Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Genetic and comparative mapping of Lupinus luteus L. highlight syntenic regions with major orthologous genes controlling anthracnose resistance and flowering time

Abstract

Anthracnose susceptibility and ill-adapted flowering time severely affect Lupinus luteus yield, which has high seed protein content, is excellent for sustainable agriculture, but requires genetic improvement to fulfil its potential. This study aimed to (1) develop a genetic map; (2) define collinearity and regions of synteny with Lupinus angustifolius; and (3) map QTLs/candidate genes for anthracnose resistant and flowering time. A few linkage groups/genomic regions tended to be associated with segregation distortion, but did not affect the map. The developed map showed collinearity, and syntenic regions with L. angustifolius. Major QTLs were mapped in syntenic regions. Alleles from the wild parent and cultivar, explained 75% of the phenotypic variance for anthracnose resistance and 83% for early flowering, respectively. Marker sequences flanking the QTLs showed high homology with the Lanr1 gene and Flowering-locus-T of L. angustifolius. This suggests orthologous genes for both traits in the L. luteus genome. The findings are remarkable, revealing the potential to combine early flowering/anthracnose resistant in fulfilling yield capacity in L. luteus, and can be a major strategy in the genetic improvement and usage of this species for sustainable protein production. Allele sequences and PCR-marker tagging of these genes are being applied in marker assisted selection.

Introduction

Food security, soil fertility and sustainable food production can be significantly improved by the greater use and improvement of various grain legumes1 and especially Lupinus spp., which have been traditionally used in human and animal diets2,3. They contribute to the sustainability of cropping systems because of their low requirement for fertilizer and positive input to soil fertility. They achieve this via their efficient mobilization of soil phosphorus and their fixation of atmospheric nitrogen (through their symbiotic relationship with Bradyrhizobium)4,5,6. Lupins belong to the genus Lupinus, of the genistoids clade of the papilionoid legumes7. There are an estimated 267 species of lupin distributed around the Mediterranean region (¨Old World¨) and North and South America (¨New World¨)8. A number of lupin species have been used in cultivation; the major ¨Old World¨ cultivated species are Lupinus albus, Lupinus angustifolius and Lupinus luteus, while Lupinus mutabilis is the major “New World” species9,10. Based on their seed coat structure, the Old World lupins are divided into two groups; the smooth-seeded and rough-seeded lupins. The smooth-seeded lupins include at least two lineages, Angustifolius-Luteus and Micranthus-Albus11. Seed protein content varies between lupin species, with L. luteus (2n = 52) showing the highest values12. Its proteins allow the production of high-quality food and feed, its isolate has functional and physicochemical properties suitable for the health-food industry13,14,15, while in feed for the aquaculture sector, it is the most prominent species16. However, this species still needs to be further adapted in order to represent a realistic alternative in supplying the demand for plant protein. This is made more complex under factors of climate change that affect many aspects of agricultural systems, including; temperature, water availability, change in pathogen spread, flowering time and host susceptibility to pests17. It is well documented that the “rusticity” of lupin species is a limitation to their expansion as a crop. Also lupin breeding has faced a new challenge in the last few decades, the lack of anthracnose resistance. Indeed, in L. luteus, no elite germplasm has been reported with resistance, despite it being accepted that one of the best sustainable practices to counteract diseases is the development and use of disease-resistant cultivars. It is true that some breeding lines, that show partial resistance to the disease have been identified18. Anthracnose resistance is critical to improve yield in lupin, the disease has become a serious problem worldwide, causing significant yield losses, as high as 100%, and a major limiting factor for lupin production. It is caused by Colletotrichum lupini (Bondar)19,20.

Another essential trait in the crop adaptation processes of L. luteus is flowering time, where it is important to breed germplasm suited to specific environmental conditions, thus directly improving yield. This trait is cited as an example of the consequence of bottlenecks, where a reduction in genetic variation for flowering time has previously occurred during domestication process21. Species like L. luteus, with its recent history of breeding, is likely to have a greater lack of diversity because its domestication has been based on a small number of founding individuals, and subsequent strong and persistent selection for key traits, all of which have resulted in severe genetic bottlenecks. Thus, the adaptation of L. luteus now needs to take this into account. In addition, the deficiency of information on which molecular and genetic is enormous compared to many other crops. Even though some work has been undertaken on these aspects22,23,24, there are still many gaps which need to be filled to help the species improvement.

Having this in mind, our strategy has been the exploration of novel genetic variability in wild germplasm, using advanced genomics, developed with model plants and the reference genome from cultivated and related species. For instance, no gene or markers associated with anthracnose resistance have been reported, and only recently some flowering time QTLs have been identified in L. luteus24. Whereas, greater progress had been achieved in L. angustifolius, which has been proposed as a reference genome in comparative studies25,26. In this species, the cultivar Tanjil, has been widely used for breeding anthracnose resistance. Its resistance is controlled by a single dominant gene, Lanr1, mapped on linkage group (LG) NLL-11, and its sequence is localised in a single Scaffold_13326,27. In flowering time, a major gene (Ku) has been identified and mapped on LG NLL-10; which controls the vernalisation requirement28,29. Recent studies showed that this major Ku gene corresponds to the FLOWERING LOCUS T homologues (LanFTc1), and that a deletion in the promoter region of it is responsible for the loss of vernalisation requirement25,30.

Synteny has been reported between the reference genome of L. angustifolius, with L. albus, and other model legume species26,29,31, which highlights the possibility of synteny and collinearity with L. luteus, since both species have been reported to share the same lineage and clade11.

In order to add more molecular and genetic information to this species, in this study a first attempt was made to develop the genetic linkage map of the L. luteus genome, and to compare it with L. angustifolius. NextRAD genotyping (Nextera fragmentation with Restriction Associated Digestion) was used. This technology utilizes a Nextera (Illumina, Inc.) reaction to fragment genomic DNA and then amplification using modified primers which are complementary; therefore, only fragments containing this recognition site are amplified. The result is a randomized collection of DNA fragments that represents a sub-fraction of the tested genome32,33. Applications using NextRAD include studies of genetic diversity of the Andean lupin, Lupinus mutabilis34, and an insect33,35. This, together with the phenotypic exploration of novel genetic variability for anthracnose resistance and flowering time genes, was used to go forward. Thus, a large mapping population from a cross between a L. luteus wild accession with an elite variety was used to explore and dissect these relevant traits. The main goal of this study was to (1) develop the genetic linkage map of L. luteus using de novo genome assembly and NextRAD genotyping; (2) define collinearity and regions of synteny with the reference genome L. angustifolius; and (3) map QTLs/candidate genes for anthracnose resistant and flowering time, in order to identify important orthologous candidate gene for both traits: key genetics factors for further developing the production of sustainable plant protein.

Materials and methods

Plant materials and mapping population

To develop the L. luteus genetic linkage map, a mapping population of 188 F2 individuals was generated by crossing AluProt-CGNA with Core 98 (PI385149). The female parent, AluProt-CGNA (cultivar developed in Chile by CGNA and now grown in southern Chile) is a very early flowering variety15, with good yield but is susceptible to anthracnose. The male parent, the CGNA’s Core 98, is a wild accession, which has shown resistant to anthracnose and very late flowering time. The cross was made under greenhouse conditions (16-h light, 23 °C and 60–75% relative humidity). The F1 was grown and selfed to obtain the F2 mapping population of 188 individuals. In order to develop the L. luteus genetic linkage map, and to carry out QTL analysis, an F2:3 population was then developed by selfing each of the F2 individuals, generating 188 F2:3 families. Each family was then divided in two, generating two populations, one was used to evaluate flowering time (188 F2:3 families) and the other (100 F2:3 families) to evaluate anthracnose resistance. Both F2:3 populations were evaluated under field condition, in order to validate the F2 phenotypic data and QTLs for these traits.

