Phenotypic variation in the RIL population
Grain number (GN), setting grain number (SN) and SF showed a significant difference among the parents (Table 1). There is a significant difference of heat tolerance index(HT) between LD5 and ZYZ8 under high temperature environments. GN, SN, SF and HT of the RILs varied exhibited an approximate continuous normal distribution under different conditions. Meanwhile, the transgressive segregation that fell beyond the parents was observed, suggesting that the set of RILs was suitable for QTL analysis (Table 2, Fig. 1). In addition, natural and artificial high temperature have different degrees of effect on SF and HT. The latter has strong heat stress, and then the heat tolerance mechanism from different treatments maybe different.
Table 2
Heat tolerance-related traits from the parents and RILs populations
Year
|
Treatments
|
Trait
|
Parents
|
|
RILs populations
|
LD5
|
ZYZ
|
Mean ± Sd
|
Range
|
Skewness
|
Kurtosis
|
2019
|
Control
|
GN
|
78.40 ± 5.60
|
256.20 ± 15.60**
|
|
143.29 ± 44.80
|
50.80-290.20
|
0.58
|
0.65
|
|
|
SN
|
65.40 ± 4.00
|
166.22 ± 10.34**
|
|
85.54 ± 34.38
|
24.80-223.20
|
0.58
|
0.65
|
|
|
SF
|
83.42 ± 5.65
|
64.88 ± 10.45**
|
|
60.22 ± 16.71
|
36.50–100.00
|
-0.47
|
0.05
|
|
Nature high temperature
|
SN
|
50.40 ± 5.65
|
127.07 ± 10.32**
|
|
111.83 ± 34.81
|
28.60–218.00
|
0.42
|
0.03
|
|
|
SF
|
63.47 ± 6.40
|
58.47 ± 12.30*
|
|
70.59 ± 14.91
|
27.68–96.32
|
-0.52
|
-0.37
|
|
Artificial high temperature
|
SN
|
24.25 ± 5.20
|
54.14 ± 8.30**
|
|
35.11 ± 16.85
|
4.53–86.54
|
0.42
|
0.19
|
|
|
SF
|
26.23 ± 5.00
|
31.24 ± 10.20*
|
|
28.44 ± 12.66
|
1.44–65.05
|
0.32
|
-0.04
|
|
|
HTN
|
0.76 ± 0.09
|
0.90 ± 0.07*
|
|
0.36 ± 0.39
|
0.08–0.87
|
-0.40
|
1.24
|
|
|
HTA
|
0.31 ± 0.06
|
0.48 ± 0.04*
|
|
0.52 ± 0.20
|
0.03–0.91
|
-0.29
|
-0.70
|
2020
|
Control
|
GN
|
112.40 ± 10.20
|
211.00 ± 20.50**
|
|
160.74 ± 45.20
|
82.20-284.20
|
0.45
|
-0.18
|
|
|
SN
|
88.80 ± 8.50
|
146.10 ± 15.00**
|
|
111.83 ± 34.81
|
28.60–218.00
|
0.42
|
0.03
|
|
|
SF
|
79.00 ± 8.20
|
69.24 ± 10.25**
|
|
74.10 ± 11.75
|
34.78–96.32
|
-0.44
|
-0.12
|
|
Nature high temperature
|
SN
|
71.81 ± 12.00
|
101.67 ± 15.00**
|
|
106.47 ± 34.26
|
10.00-210.57
|
0.28
|
0.18
|
|
|
SF
|
61.71 ± 5.50
|
54.95 ± 10.40*
|
|
61.69 ± 13.84
|
11.32–87.93
|
-0.51
|
0.46
|
|
Artificial high temperature
|
SN
|
44.75 ± 8.50
|
63.60 ± 10.50**
|
|
54.22 ± 30.83
|
7.80-193.60
|
1.45
|
3.59
|
|
|
SF
|
28.87 ± 5.50
|
35.90 ± 3.00*
|
|
36.52 ± 13.20
|
5.01–68.72
|
0.09
|
-0.37
|
|
|
HTN
|
0.78 ± 0.08
|
0.80 ± 0.05*
|
|
0.84 ± 0.18
|
0.18–1.19
|
-0.41
|
0.53
|
|
|
HTA
|
0.37 ± 0.04
|
0.51 ± 0.06*
|
|
0.50 ± 0.18
|
0.07–1.03
|
0.13
|
-0.23
|
Grain number, GN; Setting grain number, SN; Spikelet fertility, SF; Heat tolerance under nature high temperature, HTN; Heat tolerance under artificial high temperature, HTA; *and ** represent significant differences at the 5% and 1% level, respectively. |
There were significantly positive correlations for all traits between two years, as shown by correlation coefficients range from 0.146–0.417(Table 3). As expected, SF was significant positively correlated with SN, and negatively correlated with GN under different environments. There were significantly positive correlations between SF and HT under high temperature environments. In addition, strong and significant correlations were found in GN, SN, SF and HT under different temperature environments in two years. The correlation analysis demonstrated that these traits were stable and had high heritability from the control, natural and artificial high temperature environments.
