short stature were two desirable traits inherited in RILs from HY normal rice, which were required for local adaptation for early heading (Hori et al., 2016) and development of high yielding lodging resistance rice (Long et al., 2020). Improvement of yield related traits in generated breeding lines (RILs) and the resulting hybrid vigor are the most optimistic opportunities for conducting breeding programs for production of soft rice on large scale. In early flower types of inbreed plants improved AC content was supposed to be inherited from high AC containing normal rice, which also results in elevation of the RS. But RILs with more than 20% AC and white stigma failed to show RTE trait. Irrespective of zygosity for softness linked marker (RM190) RILs showed RTE trait only when AC content was low (< 20%) and stigma was purple. But one type of RILs (type VII) with white stigma did not show RTE trait despite low AC (11.57%), which indicates that in addition to low AC during post hybridization period, there may be a coinheritance with AD deposition in stigma and softness. Among eight types of EF plants, only three inbreed lines showed promising results for RTE trait. However, among them, only two (type IV and type VIII) had rare phenotypic combinations (only tip of the lemma and palea pigmented) and improved AC and RS, which indicates that soft trait is not a dominating trait, but a unique trait under polygenic control and development of hybrid soft rice with low GR will be very difficult and rare. After 30 min of soft rice consumption sharp increase and from 60 to 90 min sharp decline of PP blood glucose was noted, which is significant with respect to normal rice and showed the characteristic of high GR. On the other hand, after 30 min of consumption, type VIII inbreed lines showed less significant increment and then up to 120 min no significant decrease of PP blood glucose level as compared to normal rice showing moderate GR. These results clearly indicate that the improvement of GR in the inbreed line is due to upliftment of AC and RS content. To achieve low GR hybrid rice with RTE trait higher increase of AC and RS content is required but due to high AC content, these inbreed lines may lose their RTE trait. Thus inbreed plants (type VIII) with moderate AC, RS, GR and promising RTE trait will be best for generation improvement to produce stable population of hybrid lines for future GI test in humans.
From the present study, we can conclude that due to lower amylose content, soft rice showed less cooking time but low AC is associated with lower amount of RS which results in high glycemic response and is very unhealthy for consumption as a readymade food. Though soft rice is a japonica type of rice, it has unique type of sequence diversity in SSRGs, which is different from both Japonica and Indica rice with no RTE trait, but fewer variants in WAXY locus revealing the Japonica ancestry of soft rice. Down-regulation of GBSSI along with most of the SSRGs highlighted the possible mechanism of the unique type of grain filling present in soft rice for its high GR and RTE trait. Findings from molecular breeding indicated that anthocyanin deposition in immature spikelets, especially in stigma is a very dominant trait with RTE trait among RILs. Though different types of Waxy rice under japonica subgroups possess very less amounts of AC (0–2%) they are unable to show the RTE trait. However, the studied rice population having the AC content within a specified range (3–20%) showed RTE trait with a variable glycemic response. Thus molecular breeding of soft rice in higher yield background is a very sensitive task, where the AC content should not be higher than 20% to retain the RTE trait with moderate GR. Considering this tricky issue, instead of backcrossing, development of MAGIC population should be attempted to generate hybrid soft rice with a low glycemic index (GI) through the inclusion of an additional parental line (with low GI) to hybridize with already existing hybrid lines of this present study. Generated RILs may be used for QTL mapping linked with target trait (RTE and GR) through the involvement of more polymorphic markers in addition to RM190 and RM332. At present developed inbreed lines with moderate GR and high yielding characteristic needs to get generation improvement for large scale production of ready-to-eat future food with improved glycemic response.