*2.6. qDOM3.1 Increased Seed Endogenous ABA Content and ABA Sensitivity*

ABA plays an essential role in the regulation of seed dormancy [4,24,37]. The endogenous ABA level was measured in the near-isogenic line (NIL) of *qDOM3.1* (NIL-NIP) and the corresponding background line (NIL-ZS97). NIL-NIP had an ABA level almost five times higher than that of the NIL-ZS97 (Figure 7).

**Figure 7.** Endogenous ABA concentration in near-isogenic line (NIL)-ZS97 and NIL-NIP.

Subsequently, the ABA sensitivity was investigated in the near-isogenic lines (NIL-NIP and NIL-ZS97) for freshly harvested seeds (Figure 8A) and after-ripened seeds (Figure 8B). The germination percentage and germination speed were significantly lower and slower in NIL-NIP compared with NIL-ZS97 (Figure 8A). Freshly harvested seeds were treated in 43 ◦C for three days to break seed dormancy, and the germination behavior was almost the same for after-ripened NIL-NIP and NIL-ZS97 (Figure 8B). Then, the pair of near-isogenic lines (after-ripened) were treated in a series of ABA solution to investigate ABA sensitivity. Up to 10μM ABA had no significant effect on seed germination, whereas 20–100μM ABA significantly decreased seed germination of NIL-NIP compared with NIL-ZS97 (Figure 8C). Thus, *qDOM3.1* was sensitive to ABA treatment. Therefore, we hypothesized that the target region may contain an ABA responsive gene.

**Figure 8.** (**A**) Germination behavior of freshly harvested NIL-ZS97 and NIL-NIP seeds; (**B**) germination behavior of after-ripened NIL-ZS97 and NIL-NIP seeds; (**C**) ABA sensitivity of NIL-ZS97 and NIL-NIP.

#### *2.7. Candidate Gene Expression Changes Upon ABA Treatment*

For that reason, the gene expression of eight candidates was measured upon ABA treatment (20 μM), using after-ripened NIL-NIP and NIL-ZS97. In total, four out of eight candidate genes were ABA responsive genes (Figure 9). LOC\_Os03g01442's expression level had no significant difference in non-treated NIL-NIP and NIL-ZS97 (CK); however, upon ABA treatment, the expression level was significantly higher in NIL-NIP than NIL-ZS97. LOC\_Os03g01540 had the opposite effect, which was lower in non-treated NIL-NIP than in NIL-ZS97 and, upon ABA treatment, the expression level had no difference in NIL-NIP and NIL-ZS97. LOC\_Os03g01530's expression level was significantly increased upon ABA treatment. LOC\_Os03g01470's expression level was significantly higher in non-treated NIL-NIP than NIL-ZS97, and, after ABA treatment, the expression level was significantly lower in NIL-NIP compared with NIL-ZS97. The other four candidate genes had the same trend before and after the ABA treatment.

**Figure 9.** Relative mRNA abundance level of eight candidate genes under *qDOM3.1* using after-ripened NIL-ZS97 and NIL-NIP before (CK) and after 20 μM of abscisic acid (ABA) treatment (ABA). \* and \*\* indicate significant differences at *p* < 0.05 and *p* < 0.01 using the Student's *t*-test of NIL-NIP against NIL-ZS97, respectively. CK of NIL-ZS97 in each figure was determined as 1.

Therefore, if the expression of the candidate gene was changed by the addition of ABA, they were unlikely to be our candidate genes. Therefore, there were four candidate genes under *qDOM3.1*, including two expressed proteins (LOC\_Os03g01442 and LOC\_Os03g01470), one tubulin/FtsZ domain-containing protein (LOC\_Os03g01530), and one DNA-binding protein (LOC\_Os03g01540). However, we cannot rule out the posttranslational modifications of the candidate gene, such as phosphorylation/dephosphorylation. More experimental evidence is needed.
