Introduction

Soybean (Glycine max) is an oilseed crop of global importance, and soybean oil comprises almost a third of vegetable oils consumed around the world [1]. Genetic approaches have been used to modify all aspects of the soybean oil profile (reviewed in [2–4]). Of particular emphasis in the improvement of soybean oil is the reduction in levels of saturated fatty acids, such as stearic acid and palmitic acid. High intake of dietary palmitic acid is believed to contribute to increased cholesterol levels and increased risk of heart disease in humans [5, 6]. Commodity soybean oil has relatively low levels of palmitic acid, around 10–12%, but reductions in this level are genetically tractable, therefore 7% total saturated fats (stearic acid and palmitic acid combined) is a goal for breeding for oil composition [4]. Specifically, it has been demonstrated that mutation in the KASIII or FATB1a genes can result in reduced levels of palmitic acid. Mutations in KASIII were initially isolated as the fap1 locus, represented by a single nucleotide change that results in a premature truncation of the KASIII protein and reduces palmitic acid levels by 20–30% [7]. The fap3 locus corresponds to the FATB1a gene, and several independent deletions and single nucleotide polymorphisms have been identified in this gene that reduce palmitic acid content to as little as 6% [8–13] similar to the effect of silencing the expression of FATB gene in soybean seeds [14]. Mutations in both KASIII and FATB1a can function additively to reduce palmitic acid content in seeds [7]. The FATB1a gene encodes an acyl-acyl carrier protein thioesterase which controls export of the 16-carbon fatty acids from the plastid after synthesis, and therefore influences the balance of saturated to unsaturated fatty acids in the seed [15]. As further developments are made to produce healthier soybean oil low in trans fats, with a high oleic profile that finds broader markets and new uses, continued efforts to identify and deploy non-transgenic variation to reduce saturated fats will be required. In this work we describe the isolation and characterization of a new nonsense allele of FATB1a that can be a tool to develop improved soybean germplasm.

Materials and methods

The mutant population was generated by NMU treatment and an ongoing screen of fatty acid profiles in approximately 5,000 M₃ bulk seed samples was performed as described previously [16]. For characterization of the low palmitic acid trait, plants were harvested in bulk from 1.8 meter field plot rows in West Lafayette, Indiana during the 2016–2020 growing seasons (excluding 2019), and individual seeds from multiple individuals were phenotyped. Ten single seeds were chipped for analysis by GC as described previously [12]. Statistical significance of fatty acid content was performed by comparing wild type and mutant composition each year using two-tailed, type 2 t-tests.

In the field in 2015 the putative mutant line (known as plant 19479) was outcrossed to the purple flowered cultivar Prize (PI 548554), to facilitate genetic mapping of the low palmitic acid trait, and the F₁ plant from this cross was grown to maturity in the field during the 2016 growing season. To test for complementation of the new mutation to a previously identified FATB1a mutant, plant 19479 was crossed in the field during 2017 to a line carrying a reference mutant allele of FATB1a (FATB1a_G180D) [12], and the F₁ plants were germinated and grown to maturity in the greenhouse in 16h light/8 hour dark cycles at a temperature of 27C. For both crosses F₂ seeds were harvested and subjected to single seed gas chromatograph phenotyping to measure fatty acid content in parallel with single seed genotyping as described previously [17]. For the complementation cross, 49 single seeds were analyzed, and for the backcross population, 80 single seeds were analyzed.

Primers for amplification and sequencing of the FATB1a gene were 5’-GTCTTCTGGTGGCTTGAAGG-3’ and 5’-CCCAGACAAATTTCCAAAGC-3’ for the first and second exons, and 5’-GACATAGTTCAAGTGGACACT-3’ and 5’-TTCACAACACCAAAGTTGTTCAC-3’ to cover exons three to six. PCR amplification was performed using an initial 60 second denaturation at 95C, followed by 5 cycles of 94C for 30 seconds, 56C for 20 seconds, and 68C for 4 minutes, followed by 20 cycles of 94C for 30 seconds, 58C for 20 seconds and 68C for 4 minutes. PCR amplicons were sequenced using the same primers and dye terminator sequencing and BigDye (Life Technologies, Waltham, MA) standard kit protocols. Genotyping for the FATB1a_Q52STOP polymorphism was performed with a cleaved amplified polymorphic sequence (CAPS) marker designed to recognize the polymorphism. DNA was extracted from seed chips using the Mega EZ Plant 96-well DNA kit. Genotyping primer sequences were 5’- AACTGATGTGCTGTGCTGTT-3’ and 5’- TCAGCGGCCAAGAAAATTGT-3’. PCR and digestion reaction conditions were 1 minute at 95C, 30 cycles of 94C for 20 seconds, 25 seconds at 58C, 68C for 1 minute, and a final extension for 7 minutes at 68C. PCR amplicons were digested overnight at 37C with the restriction enzyme HhaI (New England Biolabs) which cuts the wild type sequence, and visualized by electrophoresis on 1% TBE agarose gel.