Isolation of genomic DNA and development of NextRAD markers

Young leaves were collected from each F2 individual of the mapping population and the two parental lines. DNA extraction was based on a modified CTAB protocol developed at CGNA, as described by Parra-Gonzalez et al.22. Extracted DNA was quantified using fluorometric quantification (Qubit) and diluted to 50 ng/μl. Genomic DNA was converted into NextRAD genotyping-by-sequencing libraries by SNPsaurus (SNPsaurus, LLC)32. Genomic DNA was first fragmented with Nextera reagent (Illumina, Inc), which also ligated short adapter sequences to the ends of the fragments. The Nextera reaction was scaled for fragmenting 20 ng of genomic DNA. Fragmented DNA was then amplified as follows: 26 cycles at 73 °C, with the primer matching the adapter and extending 9 nucleotides into the genomic DNA with the selective sequence GTGTAGAGC. Thus, only fragments starting with a sequence that hybridized with the selective sequence of the primer was efficiently amplified. The NextRAD libraries were sequenced on an Illumina HiSeq 4000 instrument with two lanes of 150 bp single-end reads (at the University of Oregon). Parental lines were sequenced separately to develop the short-read reference sequence to map the data from the F2 population. Thus, marker data can be collected without the need to produce a high quality reference genome. Bioinformatics analysis of data including SNP calling was provided by the NextRAD service provider using a reference-independent pipeline (SNPsaurus, LLC). Briefly, the reads were trimmed using bbduk (BBMap tools, https://sourceforge.net/projects/bbmap/) with the following parameters: bbmap/bbduk.sh in = out = ktrim = r k = 17 hdist = 1 mink = 8. The parental samples were used to identify a set of polymorphic loci and to generate a set of reference DNA sequences in the FASTA text-based format. The reads were then mapped to this FASTA file with an alignment identity threshold of 0.98, using bbmap. Genotype calling was achieved using call variants (BBMap tools). Genotype imputation was not used. A genotype table was provided for use with JoinMap (see below), along with a Variant Call Format (VCF) file of data from all samples, and a FASTA formatted file containing the short-read sequence common to both parents. These data were used to evaluate sequencing coverage of the sample. Sequencing depth (DP) at each marker for each sample was evaluated using DP values from the Variant Call Format (VCF) file that contained nucleotide variation predicted by the software. The heatmap of sequencing depth at each marker for each sample was created using DP values from the VCF file, using heatmap.bp in R package vcfR with zero values omitted (https://www.R-project.org/)36. Mean DPs for all F2 samples were plotted, using the barplot function in R.

De novo genome assembly and PCR marker development

Additional molecular markers were developed using L. luteus scaffold sequences created from low-coverage whole genome sequencing of the parental line Core 98, by sequencing-by-synthesis using Illumina Hi-Seq 2000, followed by de novo assemblage, using SOAPdenovo2. Scaffold statistics were prepared using QUAST37. Fastq files in order to prepare scaffold sequences and raw sequence data used to create NextRAD markers, which are deposited in the NCBI Sequence Read Archive (BioProject accession PRJNA594652, https://www.ncbi.nlm.nih.gov/sra/PRJNA594652). The scaffold sequence was then subjected to BLAST with the sequences of some PCR markers already mapped on the L. angustifolius map25,29. The markers where those evenly distributed over the L. angustifolius map. Analysis of the DNA sequences between the species was carried out using Geneious v.6.1.8 software (https://www.geneious.com)38. Primers were designed using the algorithm Primer3. Amplifications were carried out in 20 ul PCR reactions containing 100 ng of genomic DNA, 0.2 mM dNTPs, 2 mM MgCl2, 1X PCR buffer and 2.5% DMSO (only for SSR (Simple sequence repeats) markers), 1U Taq polymerase (GoTaq G2, Promega) and 2.5 mM of each reverse and forward SSR primer; and 1 mM of each reverse and forward SNP primer. The PCR protocol was as follows: 95 °C for 5 min, 30 cycles of 1 min at 95 °C, 1 min at 58 °C, 1 min at 72 °C, and a final extension at 72 °C was performed for 5 min. SSR PCR products were separated on 6% denaturing polyacrylamide gels (urea 8 M) run in 1X TBE buffer at 60 watts for 3–4 h and their polymorphised products visualized using silver stain procedures22,39. SNP PCR products were separated on 8% polyacrylamide gels using Asymmetric Single Stranded Conformational Polymorphism technologies (SSCP)40, run in 1X TBE buffer at 2–3 watts for 16–18 h and polymorphism visualized using silver stain procedures. The mapping population was genotyped with polymorphic co-dominant PCR markers, identifying parental loci as homozygous (allele A or B), and the heterozygous loci constitution (allele H). The allelic constitution for each F2 individual was entered on to a matrix in Excel (Microsoft Corporation) for linkage analysis.

Construction of genetic linkage map

Pairwise analysis, grouping of markers and mapping, were performed with JoinMap 4.041. Single locus analysis (JoinMap Single Locus Analysis, JMSLA) was applied to detect possible markers or regions with segregation distortion. Since distorted markers were detected, two marker datasets were defined: one set contained all markers, and the other having markers without significant segregation distortion. Linkage analysis and mapping was carried out with both datasets. This was in order to establish the most accurate map and QTL analysis. The map was constructed based on recombination frequencies and LOD values. The markers were assigned to LGs based on (modified) LOD scores of pairs of markers. In the grouping of markers, LOD values from 1 to 8 were used to detect the stability of grouping. Markers in the group were analyzed for pairwise linkages with initial set up values of REC and LOD thresholds of 0.499 and 1, respectively. Strong linkage was considered to be present with a REC smaller than 0.01 or a LOD larger than 10. A map distance was calculated using Kosambi’s mapping function42.

Comparative mapping of L. luteus and L. angustifolius along with physical location of markers

The 150 bp marker sequences from NextRAD sequencing along with PCR marker sequences were ordered according to the LGs prepared with JoinMap. These were used as queries against the Lupinus angustifolius reference genome (LupAngTanjil v1.0) in a local megablast search where the top scoring hits were retrieved (blastn-task megablast-query-db-outfmt 6-num_alignments 1-out). Chromosome name, start and end positions were extracted with awk. Synteny block size was calculated by counting the distance in base pairs between adjacent markers that mapped to the same L. angustifolius chromosome. Alignments of L. luteus markers to L. angustifolius unassigned scaffold sequences were not considered for this calculation. Circular plots were prepared using circos43, and lollipop chart using R package ggplot244. The L. angustifolius genome build contains 13,564 scaffold sequences. For clarity of graphical representation, alignments to these regions were removed from the circos plots. Linear comparisons were prepared using starting base pair positions of BLAST hits on L. angustifolius chromosomes drawn using the R package LinkageMapView and manually connected to the L. luteus genetic map45. In order to define collinearity and marker order between L. luteus map and L. angustifolius map, common markers anchored in both maps were compared along LGs. The order and map interval were then calculated.

Phenotypic evaluation and data analysis for anthracnose resistance and flowering time

The mapping population of 188 F2 individuals, together with the parents, were assessed by QTL analysis for flowering time and anthracnose resistance. Flowering time, measured as days to flowering (DTF), represents the period of time from sowing until the first whorl was fully open. The F2 population was evaluated for DTF under field conditions at Vilcún (Southern Chile, La Araucanía Region, 38° 41′ 44.04″ S and 72° 25′ 1.94″ O), over the winter growing season 2013–2014. The segregating population of 188 F2:3 families was evaluated for DTF in field conditions at Huichahue (Southern Chile, La Araucanía Region, 38° 50′ 6.30″ S and 72° 31′ 2.22″ O) over the winter growing season 2016–2017, using a completely randomized block experimental design with three replications. Each replication comprised 20 plants.

