Molecular genetic variability of Juglans regia L. and Juglans sigillata D. as revealed by fluorescent amplified-fragment length polymorphism

Abstract

Juglans regia L. and J. sigillata D. have a long cultivation history in China. With the purpose of detecting interspecific and intraspecific genetic variation, 35 clonal cultivars of J. regia, J. sigillata, J. sigillata×J. regia and J. hopeiensis Hu. (J. mandshurica Maxim. ×J. regia) were selected from a broad geographic range and applied in amplified-fragment length polymorphism (AFLP) analysis. Clear-cut AFLP profiles were produced using 9 EcoR I/Mse I primer combinations. As for J. regia, J. sigillata and their hybrids, the percentages of polymorphic bands (PPBs) averaged 93.86%, 93.94% and 60.46%; the observed number of alleles (Na) averaged 1.5370, 1.6389 and 1.2593; the effective number of alleles (Ne) averaged 1.2582, 1.2543 and 1.1646; Nei's gene diversity (H) scored 0.1560, 0.1556 and 0.0951; Shannon’s information index (I) averaged 0.2402, 0.2459 and 0.1414, respectively. These results implied a moderate level of genetic diversity within the investigated species. The total gene diversity (H_t = 0.1596) can be distributed into intra- (H_s = 0.1356) and inter- (D_st = 0.0240) species gene diversity. The proportion of the distributed genetic diversity among populations (G_st) was 0.1505, indicating that only 15% of the total gene diversity was attributed to differences between species. The factorial correspondence analysis (FCA) and unweighted pair-group method of arithmetic averages (UPGMA) analysis revealed a close genetic distance between these two species.

Keywords:

• Walnut
• genetic relationship
• genetic diversity
• AFLP

Introduction

Juglans regia L. is widely cultivated throughout temperate regions of the world for its high quality edible nuts and its timber. J. sigillata D. is another species of genus Juglans native to China [1]. Studies on their genetic diversity, structure and relationship have been based on morphology, cytological markers, isozyme markers and molecular markers [2–4]. J. sigillata was taken as an independent species for a long time [5]. However, Orel et al. [6] studied the chloroplast DNA in the trnL-trnF region and suggested that the relationship between these two species was really close. Molecular genetic diversity studies carried out by Gunn et al. [7] found it hard to discriminate J. regia from J. sigillata although genetic variations among and within villages were significant. Wang et al. [3] suggested a moderately high level of polymorphism in Tibetan walnut populations on the basis of simple sequence repeat (SSR) analysis and proposed the presence of J. regia × J. sigillata hybrids. Furthermore, phylogenetic studies using chloroplast markers revealed they were clustered together in one group [8]. A recently published study indicated they should be considered as a single species from section Juglans [9]. A chromosome level study indicates that J. regia speciated with J. sigillata about 0.84 million years ago, much later than compared with other related species (2.64 million years) and the whole genome duplication event (24.48 million years) [10]. These reports increased our understanding of walnut genetic diversity. However, genetic diversity and relationships among these two species and their hybrids are still in great need to be unveiled.

Genetic diversity and cultivar discrimination studies of non-wood-product forests are currently supported by molecular methods, such as molecular makers and genome sequencing [11]. Amplified fragment length polymorphism (AFLP) technique is a reliable genetic marker with low cost, superior repeatability and high resolution [12]. Thus, AFLP markers have been with broad application in systematics [13], genetic diversity [11,12,14], genetic linkage map construction [15] and association analysis [16] till now. Here, 35 clonal cultivars selected from the main cultivating area in China were applied in AFLP analysis. The purpose of this study was to analyze the genetic variation among and within the two species and their hybrids, which will provide essential information at the molecular level for future genetic studies and breeding.

Materials and methods

Plant materials

A total of 35 clonal cultivars from a broad geographic range (Table 1), including 15 J. regia cultivars, 14 J. sigillata cultivars, 5 hybrids of J. sigillata× J. regia (as a transitional type) and 1 J. hopeiensis Hu. cultivar (as an outgroup), were chosen for AFLP analysis (Figure 1). Cultivars introduced from abroad were sampled in the Walnut Genebank in China National Clonal Plant Germplasm Repository (Tai’an City, Shandong Province, China), whereas the others were sampled form their area of origin (Table 1). The selected cultivars are listed in Table 1. Fresh leaves were collected from a single mature, healthy tree of each cultivar and frozen quickly and stored at −80 °C until DNA isolation.

Figure 1. Typical nut characteristics for Juglans regia (a), J. sigillata (b), J. sigillata × J. regia (c), and J. hopeiensis (d).

