Phenotypic data for TSPL and TSPW
Over the 4-year study period, the average TSPL ranged from 3.25 to 3.63 (Table 1). The average TSPW ranged from 0.91 to 0.97 and the CV was between 6.2% and 7.6%. The four QTLs associated with TSPL from 2016 to 2019 were distributed on linkage groups Chrom11, Chrom13, Chrom17, Chrom04 and Gm17 (Table 2). The PVE% (
i.e., the contribution of the QTLs) for all QTLs ranged from 5.17 to 35.29 and the LOD was between 4.41 and 25.68. The QTLs varied in size from 22.24 kb to 54.82 kb. Additionally,
qTSPL-D2-1, which was detected in 2017 and 2018, was distributed between 13,261,056 bp and 13,283,299 bp on the D2 linkage group, with the highest PVE%, LOD and additive effect (35.29, 25.68 and 0.35, respectively) as well as the smallest interval (22.24 kb) in 2018.
From 2016 to 2019, four QTLs affecting TSPW were identified on four linkage groups, namely Chrom13 (1), Chrom16 (1), Chrom19 (1) and Chrom17 (2) (Table 2). In this study, the locations of four QTLs for TSPL and four QTLs for TSPW were determined. The
qTSPL-B1-1 region included a QTL for soybean pod maturity that was previously identified in 2 consecutive years (PVE of 21.1% and 13.7%)
(Lee et al., 2015). The
qTSPL-F-1 region also contained a previously detected QTL for soybean pod maturity
(Bachlava et al., 2009). The
qTSPL-D2-1 and
qTSPW-D2-1 region included a QTL underlying the number of pods
(Li et al., 2010) as well as a QTL for seed weight
(Kuroda et al., 2013). The
qTSPW-F-1 region comprised QTLs related to seed size
(Hyten et al., 2004) and grain weight
(Yan et al., 2014). The
qTSPW-L-1 region included a QTL for seed width
(Salas et al., 2006). Because of the correlation between pods and seeds, these regions may contain important gene(s) controlling pod and seed development. Two QTLs (
qTSPL-D2-1 and
qTSPW-D2-1) were localized to the same position and their additive effects and PVE were higher than those of the other QTLs. Furthermore, the additive effects were positive, implying the two traits are related at both the phenotypic and genetic levels.
Analysis of the epistatic effects on TSPL and TSPW in the CSSL population
A total of 428 loci pairs had epistatic effects on TSPL (Table 3). Some of the QTLs detected during the analysis of epistatic effects were not included among the identified additive effect QTLs (Fig 2). Six pairs of stable QTLs were detected for TSPL and TSPW in 2017, 2018 and 2019. Additionally,
qTSPL-D2-1 and
qTSPW-D2-1, which were revealed as additive effect QTLs, had epistatic effects with non-additive effect QTLs, including
qTSPW-D1b-1,
qTSPW-D1a-1 and
qTSPL-D2-2 (Table 4)
.
Analysis of candidate genes
The additive effect QTL intervals comprised 23 genes (Fig 3, Table 5). The amino sequences encoded by these 23 candidate genes in SN14 and ZYD00006 were predicted and analyzed, which revealed sequence variations between the parents in the proteins encoded by the following seven genes (Fig 4):
Glyma.13G132700,
Glyma.13G132800,
Glyma.13G133200,
Glyma.13G174400,
Glyma.13G174700,
Glyma.16G012500 and
Glyma.19G161700. Twelve genes were identified with variations in the promoter and coding region (Figs 5, 6 and Table 6).
The
Arabidopsis thaliana RGE1 gene encodes a GDSL lipase, which influences plant growth and development. An earlier study indicated this gene is expressed in the endosperm and may regulate embryo development
(Kondou et al., 2008). The
EXL4 gene, which also belongs to the GDSL lipase family, encodes an extracellular lipase localized to the pollen wall, wherein it regulates pollen tube development. Previous research confirmed
Glyma.04G188800 is a WNK family member that regulates the growth of soybean lateral roots through ABA signaling pathways
(Wang et al., 2010). In
A. thaliana, WNK3 is an important kinase for regulating the cell volume and/or intracellular chloride concentration (
Diana and Gerardo, 2011). The
Glyma.19G161700 gene was identified as a homolog of AT2G30580, which belongs to the DREB2A-interacting protein 2 family
(Sakuma et al., 2006). Moreover,
Glyma.11G164700 participates in the phenylpropanoid biosynthesis pathway (
Peer and Murphy, 2007), whereas
Glyma.13G133200 encodes a kinesin family member. A previous study indicated that kinesins help regulate cell division (
Piao et al., 2016). Furthermore,
Glyma.13G174700 is a NAC family gene. The effects of the NAC family on plant drought resistance and the regulation of ABA signaling have been characterized (
Nguyen et al., 2018). Friedrichsen identified three closely related basic helix-loop-helix (bHLH) transcription factors (BEE1, BEE2 and BEE3) encoded by early response genes required for a complete BR response. A comparison of the phenotypes of plants overexpressing
BEE1 and
bee1 bee2 bee3 triple-knockout mutant plants suggested that BEE1, BEE2 and BEE3 are functionally redundant positive regulators of BR signaling. Rybel proposed that an auxin-regulated bHLH transcription factor dimer is a critical regulator of vascular development
(Rybel et al., 2013). In
A. thaliana, the bHLH-type transcription factor AtAIB positively regulates ABA responses
(Li et al., 2007). The
Glyma.19G161800 gene is a member of the
PRA1 gene family. The fact that
PRA1 homologs are ubiquitous in human tissues reflects their importance for development
(Schweneker et al., 2005). The
A. thaliana AtPRA1 genes are expressed in stomata, trichomes and vascular tissues, implying they can contribute to rapid cell expansion and growth. The protein encoded by
Glyma.13G174400 bears a Fork-head-associated (FHA) domain, which is a phosphoprotein-binding sequence in diverse signaling proteins that bind to proteins with phosphorylated threonine or serine residues. The FHA domain of the kinase-associated protein phosphatase in
A. thaliana negatively regulates the receptor-like kinase signaling pathways, which are important for plant development. The pathway associated with
Glyma.13G132700 is K01099 (phosphatidylinositol-bisphosphatase; EC:3.1.3.56), suggesting it plays a key role in the formation of cotyledon veins (Carland and Nelson, 2004).
