Genotyping by Sequencing Reads of 20 Vicia faba Lines with High and Low Vicine and Convicine Content

Abstract

The grain faba bean (Vicia faba) which belongs to the family of the Leguminosae, is a crop that is grown worldwide for consumption by humans and livestock. Despite being a rich source of plant-based protein and various agro-ecological advantages its usage is limited due to its anti-nutrients in the form of the seed-compounds vicine and convicine (V+C). While markers for a low V+C content exist the underlying pathway and the responsible genes have remained unknown for a long time and only recently a possible pathway and enzyme were found. Genetic research into Vicia faba is difficult due to the lack of a reference genome and the near exclusivity of V+C to the species. Here, we present sequence reads obtained through genotyping-by-sequencing of 20 Vicia faba lines with varying V+C contents. For each line, ∼3 million 150 bp paired end reads are available. This data can be useful in the genomic research of Vicia faba in general and its V+C content in particular.

Dataset: The reads have been submitted to the European Nucleotide Archive (ENA) under the accession PRJEB38838.

Dataset License: CC-BY

1. Summary

The Protein Crop Strategy of the German Federal Ministry of Food and Agriculture has the goal of raising the importance of domestic protein crops e.g., legumes in Germany and Europe in order to improve ecosystem services and resource conservation as well as to reduce the dependency on imported crops [1]. The faba bean (Vicia faba) is a prime candidate for this strategy being a globally grown legume that has several agro-ecological advantages (N-symbiosis, rotation hygiene, and pollinator support) and serving as food for humans and livestock [2]. Regardless of these benefits its usage is limited due to the anti-nutrients vicine and convicine (V+C) that occur in their seeds. These compounds have negative effects to animals as well as to humans suffering from G6PD deficiency [3,4]. Despite ongoing research efforts and the discovery of a robust marker for the V+C content, the responsible genes and mechanisms remained unknown for a long time and the location of the locus could only be restricted to an interval on chromosome 1 of Vicia faba that shows conserved synteny with a region on chromosome 2 of the related species Medicago truncatula that is about 900,000 bp long [3]. Research has been exacerbated due to the lack of an annotated reference genome for Vicia faba, which is assumed to be about 13 Gbp [5]. In a recent preprint the authors have found an enzyme associated with V+C biosynthesis and identified it as a guanosine triphosphate (GTP) cyclohydrolase II, proposing the purine GTP as a precursor for vicine [6]. The breeding of novel low V+C varieties to improve the usage of Vicia faba as feed is the goal of the project Abo-Vici, which is supported by the German Federal Ministry of Food and Agriculture [7]. As part of this project we obtained reads from 20 Vicia faba lines with known V+C content through genotyping-by-sequencing (GBS). We offer this data here for the benefit of researches into Vicia faba and its V+C content in particular. We have so far successfully used this data for the prediction of regulatory regions in Vicia faba and the identification of regulatory single nucleotide polymorphisms (SNPs) that are associated with V+C content [8]. For this we built a partial genome for Vicia faba from the GBS reads that spanned ∼1% of the total genome and performed variant calling with it, which resulted in more than 600,000 high quality SNPs. This partial genome is available upon request from the corresponding author. Finally, while the data itself are not enough alone, the data can support the eventual creation of an annotated reference genome for Vicia faba.

2. Data Description

The sequence reads for 20 Vicia faba lines obtained through GBS are stored as paired end reads in two FASTQ files per Vicia faba line. These FASTQ files contain both the nucleotide sequence and its corresponding quality scores as text. Per sample ∼3 million 150 bp paired end reads are available, such that the total amount of sequence amounted to 18 Gbp. The uncompressed data required 51 GB of disk space. The sequences have been deposited at the European Nucleotide Archive (ENA) under the accession number PRJEB38838.

3. Methods

Plant Material and Sequencing

We obtained GBS data from 20 inbred lines of Vicia faba. The lines were inbred via single-seed descent from cultivars, from a gene-bank accession, from biparental crosses or from a landrace and include winter and spring types (see

Table 1. Vicine and convicine status of the 20 lines, name, sample ID, additional notes and European Nucleotide Archive (ENA) accession number under which the sample data is available.