Table 1. Walnut cultivars (Juglans regia and J. sigillata) used for the AFLP analysis.

Cultivar	Species	Origin
Chandler	J. regia	USA
Damuma	J. regia	Xinjiang, PR China
Franquette	J. regia	FRA
Hartley	J. regia	USA
Jing861	J. regia	Beijing, PR China
Liaoning1	J. regia	Liaoning, PR China
Shanhe1	J. regia	Shaanxi, PR China
Sunland	J. regia	USA
Tulare	J. regia	USA
Wusiman	J. regia	Xinjiang, PR China
Xiangling	J. regia	Shandong, PR China
Xinfeng	J. regia	Xinjiang, PR China
Xinguang	J. regia	Xinjiang, PR China
Yechenghetaowang	J. regia	Xinjiang, PR China
Zixin	J. regia	Yunnan, PR China
Huaningdabaike	J. sigillata	Yunnan, PR China
Li53	J. sigillata	Yunnan, PR China
Lushui1	J. sigillata	Yunnan, PR China
Santaihetao	J. sigillata	Yunnan, PR China
Wei2	J. sigillata	Yunnan, PR China
Xixiang	J. sigillata	Yunnan, PR China
Yangbipaohetao	J. sigillata	Yunnan, PR China
Yangjiang1	J. sigillata	Yunnan, PR China
Yangza1	J. sigillata	Yunnan, PR China
Yangza2	J. sigillata	Yunnan, PR China
Yangza3	J. sigillata	Yunnan, PR China
Yong11	J. sigillata	Yunnan, PR China
Zier	J. sigillata	Yunnan, PR China
Ziyuan	J. sigillata	Yunnan, PR China
Yunxin301	J. sigillata × J. regia	Yunnan, PR China
Yunxin303	J. sigillata × J. regia	Yunnan, PR China
Yunxin306	J. sigillata × J. regia	Yunnan, PR China
Yunxingaoyuan	J. sigillata × J. regia	Yunnan, PR China
Yunxinyunlin	J. sigillata × J. regia	Yunnan, PR China
Jilong	J. hopeiensis	Hebei, PR China

DNA isolation

DNA was extracted from the leaves and its quality determined following the protocol described by Chen et al. [17].

AFLP analysis

The AFLP protocol was carried out using an AFLP Kit (Dingguo Bio-technology Co., Ltd., Beijing, China) following the manufacturer’s instructions. Fluorescently labeled primers were used for amplification, and the amplification products were detected on an ABI PRISM 377 DNA Sequencer. A total of 9 polymorphic EcoR I + 3/Mse I + 3 primer combinations were applied in the analysis of different accessions (Table 2).

Table 2. Summary of the AFLP markers scored for J. regia, J. sigillata and the hybrids using 9 primer combinations.

Primer combinations*	Total bands (number)			Polymorphic bands (number)			Percent polymorphism (%)
	J. regia	J. sigillata	hybrids	J. regia	J. sigillata	hybrids	J. regia	J. sigillata	hybrids
E-AAC/M-CAA	119	140	80	111	131	43	93.28	93.57	53.75
E-AAC/M-CTT	143	104	75	130	92	42	90.91	88.46	56.00
E-AAG/M-CAA	178	144	75	172	139	44	96.63	96.53	58.67
E-AAG/M-CAG	176	147	98	171	142	60	97.16	96.60	61.22
E-ACA/M-CAC	111	123	89	107	111	39	96.40	90.24	43.82
E-ACT/M-CTA	156	128	86	151	121	43	96.79	94.53	50.00
E-AGC/M-CAG	138	101	75	131	96	61	94.93	95.05	81.33
E-AGC/M-CAT	123	75	49	107	73	39	86.99	97.33	79.59
E-AGC/M-CTT	132	117	87	121	109	52	91.67	93.16	59.77
Total	1276	1079	714	1201	1014	423
Average	141.78	119.89	79.33	133.44	112.67	47.00	93.86	93.94	60.46

* E seq: 5′-GACTGCGTACCAATTC-3′; MseI: 5′-GATGAGTCCTGAGTAA-3′.

Statistical procedures

Bands with strong intensity were recorded and classified as present (1) or absent (0) for all accessions to construct a binary data matrix, from which markers lower than 70 bp were excluded. Cluster analysis was carried out using the UPGMA method [11] with the NTSYSpc version 2.2 software package [18] based on Dice’s coefficient. Factorial correspondence analysis (FCA) was carried out using GENETIX 4.05 software [19,20].

