Genome-Wide Association Studies of Embryogenic Callus Induction Rate in Peanut (Arachis hypogaea L.)

The capability of embryogenic callus induction is a prerequisite for in vitro plant regeneration. However, embryogenic callus induction is strongly genotype-dependent, thus hindering the development of in vitro plant genetic engineering technology. In this study, to examine the genetic variation in embryogenic callus induction rate (CIR) in peanut (Arachis hypogaea L.) at the seventh, eighth, and ninth subcultures (T7, T8, and T9, respectively), we performed genome-wide association studies (GWAS) for CIR in a population of 353 peanut accessions. The coefficient of variation of CIR among the genotypes was high in the T7, T8, and T9 subcultures (33.06%, 34.18%, and 35.54%, respectively), and the average CIR ranged from 1.58 to 1.66. A total of 53 significant single-nucleotide polymorphisms (SNPs) were detected (based on the threshold value −log10(p) = 4.5). Among these SNPs, SNPB03-83801701 showed high phenotypic variance and neared a gene that encodes a peroxisomal ABC transporter 1. SNPA05-94095749, representing a nonsynonymous mutation, was located in the Arahy.MIX90M locus (encoding an auxin response factor 19 protein) at T8, which was associated with callus formation. These results provide guidance for future elucidation of the regulatory mechanism of embryogenic callus induction in peanut.


Introduction
Cultivated peanut (Arachis hypogaea L.) is among the most economically important oil, food, and feed crops worldwide [1].The seeds are rich in protein and oleic acid, which can regulate human physiological functions and promote growth and development.The seeds have a high nutritional value and are readily absorbed and utilized [2].In recent years, transgenic technology [3,4] has been increasingly widely applied in crop genetics and breeding, which has not only reduced the impact of diseases and pests on peanut yield and quality, but also overcome the problem of resistance in traditional breeding.Tissue culture is a basic technology essential for transgenic breeding and verification of gene function.However, in addition to the strong genotype dependence of the in vitro culture of peanut, the plant regeneration rate is low, which greatly limits biotechnology-based breeding.Therefore, there is an urgent need to elucidate the genetic mechanism controlling embryogenic callus induction in peanut to enhance the efficiency of genetic improvement and breeding.
Tissue culture response refers to the process by which any organ, tissue, or cell of a plant develops into a complete plant in a specific environment.In 1958, Steward et al. [5] proposed that plants produced disorganized cell clusters in response to adversity stress, which had the ability to develop into complete plants.Embryogenic callus induction is a crucial step in tissue culture and the initial process of somatic cells regaining totipotency, which is caused by changes in endogenous plant hormone concentrations [6,7].ARF7 and ARF19 regulate the callus formation of Arabidopsis thaliana by activating the promoter and expression of LBD16, LBD17, LBD18, and LBD29 [8] in the absence of exogenous hormones.Inhibiting the expression of these genes inhibits callus formation.Interestingly, root explants of the Arabidopsis aux1 mutant do not form callus on the standard medium, but callus formation is initiated with an increase in the auxin concentration in the medium.Moreover, the induction and differentiation of embryogenic callus show similar histological characteristics to those of the root meristem [9,10].WIND1 can promote callus formation and shoot regeneration by upregulating the expression of ESR1 [11].ESR1 is a differentially expressed gene, which is encoded by an APETALA 2/Ethylene Response Factor (AP2/ERF) transcription factor in Arabidopsis, and high abundance of the WIND1 protein can induce cell dedifferentiation [12,13].
The formation of plant embryogenic callus is controlled by heredity, which can be qualitative [14] or quantitative [15][16][17].Therefore, the elucidation of the mechanism of callus formation is of considerable importance for the optimization of plant in vitro culture systems.The GWAS approach can help determine the number of loci controlling the genetic variation of complex traits and predict candidate genes [18].Quantitative trait loci (QTLs) controlling callus differentiation have been identified in many crops, such as rice [19], Populus euphratica [20], soybean [21], maize [22], and wheat [23].In recent decades, with the development of diverse molecular markers, more accurate high-throughput sequencing technology has been applied to peanut genetics and breeding [24,25].In the present study, we compared the genotype and callus induction rate (CIR) of peanut varieties, constructing a high-density SNP-array genetic map based on the whole-genome resequencing of 353 peanut accessions, with the aim of identifying germplasm resources with a high CIR to expand the efficiency of the genetic transformation of peanut.The results are an important foundation for improvement in the genetic transformation and germplasm conservation of peanut.