Anthracnose resistance was evaluated in in vitro conditions as describe by Cuccuza and Kao46 using the 188 F2 individuals of the mapping population. Colletotrichum lupini var. setosum19 was isolated from infected plants of cultivar AluProt-CGNA, collected from different locations in the southern region of Chile, a major area of lupin production. Colletotrichum lupini was used, since in southern Chile it has been widely reported as the only detected and causal agent for anthracnose in lupin20. All fungi collected showed cultural and morphological features of Colletotrichum lupini var. setosum as reported by Nirenberg et al.19. Infected plant organs were surface-sterilized (water for 20 min, 35% ethanol, 0.5% sodium hypochlorite for 1 min respectively and three washes with sterile distilled water for 1 min) and placed onto potato dextrose agar (PDA). Cultures were then grown on PDA at 25 °C until fungus growth was observed (3–5 days). Disks from the edge of the active colony growth were transferred aseptically to new Petri dishes with PDA media. These cultures were incubated under the same conditions for 7–10 days. Two cotyledons from each of the 188 F2 and parental lines were collected. The inoculation was performed with a conidial suspension of 1 × 106 conidia per ml into an injury of 2 mm length on the upper surface of the cotyledon. The inoculated cotyledons were incubated in a moistened Petri dish at 23 ± 2 °C under 16-h photoperiod of white fluorescent light (2000 lx). The evaluation was carried out 10 days after inoculation. To evaluate the degree of damage, a scale from 1 to 5 was used. Cotyledons were given the following scores: Score 1 when they exhibited a spot of soft yellow color in the area of inoculation. Score 2, a spot of necrosis in the area of inoculation. Score 3, a little localized hypha (diameter less than 2 mm). Score 4, hyphae with diameter more than 4 mm. Score 5, abundant presence of fungus and tissue degradation. In order to validate the anthracnose resistance observed in the F2 mapping population under in vitro assays, 100 F2:3 families derived from the F2 mapping population were tested for anthracnose resistance under field conditions, as described by Fischer et al.47. A completely randomized block experimental design was used with three replications, where each replication was represented by 20 plants of each F2:3 family sown in two rows (10 plants per row planted 10 cm apart and 20 cm between rows), and one infection row of 10 plants (seed of the susceptible parent cv. AluProt-CGNA was inoculated with the suspension of strain Colletotrichum. lupini, as in vitro assay) in the middle, to obtain a high and permanent infection pressure. The resistant parent (Core 98) was also included as a control, and similarly to F2:3 families, was sowed with an infection row. Infection rows were sown two weeks after sowing the F2:3 families. The data were collected from all the plants in all the plots, scoring the resistant plants (non-infested) and susceptible plants (infested), as described by Fischer et al.47.

Analysis of variance (ANOVA) (SAS Institute, Cary, NC, USA) was carried out to explore the phenotypic variation of anthracnosis resistance and DTF in the F2:3 families. Estimates of the extent of genotypic and phenotypic variation calculated following Burton and DeVane48 as well as the broad sense heritability (H2)49. Treatment means were separated by Tukey test for P ≤ 0.05.

Validation of major QTLs

When a major QTL, either for anthracnose resistance or DTF, was detected and mapped in the F2 population, markers previously reported as co-segregating for these traits in L. angustifolius and mapped in this study in L. luteus with highest LOD score, were chosen to validate the QTLs in the F2:3 families. Thus, tracing back the genetic classification of the F2 population at both loci and validating QTLs already mapped. Markers co-segregating with these traits were selected for further application in MAS.

QTL analysis and mapping

DTF and anthracnose resistant segregation were evaluated in the F2 population, assuming the presence of one main locus and a Mendelian segregation of 3:1. Significant deviations were tested using the goodness of fit chi-square test (X2), where rejection of the 3:1 segregation was applied at P < 0.05. In the first instance an inspection for the presence of a QTL was carried out using a non-parametric approach, Kruskal–Wallis test, interval mapping (IM) was then applied. Both methods were implemented in MapQTL 650. In IM LOD thresholds with significance level (P < 0.05) were empirically determined for each trait using the permutation test (1000 iterations). Once a single QTL was detected and mapped, in order to search for others minor segregating QTLs, further analysis was then carried out using Genome-wide Composite Interval Mapping (GCIM)51, implemented by the R package QTL.gCIMapping.GUI v2.0 (https://cran.r-project.org/web/packages/QTL.gCIMapping.GUI/index.html)52. Random model was used with a walk speed for genome-wide scanning of 1 cM, and the LOD score thresholds of 3 for significant QTLs.

To identified candidate genes in the QTL regions for anthracnose resistance and DTF, a search was carried out in the Lupin Genome Portal (https://www.lupinexpress.org), through BLAST in the nucleotide and protein database (Narrow-leafed lupin genome scaffold assembly v1.0, Narrow-leafed lupin annotated gene v1.0)26 and the NCBI database (https://www.ncbi.nlm.nih.gov/) was also used in order to confirm suggested candidate genes.

Results

Genetic linkage map construction

By using the NextRAD sequencing approach, 834 polymorphic sites were identified (Supplementary Table S1), which generated 685 high quality SNP markers. A heatmap analysis revealed parental lines with high coverage (Supplementary Fig. S1). Mean coverage in F2 individuals ranged from 0.15 to 155.08 (Supplementary Fig. S2, Supplementary Table S2). 12 F2 individuals with a high percentage of missing data were identified as assay failures and removed (Supplementary Fig. S1). The adjusted average mean coverage for all F2 individuals was 63.13. Mean per marker coverage ranged from 26.47 to 254.20 (Supplementary Table S2). Mean coverage for the parental mapping population, AluProt-CGNA and Core 98 was 495.44 and 433.65, respectively.

Markers developed from the de novo assembled L. luteus genome revealed 311,789 scaffolds ≥ 500 bps, with N50 = 2495 and L50 = 49,228 (Supplementary Table S3). The BLAST results for these scaffolds with genomic sequences of some PCR markers of L. angustifolius map generated 28 additional analogous PCR markers in the L. luteus genome (Supplementary Table S4).

A genetic linkage map was constructed with high LOD score (LOD = 6), which allowed detection of all markers with strong linkage, and stable marker order per LG. The map fell into 26 LGs as expected for the L. luteus genome, with a total map length of 1,772 cM and 744 loci with a mean density of one marker every 2.8 cM. The map had the largest gap on LG 20 (33.8 cM) and the smallest gap on LG 6 (3.1 cM) (Fig. 1, Table 1). The loci were identified with co-dominant markers, SNPs, SSR and INDEL. LG length ranged from 132.7 cM to 26.1 cM, with a mean of 68.2 cM. The highest number of loci mapped per LG was 47, while the lowest was 3, with a mean of 28.6 markers per LG (Table 1). Two markers, linked together, did not link to LGs and 0.9% of markers remained unlinked. LGs showed a cluster of markers and longer intervals in specific genomic regions, which is likely to reflect an uneven distribution of recombination frequencies along L. luteus chromosomes (Fig. 1).

Figure 1
figure 1

Genetic linkage map of L. luteus genome comprising 744 loci distributed among 26 linkage group (LG1–LG26).

Table 1 Description of basic characteristics for 26 linkage groups in F2 mapping population of L. luteus.

Effect of marker segregation distortion on map construction

Single locus analysis detected 5.9% of markers with highly significant (P ≤ 0.001) distortion, and 7.3% with significant (0.001 < P < 0.05) segregation distortion, giving a total of 98 markers with distortion from the expected Mendelian segregation (Table 1). Among LGs, 61.5% had a maximum of 9.7% of markers with segregation distortion; LG2, LG6, LG7, LG21 and LG25 did not have distorted markers, and LG10, LG19 and LG23 had one distorted marker. LG17 (88.9% of markers with segregation distortion), LG26 (62.5%), LG15 (32.4%), LG18 (25%), LG5 (20%), LG13 (16.7%), LG4 (16.2%), LG24 (16%), LG8 (15.2%) and LG20 (13.8%) had the major proportion of markers with segregation distortion (Table 1, Supplementary Table S5). To see whether or not these distorted markers affected the map, a second run for linkage analysis and mapping using a marker dataset that excluded the distorted markers was carried out. As expected, 26 LGs were again generated, and map coverage was reduced by 13% (231 cM). However, LG17, LG9, LG4 and LG26 reduced their map length by 88.1%, 75.5%, 47.5% and 33%, respectively, and compromised 46% of the total map length reduction. LG17 was severely affected by removing distorted markers, it was almost unmapped, since just the minimum number of markers remained to establish linkage. This LG had the largest distorted genomic region (30 cM) with a cluster of markers (16 out of 18), with very significant (P ≤ 0.001) segregation distortion. This distortion was found to share a common feature, all allele frequencies of the 16 loci in the 188 F2 individuals were skewed toward homozygosity of the female allele (mean frequency of 42.7%), and only a mean frequency of 13.2% of homozygotes with male alleles, (Supplementary Table S6). The large map length reduction of LG9 is due to only two distorted markers. LG3 showed no reduction in its map length, and LG8, LG10 and LG23, their map lengths were reduced by 0.6%, 0.2% and 0.2%, respectively. LG24, despite 16% of its markers being removed, was almost unaffected in map length (1.3% of reduction). The largest map gap had minor changes (equal to an overall mean of 17 cM) and mean map genetic distance was only reduced in 0.1 cM. Marker order was almost unaffected, and those collinear markers and syntenic regions with L. angustifolius map and reference genome, presented equivalent map positions and order with respect to the map, including all markers (see below). Thus, the map involving all markers, was used for the further analysis of comparative mapping and QTL analysis.