POPGENE32 software [21] was employed to calculate the percentages of polymorphic bands (PPBs), observed number of alleles (Na), effective number of alleles (Ne), Nei's gene diversity (H) and Shannon’s information index (I) on the premise of the Hardy-Weinberg equilibrium. The total observed diversity estimate (H_t) was partitioned into within-group diversity (H_s) and between-group diversity (D_st) components, where $H_t=H_s+D_{st}.$ Gene diversity between groups, in relation to total population diversity, was expressed as $G_{st}=D_{st}/H_t.$ Gene flow (N_m) was calculated as $N_m=0.5(1-G_{st})/G_{st}.$

Results and discussion

AFLP profiles and analysis

Clear-cut AFLP profiles for 35 assessed walnut cultivars were produced using 9 primer combinations (Table 2). One thousand and one out of 1276 AFLP bands for J. regia were polymorphic and 1014 fragments among the total 1079 exhibited polymorphisms for J. sigillata, whereas 423 out of 714 were polymorphic fragments for the hybrids. For J. hopeiensis, totally 491 fragments were detected. The fragment size of all samples ranged from 70 to 500 bp, and most of them were distributed between 70 and 300 bp. The averaged percentage of polymorphism for J. regia, J. sigillata and the hybrids scored 93.86%, 93.94% and 60.46% respectively, showing that PPBs of J. regia was comparable to that of J. sigillata but higher than that of the hybrids.

AFLP technology was previously applied in distinguishing walnut cultivars [22]. And it was also reported that AFLP technology had been successfully applied into analyzing genetic relationships among species [23–25]. The technique was demonstrated to be effective for identifying genotypes, evaluating genetic relationships, and determining genetic structure. Here, AFLP were employed in analyzing the genetic diversity of two walnut species with long-term cultivation history, J. regia and J. sigillata. The results obtained with 35 clonal cultivars chosen for AFLP analysis were repeatable and accurate.

Genetic diversity indices

Diversity parameters, including Na, Ne, H and I, were calculated according to the allele frequencies in each primer combination (Table 3). Na averaged 1.5370, 1.6389 and 1.2593 respectively. Ne averaged 1.2582, 1.2543 and 1.1646. H scored 0.1560, 0.1556 and 0.0951. I averaged 0.2402, 0.2459 and 0.1414. The total gene diversity (H_t = 0.1596) could be distributed into intra- (H_s = 0.1356) and inter- (D_st=0.0240) species gene diversity (Table 3). G_st was 0.1505, indicating that only 15% of the total gene diversity was attributed to differences between species (Table 3).

Table 3. Na, Ne, H, and I for J. regia, J. sigillata and the hybrids based on AFLP analysis.

	Na*	Ne*	H*	I*
J. regia	1.5370 ± 0.4998	1.2582 ± 0.3390	0.1560 ± 0.1856	0.2402 ± 0.2668
J. sigillata	1.6389 ± 0.4814	1.2543 ± 0.3352	0.1556 ± 0.1791	0.2459 ± 0.2529
Hybrids	1.2593 ± 0.4392	1.1646 ± 0.3164	0.0951 ± 0.1725	0.1414 ± 0.2499
Average	1.4784 ± 0.1965	1.2257 ± 0.0530	0.1356 ± 0.0350	0.2092 ± 0.0588
	Ht*	Hs*	Gst*	Nm*
	0.1596 ± 0.0306	0.1356 ± 0.0216	0.1505	2.8219

* Na, Observed number of alleles.

* Ne, Effective number of alleles.

* H, Nei's gene diversity.

* I, Shannon's Information index.

* H_t, total heterocigosity.

* H_s, genetic diversity among populations.

* G_st, proportion of the distributed genetic diversity among populations.

* N_m, gene flow.

H was an important index in the assessment of genetic diversity and could be used to estimate genetic variation among and within populations [3,7]. In this study, H scores of different species suggested that the polymorphic level for J. regia was comparable to that of J. sigillata but for J. sigillata×J. regia it was much lower. This was mainly because the 5 hybrids had similar pedigrees. H_t, H_s and D_st, which scored 0.1596, 0.1356 and 0.0240, were calculated based on H. I suggested that about 79.24% of the total genetic diversity was attributable to variation within species with only 20.76% among species. All these indices based on AFLP data indicated a very close sibship between J. regia and J. sigillata. Moreover, the parameter values for these two species were relatively high compared with Herpetospermum caudigerum (Na = 1.352, Ne = 1.188, H = 0.111 and I = 0.168) [12], comparable to that of Glehnia littoralis (Na = 1.224, Ne = 1.235, H = 0.149, I = 0.238) [11], but much lower than thyme (Na = 1.94, Ne = 1.28, H = 0.21, I = 0.36) [13] based on AFLP data, which showed a moderate level of genetic diversity.

