Establishment of a Breeding Approach Combined with Gamma Ray Irradiation and Tissue Regeneration for Highbush Blueberry

Abstract

Blueberries are a relatively recently domesticated species, primarily bred through hybridization. Mutation breeding, which uses chemical or physical treatment to increase plant mutation, has not yet been applied to blueberries. This study introduces a mutation breeding strategy for the highbush blueberry cultivar Vaccinium corymbosum. We established a high-efficiency regeneration protocol, which was applied to leaves and stems exposed to gamma irradiation using ⁶⁰Co-γ rays at doses of 10, 20, 40, 80, and 120 gray (Gy), to increase the efficiency of mutated cells to develop into adventitious shoots. We determined that the median lethal dose (LD₅₀) was approximately 56 Gy for leaf explants and 80 Gy for stem explants. Phenotypic variations, including changes in leaf color and growth characteristics, which may be due to altered plant response to environmental factors, were successfully observed in the first-generation (M1) plants. The height of M1 plants quantitatively decreased with increasing irradiation doses. To evaluate the mutants induced by each irradiation dose, whole-genome resequencing was conducted on individuals from each dose group, revealing significant genomic alterations at the 80 Gy dose. This approach provides a valuable reference for future blueberry breeding programs aimed at enhancing genetic diversity and improving cultivar performance.

1. Introduction

Blueberries (Vaccinium spp.) have a relatively short domesticated history of approximately 120 years [1]. In recent years, rising consumer demand for healthier dietary options and increased awareness of the nutritional benefits of blueberries have driven their popularity. This interest has, in turn, spurred extensive research into the breeding of new blueberry cultivars [2]. Hybridization remains the dominant breeding method for blueberry breeding. The species of highbush blueberry (Vaccinium corymbosum) has been improved the most with hybridization breeding [2,3]. Originating from Canada and North America, V. corymbosum underwent selection breeding to develop northern highbush blueberry cultivars [4]. Hybridization with V. darrowii, an evergreen blueberry, expanded the growth range of northern highbush blueberries to subtropical/tropical regions, creating Southern highbush blueberries [5,6]. Subsequently, crosses between V. corymbosum and other species in the section Cyanococcus have further enhanced the adaptation and berry quality of highbush blueberries, improving traits such as drought tolerance, alkaline environment adaptability, berry firmness, size, taste, and yield [2,3].

Despite significant advancements through traditional hybridization [7,8,9], advanced breeding methods such as mutation breeding have rarely been utilized in blueberries. There have been no reports on blueberry breeding that have applied approaches to improve mutation occurrence. Only natural bud mutations have been subject to sport selection, which has left blueberry breeding still at the junior stage. Mutation breeding, using physical and chemical mutagens, induces random and versatile variability in plant genomes, including site mutations and large fragment changes [10]. Because of the advances of mutation breeding in shortening the breeding time for new cultivar development, it has been widely applied to numerus fruit crops, such as pears, peaches, apples, bananas, grapes, plums, and citrus species, enhancing traits like fruit appearance, taste, aroma, tree structure, and self-compatibility [11]. Although mutation breeding is very promising in breeding customer-desired cultivars, it is unexpected that there has no any successful cases reported for blueberries.

The increasing market demand has led to a global expansion of blueberry cultivation. In the absence of advanced breeding methods, hybridization between cultivars has been extensively used to screen locally adapted seedlings in various countries, including China, Chile, Japan, the UK, and New Zealand [12,13,14]. For instance, China’s blueberry cultivation area has increased over 200-fold in the past 20 years [15]. However, the soil in many regions of China is unsuitable for North American-derived blueberries, necessitating breeding for better-adapted cultivars. The Southern highbush blueberry “Lanmei No.1” (LM1) is a commercially bred cultivar with outstanding soil adaptation in China. Additionally, LM1 blueberries have high anthocyanin content [16]. Thus, LM1 has been recognized as a national superior cultivar and recommended for cultivation in southern China. Currently, LM1 is the most widely cultivated native cultivar in China, highly valued in both the fresh berry market and for industrial anthocyanin extraction [17]. For the further breeding of elite cultivars like LM1, advanced breeding methods are more suitable than hybridization. Hybridization breeding, which combines the entire genomes of two parents, can lead to undesired phenotypes in the F1 generation. In contrast, mutation breeding is preferable for improving specific traits in elite cultivars while maintaining other desirable characteristics. Although physical agents induce mutations randomly, an enlarged selection population allows for obtaining desired plants with unique phenotype variations. Thus, establishing a mutation breeding system for blueberry breeding is essential for enhancing advanced cultivars. Among the successful methods of mutation breeding, gamma irradiation is the most commonly used method, which has produced over 70% of mutation-bred cultivars, including berry plants [18]. Thus, we selected LM1 as a breeding target, employed gamma irradiation, and aimed to establish a mutation breeding approach for blueberries.