Comparative analysis of L. luteus with map and reference genome of L. angustifolius

Comparing the L. luteus map with the referential consensus L. angustifolius map, which has been highly saturated, represented about 72% of the total map length, which is interesting considering that the loci mapped in L. luteus genome represents about 25.4% of the loci mapped on the consensus map of L. angustifolius. Markers already mapped in L. angustifolius were mapped onto the L. luteus map, representing 3.1% of loci, where LG11, LG13 and LG23, have 4, 5 and 5 common loci, respectively. Collinearity was observed between both species, i.e. L. luteus LG13, LG23 and LG11, were collinear with LG11, LG10 and LG17 of the L. angustifolius reference genetic map, respectively (Fig. 2a,c,e). Markers identified loci in similar genetic positions, in the same order, but with different map length intervals. Clearly L. luteus LGs had longer intervals map than L. angustifolius in one distal genomic region, with shorter intervals length at the other end. However, LG13 and LG11 of L. luteus which compare with LG11 and LG17 of L. angustifolius, respectively (Fig. 2a,e), exhibit equal cluster features of markers at one end of both LGs, which might indicate that both could share unequal chromosome arm lengths, indicating possible submetacentric chromosomes. Moreover, LG23 of L. luteus shared the same marker distribution with LG10 of L. angustifolius, with a clustering of markers in the central region (Fig. 2c), which could suggest similar metacentric features in this chromosome.

Figure 2
figure 2

Comparative mapping and physical position of marker sequences of LGs of L. luteus in pseudochromosome (Chr) of L. angustifolius26. (a,c,e) collinearity between LGs of both lupin species. (b,d,f) Syntenic genomic regions with pseudochromosome of L. angustifolius in black lines, and in red lines regions containing genes controlling anthracnose resistance and DTF.

A comparative analysis of marker sequences was performed with the L. angustifolius genome. In total, 56.7% of the marker sequences in L luteus produced homologous alignments with L. angustifolius pseudochromosomes. Of these, 28% aligned to scaffold sequences of L. angustifolius. Considering this sequence homology, L. luteus LG would correspond to the following L. angustifolius pseudochromosomes (with their relevant proportions): LG1-Chr6 (37.5%), LG2-Chr8 (62.5%), LG3-Chr1(14.3%), LG4-Chr6 (15.8%), LG4-Chr4 (15.8), LG5-Chr2 (43.8%), LG6-Chr14 (20%), LG6-Chr17 (20%), LG7-Chr14 (100%), LG8-Chr12 (57.1%), LG9-Chr13 (45%), LG10-Chr20 (38.9%), LG11-Chr17 (50%), LG13-Chr11 (53.6%), LG14-Chr3 (20%), LG15-Chr5 (58.8%), LG16-Chr15 (30.4%), LG17-Chr3 (36.4%), LG18-Chr18 (30.8%), LG19-Chr4 (33.3%), LG20-Chr7 (31.3%), LG21-Chr1 (43.8%), LG22-Chr16 (31.3%), LG23-Chr10 (60%), LG24-Chr8 (44.4%), LG25-Chr19 (26.7%) and, LG26-Chr9 (40%) (Fig. 3, Supplementary Fig. S3, Supplementary Tables S7, S8). Within L. luteus LG, LG13 showed two syntenic regions, a short one: 2315179 bp, and a larger one: 7775501 bp (Fig. 2b). LG23 gave the largest syntenic region with 15558191 bp (Fig. 2c). LG11 showed two syntenic regions, a short one: 5294198 bp, and a larger one: 7978928 bp (Fig. 2f). LG1 and LG4; LG4 and LG19, shared syntenic region with Chr6 and Chr4 of L. angustifolius, respectively (Supplementary Tables S7, S8).

Figure 3
figure 3

Syntenic link between linkage group (LG) of L. luteus and pseudochromosomes (Chr) of L. angustifolius. Pseudochromosomes and linkage groups are not drawn to scale.

Mapping QTLs for anthracnose resistance and DTF in L. luteus

Clear and contrasting phenotypes for anthracnose resistance and DTF were observed in the parents and mapping population, allowing unambiguous phenotyping of the F2 mapping population. The mean frequency distribution for anthracnose resistance and DTF in the 188 F2 individuals followed a typical 3:1 Mendelian segregation (X2 = 0.1134, P = 0.75; X2 = 0.2553, P = 0.75 respectively, Supplementary Fig. S4), i.e. three classes of resistance level and flowering time, thus, as would be expected, they did not show normal distributions, as typifying the presence of a single dominant genes for anthracnose resistance and early flowering (Supplementary Fig. S4). The mean performance of the F2:3 population for anthracnose resistance, and the mean performance of the F2:3 population for DTF, evaluated under field condition, gave highly significant (P < 0.05) differences between families, and no significant effect due to replications (Table 2). Looking at each F2:3 family of each population for each trait, three phenotypic classes were observed, either for anthracnosis resistance or DTF: full resistance, ¾ resistance or fully susceptible and early flowering, ¾ early flowering or late type, which validated the genetic constitution of the loci under study as inferred from the F2, and the allelic dominance displayed in the next generation. High values of genotypic and phenotypic variances were observed for each trait. Broad sense heritabilities (H2) were obtained, and considering that H2 captures the proportion of the total variance due genetic effect, the high heritability values obtained confirm the strong genetic effect determining each trait (Table 2), which is in agreement with the result obtained from the F2 mapping population.

Table 2 ANOVAs for differences in anthracnose resistance and DTF, along with population parameters of the F2:3 families.

A major QTL for anthracnose resistance was identified and mapped in the L. luteus LG13 in the F2 population (Figs. 2, 4a). The non-parametric, Kruskal–Wallis test, revealed that markers mapped between 67.7 and 73.3 cM of LG13, were highly significantly associated with the phenotypic variation of the trait (P < 0.0001) (Supplementary Table S9). Interval mapping confirmed this result; a significant QTL was mapped in the same genomic region, and the marker sca82470, which mapped to a position of 67.7 cM, gave the highest LOD score (LOD = 56.5) in F2 (Table 3). As expected, the alleles from the wild, resistant parent at this locus explained 75% of the phenotypic variance of the trait in the F2 population (Table 3). Similarly, a single major QTL was also identified for DTF. The markers localized between 68.1 cM to 69.6 cM of LG23 were highly significantly associated with the variation in DTF (P < 0.0001) by Kruskal–Wallis test (Supplementary Table S9). Interval mapping detected a single major QTL, where marker sca43802, mapped at the position of 68.1 cM, presented the highest LOD score (LOD = 68.4) in the F2 population and explained 81.4% of the phenotypic variance for DTF (Fig. 2, Table 3). In both genomic regions neither clusters or distorted markers were mapped, thus giving greater confidence in the QTL positions.

Figure 4
figure 4

QTLs for anthracnose resistance (a) and DTF (b) in the F2 mapping population of L. luteus using QTL.gCIMapping.GUI. The LOD score of QTLs are marked in red line, and − log10(P value) for additive and dominance, are marked in grey and green lines, respectively. Black solid horizontal line is the LOD thresholds for significant QTLs. Candidate genes and flanking markers are shown by blue and black colors, respectively.

Table 3 Anthracnose resistance and DTF QTLs in F2 and F2:3 population of L. luteus, detected by interval mapping and genome-wide composite interval mapping.

Despite the phenotypic performances showing clear evidence of a single major QTL for each trait, a further analysis was carried out to search for any minor segregating QTLs. By using a GCIM approach, no other minor segregating QTLs were detected in this F2 mapping population for anthracnose resistance, and the major QTL was detected (LOD = 61.9) and mapped to the same position as Interval Mapping (Fig. 4a, Table 3). However, for DTF three minor QTLs were detected and mapped on LG4 (LOD = 4.7), LG10 (LOD = 5.1) and LG11 (LOD = 12.5), together with the major QTL which was detected (LOD = 94.5) and mapped in the same position as Interval Mapping (Fig. 4b, Table 3).

In order to validate these major QTLs confirmed by three mapping methods, the marker sca82470 for anthracnose resistance and marker sca43802 for DTF, were used to validate each QTL (Table 3). These markers were then used to trace back the allelic constitution in both traits in each F2:3 family. The results showed that F2 plants had a good correlation between genotypic and the phenotypic values; it was noted that as expected for a single dominant gene, the LOD values, when based on mean performance and classes of F2:3 families, gave lower relative values to those based on individual F2 plant (Tables 2, 3). Thus, validating both QTL position on LG13 and LG23 in the L. luteus map.