Plant Material
A panel of 353 peanut accessions from 26 countries was used in this study, which represented five botanical varieties and two irregular types.The irregular morphologies were formed by the hybridization of different botanical varieties, which were classified as irregular types [26].These accessions comprise 44 irregular fastigiata, 100 irregular hypogaea, 26 subsp.fastigiata var.fastigiata, 2 subsp.fastigiata var.peruviana, 84 subsp.fastigiata var.vulgaris, 12 subsp.hypogaea var.hirsuta, and 85 subsp.hypogaea var.hypogaea accessions (Table S1).The CIR of each accession was observed and recorded; however, only data for 335 accessions were obtained and analyzed, as some accessions were excluded due to contamination.

Embryogenic Callus Induction
Mature pods of the 353 accessions were artificially shelled, and then clean and mature seeds without plaque were selected.The seeds were disinfected with 75% ethanol for 30 s, followed by 1% (w/v) NaClO solution for 8 min, and finally rinsed five times with sterile water.Next, the testa was peeled off with sterile tweezers after the seeds were removed from the water; then, the embryonic axes were excised and inoculated on a SEM medium composed of MS salts, B5 vitamins, 0.088 M sucrose, 12.4 µM picloram, and 0.8% agar (the pH was adjusted to 5.8) [27].The embryos were cultured on the medium and were subcultured at 4-week intervals.Each subculture of each accession comprised five replicates, and the experiment was repeated three times.

Phenotypic Identification and Statistics
After the sixth subculture, the physiological status of the callus began to stabilize.The number of calli was recorded in each of the seventh, eighth, and ninth subcultures (T7, T8, and T9, respectively).The CIR was calculated as follows [28]: where N is the final callus number and N 0 is the number of inoculations.The CIR value was used for GWAS analysis (Table S1).All phenotypic data were analyzed using Graph Pad Prism 7 and IBM SPSS Statistics 25, including correlation analysis, analysis of variance (ANOVA), and frequency distribution analysis.All phenotypic evaluations were conducted by the same person to maintain consistency between tissue-culture experiments and phenotypic evaluations, and to minimize artificial errors caused by different operational behaviors.

Genome-Wide Association Study
DNA extraction, library preparation, and resequencing were performed as described in detail in a previous study [26].Whole-genome resequencing of the 353 peanut accessions was conducted using the Illumina HiSeq platform (San Diego, CA, USA).A total of 864,179 single-nucleotide polymorphisms (SNPs) and 71,052 insertions/deletions (InDels) were obtained after quality control.The mixed linear model (MLM) GWAS method [29] was implemented in R package GAPIT (v 3.0) based on the high-density markers.In this study, the significance of traits and markers in association was determined using a threshold of −log 10 (p) = 4.5.Manhattan and Q-Q plots were generated using the 'qqman' R package [30].We used the 200 kb genomic regions upstream and downstream of the SNPs that were significantly associated with the peanut CIR as the candidate genomic region, and we used heterotetraploid cultivated peanut genomes for functional annotation.Based on previous reports, we predicted candidate genes associated with embryogenic callus induction in peanut.

Callus Formation
On the germination medium, 324 of the 335 peanut accessions formed callus, and the growth status of the callus was recorded at T7, T8, and T9 (Figure 1).The CIR phenotype values of the 335 accessions were normally distributed at T7, T8, and T9, with skewness ranging from −0.17 to 0.36 and kurtosis ranging from 2.76 to 4.15 (Figure 2, Table S2).The average CIR ranged from 1.58 to 1.66, and the coefficient of variation was 33.06%, 34.18%, and 35.54% at T7, T8, and T9, respectively (Table S2).A significance analysis showed that the CIR was significant, and a moderately high correlation coefficient was observed between embryogenic callus induction in the three subcultures (0.56 to 0.61) (Figure 2).Given that only two accessions of subsp.fastigiata var.peruviana were included in the study panel, the phenotypic variation of the remaining four varieties and two irregular types was analyzed.At T7, the CIR varied among the six types, but the differences were not significant.The CIR values of the six types were ranked in the following order: var.hirsuta > irregular hypogaea > irregular fastigiata > var.fastigiata > var.vulgaris > var.hypogaea (Figure 3a).However, at T8 and T9, significant differences in CIR were observed among the six types, among which the highest CIR was recorded for the irregular hypogaea type, and the lowest CIR was that of var.vulgaris (Figure 3b,c).Given that only two accessions of subsp.fastigiata var.peruviana were included in the study panel, the phenotypic variation of the remaining four varieties and two irregular types was analyzed.At T7, the CIR varied among the six types, but the differences were not significant.The CIR values of the six types were ranked in the following order: var.hirsuta > irregular hypogaea > irregular fastigiata > var.fastigiata > var.vulgaris > var.hypogaea (Figure 3a).However, at T8 and T9, significant differences in CIR were observed among the six types, among which the highest CIR was recorded for the irregular hypogaea type, and the lowest CIR was that of var.vulgaris (Figure 3b,c).
Interestingly, during the three subcultures, SNPs on chr13 formed one small clu