Table 3. Correlation coefficients among SF in RILs populations under different environments
Treatment
|
Traits
|
Control
|
|
Natural high temperature
|
|
Artificial high temperature
|
HTN
|
HTA
|
|
GN
|
SN
|
SF
|
SN
|
SF
|
|
SN
|
SF
|
HTN
|
Control
|
GN
|
0.310**
|
0.715**
|
-0.244**
|
|
0.155*
|
-0.172*
|
|
0.171*
|
-0.098
|
-
|
-
|
|
SN
|
0.685**
|
0.245**
|
0.335**
|
|
0.208**
|
0.251**
|
|
0.215**
|
0.001
|
-
|
-
|
|
SF
|
-0.085
|
0.307**
|
0.146*
|
|
0.116
|
0.445**
|
|
0.058
|
0.213**
|
-0.310**
|
-0.229**
|
|
Natural high temperature
|
SN
|
0.017
|
0.159*
|
0.238**
|
|
0.189**
|
0.355**
|
|
0.387**
|
0.059
|
-
|
-
|
|
SF
|
-0.040
|
0.452**
|
0.282**
|
|
0.245**
|
0.194**
|
|
0.147*
|
0.273**
|
0.702**
|
0.282**
|
|
Artificial high temperature
|
SN
|
0.131
|
0.210**
|
0.132
|
|
0.169*
|
0.138
|
|
0.173*
|
0.509**
|
-
|
-
|
|
SF
|
-0.195**
|
0.141
|
0.270**
|
|
0.027
|
0.280**
|
|
0.671**
|
0.330**
|
0.485**
|
0.888**
|
|
HTN
|
-
|
-
|
-0.426**
|
|
-
|
0.730**
|
|
-
|
0.061
|
0.417**
|
0.268**
|
|
HTA
|
-
|
-
|
-0.149*
|
|
-
|
0.132
|
|
-
|
0.874**
|
0.223**
|
0.197**
|
|
Lower and upper triangle data represent for 2019 and 2020; *and ** represent significant differences at the 5% and 1% level, respectively.
QTL mapping within the RIL population
A total of sixty-one QTLs were detected using the ICIM-ADD method, 25, 27 and 14 additive QTLs were identified under the control, natural and artificial high temperature conditions, respectively(Table 4, Fig. 2). These QTLs were located on chromosomes 1–8, 10 and 12, explained around 3.18–42.31% of the phenotype variation. Twenty-five QTL were associated with the SF-related traits under control environments, 13 and 12 QTLs for 2019 and 2020, respectively. Among these QTLs, six QTLs were commonly detected in both years, the other QTLs only can be found in one environment. Ten QTLs associated with GN were identified on chromosomes 1, 4, 6, 8 and 10, these QTLs individually accounted for 4.82–15.84% of the phenotypic variance. Among these QTLs, only qGN1a, qGN4b and qGN6 were identified in different years, while the major QTL qGN1a has the highest phenotypic variation and additive effect. Nine putative QTLs associated with SN were detected on chromosomes 1, 4, 5, 6 and 8; among these, qSN1a, qSN1b and qSN4 were identified across two years. Six QTLs for SF were identified, and all of these QTLs only can be found in one environment; qSF2 and qSF4 have a higher phenotypic variation score, which contributed to 13.80% and 37.92% of phenotypic variation, respectively.