ENA Accession	Sample ID	V+C	Line	Notes
ERS4652931	Sample_8	Low	Line 1268-4-1	Ancestor of low V+C content
ERS4652926	Sample_3	Low	Mélodie/2	cv. Mélodie; minor, spring bean
ERS4652927	Sample_4	Low	F7(Mélodie/2 x ILB938/2)-139-1-1	Near isogenic lines (ILB938/2 is from Ecuador)
ERS4652928	Sample_5	Low	F7(Mélodie/2 x ILB938/2)-201-3-1
ERS4652932	Sample_9	High	F7(Mélodie/2 x ILB938/2)-139-2-1
ERS4652933	Sample_10	High	F7(Mélodie/2 x ILB938/2)-201-4-1
ERS4652929	Sample_6	Low	F7[VC.14.8099-843-2-1]	Near isogenic lines from a breeder’s cross, spring beans
ERS4652930	Sample_7	Low	F7[VC.14.8099-848-3-1]
ERS4652934	Sample_11	High	F7[VC.14.8099-843-3-3]
ERS4652935	Sample_12	High	F7[VC.14.8099-848-4-1]
ERS4652924	Sample_1	High	HediLin-1	cv. Hedin; minor, spring bean
ERS4652936	Sample_13	High	PietraLin	Major, Mediterranean bean
ERS4652937	Sample_14	High	(HediLin/1 x PietraLin)-2-4	Near isogenic lines
ERS4652938	Sample_15	High	(HediLin/1 x PietraLin)-4-4
ERS4652939	Sample_16	High	S_281	Academic winter bean lines
ERS4652940	Sample_17	High	S_301
ERS4652941	Sample_18	High	S_034
ERS4652942	Sample_19	High	S_290
ERS4652925	Sample_2	High	Hiverna/2	cv. Hiverna; minor, winter bean
ERS4652943	Sample_20	High	Côte d’Or/1	Côte d’Or; minor, winter bean

for more information). Six of the lines had a low V+C content and 14 had high V+C content. DNA extraction, sequencing and filtering were carried out by LGC Genomics GmbH (Berlin, Germany). The DNA was extracted via LGC’s sbeadex livestock kit following the lysis protocol L for plant tissue from the grains of the plants. From each line two pooled grains were used. An extraction using the sbeadex plant kit was tested but provided poorer results than the livestock kit. For each library construction 100–200 ng of genomic DNA were used, which was quantified with a NanoDrop. The DNA was digested with 2 units of the restriction enzyme MslI (NEB, recognition sequence: CAYNN⌃NNRTG) in NEB4 buffer in 20 $\mathsf{\mu }$L volume for 2 $\mathrm{h}$ at 37 ${}^{\circ }\mathrm{C}$. The restriction enzyme was heat inactivated by incubation at 80 ${}^{\circ }\mathrm{C}$ for 20 $\min$. The TrueSeq adapter sequences used were:

• adapter_prefix_R1 ‘AGATCGGAAGAGCGGTTCAGCAGGAATGCCGAGACCGATC’
• adapter_prefix_R2 ‘AGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGTAG’

For the ligation 10 $\mathsf{\mu }$L of each restriction digest were transferred to a new 96well PCR plate, mixed on ice first with 1.5 $\mathsf{\mu }$L of one of 96 inline-barcoded forward blunt adaptors (pre-hybridized, concentration 5 $\mathrm{p}\mathrm{mol}/\mathsf{\mu }$L), followed by addition of 20 $\mathsf{\mu }$L Ligation master mix (contains: 15 $\mathsf{\mu }$L NEB Quick ligation buffer, 0.4 $\mathsf{\mu }$L NEB Quick Ligase, 7.5 $\mathrm{p}\mathrm{mol}$ pre-hybridized common reverse blunt adaptor). Ligation reactions were incubated for 1 $\mathrm{h}$ at room temperature, followed by heat inactivation for 10 min at 65 ${}^{\circ }\mathrm{C}$. After the ligation reactions the libraries were purified using Agencourt XP beads. Following that the libraries were size selected by a size selection on a LMP-agarose gel, removing fragments smaller than 300 bp or larger than 400 bp. For the final quality control, Fragment Analyzer and Qubit were used. The libraries were then amplified in 20 $\mathsf{\mu }$L PCR reactions using MyTaq (Bioline) and standard Illumina TrueSeq amplification primers. The number of cycles was limited to 14. Having each line uniquely barcoded the samples were pooled and ran on the same sequencing run. An Illumina NextSeq 500 V2 platform was then used for genotyping-by-sequencing. Demultiplexing of the libraries was done using the Illumina bcl2fastq 2.17.1.14 software. Finally sequencing adapter remnants were clipped using cutadapt 1.13+18 [9] and reads whose 5${}^{\prime }$ ends did not match the restriction enzyme site were discarded. As a last step FastQC reports (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/) were generated for the FASTQ files.