Identification of major QTLs associated with agronomical traits and candidate gene mining in soybean

Abstract

Soybean is valued in foods, animal feed and some industrial applications, and agronomic traits are important targets for soybean breeding. In this study, two sets of recombinant inbred line (RIL) populations were used to map yield and quality trait quantitative trait loci (QTLs) across three environments. In total, we detected 37 QTLs in two RIL populations grown in three locations. In the 3613 RIL population, three QTLs related to quality traits and 13 related to yield traits were identified among the 16 QTLs. In the 6013 population, we identified six QTLs related to quality traits and 15 related to yield traits. Among these QTLs, some genomic regions showed pleiotropic effects on yield and quality traits. We further identified candidate genes from these regions. Our results lay the foundation for producing high-yielding soybean with excellent quality traits via molecular breeding.

Keywords:

• Soybean
• agronomic traits
• QTL mapping
• candidate gene mining

Introduction

Soybean [Glycine max (L.) Merr.] is one of the most important legume crops worldwide, with extensive acreage in the USA, Brazil, China and other countries [1]. Soybean has several beneficial properties, including a 20% oil content, 40% protein content and high fatty acid levels [2]. Therefore, soybean represents an important source of raw materials for the production of plant protein and vegetable oil products, as well as industrial and pharmaceutical products [3]. Although soybean breeders focus on increasing yield traits and improving quality traits, these quantitative traits are controlled by multiple genes, and the underlying genetic mechanisms are relatively complex [4]. Traditional breeding methods based on phenotypic screening have several disadvantages, such as the large amount of labour and materials required, low efficiency and long operating cycles. Therefore, the use of molecular marker technology to construct soybean genetic linkage maps has become a growing area of study. This method, which involves detecting the degree of linkage between molecular markers and related quantitative trait loci (QTLs) underlying useful traits, can greatly facilitate soybean breeding efforts [5].

Molecular genetic maps of soybean have been continuously improving, as they are required for QTL mapping of soybean quality- and yield-related traits. In the late 1980s, Apuya et al. [4] constructed the first molecular genetic linkage map of soybean, containing 11 restriction fragment length polymorphism (RFLP) markers. In 1990, Keim et al. [5] constructed a second soybean linkage map, which covered 1200 cM of the soybean genome with 130 RFLP markers. A map was constructed with 490 markers covering 3000 cM of the soybean genome. However, the lack of integration of conventional markers limited the use of these of molecular marker maps for QTL localisation. In 1995, Shoemaker and Specht [6] used an F₂ segregating population obtained via hybridisation of the near-isogenic soybean lines Clark and Harsoy and constructed the first soybean ‘consensus linkage map’. Cregan et al. [7] integrated the genetic linkage maps of three populations and obtained a common map consisting of 26 classical and 10 isozyme loci along with 79 random amplified polymorphic DNA (RAPD), 689 RFLP and 606 simple sequence repeat (SSR) loci. Song et al. [8] generated a genetic linkage map by integrating five soybean genetic maps, including those for two F2 populations (‘A81-56022’ × ‘PI468916’ and ‘Clark’ × ‘Harosoy’) and three recombinant inbred line (RIL) populations (‘Minsoy’ × ‘Noir1’, ‘Minsoy’ × ‘Archer’, and ‘Noir1’ × ‘Archer’). In the past 10 years, as genome sequencing efforts have advanced, many high-density genetic maps have been constructed for soybean [9–11].

To date, many QTLs underlying agronomic traits in soybean have been mapped and deposited in the SoyBase database (www.soybase.org), including 322 seed oil QTLs, 240 seed protein QTLs, 318 seed weight QTLs, 255 plant height QTLs, 24 seed weight per plant QTLs and 48 pod number QTLs. However, to date, only a few QTLs affecting multiple agronomic traits have been mapped in crops. Gahlaut et al. [12] performed QTL mapping of wheat double haploid populations planted in four different locations in India for 3 years and identified essential genomic regions carrying multiple QTLs for different traits on the 5A and 7A linkage groups. The locus associated with lodging in maize identified by Larsson et al. [13] coincides with the positions of previously identified QTLs for stem strength. Using two F_2:3 populations of maize under drought stress, Zhao et al. [14] identified QTLs controlling multiple traits at several locations. Wu et al. [15] integrated QTLs from soybean at different growth stages and identified 10 QTLs controlling multiple growth stages in five linkage groups (C2, D1a, D1b, F and J). Hacisalihoglu et al. [16] performed QTL mapping of a soybean RIL population grown in environments with different levels of P availability, and found that QTLs controlling seed quality and volume on Gm06 shared overlapping regions. The use of these types of loci could help improve multiple agronomic traits in soybean, thereby increasing the yield and quality of this important crop.

In this study, we genotyped two sets of RIL populations and investigated their yield and quality traits in three environments. The aims of this study were to (1) construct a genetic map of two RILs, (2) map the main-effect QTLs of agronomic traits and (3) identify the key loci that affect multiple agronomic traits in soybean. The results lay the foundation for producing high-yielding soybean with excellent quality traits via molecular breeding.

Materials and methods

Construction of RIL populations and experimental design

Three soybean species G. max (from Northeast Agriculture University Soybean Institute) with clearly different morphological traits were employed as the parents for mapping. Two RIL populations were constructed using Dongnong L13 as the female parent and Heihe 36 and Heinong 60 as the male parents. An RIL mapping population was obtained by multiple-seed descent using single-seed descent until the F₅ generation, and two F₆ RIL populations were constructed: RIL3613 and RIL6013 [17].

Field experiment

On 10 May 2016, soybean seeds were planted in the experimental fields of Northeast Agricultural University in Harbin (E128°N45°), S’cheng (E125°N45°) and Acheng (E126°N45°), China. Experimental treatments were applied in a randomised block design, and the planting density was 220,000 plants/hm². The plants were arranged in a randomised experimental three-row plot (5 m long and 0.67 m wide) at a spacing of 13.89 cm via ridge cultivation.