Candidate genes anthracnose resistance and DTF

By searching in the Lupin Genome Portal, nucleotide BLAST results on marker sca82470, which was highly associated with anthracnosis resistance in L. luteus in this study, produced a hit of 90.86% identity (E value = 0.00) to Scaffold_133 (Narrow-leafed lupin genome scaffold assembly v1.0) on pseudochromosome 11, 3376319–4744591 bp (Fig. 2b), where the Lanr1 gene for anthracnose resistance has been localized26. The marker sequence of sca82470 predicted gene Lup005048. A further BLAST search on NCBI predicted L. angustifolius uncharacterized LOC 109360795 (LOC109360795) in this marker, with a 91.71% identity (E vale = 0.0).

DTF nucleotide BLAST result for marker sca43802 of L. luteus produced a 77.97% identity (E value = 0.00) to Scaffold_276_44 (Narrow-leafed lupin genome scaffold assembly v1.0) localized to pseudochromosome 10 from base positions 8,013,021 to 8,250,716 in L. angustifolius (Fig. 2d). The LanFTc1 gene is localized to Scaffold_276_44 in L. angustifolius26. The marker sequence of sca43802 was homologous to LanFTc1, Lup015264. NCBI BLAST search in this marker predicted the presence of L. angustifolius protein TWIN SISTER of FT-like (LOC109357767), with a 93.13% identity (E value = 0.0). Out of three minor QTLs detected by GCIM, marker sca409063 flanking QTL on LG4, predicted the L. angustifolius floral homeotic protein AGAMOUS (LOC109345242) with a 97.91% identity (E value 2e-110) to L. angustifolius. With the marker 13561_55_79 flanking QTL on LG10, no significant similarity was found in NCBI database. The marker sca172949 flanking QTL on LG11, predicted a L. angustifolius CBS domain-containing protein CBSX1, chloroplastic-like (LOC109330470).

Discussion

Lupinus luteus has high seed protein content with values 60% (DM) in dehulled seed have been achieved in south macro area of Chile15. This large area provides a favorable environment for lupins, with deep volcanic soil, deficient in phosphorus (P) but with abundant organic matter and high rainfall53, which helps explain the good performance of L. luteus. Lupins have the potential to mobilize scarcely available nutrients, in particular P and micronutrients, for themselves or subsequent crops. Their ability to symbiotically fix gaseous nitrogen is widely acknowledged as a factor contributing to soil improvement4,5,6. This natural adaptation for nutrient acquisition, not present in most major crops, is highly relevant for a sustainable agriculture, helping to face the global challenge of food security, with lower fertilizer and water footprints. However, elite germplasm of L. luteus is susceptible to anthracnose disease in high rainfall areas, meaning its yield can be dramatically reduced19,54,55. Even more, fluctuating temperatures, frost, day length, and other climate changes, means flowering time is another important trait. Breeding efforts toward combining optimal expression of these two key traits: anthracnose resistance and early flowering time/no vernalisation requirement, is indeed essential to allow expression of the yield potential in elite germplasm of this species. In L. albus this strategy has proved successful, despite the complexity, because of a low frequency of early flowering progenies and the apparent quantitative nature of anthracnose resistance in this species20. Flowering is a vital stage in plant development; it plays an important role in the initiation of grain setting and is highly sensitive to stresses56. The existence of DTF variability allows selection to maximize yield by optimizing plant phenology in different environments. It allows selection for better crop adaptation in different climatic conditions and different geographical regions57. Common flowering pathways and a number of highly conserved genes described across species have suggested a tight genetic control for this trait58.

Here, we report novel QTLs harboring a very early DTF gene from elite germplasm, and a single dominant gene Lanr1, from a wild accession, for anthracnose resistance in the L. luteus genome. These QTLs were mapped in a large F2 segregating population, with three QTL mapping approaches, consistent with previous studies of these traits in model legume plants and in the reference genome30,59. It is interesting that three significant minor QTLs associated with DTF, were detected only with the GCIM approach, but only one of these QTLs was harboring a candidate gene associated with flowering. The sequence of the marker sca409063 flanking this QTL, predicted the L. angustifolius Floral homeotic protein AGAMOUS (AG). The AG gene encoding a MADS-box transcription factor has been reported in plant flowering regulation in different species60,61. It is also relevant to point out that these significant minor QTLs were detected with GCIM-random methods, which is in agreement with Wen et al.51. Who have demonstrated the power of this method in the detection of QTL in F2 population.

Anthracnose resistance and DTF are a key combination of genes, and will allow better adaptation and thus fulfilment of the yield potential of L. luteus elite germplasm. This is even more relevant if it is noted that this species already has important natural adaptation for nutrient acquisition, allowing more sustainable agriculture. The mapping of Lanr1 gene in L. luteus, allows interesting projections, since it was first identified in L. angustifolius, in which the cultivar Tanjil has been widely used for breeding anthracnose resistance in this species, because this single dominant gene had proved to show durable resistance26,59.

Our key result was the development of the genetic linkage map of L. luteus genome, which showed collinearity with the L. angustifolius reference map and syntenic genomic regions harboring major QTLs for these important traits. The advance of Next Generation Sequence technology and its wide application, together with genomic knowledge of the L. angustifolius reference genome, have greatly reduced the cost of nucleotide sequencing and facilitated our work. The work presented here may be the first example of using NextRAD markers to develop a genetic map in plants. The NextRAD markers were developed with a PCR step utilizing an oligo with a nine-nucleotide selective sequence to further reduce genome complexity. It is therefore expected that a higher density map of L. luteus can be directly prepared with more markers by simply removing or reducing this base-discrimination step. While the current map was suitable to identify candidate genes in two traits, a higher density map may be desirable for breeding future traits. We therefore conclude that NextRAD is a suitable approach for SNP discovery for genetic map development. The markers developed from the de novo assembly of L. luteus genome helped the comparative mapping studies with the reference map of L. angustifolius, by anchoring common markers in the maps of both lupin species, which confirmed LGs, map and syntenic regions found in this species.

Iqbal et al.24 recently described GBS for L. luteus whereby low-coverage sequencing data and imputation resulted in 948 selected SNP markers. The GBS markers from that study were combined with 2,006 DArT markers to produce a genetic map containing 40 LGs. Cytological studies of L. luteus support 2n = 52, or 26 LGs11, 62, thus the comparisons of this L. luteus map with the map generated in our study is complex. The map generated in this study represents 78.4% of the map length generated by Iqbal et al.24. Who covered a total length of 2,261 cM.

The mapping population used in this study had a wild male parent, containing valuable genetic variation for many traits. Interestingly, our results highlight that this cross with the wild accession also had the natural phenomenon of segregation distortion as observed in other species and crops63. According to the distribution of distorted markers along linkage groups of L. luteus, two types were identified: segregation distortion loci (SDL), i.e. loci widespread between chromosomes, with alleles in either parental class, and a segregation distortion region (SDR), i.e. markers showing highly significant distortion skewed in the same direction, clustered in a specific genomic region. SDRs, as found in LG17, with markers highly skewed toward the female parent, and in LG15, with markers highly skewed toward homozygosity i.e. to both parental types. An example of SDR is showed in Supplementary Table S6. This is the first reported distribution and direction of distorted markers in L. luteus, and shares the same features as other species and crops widely reported64,65. However, its genetic causation remains to be studied in L. luteus, as they are important when it is considered that the introduction of further wild accessions will be needed in order to reduce the genetic variation bottleneck apparent in this species.

Recently in soybean the effect of segregation distortion has been elucidated in mapping and QTL analysis, across diverse mapping populations and genetic backgrounds. Few chromosomes and clusters of markers generates complexity, and when distorted markers were included a more accurate map was obtained65. Interestingly, our results in L. luteus were very much in agreement with the results found in soybean, few chromosomes and genomic regions of L. luteus were mainly associated with segregation distortion. When these distorted markers were removed, the map generated was not entirely consistent with the map including all markers, that shared good collinearity and syntenic regions with the map and reference genome of L. angustifolius. In our study neither SDL or SDR were located in the genomic regions harboring QTLs for anthracnose resistance or DTF.