Analysis of Candidate Genes for SNP Loci Associated with CIR in Peanut
Genomic regions within 200 kb upstream and downstream of the significant SNPs were selected as candidate regions.Based on the SNPs significantly associated with CIR and functional annotations in these regions from the peanut reference genome, 600 annotated genes were obtained (Table S3).Thirty-six genes were potentially associated with embryogenic callus induction (Table 2).At T7, 17 genes were located near the nine significant SNPs, of which the most highly significant SNP was on chr13 at position 83801701, corresponding with the gene Arahy.LC8K5G that encodes a peroxisomal ABC transporter 1.At T8, five significant SNPs associated with CIR were detected, of which three SNPs were located on chr13.These SNP regions contained genes that affect embryogenic callus induction, and which are directly involved in callus formation or related developmental processes, and they comprised a peroxisomal ABC transporter 1, Pentatricopeptide repeat (PPR) superfamily protein, SAUR-like auxin-responsive protein, MYB transcription factor, and Homeodomain-like transcriptional regulator.The candidate region centered on the SNPA05-94095749 locus contained six candidate genes on chr05, of which four encoded an auxin response factor.At T9, the locus SNPB03-87089142 had the highest phenotypic variance (Table 1).

Discussion
This study is the first to use GWAS to explore the genetic basis of embryogenic callus induction in peanut.Here, we presented CIR data from 335 peanut accessions in three subcultures, identified the candidate genomic regions, and predicted genes associated with callus formation.The results contribute to an improved understanding of the mechanism of embryogenic callus induction and provide insights useful for further studies of tissue culture in peanut.

Induction of Callus from 353 Peanut Genotypes
Many studies have successfully regenerated plants through somatic and organogenic pathways, accelerating research progress in tissue culture and providing a convenient method for the preservation of rare or precious wild-type germplasm and the rapid propagation of superior germplasm in vitro.The source and genotype of explants are the main factors that regulate somatic embryo formation [31,32].Callus culture is an essential requirement for crop genetic transformation, but most studies of peanut callus culture to date have focused on comparisons among a small number of varieties, thus knowledge of the genetic mechanism of peanut callus induction is incomplete.The CIR is the most intuitive trait that characterizes the strength of callus induction ability [33].To explore the genetic basis of variation in callus differentiation, we studied the CIR of 353 peanut accessions representing five botanical varieties and two irregular types, and identified candidate genomic regions and associated genes that control callus differentiation.The present study used the largest number of peanut genotypes of any published study to date to evaluate single-callus formation.
The present data showed that the CIR in the three subcultures was continuous and normally distributed (Figure 2), indicating that CIR was controlled by multiple loci in the study population.In addition, differences in CIR were observed among different botanical varieties and irregular types, for which the difference was not significant at T7, but was significant at T8 and T9 (Figure 3).This variation may be caused by the unstable physiological state of the calli in the early period of embryogenic callus induction in peanut.In addition, the CIR phenotype values in the three subcultures were strongly correlated (Figure 2).These results reflected that the CIR phenotype may be partially controlled by the same genetic factors.In a previous report, the CIR of three peanut market types was ranked as Spanish type > Valencia type > Virginia type [34].In the present study, at T8 and T9, the CIR of the irregular hypogaea type was the highest, and that of var.vulgaris was the lowest, which is inconsistent with previous findings.This difference may reflect the different numbers and genotypic representations of materials in the study.Nevertheless, the results confirm that the population structure plays an important role in regulating callus formation.