Phenotypic measurements

At the maturity stage, five individual plants were randomly selected from the middle row to measure quality and yield traits. A FOSS Infratec™1241 Grain Analyser was used to analyse seed protein and oil content, and post-harvest surveys were performed to investigate yield-related traits

Genetic map construction and QTL mapping

Genetic maps have been constructed previously [17]. The inclusive composite interval mapping (ICIM) method was employed to analyse 1-year multilocation RIL data using IciMapping software (http://www.isbreeding.net/software/). The model was used with a mapping step of 2.0 cM, a LOD significance threshold determined by 1000 permutation tests with a minimum value of 2.5, and significance set at the p = .05 level. The positive additive effect values were derived from the female parent, ‘Dongnong L13’, and the negative additive effect values were derived from the male parents, ‘Heihe 36 and Heinong 60’.

Gene mining involving consensus QTL intervals

Several candidate genes related to protein and oil accumulation and plant growth and development were selected based on genomic information from the Williams 82 Assembly two Genomic Sequence (Wm82.a2) and physical locations of the QTL intervals. These genes were annotated using information from various databases, primarily SoyBase (Integrating Genetics and Genomics to Advance Soybean Research) (https://www.soybase.org/), Phytozome (https://phytozome.jgi.doe.gov/pz/portal.html), QuickGO (Gene Ontology [GO] and GO Annotations) (https://www.ebi.ac.uk/QuickGO/) and the Kyoto Encyclopaedia of Genes and Genomes (KEGG) (https://www.kegg.jp/).

Results and discussion

Linkage map and traits of two RIL populations grown in three environments

Several studies have highlighted the importance of plant height and plant weight to soybean yield. Lee et al. [18] used an RIL population to construct a high-density single-nucleotide polymorphism (SNP) map, and mapped QTLs associated with seven plant height traits in soybean via composite interval mapping in three environments. Liang et al. [19] located two QTLs for plant weight in an RIL population derived from a cross between soybean lines BD2 and BX10. However, the RIL populations utilised in these studies were derived from two parents, and the linkage analysis involved only two alleles at the same locus. Due to the use of populations with limited genetic backgrounds, the low molecular marker polymorphism of such single populations leads to the identification of fewer markers for genetic map integration and low saturation, which limits the utility of these maps [20].

Here, we used three soybean species differing in morphological traits to generate two RIL populations, RIL3613 and RIL6013 [17], so that we could identify QTLs affecting yield-related traits. The RIL populations used in this study were designed by four-way hybridisation [20–22]. The use of populations produced from four parental double crosses increased not only the number of polymorphic markers, but also the density of the genetic map. The total length of the RIL 3613 genetic map was 2848.56 cM, including 150 SSRs. The average distance between markers was 18.99 cM, the length of a single linkage group ranged from 1.15 to 283.42 cM, and the number of markers ranged from 2 to 13. The RIL 6013 genetic map was 1886.8 cM, with an average length between markers of 13.77 cM, including 137 pairs of SSR primers. The length of a single linkage group ranged from 19.68 to 163.67 cM, and the number of markers ranged from 4 to 11 [17].

In recent years, most studies focused on QTL mapping of soybean quality and yield traits have been based on single-environment analyses, which makes the mapping results difficult to confirm. We reasoned that results from plants grown in different environments could be compared to increase the accuracy of the analysis. Phenotypic data for the two RIL populations grown in three environments are therefore provided in Table 1, Figures 1 and 2. Thirty-seven QTLs were detected in the two RIL populations from three environments. Table 1 lists information about these populations, including the population name, trait ID, trait name, chromosome, position, left marker, right marker, LOD, phenotypic variability explained (PVE), additive effect, left confidence interval (CI), right CI and environment.

Figure 1. Phenotype distribution of RIL3613.

Figure 2. Phenotype distribution of RIL6013.

Table 1. QTL mapping information.

Population	Trait ID	Trait name^a	GM	Position	Left marker	Right marker	LOD	PVE	Add	Left CI	Right CI	Location^b
6013	10	E1N1HSW	1	60	Sat_413	Sat_160	2.6259	8.8559	2.0224	51.5	64.5	AC
6013	31	E1PNPPRE	16	76	sat_394	Sat_366	4.6962	1.3982	−0.8117	74.5	77.5	AC
6013	31	E1PNPPRE	11	45	satt197	sat_123	3.5981	1.5875	−0.5678	41.5	47.5	AC
6013	32	E1SNPPRE	16	76	sat_394	Sat_366	4.8546	1.3969	−0.8848	74.5	77.5	AC
6013	32	E1SNPPRE	11	45	satt197	sat_123	4.569	1.7728	−0.6621	41.5	47.5	AC
6013	33	E1SWPPRE	19	50	sat_195	satt448	3.6414	0	−0.8401	49.5	50.5	AC
6013	33	E1SWPPRE	16	76	sat_394	Sat_366	4.9442	0	−1.1629	74.5	77.5	AC
6013	28	E1P2NRE	3	16	Sat_166	Satt584	5.7475	0.4592	−1.3295	15.5	16.5	AC
6013	28	E1P2NRE	11	45	satt197	sat_123	8.063	0.5084	−1.1342	42.5	46.5	AC
6013	28	E1P2NRE	3	12	satt237	Sat_166	6.1778	0.4872	−1.2307	10.5	13.5	AC
6013	28	E1P2NRE	18	122	satt352	satt564	7.5296	0.5003	−1.1451	121.5	123.5	AC
6013	2	E1N1NN	16	76	sat_394	Sat_366	3.8267	3.1054	−4.585	74.5	77.5	AC
6013	26	E1NNRE	12	19	Satt181	satt434	3.752	10.7616	−0.1018	12.5	19	AC
6013	29	E1P3NRE	10	70	Sat_341	Sat_221	9.8016	0.9753	−1.1068	66.5	71.5	AC
6013	29	E1P3NRE	16	79	Sat_366	sat_228	7.9241	0.9402	−1.1696	78.5	80.5	AC
6013	29	E1P3NRE	16	75	sat_394	Sat_366	10.1074	0.9089	−1.1983	73.5	76.5	AC
6013	29	E1P3NRE	1	58	Sat_413	Sat_160	8.6253	0.971	−1.0761	54.5	59.5	AC
6013	29	E1P3NRE	8	29	satt119	sat_406	6.8947	0.9559	−1.1561	26.5	31.5	AC
6013	29	E1P3NRE	11	44	satt197	sat_123	11.7883	0.9586	−1.1375	41.5	45.5	AC
6013	30	E1P4NRE	2	41	sat_096	sat_289	3.1584	3.5463	−1.7841	31.5	44.5	AC
6013	6	E1N1P4N	1	68	Sat_160	satt482	13.4888	1.3733	−4.1752	66.5	69.5	AC
6013	6	E1N1P4N	3	18	Sat_166	Satt584	11.7783	1.3986	−4.4013	16.5	19.5	AC
6013	30	E1P4NRE	1	60	Sat_413	Sat_160	4.0797	3.5683	−1.8254	54.5	63.5	AC
6013	6	E1N1P4N	1	61	Sat_413	Sat_160	13.1038	1.4914	−3.8273	57.5	63.5	AC
6013	6	E1N1P4N	3	11	satt237	Sat_166	13.584	1.368	−4.8919	8.5	12.5	AC
6013	6	E1N1P4N	1	76	satt482	satt254	11.0186	1.5665	−3.4414	73.5	78.5	AC
6013	6	E1N1P4N	15	28	satt685	sat_136	9.7868	1.3776	−4.5284	26.5	29.5	AC
6013	27	E1P1NRE	3	17	Sat_166	Satt584	6.5171	0.3473	−1.4191	16.5	18.5	AC
6013	27	E1P1NRE	19	48	sat_195	satt448	9.2339	0.3648	−1.4688	47.5	51.5	AC
6013	27	E1P1NRE	13	42	sat_240	sat_039	4.6467	0.4047	−1.1545	41.5	42.5	AC
6013	27	E1P1NRE	16	76	sat_394	Sat_366	8.002	0.3285	−1.5769	74.5	77.5	AC
6013	27	E1P1NRE	8	30	satt119	sat_406	4.0152	0.4125	−1.0728	26.5	31.5	AC
6013	27	E1P1NRE	3	12	satt237	Sat_166	8.7084	0.3872	−1.2179	9.5	13.5	AC
6013	27	E1P1NRE	20	46	satt330	satt367	5.7544	0.4228	−1.0976	43.5	48.5	AC
6013	36	E1OCRE	12	15	Satt181	satt434	2.7307	10.2949	−0.0156	5.5	19	AC
6013	1	E1N1PH	13	62	sat_039	Sat_417	3.2489	3.21	15.7834	59.5	64.5	AC
6013	25	E1PHRE	10	64	Sat_341	Sat_221	5.7497	8.8137	−0.127	48.5	69.5	AC
6013	25	E1PHRE	16	76	sat_394	Sat_366	4.8228	4.2704	−0.2529	74.5	77.5	AC
6013	1	E1N1PH	1	56	Sat_413	Sat_160	5.9393	2.8691	17.9361	51.5	59.5	AC
6013	130	E3HSWRE	2	31	sat_096	sat_289	2.7429	13.0708	−0.1341	23.5	38.5	HB
6013	130	E3HSWRE	18	120	satt352	satt564	5.4552	4.6139	−0.2455	117.5	121.5	HB
6013	129	E3SWPPRE	15	34	sat_136	satt651	2.8179	3.6006	−0.3259	31.5	37.5	HB
6013	107	E3N1PC	18	119	sat_117	satt352	3.1912	8.0104	0.9871	115.5	120.5	HB
6013	124	E3P2NRE	15	34	sat_136	satt651	4.449	0	−0.495	32.5	36.5	HB
6013	124	E3P2NRE	10	71	Sat_341	Sat_221	4.4361	0	−0.5495	68.5	72.5	HB
6013	124	E3P2NRE	16	75	sat_394	Sat_366	6.9752	0	−0.5992	73.5	76.5	HB
6013	124	E3P2NRE	1	77	satt482	satt254	3.3969	0	−0.4911	74.5	80.5	HB
6013	102	E3N1P4N	1	68	Sat_160	satt482	5.2332	0.4891	−5.0586	66.5	69.5	HB
6013	102	E3N1P4N	3	18	Sat_166	Satt584	3.7324	0.4722	−4.8167	16.5	19.5	HB
6013	102	E3N1P4N	16	78	Sat_366	sat_228	5.3316	0.3578	−6.8622	77.5	79.5	HB
6013	102	E3N1P4N	16	76	sat_394	Sat_366	5.3055	0.3673	−6.8609	74.5	76.5	HB
6013	102	E3N1P4N	1	56	Sat_413	Sat_160	16.8139	0.9905	−4.2785	52.5	58.5	HB
6013	102	E3N1P4N	3	9	satt237	Sat_166	11.4361	0.9795	−4.3196	6.5	10.5	HB
6013	102	E3N1P4N	1	79	satt482	satt254	15.0703	0.9982	−4.2216	76.5	81.5	HB
6013	123	E3P1NRE	15	35	sat_136	satt651	3.3699	1.1363	−0.6198	32.5	40.5	HB
6013	123	E3P1NRE	13	41	sat_240	sat_039	2.9181	1.2574	−0.7569	38.5	45.5	HB
6013	123	E3P1NRE	16	78	Sat_366	sat_228	5.2576	0.9542	−1.0388	77.5	79.5	HB
6013	123	E3P1NRE	16	76	sat_394	Sat_366	3.7028	1.0273	−0.9526	74.5	76.5	HB
6013	123	E3P1NRE	3	12	satt237	Sat_166	4.6832	1.1662	−0.8655	8.5	13.5	HB
6013	132	E3OCRE	11	50	sat_123	BE806308	2.8405	0	−0.0592	49.5	51.5	HB
6013	132	E3OCRE	16	76	sat_394	Sat_366	2.8559	0	−0.0648	75.5	77.5	HB
6013	97	E3N1PH	13	61	sat_039	Sat_417	3.9871	3.3349	15.6442	57.5	65.5	HB
6013	97	E3N1PH	15	34	sat_136	satt651	2.9831	3.1655	23.5375	32.5	37.5	HB
6013	97	E3N1PH	3	18	Sat_166	Satt584	3.2475	2.6088	26.7388	16.5	19.5	HB
6013	121	E3PHRE	19	48	sat_195	satt448	5.0553	1.7132	−0.2224	46.5	50.5	HB
6013	121	E3PHRE	10	71	Sat_341	Sat_221	3.5519	1.7304	−0.224	68.5	71.5	HB
6013	97	E3N1PH	10	72	Sat_341	Sat_221	2.5131	3.2609	23.314	71.5	74.5	HB
6013	97	E3N1PH	11	44	satt197	sat_123	3.8804	2.9228	23.8381	40.5	46.5	HB
6013	97	E3N1PH	20	46	satt330	satt367	2.9322	3.4756	22.0246	42.5	47.5	HB
6013	121	E3PHRE	18	122	satt352	satt564	4.0523	1.7021	−0.2252	120.5	122.5	HB
6013	79	E2PNPPRE	10	70	Sat_341	Sat_221	3.1856	0	−0.397	66.5	70.5	SC
6013	79	E2PNPPRE	16	78	Sat_366	sat_228	3.1306	0	−0.6613	77.5	78.5	SC
6013	79	E2PNPPRE	3	10	satt237	Sat_166	2.8778	0	−0.4669	8.5	10.5	SC
6013	80	E2SNPPRE	16	77	sat_394	Sat_366	3.5778	0	−1.0032	75.5	77.5	SC
6013	80	E2SNPPRE	3	11	satt237	Sat_166	4.5114	0	−0.5235	7.5	12.5	SC
6013	81	E2SWPPRE	16	77	sat_394	Sat_366	3.346	0	−1.0579	76.5	77.5	SC
6013	81	E2SWPPRE	3	10	satt237	Sat_166	3.81	0	−0.5229	7.5	11.5	SC
6013	59	E2N1PC	16	78	Sat_366	sat_228	3.8807	0	2.6519	77.5	79.5	SC
6013	76	E2P2NRE	10	73	Sat_341	Sat_221	7.9219	0.6868	−1.0601	69.5	74.5	SC
6013	76	E2P2NRE	16	76	sat_394	Sat_366	10.0807	0.6962	−1.3098	75.5	76.5	SC
6013	76	E2P2NRE	1	58	Sat_413	Sat_160	4.6639	0.702	−1.0205	55.5	58.5	SC