2. Materials and Methods

2.1. Plant Materials and Growth Condition

The blueberry cultivar Lanmei No. 1 (LM1) was provided by Zhejiang Lanmei Technology Co., Ltd. (Zhuji, Shaoxing, China). Tissue culture seedlings were initially used, which were sub-cultured in vitro under conditions of 100 μmol m⁻² s⁻¹ light, 60% humidity, a 16/8 h (light/dark) photoperiod, and temperatures of 24/24 °C (day/night). Seedlings longer than 10 cm lengths were planted in a peat and vermiculite mixture (1:1) for root setting in a growth room and then transferred to a greenhouse for further development. Phenotypic observations on plant height and leaf morphology were made during the seedling growth period.

2.2. Plant Regeneration and Adventitious Shoot Cultivation

The regeneration medium used was Woody Plant Basal Medium with Vitamins (WPM; Phyto Tech, Lenexa, KS, USA, L449), supplemented with 8.5 g/L agar and specified concentrations of plant hormones and sucrose. The hormones commonly used for tissue culture including zeatin (ZT; Sangon Biotech, Shanghai, China, 1637-39-4), thidiazuron (TDZ; Sangon Biotech, 51707-55-2), kinetin (KT; Coolaber, Beijing, China, 525-79-1), 3-indole butyric acid (IBA; Sangon Biotech, 133-32-4), and 3-indole acetic acid (IAA; Sangon Biotech, 87-51-4) were used in this study. ZT, TDZ, and IBA were used for regeneration; similar to our previous study [19], KT and IAA hormones were used for rooting. Explants, including stems and leaves from in vitro seedlings, were used. Stems were cut into 1 cm fragments and placed on the medium, while leaves were halved and placed adaxial side down.

The optimized shoot-induction medium consisted of WPM with 5 g/L sucrose, 8.5 g/L agar, and 0.05 mg/L TDZ, at pH 5.2. Shoots longer than 2 cm were transferred from petri dishes to jars containing shoot growth medium of WPM with 30 g/L sucrose, 8.5 g/L agar, 0.05 mg/L ZT, and 0.05 mg/L IBA at pH 5.2 to promote elongation. To promote root setting, a medium containing 30 g/L sucrose, 8.5 g/L agar, 0.5 g/L charcoal, 0.3 mg/L IAA, and 1 mg/L KT was used. Plant materials were sub-cultured every two weeks. Each jar contains 18 seedlings.

2.3. Mutagenesis with Gamma Radiation

Uniformly growing and robust tissue culture seedlings were used for irradiation mutagenesis. ⁶⁰Co-γ rays equipped in the Crop and Nuclear Technology Research Institute of the Zhejiang Academy of Agricultural Sciences were used to irradiate LM1 tissue culture seedlings. The irradiation doses applied were 10, 20, 40, 80, 120, and 150 Gy, with a dose rate of 1 Gy min⁻¹. Unirradiated tissue culture seedlings served as controls. Each irradiation dose treatment involved 15 bottles of tissue culture seedlings. Following irradiation, the LM1 tissue culture seedlings were cut and prepared for regeneration using both stems and leaves. The explants were initially cultured in a shoot-induction medium. Once adventitious shoots emerged, the explants were transferred to jars containing shoot growth medium. When the seedlings reached a height of over 10 cm, the regenerated seedlings were transplanted into soils for root induction and seedling development. Three months later, seedlings showing significant growth differences from other plants were evaluated. This work is focused exclusively on growth phenotypes.