Since anthracnose resistance was studied in a large F2 mapping population and F2:3 families, in in vitro and field conditions, allowing the validation of methods and data. This supported the high LOD score found for this QTL, and identified a linked and fully co-segregating marker in the target syntenic region. High correlations between in vitro assay and field evaluation for anthracnose resistance have been reported previously66,67. Anthracnose resistance in several other legumes has been reported as being genetically control by a single dominant gene59,68,69, which is coherent with the single dominant gene mapped in this study. This also reflect the narrow range of pathogenic races of this fungus and follow the gene-for-gene model70. The assembly of homologous sequences with L. angustifolius in the syntenic region for anthracnose resistance QTL, suggested Lanr1 as an orthologous candidate gene in the L. luteus genome (Figs. 2, 4). Moreover, BLAST results of genomic DNA sequences in the gene region of L. angustifolius chromosome 11 to EST sequences of L. luteus (SRA055806) produced a significant alignment, suggesting the gene encoded in this region was expressed23, supporting the QTL results of this study. In L. angustifolius, in a different genetic background, a fine map location of Lanr1 showed complete conservation, allowing the proposal of an NLR gene as a likely candidate for Lanr126. Flowering time was also evaluated in large F2 and F2:3 populations, in different environments and seasons, allowing the identification and validation, for the first time in this species, of a major QTL for this trait in the L. luteus genome. Like the Lanr1 gene, it was also mapped in a syntenic and collinear genomic region of L. angustifolius reference genome. From previous studies in L. angustifolius, the LanFTc1 gene has a large effect in terms of the regulation of DTF. FTc1 expression is increased with vernalisation in late genotypes of L. angustifolius30. In other legumes, such as Glycine max, Pisum sativum and Medicago truncatula, FT genes have been described as regulators of flowering time71,72,73.

The results of this study are in close agreement with previous studies in legumes; strongly suggesting that the QTL found in the L. luteus genome is harboring an orthologous FTc1 gene, with highly probably of sharing similar functional effects as L. angustifolius and other legumes species.

The L. luteus map reported here, its collinearity and synteny of specific genomic regions allowed the discovery of important domestication traits. Considering these results, it is highly possible that L. luteus is also preserving collinearity with the reference legume genomes of Arachis duranensis, Glycine max and Medicago truncatula25,74.

Thus, the results of comparative mapping, synteny analysis, and presence of orthologous genes between L. luteus and the L. angustifolius reference genome, are suggesting that L. luteus LG13, LG23 and LG11 correspond to the pseudochromosomes: Chr11, Chr10 and Chr17, respectively.

The discoveries in this study provide strong validation of the synteny approach for transferring genomic knowledge from a model genome to a less well-resourced crop genome, with reduced time and economy of effort. It is proving that the sequence information in model genome plants, advanced molecular and functional information from species like L. angustifolius, are helping enormously to target and explore important genes for breeding improvement, especially under the pressure of sustainable production and food security.

References

  1. Foyer, C. et al. Neglecting legumes has compromised human health and sustainable food production. Nat. Plants 2, 16112. https://doi.org/10.1038/nplants.2016.112 (2016).

    Article  PubMed  Google Scholar 

  2. Gladstone, J. S. Lupin as a crop plats. Field Crop Abstr. 23, 123–148 (1970).

    Google Scholar 

  3. Zalewski, D., Galek, R., Kozak, B. & Sawicka-Sienkiewicz, E. Pheno-morphological and agronomic diversity in a collection in a collection of wild and domesticated species of the genus Lupinus. Turk J. Field Crops 20, 43–48. https://doi.org/10.17557/.08197 (2015).

    Article  Google Scholar 

  4. Gladstone, J. S. Distribution, origin, taxonomy, history and importance. In Lupins as Crop Plants. Biology: Production and Utilization (eds Gladstone, J. S. et al.) 1–37 (CAB International, Cambridge, 1998).

    Google Scholar 

  5. Lambers, H., Clements, J. C. & Nelson, M. N. How a phosphorus-acquisition strategy based on carboxylate exudation powers the success and agronomic potential of lupines (Lupinus, Fabaceae). Am. J. Bot. 100, 263–288. https://doi.org/10.3732/ajb.1200474 (2013).

    CAS  Article  PubMed  Google Scholar 

  6. Lucas, M. M. et al. The future of lupin as a protein crop in Europe. Front. Plant Sci. 6, 1–6. https://doi.org/10.3389/fpls.2015.00705 (2015).

    Article  Google Scholar 

  7. Lavin, M., Herendeen, P. S. & Wojciechowski, M. F. Evolutionary rates analysis of Leguminosae implicates a rapid diversification of lineages during the Tertiary. Syst. Biol. 54, 575–594. https://doi.org/10.1080/10635150590947131 (2005).

    Article  PubMed  Google Scholar 

  8. Drummond, C. S., Eastwood, R. J., Miotto, S. T. & Hughes, C. E. Multiple continental radiations and correlates of diversification in Lupinus (Leguminosae); testing for key innovation with incomplete taxon sampling. Syst. Biol. 61, 443–460. https://doi.org/10.1093/sysbio/syr126 (2012).

    Article  PubMed  PubMed Central  Google Scholar 

  9. Petterson, D. Composition and food uses of lupins. In Lupin as Crop Plants: Biology, Production and Utilization (eds Gladstones, J. et al.) 353–384 (CAB International, London, 1998).

    Google Scholar 

  10. Wolko, B., Clements, J. C., Naganowska, B., Nelson, M. N., & Yang, H. Lupinus. In Wild Crop Relatives: Genomic and Breeding Resources, Legume Crops and Forages. Berlin, Heidelberg XXV, 153–206. https://doi.org/10.1007/978-3-642-14387-8 (2011).

  11. Aïnouche, A., Bayer, R. J. & Misset, M. T. Molecular phylogeny, diversification and character evolution in Lupinus (Fabaceae) with special attention to Mediterranean and African lupines. Plant Syst. Evol. 246, 211–222. https://doi.org/10.1007/s00606-004-0149-8 (2004).

    CAS  Article  Google Scholar 

  12. Glencross, B., Curnow, J., & Hawkins, W. Assessment of the nutritional variability of lupins as an aquaculture feed ingredient. Final Report for the Grains Research Committee of WA Project Fisheries Research Contract Report No. 6, Department of Fisheries, Western Australia (2003).

  13. Ogura, T. et al. Proteomic characterization of seeds from yellow lupin (Lupinus luteus L). Proteomics 12, 1543–1554. https://doi.org/10.1002/pmic.201300511 (2014).

    CAS  Article  Google Scholar 

  14. Piornos, J. A. et al. Functional and physicochemical properties of a protein isolate from AluProt-CGNA: A novel protein-rich lupin variety (Lupinus luteus). Food Res. Int. 76, 719–724. https://doi.org/10.1016/j.foodres.2015.07.013 (2015).

    CAS  Article  PubMed  Google Scholar 

  15. Burgos-Díaz, C. et al. Food-grade Pickering stabilizers obtained from a protein-rich lupin cultivar (AluProt-CGNA): Chemical characterization and emulsifying properties. Food Hydrocoll. 87, 847–857. https://doi.org/10.1016/j.foodhyd.2018.09.018 (2019).

    CAS  Article  Google Scholar 

  16. Glencross, B., Evans, D., Hawkins, W. & Jones, B. Evaluation of dietary inclusion of yellow lupin (Lupinus luteus) kernel meal on the growth, feed utilization and tissue histology of rainbow trout (Oncorhynchus mykiss). Aquaculture 235, 411–422. https://doi.org/10.1016/j.aquaculture.2003.09.022 (2004).

    Article  Google Scholar 

  17. Velásquez, A. C., Danve, C. & Yang, S. Plant and pathogen warfare under changing climate conditions. Curr. Biol. 28, 619–634. https://doi.org/10.1016/j.cub.2018.03.054 (2018).

    CAS  Article  Google Scholar 

  18. Adhikari, K. N., Thomas, G., Buirchell, B. J. & Sweetingham, M. W. Identification of anthracnose resistance in yellow lupin (Lupinus luteus L.) and its incorporation into breeding lines. Plant Breed. 130, 660–664. https://doi.org/10.1111/j.1439-0523.2011.01880.x (2011).

    Article  Google Scholar 

  19. Nirenberg, H. I., Feiler, U. & Hagedorn, G. Description of Colletotrichum lupini comb. nov. in modern terms. Mycologia 94, 307–320. https://doi.org/10.2307/3761809 (2002).