Table 4
Putative QTLs for spikelet fertility related traits were detected in RILs population
Treatment
|
Trait
|
Locus
|
Marker
|
LOD value
|
|
PVE (%)
|
|
Additive effect
|
2019
|
2020
|
|
2019
|
2020
|
|
2019
|
2020
|
Control
|
GN
|
qGN1a
|
RM428-RM323
|
6.13
|
5.80
|
|
11.48
|
15.84
|
|
20.07
|
18.13
|
|
|
qGN1b
|
RM1198-RM1361.1
|
|
2.85
|
|
|
7.75
|
|
|
-16.91
|
|
|
qGN4a
|
RM471-RM1359
|
|
5.86
|
|
|
14.99
|
|
|
-19.47
|
|
|
qGN4b
|
RM2441-R4M50
|
3.50
|
2.48
|
|
8.44
|
4.59
|
|
-13.55
|
-9.80
|
|
|
qGN6
|
RM111-STS6.2
|
2.49
|
4.12
|
|
7.97
|
13.26
|
|
-16.61
|
18.08
|
|
|
qGN8
|
RM3754-RM3496
|
|
3.80
|
|
|
10.83
|
|
|
-14.85
|
|
|
qGN10
|
R10M40-STS10.3
|
3.03
|
|
|
4.82
|
|
|
13.19
|
|
|
SN
|
qSN1a
|
RM6902-RM428
|
6.89
|
2.91
|
|
9.41
|
6.88
|
|
18.39
|
9.13
|
|
|
qSN1b
|
R1M37-RM3738
|
3.94
|
3.22
|
|
7.39
|
9.25
|
|
15.33
|
-13.92
|
|
|
qSN4
|
RM1359-R4M43
|
2.82
|
5.82
|
|
3.18
|
18.84
|
|
-10.35
|
-16.90
|
|
|
qSN5
|
R5M13-RM3476
|
4.45
|
|
|
10.86
|
|
|
18.29
|
|
|
|
qSN6
|
RM217-RM111
|
|
2.66
|
|
|
5.85
|
|
|
9.39
|
|
|
qSN8
|
RM3754-RM3496
|
|
2.70
|
|
|
7.79
|
|
|
-9.70
|
|
SF
|
qSF1
|
RM8097-STS1.4
|
3.14
|
|
|
7.19
|
|
|
-5.05
|
|
|
|
qSF2
|
STS2.4-RM13603
|
6.58
|
|
|
13.80
|
|
|
-9.00
|
|
|
|
qSF3
|
MM3720-MM3778
|
|
3.07
|
|
|
9.52
|
|
|
4.67
|
|
|
qSF4
|
RM5688-RM471
|
|
2.59
|
|
|
37.92
|
|
|
-9.97
|
|
|
qSF10
|
RM3451-RM590
|
3.05
|
|
|
4.37
|
|
|
5.00
|
|
|
|
qSF12
|
RM247-RM7003
|
4.18
|
|
|
5.78
|
|
|
5.66
|
|
Nature high temperature
|
SN
|
qSN1
|
RM7180-RM6703
|
2.68
|
3.45
|
|
4.27
|
9.89
|
|
-9.82
|
-14.08
|
|
|
qSN2a
|
RM3732-RM1361.2
|
|
3.06
|
|
|
6.88
|
|
|
12.07
|
|
|
qSN2b
|
STS2.3-STS2.4
|
|
2.60
|
|
|
6.04
|
|
|
8.55
|
|
|
qSN3
|
STS3.6-STS3.7
|
|
4.20
|
|
|
11.42
|
|
|
13.41
|
|
|
qSN6
|
RM111-STS6.2
|
2.51
|
|
|
4.28
|
|
|
9.47
|
|
|
|
qSN12
|
RM101-STS12.1
|
2.59
|
|
|
8.42
|
|
|
-12.03
|
|
|
SF
|
qSF2
|
STS2.4-RM13603
|
2.74
|
|
|
5.98
|
|
|
7.55
|
|
|
|
qSF3
|
RM1350-RM3199