Results and discussion

A line (referred to as plant 19479) with low levels of palmitic acid was identified from an ongoing mutant screen for altered fatty acid composition in the Williams-82 cultivar. Based on the uniform low levels of palmitic acid in seeds from self-pollinated individual M₃ plants, it was inferred that this line was homozygous. We observed that levels of palmitic acid were reproducibly and statistically significantly low, ranging from 5.9% to 6.4% over four seasons in the field, nearly a 50% reduction in palmitic acid (Table 1). We also observed statistically significant increases in oleic acid, particularly in 2016–2018, however the extent of increase in oleic acid levels was not consistent for all seasons examined.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Fatty acid profiles of FATB1a_Q52STOP mutants and wild type Williams-82 field grown seed.

https://doi.org/10.1371/journal.pone.0262327.t001

To determine if this mutant represented a new or previously characterized locus affecting palmitic acid content, plant 19479 was crossed to a line carrying a known allele of FATB1a (FATB1a_G180D) in which seeds typically contain 6.8% palmitic acid [12]. The F₁ plant was grown to maturity and fatty acid composition was measured in individual F₂ seeds. No wild type individuals were identified, and all individuals had low palmitic acid levels below 9%, which suggested that the lesion in the low palmitic line did not complement the reference mutation in FATB1a (Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Palmitic acid content in F₂ seeds from complementation test cross to FATB1a_G180D.

Fatty acid content was measured in 49 F₂ individual seeds.

https://doi.org/10.1371/journal.pone.0262327.g001

The FATB1a gene (Glyma.05g012300, version Glyma2.0) was sequenced as a candidate for causing the low palmitic acid phenotype. A single nucleotide polymorphism (C to T) consistent with NMU mutagenesis was identified at base position 154 in the predicted FATB1a transcript, which resulted in the introduction of an early termination signal in place of amino acid Q52 (Fig 2A). The full-length protein is 417 amino acids long, and this mutation is expected to result in a truncation early in the first exon (Fig 2B). We will refer to this allele henceforth as FATB1a_Q52STOP.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Polymorphism in FATB1a.

a. DNA and predicted amino acid sequence (blue) of the FATB1a gene (Glyma.05g012300) in Williams 82 soybean and the mutant. b. Early termination signal (*) occurs within the first exon of the predicted protein.

https://doi.org/10.1371/journal.pone.0262327.g002

We took advantage of the fact that the polymorphism caused the loss of a HhaI restriction site in the mutant to design a codominant polymorphic marker to genotype for the presence of the FATB1a_Q52STOP mutation. To determine if the mutation co-segregated with the low palmitic acid phenotype, the mutant line was crossed to the cultivar Prize and individual F₂ seeds were chipped for fatty acid phenotyping and genotyped. Palmitic acid values in individuals seeds in the F₂ population ranged from 5.6% to 16.8% palmitic acid. The FATB1a_Q52STOP mutation co-segregated with the low palmitic acid phenotype, and thus is likely causative (Fig 3). Homozygous mutant individuals ranged from 5.6 to 8.6% palmitic acid with a mean of 6.9%, which was slightly higher than the levels observed in field grown plant populations, and may be a result of differences between the field and greenhouse environments. Heterozygous individuals averaged 9.4% palmitic acid. Wild type individuals contained 11.3% palmitic acid, which was similar to the levels observed in field-grown seeds.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Co-segregation of FATB1a_Q52STOP and low palmitic acid phenotype.

Distribution of palmitic acid content and genotype for 80 F₂ individuals from an outcross. (-/-) homozygous for FATB1a_Q52STOP, (+/-) heterozygous for FATB1a_Q52STOP, (+/+) wild type FATB1a.

https://doi.org/10.1371/journal.pone.0262327.g003

The FATB1a_Q52STOP polymorphism results in a low level of seed palmitic acid, comparable to the low palmitic acid fap3 mutants previously described in soybean [8, 10, 12, 18, 19]. This mutation represents an early termination of the FATB1A protein. We observed in some years a significant increase in oleic acid levels in FATB1a_Q52STOP mutants, which has been observed in previous studies of fap3 mutants [19, 20], however this was not consistent for all years in this study, and elevated oleic acid levels were not associated with the FATB1a_Q52STOP genotype within the segregating population, and therefore may be a result of secondary mutations in this line. No agronomic or physiological abnormalities were observed in the FATB1a_Q52STOP plants; however, yield was not directly tested in large scale field experiments. Reduction in palmitic acid levels or mutation in KASIII has previously been associated with negative effects on seed yield [20, 21]. Further studies will determine the extent to which this allele can be used by breeders to reduce levels of palmitic acid to develop healthier soybean oils for the edible oil market.

The single nucleotide polymorphism in FATB1a can be readily followed with a PCR-based genotyping marker to facilitate introgression into elite germplasm. This stable and non-transgenic mutation can be used in the development of conventional soybean lines with reduced saturated fat content, alone or in combination with other seed composition traits.

Acknowledgments

The authors are grateful to James Held and Carrie Winklepleck for technical assistance. Mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the US Department of Agriculture. The USDA is an equal opportunity provider and employer.