    Article  PubMed  Google Scholar 

  20. Talhinhas, P., Baroncelli, R. & Floch, G. Anthracnose of lupins caused by Colletotricum lupini: A recent disease and a successful worldwide pathogen. J. Plant Pathol. 98, 5–14. https://doi.org/10.4454/JPP.V98I1.040 (2016).

    Article  Google Scholar 

  21. Berger, J. D., Buirchell, B. J., Luckett, D. J. & Nelson, M. N. Domestication bottlenecks limit genetic diversity and constrain adaptation in narrow-leafed lupin (Lupinus angustifolius L.). Theor. Appl. Genet. 124, 637–652. https://doi.org/10.1007/s00122-011-1736-z (2012).

    CAS  Article  PubMed  Google Scholar 

  22. Parra-Gonzalez, L. et al. Development of microsatellites in Lupinus luteus (Fabaceae) and cross-species amplification in other lupine species. Am. J. Bot. 97, e72-74. https://doi.org/10.3732/ajb.1000170 (2010).

    CAS  Article  Google Scholar 

  23. Parra-Gonzalez, L. et al. Yellow lupin (Lupinus luteus L.) transcriptome sequencing: Molecular marker development and comparative studies. BMC Genom. 13, 425. https://doi.org/10.1186/1471-2164-13-425 (2012).

    CAS  Article  Google Scholar 

  24. Iqbal, M. M. et al. The first genetic map for yellow lupin enables genetic dissection of adaptation traits in an orphan grain legume crop. BMC Genet. 20, 68. https://doi.org/10.1186/s12863-019-0767-3 (2019).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  25. Kroc, M., Koczyk, G., Święcicki, W., Kilian, A. & Nelson, M. N. New evidence of ancestral polyploidy in the Genistoid legume Lupinus angustifolius L. (Narrow-Leafed Lupin). Theor. Appl. Genet. 127, 1237–1249. https://doi.org/10.1007/s00122-014-2294-y (2014).

    Article  PubMed  Google Scholar 

  26. Hane, J. et al. A comprehensive draft genome sequence for lupin (Lupinus angustifolius), an emerging health food: Insights into plant-microbe interactions and legumes evolution. Plant Biotechnol. J. 15, 318–330. https://doi.org/10.1111/pbi.12615 (2017).

    CAS  Article  PubMed  Google Scholar 

  27. Yang, H. et al. Draft genome sequence, and a sequence-defined genetic linkage map of the legume crop species Lupinus angustifolius L. PLoS One 8, e64799. https://doi.org/10.1371/journal.pone.0064799 (2013).

    ADS  CAS  Article  PubMed  PubMed Central  Google Scholar 

  28. Boersma, J., Buirchell, B., Sivasithamparam, K. & Yang, H. Development of a sequence-specific PCR marker linked to the Ku gene which removes the vernalization requirement in narrow-leafed lupin. Plant Breed. 126, 306–309. https://doi.org/10.1111/j.1439-0523.2007.01347.x (2007).

    CAS  Article  Google Scholar 

  29. Nelson, M. N. et al. Aligning a new reference genetic map of Lupinus angustifolius with the genome sequence of the model legume, Lotus japonicus. DNA Res. 17, 73–83. https://doi.org/10.1093/dnares/dsq001 (2010).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  30. Nelson, M. N. et al. The loss of vernalization requirement in narrow-leafed lupin is associated with a deletion in the promoter and de-repressed expression of a Flowering Locus T (FT) homologue. New Phytol. 213, 220–232. https://doi.org/10.1111/nph.14094 (2017).

    CAS  Article  PubMed  Google Scholar 

  31. Nelson, M. N. et al. The first gene-based map of Lupinus angustifolius L.—location of domestication genes and conserved synteny with Medicago truncatula. Theor. Appl. Genet. 113, 225–238. https://doi.org/10.1007/s00122-006-0288-0 (2006).

    CAS  Article  PubMed  Google Scholar 

  32. Russello, M. A., Waterhouse, M. D., Etter, P. D. & Johnson, E. A. From promise to practice: Pairing non-invasive sampling with genomics in conservation. PeerJ. 3, e1106. https://doi.org/10.7717/peerj.1106 (2015).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  33. Winkler, D. E. et al. Multiple introductions and population structure during the rapid expansion of the invasive Sahara mustard (Brassica tournefortii). Ecol. Evol. 9, 7928–7941. https://doi.org/10.1002/ece3.5239 (2019).

    Article  PubMed  PubMed Central  Google Scholar 

  34. Atchison, G. W. et al. Lost crops of the Incas: Origins of domestication of the Andean pulse crop tarwi, Lupinus mutabilis. Am. J. Bot. 103, 1592–1606. https://doi.org/10.3732/ajb.1600171 (2016).

    CAS  Article  PubMed  Google Scholar 

  35. Chen, W. et al. Genome of the African cassava whitefly Bemisia tabaci and distribution and genetic diversity of cassava-colonizing whiteflies in Africa. Insect Biochem. Mol. Biol. 110, 112–120. https://doi.org/10.1016/j.ibmb.2019.05.003 (2019).

    CAS  Article  PubMed  Google Scholar 

  36. Knaus, B. & Grünwald, K. vcfr: A package to manipulate and visualize variant call format data in R. Mol. Ecol. Resour. 17, 44–53. https://doi.org/10.1111/1755-0998.12549 (2016).

    CAS  Article  PubMed  Google Scholar 

  37. Gurevich, A., Saveliev, V., Vyahhi, N. & Tesler, G. QUAST: Quality assessment tool for genome assemblies. Bioinformatics 29, 1072–1075. https://doi.org/10.1093/bioinformatics/btt086 (2013).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  38. Kearse, M. et al. Geneious basic: An integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics 28, 1647–1649. https://doi.org/10.1093/bioinformatics/bts199 (2012).

    Article  PubMed  PubMed Central  Google Scholar 

  39. Tixier, M. H., Sourdilie, R. M., Levoy, P. & Bernard, M. Detection of wheat microsatellite using a non radioactive silver-nitrate staining method. J. Genet. Breed. 51, 175–177 (1997).

    CAS  Google Scholar 

  40. Rodriguez, F., Cai, D., Teng, Y. & Spooner, D. Asymmetric single-strand conformation polymorphism: An accurate and cost-effective method to amplify and sequence allelic variants. AM. J. Bot. 98, 1061–1067. https://doi.org/10.3732/ajb.1000251 (2011).

    CAS  Article  PubMed  Google Scholar 

  41. Van Ooijen, J. W. JoinMap 4, Software for the Calculation of Genetic Linkage Maps in Experimental Populations (Kyazma B.V., Waveningen, 2006).

    Google Scholar 

  42. Kosambi, D. D. The estimation of map distances from recombination values. Ann. Eugen. 12, 172–175. https://doi.org/10.1007/978-81-322-3676-4_16 (1944).

    Article  Google Scholar 

  43. Krzywinski, M. et al. Circos: An information aesthetic for comparative genomics. Genome Res. 19, 1639–1645. https://doi.org/10.1101/gr.092759.109 (2009).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  44. Wickham, H. ggplot2: Elegant Graphics for Data Analysis (Springer, New York, 2016) ((ISBN 978-3-319-24277-4)).

    Book  Google Scholar 

  45. Ouellette, L. A., Reid, R. W., Blanchard, S. G. & Brouwer, C. R. LinkageMapView-rendering high resolution linkage and QTL maps. Bioinformatics 34, 306–307. https://doi.org/10.1093/bioinformatics/btx576 (2018).

    CAS  Article  PubMed  Google Scholar 

  46. Cucuzza, J. & Kao, K. In vitro assay of excised cotyledons of Alfalfa (Medicago sativa) to screen for resistance to Colletotrichum trifoli. Plant Dis. 70, 111–115 (1986).

    Article  Google Scholar 

  47. Fischer, K. et al. Characterization and mapping of LanrBo: A locus conferring anthracnose resistance in narrow-leafed lupin (Lupinus angustifolius L.). Theor. Appl. Genet. 128, 2121–2130. https://doi.org/10.1007/s00122-015-2572-3 (2015).

    CAS  Article  PubMed  Google Scholar 

  48. Burton, G. W. & DeVane, E. H. Estimating heretability in tall fescue (Festuca arundinacea) from replicated clonal material. Agron. J. 45, 478–481. https://doi.org/10.2134/agronj1953.00021962004500100005x (1954).