|
3.26
|
|
|
9.39
|
|
|
4.98
|
|
|
|
qSF7
|
R7M37-RM1132
|
2.90
|
|
|
7.43
|
|
|
-4.16
|
|
|
|
qSF12
|
RM7003-RM101
|
2.89
|
2.50
|
|
6.40
|
11.20
|
|
-7.57
|
-9.32
|
|
HTN
|
qHTN2
|
RM13603-ID2
|
2.66
|
|
|
7.06
|
|
|
0.11
|
|
|
|
qHTN12
|
RM247-RM7003
|
3.34
|
2.60
|
|
7.13
|
11.45
|
|
-0.10
|
-0.15
|
Artificial high temperature
|
SN
|
qSN4
|
R4M43-RM3288
|
2.90
|
3.73
|
|
10.54
|
9.88
|
|
-9.50
|
-8.83
|
|
|
qSN7
|
RM5055-RM5711
|
|
2.67
|
|
|
8.20
|
|
|
9.57
|
|
|
qSN10
|
RM3451-RM590
|
2.74
|
|
|
6.68
|
|
|
4.72
|
|
|
SF
|
qSF2
|
STS2.4-RM13603
|
|
2.92
|
|
|
42.31
|
|
|
-10.04
|
|
|
qSF3
|
RM3199-RM1352
|
|
2.53
|
|
|
6.61
|
|
|
3.95
|
|
|
qSF6
|
RM1340-R6M44
|
2.79
|
|
|
6.55
|
|
|
3.80
|
|
|
|
qSF7
|
RM8261-RM1209
|
|
2.53
|
|
|
6.17
|
|
|
-3.84
|
|
|
qSF8
|
RM3754-RM3496
|
|
2.88
|
|
|
10.11
|
|
|
-4.87
|
|
|
qSF12
|
RM7003-RM1337
|
3.01
|
2.64
|
|
9.50
|
10.23
|
|
4.75
|
4.91
|
|
HTA
|
qHTA1
|
R1M30-RM3240
|
2.70
|
2.65
|
|
7.54
|
5.07
|
|
0.10
|
0.07
|
|
|
qHTA3
|
MM3720-MM3778
|
|
2.81
|
|
|
6.73
|
|
|
-0.08
|
|
|
qHTA6
|
RM111-STS6.2
|
2.63
|
|
|
5.31
|
|
|
-0.09
|
|
|
|
qHTA12
|
RM7003-RM1337
|
|
2.59
|
|
|
12.64
|
|
|
-0.20
|
Twenty-seven QTLs were associated with heat tolerance-related traits under natural high temperature environments. Seven additive QTLs identified for SN were detected on chromosomes 1, 2, 3, 6 and 12, except for qSN1 which was found in both years and explaining − 9.82 and − 14.08 of the phenotypic variation, all other QTLs were only detected in one environment. Four QTLs associated with SF, namely qSF2, qSF3, qSF7 and qSF12, were detected on chromosomes 2, 3, 7, and 12 respectively. Among these QTLs, the stable expressed major QTL qSF12 has a higher LOD value and phenotypic variation, the other QTLs can be detected only in one environment. Two QTLs associated with HTN were detected, among which qHTN12 with the greatest effect on phenotypic variation was collocated with the QTL qSF12.