2.4. Phenotypic Evaluation of M1 Generation

All M1 plants were first placed in the growth room, and those with similar sizes were grouped together. Phenotypic evaluation was conducted after three months of growth. Leaf morphology was observed initially. As the plants grew larger, they were transferred to the greenhouse. Then the plant height was measured using a ruler, and the most outstanding plants were photographed, as shown in the figures.

2.5. Re-Sequencing

Individual plants of M1 generation (five plants in each irradiation group) were randomly selected for re-sequencing. Total DNA was extracted from leaves using the cetyltrimethylammonium bromide (CTAB) method [5]. A total amount of 0.2 μg DNA of each sample was used to construct the DNA library and then sequenced under DNBSEQ-T7 to produce 150 bp paired-end reads in Novogene Bioinformatics Technology Co., Ltd. (Beijing, China). The short reads (raw data) were obtained by transforming original fluorescence image files by base calling and recorded in FASTQ format. Then, the reads that contained undetermined, adaptor contamination and low-quality bases were removed using Fastp version 0.23.1 [20].

2.6. SNP Calling and Statistics

Paired-end resequencing reads were mapped to the Vaccinium corymbosum cv. Draper v1.0 genome [21] using BWA [22] with the default parameters. SAMtools V 0.1.18 software [23] was used to convert mapping data to BAM format. Unmapped, non-unique, and duplicated reads were filtered with the Picard package (picard.sourceforge.net (accessed on 11 December 2018), Version: 1.87). Variant detection for each sample was performed using the “HaplotypeCaller” function of the GATK4 version 4.3.0.0 [24], generating one gVCF file per sample. Subsequently, the gVCF file of four control samples and each Gy-treated sample were merged into a single VCF file using the “GenotypeGVCFs” function. The SNPs were extracted using the “SelectVariant” function and then filtered with QD < 2.0”, MQ < 40.0, FS > 60.0, SOR > 3.0, MQRankSum < −12.5, and ReadPosRankSum < −8.0 through the “VariantFiltration” function. Finally, the SNPs with the same allele among four control samples and each irradiation-treated sample were filtered using a Perl script, and the remaining SNPs were used for calculation.

2.7. Statistical Analysis

One-way ANOVA followed by Tukey’s post hoc tests for multiple comparisons was performed to compare different groups using GraphPad Prism 7 software [19]. p ≤ 0.05 was regarded as statistically significant. All experiments were repeated three times unless stated otherwise.

3. Results

3.1. Establishing a Regeneration Method for Blueberry Cultivar LM1

Gamma irradiation induces random mutations in plant cells, with each cell potentially exhibiting different mutations. A regenerated plant almost originated from a single progenitor cell [25]. Therefore, increasing the number of embryonic cells can enhance the diversity of mutated lines. Traditionally, fruit crop branches are commonly irradiated, and then the cuttings are used to generate screening populations [26]. However, this method relies on the number of surviving cuttings, making it less efficient compared to cell regeneration. Therefore, we opted for a tissue regeneration approach to propagate irradiated plant tissues

To develop a regeneration system for LM1, we initially screened two commonly used cytokinins for the Vaccinium species: thidiazuron (TDZ) and zeatin (ZT) [19]. It was found that TDZ was more effective in inducing adventitious shoots compared to ZT. Specifically, a low concentration of TDZ at 0.05 mg/L achieved the highest induction efficiency, while ZT failed to induce shoots at concentrations ranging from 0.5 to 2 mg/L (Figure 1A,B). We initially used 3% sucrose as the carbon supplement. To optimize the sugar concentration, we tested sucrose doses ranging from 5 to 40 g/L, corresponding to sucrose concentrations of 0.5% to 4%. Low concentrations of sucrose, specifically 5 and 10 g/L, performed better in shoot induction (Figure 1C). Shoots grown on 5 g/L sucrose exhibited more vigor compared to those on 10 g/L during later stages of growth. Therefore, the optimized regeneration medium is WPM medium with 5 g/L sucrose, 8.5 g/L agar, and 0.05 mg/L TDZ. Adventitious shoots were transferred to shoot elongation medium and subsequently to root induction medium to promote further growth.