    Article  Google Scholar 

  49. Allard, R. W. Principles of Plant Breeding (Wiley, New York, 1960).

    Google Scholar 

  50. Van Ooijen, J. W. MapQTL 6, Software for the Mapping of Quantitative Trait Loci in Experimental Populations of Diploid Species (Kyazma B.V, Waveningen, 2009).

    Google Scholar 

  51. Wen, Y. J. et al. An efficient multi-locus mixed model framework for the detection of small and linked QTLs in F2. Brief. Bioinform. 20, 1913–1924. https://doi.org/10.1093/bib/bby058 (2018).

    CAS  Article  PubMed Central  Google Scholar 

  52. Zhang, Y. W., Wen, Y. J., Dunwell, J. M. & Zhang, Y. M. QTL.gCIMapping.GUI v2.0: An R software for detecting small-effect and linked QTLs for quantitative traits in bi-parental segregation populations. Comput. Struct. Biotech. 18, 59–65. https://doi.org/10.1016/j.csbj.2019.11.005 (2020).

    CAS  Article  Google Scholar 

  53. Borie, F. & Rubio, R. Total and organic phosphorus in Chilean volcanic soils. Gayana Bot. 60, 69–73. https://doi.org/10.4067/S0717-66432003000100011 (2003).

    Article  Google Scholar 

  54. Thomas, G. J. & Sweetingham, M. W. Cultivar and environment influence the development of lupin anthracnose caused by Colletotrichum lupini. Australasian Plant Pathol. 33, 571–577. https://doi.org/10.1071/AP04060 (2004).

    Article  Google Scholar 

  55. Riegel, R., Véliz, D., von Baer, I., Quitral, Y. & Muñoz, M. Genetic diversity and virulence of Colletotrichum lupini isolates collected in Chile. Trop. Plant Pathol. 35, 144–152. https://doi.org/10.1590/S1982-56762010000300002 (2010).

    Article  Google Scholar 

  56. Kumar, A. A., Mishra, P., Kumari, K. & Panigrahi, K. C. Environmental stress influencing plant development and flowering. Front Biosci. 4, 1315–1324. https://doi.org/10.2741/s333 (2012).

    Article  Google Scholar 

  57. Jung, C. & Müller, A. E. Flowering time control and applications in plant breeding. Trends Plant Sci. 14, 563–573. https://doi.org/10.1016/j.tplants.2009.07.005 (2009).

    CAS  Article  PubMed  Google Scholar 

  58. Weller, J. L. & Ortega, R. Genetic control of flowering time in legumes. Front Plant Sci. 6, 207. https://doi.org/10.3389/fpls.2015.00207 (2015).

    Article  PubMed  PubMed Central  Google Scholar 

  59. You, M. et al. A PCR-based molecular marker applicable for marker-assisted selection for anthracnose disease resistance in lupin breeding. Cell. Mol. Biol. Lett. 10, 123–134 (2005).

    CAS  PubMed  Google Scholar 

  60. Heijmans, K., Morel, P. & Vandenbussche, M. MADS-box genes and floral development: The dark side. J. Exp. Bot. 63, 5397–5404. https://doi.org/10.1093/jxb/ers233 (2012).

    CAS  Article  PubMed  Google Scholar 

  61. Chi, Y. et al. GmAGL1, a MADS-Box gene from Soybean, is involved in floral organ identity and fruit dehiscence. Front. Plant Sci. 8, 175. https://doi.org/10.3389/fpls.2017.00175 (2017).

    Article  PubMed  PubMed Central  Google Scholar 

  62. Kazimierski, T. & Kazimierska, E. M. Morphological and cytological within the species Lupinus luteus L.. Acta Soc. Bot. Pol. XLIV, 4 (1975).

    Google Scholar 

  63. Dai, B., Guo, H., Huang, C., Ahmed, M. M. & Lin, Z. Identification and characterization of segregation distortion loci on cotton chromosome 18. Front Plant Sci. 7, 2037. https://doi.org/10.3389/fpls.2016.02037 (2016).

    Article  PubMed  Google Scholar 

  64. Song, X. L., Sun, X. Z. & Zhang, T. Z. Segregation distortion and its effect on genetic mapping in plants. Chin. J. Agric. Biotechnol. 3, 163–169. https://doi.org/10.1079/CJB2006110 (2006).

    Article  Google Scholar 

  65. Zuo, J. et al. Effect of marker segregation distortion on high density linkage map construction and QTL mapping in Soybean (Glycine max L.). Heredity 123, 579–592. https://doi.org/10.1038/s41437-019-0238-7 (2019).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  66. Fokunang, C. N. et al. In vitro, greenhouse and field assessments of cassava lines for resistance to anthracnose disease caused by Colletotrichum gloeosporioides f. sp. manihotis. Mycopathologia 154, 191–198. https://doi.org/10.1023/A:1016372208982 (2001).

    Article  Google Scholar 

  67. Poolsawat, O., Tharapreuksapong, A., Wongkaew, S., Chaowiset, W. & Tantasawat, P. Laboratory and field evaluations of resistance to Sphaceloma ampelinum causing anthracnose in grapevine. Austral. Plant Pathol. 41, 263–269. https://doi.org/10.1007/s13313-012-0127-5 (2012).

    Article  Google Scholar 

  68. Yang, S. et al. Genetic and physical localization of an anthracnose resistance gene in Medicago truncatula. Theor. Appl. Genet. 116, 45–52. https://doi.org/10.1007/s00122-007-0645-7 (2007).

    CAS  Article  PubMed  Google Scholar 

  69. Meziadi, C. et al. Development of molecular markers linked to disease resistance genes in common bean based on whole genome sequence. Plant Sci. 242, 351–357. https://doi.org/10.1016/j.plantsci.2015.09.006 (2016).

    CAS  Article  PubMed  Google Scholar 

  70. Flor, H. H. Current status of the gene-for-gene concept. Annu. Rev. Phytopathol. 9, 275–296. https://doi.org/10.1146/annurev.py.09.090171.001423 (1971).

    Article  Google Scholar 

  71. Kong, F. et al. Two coordinately regulated homologs of FLOWERING LOCUS T are involved in the control of photoperiodic flowering in soybean. Plant Physiol. 154, 1220–1231. https://doi.org/10.1104/pp.110.160796 (2010).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  72. Laurie, R. E. et al. The Medicago FLOWERING LOCUS T homolog, MtFTa1, is a key regulator of flowering time. Plat Physiol. 156, 2207–2224. https://doi.org/10.1104/pp.111.180182 (2011).

    CAS  Article  Google Scholar 

  73. Hecht, V. et al. The pea GIGAS gene is a FLOWERING LOCUS T homolog necessary for graft-transmissible specification of flowering but not for responsiveness to photoperiod. Plant Cell 23, 147–161. https://doi.org/10.1105/tpc.110.081042 (2011).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  74. Książkiewicz, M. et al. A high-density consensus linkage map of white lupin highlights synteny with narrow-leafed lupin and provides markers tagging key agronomic traits. Sci Rep. 7, 1–15. https://doi.org/10.1038/s41598-017-15625-w (2017).

    CAS  Article  Google Scholar 

Download references

Acknowledgements

This research was funded by the Agriaquaculture Nutritional Genomic Center (CGNA), CONICYT-Regional Program; GORE La Araucanía, Chile, FONDECYT N°1080520, and CONICYT Chilean Ph. D. Scholarship N°21130132.

Author information

Authors and Affiliations

Authors

Contributions

N.L. and H.S.G. designed, performed the experiment, mapping, QTL analysis, and wrote the paper. B.T. participated in genomic data analysis and writing. A.R. participated in the phenotyping data collection. F.W. established and coordinated the field trials. P.D.S.C. provided extensive revision of the manuscript. M.O. contributed with computing data base and automatic control maintenance. All authors read the final version of the manuscript and approved its publication.

Corresponding author

Correspondence to Haroldo Salvo-Garrido.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Publisher's note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Lichtin, N., Salvo-Garrido, H., Till, B. et al. Genetic and comparative mapping of Lupinus luteus L. highlight syntenic regions with major orthologous genes controlling anthracnose resistance and flowering time. Sci Rep 10, 19174 (2020). https://doi.org/10.1038/s41598-020-76197-w

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1038/s41598-020-76197-w

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing