*3.3. Regeneration Rate from Microspore Culture*

The microspore cultures were repeated twice. After completing the microspore culture experiment, each embryonic body was induced for about a month, and the induced embryos were confirmed. The average number of embryos was 3.52 in 62 F1 individuals. Only embryos that were not contaminated were selected and cultured on the MS medium. The average number of embryos cultured on the MS medium was 1.71 (Table 3). The regeneration rate was calculated using the formulas shown in the Materials and Methods. As we can see from the averages, the overall regeneration rate was quite low. The individuals without any embryogenesis totaled 31 of the F1 individuals.



1 The average number of embryos from the entire 62 F1 individuals with two replications. 2 The average number of culturable embryos (not contaminated) from the total amount of embryos. 3 Rounded to two decimal places.

### *3.4. QTL Mapping Analysis of Regeneration Rate and the Candidate Genes Related to Microspore Culture*

Using SNP markers obtained through the GBS analysis, linkage maps of 439.2 cM and 420.8 cM, respectively, were constructed, and a QTL analysis was performed by ICIM (inclusive composite interval mapping) by inputting individual phenotypic values corresponding to traits related to the regeneration rate. For the QTL peak of the trait, only those above the threshold value of the trait were considered as significant peaks. From the results of the QTL analysis, four QTL peaks in GX50 were observed (one in chromosome 3, one in chromosome 8, and two in chromosome 9). The LOD values for each QTL are 1.86, 1.79, 1.87, and 1.68, respectively. The additive effect appears to be

negative, indicating that those alleles originate from GX71. The phenotypic variations of peaks were 5.81%, 5.54%, 6.25%, and 5.24%, respectively, and all the values are within a 5–7% range and represent small values (Table 4, Figure 5). In the QTL analysis of GX71, only one QTL peak was observed in chromosome 9. The LOD value is 1.75 and the additive effect is positive, indicating that the allele also originates from GX71. In addition, the phenotypic variation was 13.19%, which was somewhat higher than the GX50 (Table 5, Figure 6). A candidate gene approach was used to determine whether specific loci may explain the responses for the regeneration rate observed in two lines selected for a high regeneration rate and a low regeneration rate. At the position of each QTL, we looked up genes corresponding to 1Mb up and down stream sequences based on left and right markers. Genes were identified in the Radish Genome Database (http://radish-genome.org/), and candidate genes for regeneration were selected based on previous studies (Table 6).

**Table 4.** Effects of the QTLs associated with the regeneration rate detected in F1 populations (GX50).


1 Estimated additive effect of QTL. Negative values indicate effects from GX71; Positive values indicate effects from GX50. 2 Phenotypic variation explained by QTL.



1 Estimated additive effect of QTL. Negative values indicate effects from GX50; Positive values indicate effects from GX71. 2 Phenotypic variation explained by QTL.

**Table 6.** List of *Arabidopsis* orthologs associated with plant regeneration near the QTL region.


1P1 is parent 1, meaning GX50. 2 P2 is parent 2, meaning GX71.

*Genes* **2020**, *11*, 337

**Figure 6.** LOD score plots for chromosome 9 containing QTL with LOD score of 1.75 (GX71). The horizontal lines indicate the thresholds for the LOD.