Sixteen QTLs associated with the heat tolerance-related traits were identified under artificial high temperature environments, 6 and 10 QTL in 2019 and 2020, respectively. Three QTLs associated with SN were detected on chromosomes 4, 7 and 10, and these QTLs can be identified in one environment, which explained 6.68–9.88% of phenotypic variance. Seven putative QTLs associated with SF were detected on chromosomes 2, 3, 6, 7, 8, and 12. Among these QTLs, only qSF12 was located in two years. Four QTLs associated with HTA were detected, that is qHTA1, qHTA3, qHTA6 and qHTA12 located on chromosomes 1, 3, 6 and 12, respectively. Among these QTLs, only qHTA1 can be found in two years.
Our results reveal that pleiotropic effects and QTL hotspots are the key factors affecting GN and SF in rice, and then these traits may have a similar genetic regulation. We could assign the 64 QTLs to 15 chromosome regions, each corresponding region contains about 4 QTLs. Seven major QTL clusters were assembled on the chromosomes 1, 2, 3, 4, 6 and 12, which contain 38 QTLs, accounting for 59.38% of the whole locus. These related QTLs were identified in previous studies. Among these QTLs, clusters qHTSF1a, qHTSF1b, qHTSF4 and qHTSF12 were stably detected (Table 5, Fig. 2).
Table 5
Heat tolerance-related traits of QTL hotspot and pleiotropic region
Major QTL cluster
|
Chr.
|
Interval
|
Pleiotropic QTL
|
Related reports
|
qHTSF1a
|
1S
|
RM3740-RM323
|
qGNC1a,qSNC1a,qGNN1,qGNA1
|
Gn1a (Ashikari et al.,2005)
|
qHTSF1b
|
1L
|
RM3240-RM1198
|
qGNC1b,qSNC1b,qSFC1,qSNN1,qHTA1
|
qHBT1-1 (Cao et al., 2020)
|
qHTSF2
|
2
|
STS2.4-RM13603
|
qSNC2,qSFC2,qSNN2,qSFN2,qHTN2,qSFA2
|
-
|
qHTSF3
|
3
|
R3M30-RM1350
|
qSFC3,qSFN3,qSFA3,qHTA3
|
qSF3.2 (Nubankoh et al., 2020)
|
qHTSF4
|
4
|
RM471-RM3288
|
qGNC4,qSNC4,qSFC4,qGNN4,qGNA4,qSNA4
|
qHTS4.1 (Ye et al., 2015b)
|
qHTSF6
|
6
|
RM111-STS6.2
|
qGNC6,qSNC6,qGNN6,qSNN6,qHTA6
|
qHTSF6.1 (Raza et al., 2020)
|
qHTSF12
|
12
|
RM247-RM101
|
qSFC12,qSNN12,qSFN12,qHTN12,qSFA12
|
-
|
The bold represent the stable QTL under different environments, respectively. |
In this study, 17 pairs of epistatic interaction QTLs for heat tolerance-related traits were detected (Table 6). All these QTLs are located on chromosomes 1, 2, 3, 4, 5, 6, 7, 8, 9 and 12, and LOD value of these QTLs range from 6.7 to 12.9, accounting for a wide phenotypic variation ranging from 8.32–28.09%. 13 out of 17 epistatic QTL pairs were nearby the additive QTL. For example, the genomic region RM5688-RM471 on the chromosome 4 is interacted with the chromosomes 2, 3, 5,7, 8,11 and 12; the genomic region RM1361-RM6321 on chromosome 1 epistatic effection on the chromosomes 2,4,8 and 12. There were 8 pairs of epistatic interaction QTL for HT (HTN and HTA), no epistatic QTL of epistatic interaction for SF under natural and artificial high temperature stress. Therefore, epistatic QTLs played an essential role in controlling the genetic expression of the heat tolerance gene.
Table 6
Identification and analysis of the epistatic interaction QTL under different temperature environments
Environments
|
Trait
|
Chr.i
|
Marker1
|
Chr.j
|
Marker2
|
LOD
|
PVE(%)
|
AA
|
Control
|
GN
|
1
|
STS1.4-RM7180
|
9
|
RM7048-STS9.1
|
5.91
|
12.39
|
17.90
|
|
SF
|
1
|
RM1361-RM6321#
|
4
|
R4M43-RM3288
|
7.29
|
23.35
|
-9.70
|
Nature high temperature
|
SN
|
12
|
RM247-RM7003#
|
6
|
RM587-RM217
|
5.95
|
16.62
|
-30.98
|
|
HTN
|
1
|
RM1198-RM1361#
|
2
|
STS2.4-RM13603
|
9.64
|
15.09
|
-0.43
|
|
|
1
|
RM1198-RM1361#
|
12
|
STS12.2-RM1226
|
12.19
|
13.69
|
0.43
|
|
|
4#
|
RM5688-RM471
|
2
|
RM5699-RM1358
|
11.15
|
8.32
|
-0.43
|
|
|
4#
|
RM5688-RM471
|
11
|
RM21-STS11.4
|
10.21
|
18.23
|
0.39
|
Artificial high temperature
|
SN
|
1
|
STS1.3-RM306
|
2
|
RM7451-STS2.1
|
6.33
|
13.36
|
-19.89
|
|
|
1
|
R1M30-RM3240
|
3
|
RM3199-RM1352
|
6.38
|
15.69
|
-23.34
|
|
|
4#
|
RM5688-RM471
|
5
|
R5M13-RM3476
|
7.09
|
19.92
|
26.81
|
|
|
4#
|
RM5688-RM471
|
7
|
RM1132-RM8261
|
6.85
|
23.04
|
-31.10
|
|
|
4#
|
RM5688-RM471
|
12#
|
RM19-RM247
|
7.34
|
25.68
|
-31.36
|
|
|
7
|
RM5711-STS7.1
|
8
|
RM5556-RM6208
|
6.45
|
21.45
|
-24.61
|
|
HTA
|
1
|
RM1361-RM6321#
|
8
|
RM1376-RM4085
|
20.71
|
26.72
|
-0.72
|
|
|
4#
|
RM5688-RM471
|
7
|
RM3826-R7M37
|
22.56
|
25.12
|
-0.70
|
|
|
4#
|
RM5688-RM471
|
8
|
RM6208-RM3395
|
21.26
|
22.85
|
-0.69
|
|
|
8
|
RM1376-RM4085
|
12#
|
RM19-RM247
|
18.55
|
28.09
|
-0.71
|
# represents that the additive QTL chromosome region have an epistatic interaction. |
Identify the major QTLs through BSA
BSA method was proved to be a convenient method to the detection of major QTLs, in particular to identification of QTLs for adversity stresses such as drought, salinity, heat and cold (Zhang et al., 2009; Sandhu et al.,2014; Sun et al. 2018; Wu et al., 2019). In this study, SF and HT were used to evaluate the heat tolerance, two extreme bulks, including the heat-tolerant bulk with higher SF and HTA, and the heat-sensitive bulk with lower SF and HT under artificial high temperature conditions. 10 lines with heat tolerance and sensitivity were selected in the F5:6 populations (510 lines), respectively. The SF and HTA were significantly lower in the S-bulk than in the T-bulk, indicating that the overall phenotypic of the S-bulk was greatly affected by high temperature (Fig. 3). 24 markers have polymorphism from two bulks and the parents, which have clear correspondence with phenotype. Out of these markers, 11 markers were together located on the chromosomes 1, 2 and 12, namely three putative QTLs, qSF1, qSF2 and qSF12, these QTLs exhibited the contrasting genotype between the T-bulk and S-bulk, respectively (Fig. 4A). Through BSA, it was found that these QTLs revealed the highly significant correlation among these addictive QTLs through ICMI mapping methods, which further confirmed the reliability of the detected QTLs (Fig. 4B).