FCA analysis and UPGMA clustering

FCA analysis revealed that the first axis (axis 1) explained 8.88% of the percentage variation, whereas the second axis (axis 2) explained 7.02% (Figure 2). To understand overall the genetic relationships among the 35 walnut cultivars, a dendrogram was constructed by UPGMA analysis (Figure 3). According to the results of FCA and the dendrogram, the 35 walnut cultivars were divided into 2 main clusters (Cluster A and Cluster B). Cluster A comprised the control cultivar ‘Jilong’ (J. hopeiensis), whereas the other 34 cultivars were grouped into Cluster B. Nei's genetic similarity coefficients among these 4 populations are shown in Table 4. According to the UPGMA analysis, the 35 cultivars were separated into 6 groups (I–VI) at a Dice's coefficient level of 0.5133 (the averaged value). Group I contained 25 cultivars (10 J. regia, 10 J. sigillata and 5 hybrids). Group III consisted of 4 cultivars (J. regia) from Xinjiang Province. Group V had the cultivars ‘Xinguang’ (J. regia), ‘Zier’ and ‘Ziyuan’ (J. sigillata). Group II contained the cultivar ‘Santaihetao’ (J. sigillata). And group VI consisted of the cultivar ‘Jilong’ (Figure 3). ‘Franquetty’ and ‘Xiangling’ were most closely related (with a Dice's similarity of 0.8081). ‘Ziyuan’ and ‘Jilong’ had the lowest genetic similarity, with a Dice's coefficient of 0.2784.

Figure 2. Factorial correspondence analysis (FCA) based on polymorphisms at AFLP loci for 35 walnut cultivars.

Figure 3. Dendrogram of 35 walnut cultivars based on AFLP analysis conducted using 9 primer combinations.

Table 4. Nei’s genetic identity (above diagonal) and genetic distance (below diagonal) among 4 groups.

Population	J. regia	J. sigillata	J. sigillata×J. regia	J. hopeiensis
J. regia	****	0.9952	0.9954	0.9904
J. sigillata	0.0048	****	0.9985	0.9908
J. sigillata×J. regia	0.0046	0.0015	****	0.9907
J. hopeiensis	0.0096	0.0092	0.0094	****

‘Jilong’ was a clonal hybrid cultivar of J. hopeiensis and Mu et al. [9] reported that J. hopeiensis was a hybrid of J. regia and J. mandshurica using 2b-RAD technology. They also considered J. hopeiensis, J. regia and J. sigillata to be a single species. In the present study, ‘Jilong’ was an outgroup and it was separated first, as expected. For cultivars of J. regia, J. sigillata and the hybrids, neither FCA nor the dendrogram could distinguish these 3 types distinctly. Most of the cultivars (25) referring to 3 types clustered in Group I. Five interspecific hybrids were clustered firstly with 10 J. sigillata cultivars and then formed a clade together with 10 J. regia cultivars. Group V comprised of ‘Xinguang’ (J. regia), ‘Zier’ and ‘Ziyuan’ (J. sigillata). A crossing result of J. sigillata×J. regia showed that their offspring were fertile and hybrids of J. regia×J. sigillata exhibited morphological characters of J. sigillata [26]. Phylogenetic studies carried out based on chloroplast markers clustered them together in one clade [8,27]. Consistent with these studies, the results in the present study also suggested a close genetic distance among these two species and the hybrids. In addition, the cultivars investigated here are now popular in the walnut production of China, and the established genetic similarity could help future walnut germplasm identification, development and utilization.

Conclusions

In the present study, the genetic variations among and within J. regia, J. sigillata and their hybrids (J. sigillata×J. regia) were detected using the AFLP method. Genetic parameter values and cluster analysis implied a moderate level of genetic diversity within species and suggested a really close genetic relationship of interspecies. These results could provide essential information for further understanding the genetic differentiation and utilization strategies for walnut germplasm.

Disclosure statement

No potential conflict of interest was reported by the authors.

Data availability statement

The authors confirm that the data supporting the findings of this study are available within the article [and/or] its supplementary materials.

Additional information

Funding

This study was supported by Open Fund of State Key Laboratory of Tree Genetics and Breeding (TGB2015008) and the Fundamental Research Funds of Chinese Academy of Forestry (CAFYBB2017ZA004-2).