GWAS Analysis
To date, many studies have been conducted on genes associated with embryogenic callus induction in plants [35][36][37].In the present study, several significant SNPs merit discussion in detail (Table 2).The peak SNPB03-83801701 was detected in each of the three subcultures and was located 53 kb from Arahy.LC8K5G, which encodes a peroxisomal ABC transporter 1. Peroxisome transporters are crucial for the synthesis of certain bioactive molecules, such as docosahexaenoic acid in mammals and jasmonic acid [38].This protein may regulate plant growth and development by mediating the synthesis of specific bioactive molecules [39].
In the present study, a gene with nonsynonymous mutations was detected in proximity to SNPA05-94095749, namely, Arahy.MIX90M, which encodes ARF19 [40], a transcriptional activator of auxin early-response genes.In Arabidopsis, ARF19 regulates lateral root formation by activating LBD/ASL genes, reflecting that auxin-induced callus formation has similar characteristics to the root formation metabolic pathway [41].Moreover, it was observed that the candidate genes Arahy.SI4FKG, Arahy.HDV1A7, Arahy.F6IJX6, Arahy.X8R2JI, and Arahy.K8JH1A encode ARF19, an auxin canalization protein, and ARF19-like isoform X1.Auxin signaling activates the expression of downstream genes and regulates plant tissue meristem and embryonic development by mediating these auxin transcription factors and auxin channel protein genes [42].Furthermore, we located several genes on chr13 that were significantly associated with embryogenic callus induction, most of which were associated with a PPR superfamily protein, Homeodomain-like transcriptional regulator, and Saur-like auxin responsive protein.
The peak SNPB03-17372387 was located 119 kb from Arahy.2NE4Q7, which encodes a SAUR-like auxin-responsive protein family member.SAUR genes are the largest early auxin-responsive gene family in plants [43].Active auxin can rapidly induce the expression of SAUR genes [44].A SAUR gene was first identified in soybean hypocotyls [45], which had been confirmed to play a role in different processes in plant growth, development, and stress response [46][47][48].Therefore, it was speculated that candidate genes may affect plant tissue differentiation by regulating cell growth and development.The SNP marker SNPA05-94095749 is located in Arahy.KFS3KW, which encodes a B3 domain-containing VRN1-like transcription factor.VRN1, a critical protein that responds to long-term cold treatment to promote flowering, was detected in the vernalization response of Arabidopsis and is crucial for development in Arabidopsis [49].Overexpression of VRN1 leads to early flowering and phenotypic abnormalities [50].
Recently, marker-trait associations have been analyzed for embryogenic callus induction in a number of plants; for example, SNPB03-17372387 and SNPB07-124709181 are derived from genes that encode an MYB transcription factor.Ge et al. [51] located zmMYB138, which is a nucleus-localized member of the MYB transcription factor family of maize, and determined that zmMYB138 could promote the formation of maize embryogenic callus through gibberellin signal transduction.The MYB75 transcription factor in Arabidopsis regulates anthocyanin biosynthesis in Nicotiana callus [52].Significantly, we found that one SNPB07-124709181 was strongly associated with CIR, and this was consistently across genotypes with low or NO callus induction (Figure S1).The linkage analysis suggested that the accessions with CAT/CAT have a high CIR, while C/C have a low or zero CIR at this locus at T7, T8, and T9.Due to the small difference in CIR between the two genotypes of the material, this SNP was not identified at T8 and T9, while there was a significant difference in CIR between the two genotypes at the level of 0.05.
The present study was focused on the prediction of candidate genes associated with embryogenic callus induction.Further functional analysis and verification are needed.At present, although many QTLs of embryogenic callus induction traits have been identified in other crops, there are currently few reports on genes associated with peanut embryo-genic callus induction, and the genetic mechanism of peanut embryogenic callus induction remains unclear.The SNPs and candidate genes identified in the current study lay the foundation for further cloning of the functional genes that regulate peanut embryogenic callus induction and analysis of the genetic mechanism of peanut embryogenic callus induction.

Conclusions
In this study, we conducted the embryogenic callus induction of 353 peanut accessions and observed and scored their phenotypes.The results showed that callus induction might be partially controlled by the same genetic factors.In addition, we identified the genomic regions associated with callus formation and located candidate genes associated with callus formation based on functional annotations.

Figure 2 .
Figure 2. Correlation between callus induction rate in peanut at the T7, T8, and T9 subcultures.The correlation coefficients are presented above the diagonal, and the distributions of the number of calli data at T7, T8, and T9 are presented on and below the diagonal.*** p < 0.001.

Figure 2 .
Figure 2. Correlation between callus induction rate in peanut at the T7, T8, and T9 subcultures.The correlation coefficients are presented above the diagonal, and the distributions of the number of calli data at T7, T8, and T9 are presented on and below the diagonal.*** p < 0.001.

Figure 3 .
Figure 3. Analysis of variance of differences in callus induction rate (CIR) among six peanut b ical varieties and irregular types at the T7, T8, and T9 subcultures.Boxplots show the variati the CIR in each peanut type at T7 (a), T8 (b), and T9 (c).Different lowercase letters above the b indicate a significant difference within a subculture (p ≤ 0.05; Tukey's HSD test).

Figure 3 .
Figure 3. Analysis of variance of differences in callus induction rate (CIR) among six peanut botanical varieties and irregular types at the T7, T8, and T9 subcultures.Boxplots show the variation in the CIR in each peanut type at T7 (a), T8 (b), and T9 (c).Different lowercase letters above the boxes indicate a significant difference within a subculture (p ≤ 0.05; Tukey's HSD test).

Table 1 .
Single-nucleotide polymorphisms (SNPs) significantly associated with peanut callus induction rate identified at the T7, T8, and T9 subcultures through genome-wide association study.

Table 1 .
Single-nucleotide polymorphisms (SNPs) significantly associated with peanut callus induction rate identified at the T7, T8, and T9 subcultures through genome-wide association study.

Table 2 .
Single-nucleotide polymorphisms (SNPs) and candidate genes significantly associated with callus induction rate in peanut at the T7, T8, and T9 subcultures.