6013	76	E2P2NRE	8	31	satt119	sat_406	5.2426	0.7017	−1.2073	29.5	33.5	SC
6013	76	E2P2NRE	11	43	satt197	sat_123	7.7255	0.6818	−1.3852	41.5	43.5	SC
6013	76	E2P2NRE	20	47	satt330	satt367	3.4248	0.6043	−0.7486	43.5	49.5	SC
6013	76	E2P2NRE	1	76	satt482	satt254	5.7575	0.686	−0.7903	73.5	79.5	SC
6013	74	E2NNRE	16	76	sat_394	Sat_366	3.398	0	−0.2415	74.5	76.5	SC
6013	77	E2P3NRE	13	41	sat_240	sat_039	4.7824	0.9203	−0.8995	38.5	43.5	SC
6013	77	E2P3NRE	16	78	Sat_366	sat_228	6.9941	0.7784	−1.1106	77.5	79.5	SC
6013	77	E2P3NRE	16	76	sat_394	Sat_366	7.2317	0.7899	−1.1055	74.5	76.5	SC
6013	77	E2P3NRE	3	11	satt237	Sat_166	9.679	0.9071	−0.9429	7.5	12.5	SC
6013	53	E2N1P3N	1	71	satt482	satt254	2.8642	8.4198	2.9703	68.5	77.5	SC
6013	54	E2N1P4N	18	119	sat_117	satt352	6.7429	0.6528	−4.0743	118.5	119.5	SC
6013	78	E2P4NRE	15	34	sat_136	satt651	2.6218	0.9525	−1.6248	33.5	36.5	SC
6013	78	E2P4NRE	11	41	satt197	sat_123	4.0718	1.0099	−1.5004	38.5	41.5	SC
6013	54	E2N1P4N	3	11	satt237	Sat_166	8.6524	0.7725	−3.6415	8.5	13.5	SC
6013	54	E2N1P4N	18	121	satt352	satt564	5.4692	0.5888	−4.1277	120.5	121.5	SC
6013	54	E2N1P4N	15	28	satt685	sat_136	6.3609	0.5793	−4.3021	26.5	29.5	SC
6013	75	E2P1NRE	11	51	sat_123	BE806308	20.173	1.1345	−3.1092	49.5	52.5	SC
6013	51	E2N1P1N	10	65	Sat_341	Sat_221	2.8599	5.1419	−1.8009	58.5	70.5	SC
6013	75	E2P1NRE	10	73	Sat_341	Sat_221	18.7469	1.1379	−3.0445	70.5	74.5	SC
6013	75	E2P1NRE	16	77	sat_394	Sat_366	18.8081	1.1342	−3.1156	76.5	77.5	SC
6013	75	E2P1NRE	1	60	Sat_413	Sat_160	14.2334	1.1501	−2.9079	56.5	62.5	SC
6013	75	E2P1NRE	8	32	satt119	sat_406	15.7266	1.1375	−3.0514	29.5	34.5	SC
6013	75	E2P1NRE	11	45	satt197	sat_123	20.8606	1.1606	−2.8411	41.5	46.5	SC
6013	75	E2P1NRE	20	48	satt330	satt367	11.6244	1.1372	−3.0543	45.5	48.5	SC
6013	75	E2P1NRE	1	75	satt482	satt254	12.4608	1.1392	−3.0243	73.5	78.5	SC
6013	84	E2OCRE	15	26	satt685	sat_136	2.8932	7.9465	0.0351	18.5	28.5	SC
6013	49	E2N1PH	11	46	satt197	sat_123	2.6629	6.3575	16.8258	42.5	47.5	SC
3613	32	E1SNPPRE	10	56	Satt243	Sat_307	3.5568	0.1374	0.6123	53.5	59.5	AC
3613	33	E1SWPPRE	10	57	Satt243	Sat_307	2.508	0.0565	0.6041	56.5	58.5	AC
3613	11	E1N1PC	19	193	Sat_099	Satt229	2.5113	8.6987	−0.7244	180.5	201.5	AC
3613	35	E1PCRE	16	167	Sat_224	Satt654	3.5581	4.4846	−0.0295	163.5	169.5	AC
3613	35	E1PCRE	20	20	Satt292	Satt571	3.4365	6.9647	−0.0254	12.5	27.5	AC
3613	28	E1P2NRE	5	92	Sat_171	SOYNOD26A	4.0301	11.382	0.4535	90.5	92.5	AC
3613	2	E1N1NN	13	25	GMRUBP	Sat_262	4.2039	2.1344	−2.8635	23.5	28.5	AC
3613	2	E1N1NN	10	47	Sat_291	Satt500	3.6587	1.1256	−3.3893	45.5	47.5	AC
3613	2	E1N1NN	13	19	Satt114	GMRUBP	4.7479	2.3956	−3.0737	15.5	21.5	AC
3613	2	E1N1NN	17	94	Satt669	Sat_292	2.6334	2.7664	−2.9746	92.5	97.5	AC
3613	6	E1N1P4N	13	26	GMRUBP	Sat_262	4.2003	0.5343	−5.3912	24.5	26.5	AC
3613	30	E1P4NRE	5	92	Sat_171	SOYNOD26A	3.8379	1.4237	2.0212	91.5	93.5	AC
3613	30	E1P4NRE	12	2	Sat_200	Satt353	5.0031	1.2503	−3.5477	1.5	2.5	AC
3613	30	E1P4NRE	8	161	Satt424	Sat_294	5.804	1.4733	−3.0799	157.5	161.5	AC
3613	27	E1P1NRE	13	17	Satt114	GMRUBP	3.6157	0.8592	−1.5124	12.5	19.5	AC
3613	130	E3HSWRE	12	4	Satt353	Satt434	6.7912	0.4477	−0.5026	3.5	5.5	HB
3613	104	E3N1SNPP	12	8	Satt353	Satt434	3.005	1.7705	−43.1966	5.5	11.5	HB
3613	128	E3SNPPRE	8	162	Satt424	Sat_294	4.0822	2.7162	−0.5006	157.5	164.5	HB
3613	105	E3N1SWPP	5	93	Sat_171	SOYNOD26A	4.0895	6.2224	7.9546	92.5	93.5	HB
3613	129	E3SWPPRE	5	93	Sat_171	SOYNOD26A	3.2159	4.6696	0.4623	92.5	94.5	HB
3613	105	E3N1SWPP	13	19	Satt114	GMRUBP	2.5107	5.512	−7.3291	15.5	21.5	HB
3613	129	E3SWPPRE	8	163	Satt424	Sat_294	5.4988	5.034	−0.5279	159.5	165.5	HB
3613	124	E3P2NRE	12	4	Satt353	Satt434	6.3748	1.2652	−1.8513	3.5	5.5	HB
3613	125	E3P3NRE	13	25	GMRUBP	Sat_262	3.7049	3.151	−0.8018	23.5	25.5	HB
3613	125	E3P3NRE	16	167	Sat_224	Satt654	3.7061	3.1644	−0.7601	164.5	167.5	HB
3613	125	E3P3NRE	8	163	Satt424	Sat_294	3.5982	3.3061	−0.76	161.5	163.5	HB
3613	126	E3P4NRE	19	193	Sat_099	Satt229	12.7854	1.1581	−2.597	190.5	196.5	HB
3613	78	E2P4NRE	19	194	Sat_099	Satt229	9.236	2.5313	−11.0111	191.5	199.5	HB
3613	126	E3P4NRE	16	168	Sat_224	Satt654	11.4729	1.1269	−3.7896	165.5	168.5	HB
3613	126	E3P4NRE	10	46	Sat_291	Satt500	14.1652	1.1267	−3.7956	45.5	47.5	HB
3613	126	E3P4NRE	17	102	Sat_292	Sat_284	13.2745	1.1268	−3.7923	100.5	104.5	HB
3613	126	E3P4NRE	14	3	Satt063	Sct_094	13.5361	1.1606	−3.3655	0	5.5	HB
3613	126	E3P4NRE	8	162	Satt424	Sat_294	13.0507	1.1563	−2.9371	158.5	164.5	HB
3613	99	E3N1P1N	9	1	Sat_119	Satt552	2.6302	2.1181	−4.669	0	1	HB
3613	123	E3P1NRE	17	106	Sat_284	Sct_192	4.6727	10.5443	−1.0909	105.5	106.5	HB
3613	99	E3N1P1N	13	18	Satt114	GMRUBP	3.5514	2.4768	−2.4539	14.5	21.5	HB
3613	99	E3N1P1N	8	163	Satt424	Sat_294	3.7464	1.9846	−2.5165	158.5	165.5	HB
3613	123	E3P1NRE	17	93	Satt669	Sat_292	3.9094	10.4963	−1.0857	87.5	96.5	HB
3613	121	E3PHRE	12	4	Satt353	Satt434	5.61	0.3448	0.2517	3.5	6.5	HB
3613	55	E2N1PNPP	12	2	Sat_200	Satt353	4.0597	5.9731	−28.1164	1.5	2.5	SC
3613	79	E2PNPPRE	12	2	Sat_200	Satt353	6.5755	2.6751	−1.1167	0	2.5	SC
3613	79	E2PNPPRE	20	20	Satt292	Satt571	5.7053	3.533	−0.7905	16.5	23.5	SC
3613	56	E2N1SNPP	12	2	Sat_200	Satt353	6.8622	5.8484	−98.3181	0	2.5	SC
3613	80	E2SNPPRE	12	2	Sat_200	Satt353	7.178	5.4049	−1.311	0	2.5	SC
3613	57	E2N1SWPP	12	1	Sat_200	Satt353	4.6159	8.9893	−17.0107	0	2.5	SC
3613	81	E2SWPPRE	12	2	Sat_200	Satt353	8.4887	8.7575	−1.4271	0	2.5	SC
3613	76	E2P2NRE	14	3	Satt063	Sct_094	4.0324	0.5635	−1.1463	1.5	4.5	SC
3613	77	E2P3NRE	12	2	Sat_200	Satt353	3.8491	4.9341	−1.5338	0	2.5	SC
3613	78	E2P4NRE	19	190	Sat_099	Satt229	9.2016	2.5032	−19.0363	189.5	190.5	SC
3613	54	E2N1P4N	12	2	Sat_200	Satt353	8.1007	2.4818	−16.0368	0	2.5	SC
3613	78	E2P4NRE	17	103	Sat_292	Sat_284	9.087	2.5481	−9.9314	102.5	103.5	SC
3613	78	E2P4NRE	13	14	Satt114	GMRUBP	10.1832	2.54	−11.1779	11.5	14.5	SC
3613	54	E2N1P4N	13	17	Satt114	GMRUBP	6.6169	2.4357	−13.5508	13.5	17.5	SC
3613	78	E2P4NRE	20	24	Satt292	Satt571	9.6043	2.5363	−11.2028	18.5	26.5	SC
3613	75	E2P1NRE	10	53	Satt243	Sat_307	14.5463	0.5012	5.4692	52.5	53.5	SC

^a E3HSWRE and E1N1HSW represent 100-seed weight detected from HB and AC. E2PNPPRE and E2N1PNPP represent pods per plant detected from SC; and E1PNPPRE represents pods per plant detected from AC. E1SNPPRE represents grain number of single plant detected from AC; E2SNPPRE and E2N1SNPP represent grain number of single plant detected from and E3SNPPRE and E3N1SNPP represent grain number of single plant detected from HB. E1SWPPRE represents grain weight per plant detected from AC; E2N1SWPP and E2SWPPRE represent grain weight per plant detected from SC; E3N1SWPP and E3SWPPRE represent grain weight per plant detected from HB. E1N1PC and E1PCRE represent protein detected from AC; E2N1PC and E3N1PC represent protein detected from SC and HB, respectively. E1P2NRE, E2P2NRE and E3P2NRE represent seed set two per pod detected from AC, SC and HB, respectively. E1N1NN and E1NNRE represent node number detected from AC; E2NNRE represents node number detected from SC. E1P3NRE represents seed set three per pod detected from AC; E2N1P3N and E2P3NRE represent seed set three per pod detected from SC; and E3P3NRE represents seed set three per pod detected from HB. E1N1P4N and E1P4NRE represent seed set four per pod detected from AC; E2N1P4N and E2P4NRE represent seed set four per pod detected from SC; E3N1P4N and E3P4NRE represent seed set four per pod detected from HB. E1P1NRE, E2N1P1N and E3N1P1N represent seed set one per pod detected from AC, SC and HB, respectively. E1OCRE, E2OCRE and E3OCRE represent oil detected from AC, SC and HB, respectively. E1N1PH and E1PHRE represent plant height detected from AC; E3N1PH and E3PHRE represent plant height detected from HB; and E2N1PH represents plant height detected from SC.

^b AC: A’cheng. SC: Shuangcheng. HB: Harbin.

The high or low LOD values obtained in previous studies are another important factor limiting the reproducibility of those results; when LOD values are too high, QTLs might not be detected and data will be lost. By contrast, using an excessively low LOD value may result in a false positive for the detected QTL. In this study, we used multiyear and multipoint data for joint analysis, which not only increased the detection intensity of QTLs, but also greatly improved the accuracy of the detected locations and effects of the QTLs.

QTL mapping analysis of the RIL populations

The chromosome number was 17 in both populations. Chromosomes Gm04, Gm06 and Gm07 had 37 left markers but only 21 right markers, indicating that the QTLs in these populations contained overlapping regions, a phenomenon that deserves further investigation. The minimum LOD value was only 2.508, which was detected in the 3613 RIL population and is related to grain weight per plant. The maximum LOD value was 20.8606, which was detected in the 6013 RIL population and is associated with seed set one per pod. Eight QTLs (all from RIL population 6013) had PVE values of 0. The highest PVE value was also obtained from this population, which reached 13.0708%. The positive and negative additive effect values were derived from the female parent and male parent, respectively. The highest absolute additive effect value was 98.3181. The highest absolute additive effect values were all detected from RIL population 3613, and were all associated with grain number of a single plant; these QTLs had additive effects: −98.3181 − 43.1966 and −28.1164. The highest positive additive effect values were all related to plant height; these values were 26.7388, 23.8381, 23.5375, 23.314, 22.0246, 17.9361, 16.8258, 15.7834 and 15.6442. The QTLs with positive additive effect values were all mapped from RIL population 6013.

We detected 37 QTLs in the two RIL populations from three locations (Table 1). Sixteen QTLs were detected in RIL population 3613, and 21 were detected in RIL population 6013. In all three locations, the number of QTLs ranged from 22 to 32, and the number of traits ranged from 11 to 12, with 12 traits identified only in A’cheng City. These QTLs were related to various soybean traits, including protein, oil and other yield traits, and were distributed on all 17 chromosomes except Gm04, Gm06 and Gm07. Among these QTLs, five were associated with protein content, four with oil content, and the remaining QTLs with 10 yield traits. These yield-related QTLs were mainly located on chromosomes Gm01 and Gm16. The QTLs for both quality and yield traits were mainly located on Gm16, a key soybean chromosome. These QTLs were related to 12 types of traits. The largest number of QTLs (24) was associated with seed set per pod and the smallest number (4) was associated with 100-seed weight and oil content.

Warrington et al. [23] previously located four QTLs related to protein content based on an RIL population. Wang et al. [24] combined the positioning results from two RIL populations and conducted joint analysis, identifying 12 seed yield QTLs, 16 oil-related QTLs and 11 QTLs controlling protein content. Zhang et al. [25] employed an RIL population derived from a cross between Bogao and Nannong94-156 to locate 10 QTLs that regulate the number of soybean pods. Here, we identified nine QTLs for protein and oil content, found on seven chromosomes: Gm11, Gm12, Gm15, Gm16, Gm18, Gm19 and Gm20. The QTLs related to protein content that were mined from the 3613 population all had negative additive effects, but the protein content-related QTLs mined from the other RIL population all had positive additive effects. The protein content QTLs had additive effect values ranging from −0.7244 to 2.6519.

Of the five QTLs associated with protein content identified in this study, four were mapped to QTL regions associated with protein content in previous studies. One protein QTL has been mapped to the region 23.91 cM (Sat_228) to 52.84 cM (Sat_366) located on Gm16 [26]. The 50.52 cM (Satt352) to 100 cM (sat_117) region on Gm18 contains eight protein content-related QTLs identified in previous studies [8,27–30]. Only two protein content-related QTLs have been mapped to region 78.23 cM (Sat_099) to 93.88 cM (Satt229) on Gm19 [31]. A region located on chromosome 20 is thought to be important, as it has been repeatedly associated with protein content by different breeders in several studies in the past. Twenty QTLs were mapped to the 18.5 cM (Satt571) to 82.77 cM (Satt292) region [8,24,27,30–36].

We identified four QTLs related to oil only in the 6013 population, with additive effects of −0.0592, −0.0156, 0.0351 and −0.0648. One of these QTLs (Gm15; 39.16–56.7 cM) associated with the soybean oil index is located in the same segment as a previously identified oil QTL [24,37–40], indicating that there is a high probability that this region contains one or several important genes that control oil traits.

The QTL related to 100-seed weight mined from the 3613 population had an additive effect value of −0.5026 and the highest LOD value (6.7912), indicating that it is located in a concentrated area of QTLs associated with 100-seed weight. Number of grains per pod was the most common yield trait among the QTLs identified from the two RIL populations, especially seed set four per pod, involving 24 QTLs. Among the QTLs related to seed set four per pod, two QTLs were detected, with additive effects of −4.2785 and −4.2216 and relatively large LOD values of 16.8139 and 15.0703, respectively, indicating that a concentrated area of QTLs is associated with seed set four per pod. A number of QTLs associated with pod number at the same position on chromosome Gm11 (satt197–sat_123) were also found to be related to pod numbers in previous studies [40]. The QTLs related to plant height had the highest additive values and LOD values.

Identification of QTLs with pleiotropic effects

A total of 32 QTLs were found to control multiple traits (i.e. QTLs with pleiotropic effects), whereas only five were associated with a single trait, including seed set one per pod, seed set four per pod and plant height. In the 3613 population, the QTLs for 11 traits are distributed on 11 chromosomes, with most QTLs associated with seed set four per pod. In the 6013 population, the QTLs for 12 traits are distributed on 13 chromosomes, with most QTLs also related to seed set four per pod. QTLs associated with more than four traits were designated as multi-effect QTLs in this study.

In the 6013 RIL population, a QTL associated with markers sat_394 and Sat_366 is related to nine traits, including one quality trait and eight yield traits, making it the QTL associated with the most traits. This QTL has negative additive effects, with a maximum PVE value of 0. The multi-effect QTL related to the second highest number of traits was that associated with markers satt237 and Sat_166, which is related to seven traits, including pod number per plant, grain weight per plant and grain number of a single plant, with a PVE of 0. The QTL associated with markers satt197 and sat_123 is related to six yield traits, with a maximum LOD value of 20.8606. The QTL associated with markers Sat_413 and Sat_160 is related to six traits, including all seed pod numbers, plant height and 100-seed weight. The QTL related to seed pod number had negative additive effects and large LOD values of 13.1038, 14.2334 and 16.8139. The QTL associated with markers Sat_341 and Sat_221 was related to five yield traits and has a large LOD value of 18.7469, and the QTL related to the pod number per plant and seed set two per pod had a PVE of 0. The QTL associated with markers Sat_366 and sat_228 was also related to quality and yield traits, with positive additive effects for the quality trait and negative additive effects for the other yield traits. The QTLs associated with marker Sat_366 are all related to quality traits.

In the 3613 RIL population, we identified two major QTLs that were only related to yield traits. One QTL, with markers Sat_200 and Satt353, was related to five yield traits, and a QTL associated with grain number of a single plant was detected with a maximum additive effect of −98.3181. Another QTL associated with markers Satt424 and Sat_294 was related to five yield traits, with a large LOD value of 13.0507.

In summary, we detected 37 QTLs in two RIL populations from three locations. In the 3613 RIL population, three QTLs related to quality traits and 13 QTLs related to yield traits were identified among the 16 QTLs. In the 6013 population, we identified six QTLs related to quality traits and 15 related to yield traits. Seeds per pod was the main yield trait associated with QTLs in both populations, especially seed set four per pod. We identified one essential chromosome, Gm16, harbouring one major QTL, containing sat_394 as the left marker and Sat_366 as the right marker, related to quality and yield traits.

Candidate gene mining for multiple-effect QTLs

We used three types of QTLs for candidate gene annotation, including QTLs related to quality traits and yield traits, QTLs related to multiple quality traits and QTLs related to multiple yield traits. We identified two QTL intervals related to quality traits and yield traits and six QTL intervals associated only with yield traits; no QTL intervals were related only to quality traits. Based on information about the markers obtained using the SoyBase online program, in the eight consensus QTL intervals, 1215 candidate genes were screened and annotated using the GO and GO Annotations and KEGG databases. Among these, 344, 396, 218, 52 and 205 genes were located in consensus QTL regions on GM01, GM8, GM11, GM12 and GM16, respectively (Table 2). The following enriched GO categories were identified: GO:0006807 (nitrogen compound metabolic process), GO:0034641 (cellular nitrogen compound metabolic process), GO:0006886 (intracellular protein transport), GO:0008565 (protein transporter activity) and GO:0015031 (protein transport). Among the enriched KEGG pathways, Ko04075 is involved in the plant hormone signal transduction pathway [41], including gibberellins, auxin, abscisic acid and other hormones. Plant hormones play important roles in modulating stem growth, plant growth and seed development. The other enriched KEGG pathways included K14484, K14493, K14496, K14486, K14488, K14498 and K13946, which are involved in plant hormone signal transduction (pathway ID ko04075).

Table 2. Consensus QTL intervals and distribution of candidate genes for related traits on five linkage groups.

Code.	Linkage group	Consensus QTL interval (Mbp)	Left marker	Right marker	QTL number	LOD range	ADD range	Mean PVE (%)	Candidate gene number
1	GM01	0.35–53.24	Sat_413	Sat_160	8	2.66–16.81	−4.28 to 17.94	2.57	344
2	GM08	10.63–43.49	Satt424	Sat_294	6	3.60–13.05	−3.08 and −0.50	2.61	396
3	GM11	8.90–32.19	Sat_123	satt197	9	2.66–20.86	−2.84 to 23.84	1.88	218
4	GM12	0.27–1.69	Sat_200	Satt353	9	2.77–5.16	3.85 to 8.49	5.16	52
5	GM16	3.05–30.4	sat_228	Sat_366	6	3.13–7.92	−6.86 to 2.65	0.51	205

Based on pathway and functional annotation, we ultimately identified 36 candidate genes (Table 3). Glyma.01G052800 was annotated as a vesicle transport v-SNARE family protein gene, which functions in intracellular vesicle transport processes in protein storage vacuoles in plants [42,43]. Glyma.11G118600, Glyma.08G222400 and Glyma.08G207300 were annotated as Coatomer with beta subunit and gamma-2 subunit genes, which function in intracellular biosynthetic protein transport mediated by non-clathrin-coated vesicles [44] and might be related to protein accumulation. Among the candidate genes, 13 were annotated as SAUR-like auxin-responsive protein family genes; SAUR proteins are involved in cell division and auxin biosynthesis [45]. Glyma.08G188300, Glyma.01G004500 and Glyma.01G183500 were annotated as sucrose non-fermenting-related protein kinase (SnRK) genes, which are closely related to sugar signalling, biotic and abiotic stress responses, seed germination and seedling growth [46,47].