3.2. Establishing a Mutation Breeding Approach Combining Gamma Irradiation Mutagenesis and Tissue Regeneration

To determine the optimal irradiation dose for the tetraploid blueberry LM1, the doses tested were 10, 20, 40, 80, and 120 Gy using ⁶⁰Co-γ as the irradiation resource. Seedlings exceeding 10 cm in height were exposed to gamma radiation. Post-irradiation, seedlings were separated into groups of leaves and stems for regeneration. Leaf discs were cultured in the dark for two weeks before being transferred to light conditions. Unirradiated leaves were used as the control, which successfully developed calluses and some adventitious shoots after two weeks, while the regeneration process of irradiated leaves was significantly delayed, and they exhibited yellowish symptoms (Figure 2A). The surviving part of irradiated leaves could develop into adventitious shoots with decreased efficiency along with the increase in irradiation doses (Figure 2B,C). Similarly, the growth rate of plants also diminished as irradiation dose increased (Figure 2D). These results confirm the effectiveness of our irradiation treatment.

The irradiated stems were cut into 2 cm fragments and horizontally placed on a shoot-induction medium. Initial regeneration was achieved for stems across all irradiation doses (Figure 3A). However, some regenerated shoots died after two months of cultivation (Figure 3A; bottom panel). This phenotype differs from the adventitious shoots derived from leaves, where no lethality was observed. Most adventitious shoots from irradiated stems originated from dormant buds. The lethal symptoms appearing after shoot sprouting suggest that high irradiation doses may cause fatal damage to these buds. Despite the lethality, viable shoots were obtained, although the regeneration rate decreased with increasing irradiation dose (Figure 3B). In conclusion, regenerated seedlings were successfully obtained from both leaves and stems irradiated with all doses of gamma rays.

The half-lethal dose (LD₅₀) is a crucial index for determining an optimal dosage that maximizes mutation rates while minimizing excessive lethality in mutation breeding [27]. Polypoid plants were considered to have a higher LD₅₀ than the diploid ones. To determine the LD₅₀ of tetraploid blueberry LM1, mortality rates for each irradiation dose treatment were recorded after two months of cultivation. To improve the accuracy of the LD₅₀, an additional treatment at 150 Gy, which resulted in nearly total lethality, was included. A linear regression equation was used to calculate the LD₅₀, which is 56.49 Gy for leaves (Figure 4A) and 81.83 Gy for stems (Figure 4B). Thus, the optimal gamma irradiation doses for tetraploid blueberry LM1 were determined to be 56.49 Gy for leaves and 81.83 Gy for stems. Overall, a mutation breeding method using gamma irradiation and tissue regeneration has been established for LM1.

3.3. Phenotypic Variation in the M1 Generation

For diploid plants, mutation breeding successfully yielded lines with desired traits in the M1 generation [28]. However, reports on polyploid plants are considerably fewer due to their complex genetic structures [29]. We screened the M1 seedlings after 6–8 months of growth in a greenhouse. Among the 4301 M1 lines, one notable phenotype was leaf color variation. Seven individuals exhibited significantly increased pigmentation, while one individual showed a dramatic decrease in pigments (Figure 5B,C). Significant variation in plant size was also observed. Many small individuals exhibited severe growth defects, rendering them of no breeding value. However, one exceptional individual displayed dramatically enhanced growth vigor, achieving more than twice the height of the average plants (Figure 5D). To evaluate the effects of irradiation dose on growth, we measured the heights of M1 generation lines. It was found that plant height decreased with increasing irradiation doses (Figure 5E). An exception was the 80 Gy group, which showed no decrease in height compared to the control (Figure 5E). No other significant phenotypic differences were observed between the dose groups. Overall, our breeding method successfully produced individuals with mutant traits in the M1 generation.

The optimization of the irradiation dose mainly uses LD₅₀ as an index. However, whether the number of mutation sites on the genome are linearly correlated with the increase in irradiation doses has not been examined. Nowadays, resequencing techniques make the examination of the mutation sites in each individual available, which allows us to check the trend of mutation along with irradiation increase. A total of 29 plants were used, including five plants in each irradiation dose and four plants without irradiation as the control. There are 507.76 G raw data generated by DNBSEQ-T7 sequencing with 150 base pair-end reads. After cleaning and quality checks, 504.47 G clean data were obtained, with an average of 17.40 G clean data per plant. The Q30 percentage of all sequences across the 29 libraries exceeded 96.48% (Table S1). The clean data were mapped to the reference genome of Vaccinium corymbosum cv. Draper v1.0 genome [21], with high mapping ratios ranging from 96.37% to 99.15% (Table S2). After single nucleotide polymorphism (SNP) calling and filtration, the SNP number of each individual and the mean SNP number of each irradiation group were calculated and exhibited in Figure 6. The trends of the SNP mean of each group are clear, which first increased along with the irradiation dose and then turned over at the 80 Gy dose (Figure 6). However, no significant difference was found between groups (Figure 6). Some individuals of the high dose groups, such as 20 Gy, 40 Gy, and 120 Gy, even have less SNP than the mean SNP number of the 10 Gy group (Figure 6). This indicates that the number of irradiations causing SNP in the regenerated plants is not fully dependent on the irradiation dose, which may serve as a reference for future breeding applications.

4. Discussion

Compared to genomic editing tools, mutation breeding offers the distinct advantage of avoiding regulatory restrictions associated with genetically modified organisms (GMOs) [30]. This makes mutation breeding a promising approach for blueberry improvement in agriculture, especially for elite cultivars. Mutation breeding can overcome undesired traits arising from traditional hybridization and introduce new characteristics in the M1 generation without genome exchange. In this study, our meticulously designed mutation breeding strategy successfully induced phenotypic changes in the M1 generation (Figure 5), providing a valuable reference for the breeding of polyploid blueberry cultivars. This method is much more efficient than the traditional sport selection that is merely dependent on natural mutation. As no other mutation breeding has been published on blueberries, efficiency comparison is not available yet. The cultivar LM1 is one of the most widely cultivated blueberries in China, valued for its adaptability to the clay soils of southern regions and its exceptional anthocyanin profiles [16,17]. While LM1 berries are highly favored for anthocyanin extraction, the absence of anthocyanins in fresh berries remains a significant breeding goal. Although precise gene editing could address this challenge, its application is currently restricted by regulatory limitations. Thus, mutation breeding emerges as the only legally viable approach under current conditions. We established an efficient regeneration method (Figure 1), with which we applied irradiation treatments and examined the LD₅₀ and the optimal irradiation dosage (Figure 4). This method successfully produced individuals with enhanced pigment accumulation in the leaves (Figure 5). In the future, this approach could facilitate the development of desired traits in LM1 without compromising its soil adaptability and anthocyanin production. Similar success has been reported in citrus, where mutation breeding led to seedless varieties or enhanced flavonoid profiles without affecting other traits [31,32]. Furthermore, our study explored the relationship between single nucleotide polymorphism (SNP) generation and irradiation dosage. At lower doses ranging from 10 to 80 Gy, the number of SNPs increased proportionally with the dose (Figure 6). However, an unexpected decline was observed at the high dose of 120 Gy. This could be attributed to the lethality of high-dose irradiation, where only minimally mutated cells survived and regenerated into shoots. Interestingly, the variation in SNP number between individual plants was larger than that observed between irradiation groups. Even in the low dose group, some individuals had bigger SNP numbers than those of high dose groups, though the highest SNP count was recorded in a plant exposed to the highest dose (Figure 6). This suggests that even low-dose irradiation can produce substantial SNP variation. Despite these findings, the correlation between irradiation dose and SNP generation requires further investigation, particularly with a larger sample size and more comprehensive analysis. Expanding this line of research would provide critical insights into optimizing irradiation doses for mutation breeding. Overall, this study represents the first application of mutation breeding for blueberries, demonstrating the feasibility of integrating gamma-ray irradiation with tissue regeneration to induce SNP and phenotypic variations in the M1 generation. These findings underscore the potential of mutation breeding as a powerful tool for developing improved blueberry cultivars with desirable traits.

5. Conclusions

This study explored the use of mutation breeding with gamma irradiation in highbush blueberries (Vaccinium corymbosum). The LD₅₀ were determined to be approximately 56 Gy for the leaves and around 80 Gy for the stems. With the high-efficiency regeneration protocol developed here, phenotypic changes were successfully observed in the M1 generation. This approach offers a valuable reference for future blueberry breeding programs. By using this strategy, researchers and breeders can potentially accelerate the development of blueberry varieties with desirable traits.