Table 3. Annotation of candidate genes associated with consensus QTLs.

Gene ID	Chr.	KEGG^a	GO^b	Functional annotation
Glyma.01G153200	GM 01	K12400	GO:0030117, GO:0016192, GO:0006886	Adaptin family protein
Glyma.12G021600	GM 12	K12392	GO:0030131, GO:0016192, GO:0006886, GO:0030117	Adaptin family protein
Glyma.16G076500	GM 16	K12400	GO:0030117, GO:0016192, GO:0006886	Adaptin family protein
Glyma.16G109200	GM 16	K12400	GO:0030117, GO:0016192, GO:0006886	Adaptin family protein
Glyma.01G032100	GM 01	K12391	GO:0030131, GO:0016192, GO:0006886, GO:0030117	Adaptor protein complex AP-1, gamma subunit
Glyma.08G289100	GM 08	K12391	GO:0030131, GO:0016192, GO:0006886, GO:0030117	Adaptor protein complex AP-1, gamma subunit
Glyma.01G103500	GM 01	K14486	GO:0006355, GO:0005634, GO:0003677, GO:0009725	Auxin response factor 9
Glyma.01G179300	GM 01	K04646	GO:0030132, GO:0030130, GO:0016192, GO:0006886, GO:0005198	Clathrin, heavy chain
Glyma.11G118600	GM 11	K17267	GO:0030126, GO:0016192, GO:0006886, GO:0005198, GO:0030117	Coatomer gamma-2 subunit, putative / gamma-2 coat protein, putative / gamma-2 COP, putative
Glyma.08G222400	GM 08	K17301	GO:0030126, GO:0016192, GO:0006886, GO:0005198, GO:0030117, GO:0005737	Coatomer, beta subunit
Glyma.08G207300	GM 08	K17302	GO:0005515, GO:0030117, GO:0016192, GO:0006886, GO:0005198	Coatomer, beta\' subunit
Glyma.12G018500	GM 12	K14290	GO:0008536, GO:0006886	Exportin 1A
Glyma.11G215500	GM 11	K01915	GO:0006807, GO:0006542, GO:0004356	Glutamine synthase clone R1
Glyma.08G207900	GM 08	K14484	GO:0006355, GO:0005634	Indole-3-acetic acid inducible 19
Glyma.01G019400	GM 01	K14484	GO:0006355, GO:0005634	Indole-3-acetic acid inducible 9
Glyma.08G273500	GM 08	K14484	GO:0006355, GO:0005634	Indole-3-acetic acid inducible 9
Glyma.11G180900	GM 11	K13035	GO:0016810, GO:0006807	Nitrilase 4
Glyma.08G203100	GM 08	K14484	GO:0006355, GO:0005634	Phytochrome-associated protein 2
Glyma.01G078200	GM 01	K14488	GO:0009733	SAUR-like auxin-responsive protein family
Glyma.01G137500	GM 01	K14488	GO:0009733	SAUR-like auxin-responsive protein family
Glyma.01G167000	GM 01	K14488	GO:0009733	SAUR-like auxin-responsive protein family
Glyma.08G155300	GM 08	K14488	GO:0009733	SAUR-like auxin-responsive protein family
Glyma.08G155500	GM 08	K14488	GO:0009733	SAUR-like auxin-responsive protein family
Glyma.08G155600	GM 08	K14488	GO:0009733	SAUR-like auxin-responsive protein family
Glyma.08G155700	GM 08	K14488	GO:0009733	SAUR-like auxin-responsive protein family
Glyma.08G155800	GM 08	K14488	GO:0009733	SAUR-like auxin-responsive protein family

^a Kyoto Encyclopedia of Genes and Genomes.

^b GO, gene ontology.

In addition, SnRK2 and SnRK3 are directly related to abscisic acid signal transduction [48]. Glyma.01G019400, Glyma.08G207900 and Glyma.08G273500 were annotated as indole-3-acetic acid-inducible genes; indole-3-acetic acid is a ubiquitous endogenous auxin in plants. Glyma.11G180900 was annotated as encoding nitrilase 4, which converts indole-3-acetonitrile to auxin, as revealed in Arabidopsis thaliana [49]. Glyma.08G203100 was annotated as encoding a phytochrome-associated protein related to flowering [50]. Glyma.01G103500 was annotated as encoding auxin response factor 9; ARF9 is expressed during early fruit development and is auxin responsive [51]. Six candidate genes were annotated as encoding Adaptin family proteins or Adaptin family proteins with gamma subunits, which are involved in the formation of intracellular transport vesicles and the selection of cargo for incorporation into the vesicles [52]. Glyma.01G179300 was annotated as Clathrin heavy chain, which functions in transport from the Golgi to vacuoles [53]. Glyma.11G215500 was annotated as glutamine synthase, a highly regulated enzyme at the core of nitrogen metabolism [54]. Glyma.16G082400 was annotated as the Sec23/Sec24 protein transport family protein; the Sec23 subunit is essential for linking COPII vesicle formation to anterograde transport events [55]. Finally, Glyma.12G018500 was annotated as exportin 1A, which is essential for various developmental pathways in plants [56].

Conclusions

We identified 37 QTLs for yield traits using two RIL populations grown in three environments. Some of these QTLs corresponded to those previously reported in other studies. Notably, most of the QTLs identified in this study are clustered in seven locations in the genome (Gm01, Gm03, Gm08, Gm10, Gm11, Gm12 and Gm16). These seven regions contain multiple QTLs controlling different traits. Additional studies on the multi-effect QTLs at these seven positions should be conducted in the future.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The authors gratefully acknowledge the financial support for this study provided by grants from the Natural Science Foundation of Heilongjiang Province, China (LC2016010) to Hailong Ning.