Characteristics of Interspecific Hybridization and Inbred Progeny of Pumpkin (Cucurbita moschata Duch.) and Winter Squash (Cucurbita maxima Duch.)

Abstract

Hybrid incompatibility and F₁ hybrid dysgenesis in the interspecific hybridization between pumpkin (Cucurbita moschata Duch.) and winter squash (Cucurbita maxima Duch.) have been previously noted. For this reason, it is difficult to obtain F₂ generations due to F₁ sterility. However, back-crossing and add-crossing can be utilized to overcome these incompatibility barriers in interspecific hybridization. To date, few studies have focused on parental selection, the back-crossing process, and offspring characteristics related to interspecific hybridization. Hence, we explored the F₁ seed acquisition rate, plant characteristics, and F₂ generation fertility after interspecific back-crossing and add-crossing hybridization of C. moschata and C. maxima. Moreover, as a female parent, X-1 (C. moschata) yielded F₁ seeds when crossed with winter squash (C. maxima) 10-04-3, 10-37, or 10-05-2. BC₁F₁ seeds and plants could be obtained with winter squash (C. maxima) as the recurrent parent. Some healthy F₂ seeds and plants were obtained from the interspecific hybrids, including ZJ-13, ZJ-8, and ZJ-11. Further fruit nutrition quality analysis showed that the starch, polysaccharide, cellulose, and dry material contents of ZJ-7 and ZJ-8 were significantly higher than those of the parent pumpkin and winter squash lines. The bioflavonoid content of ZJ-8 was lower than that of its parents, and its soluble protein was at the median value. Meanwhile, the bioflavonoid content of ZJ-1 was lower than that of its parents, but its soluble protein was significantly higher. These results suggest that back-crossing and multi-crossing can overcome the barriers to interspecific crossing of C. moschata × C. maxima. Some interspecific hybrid fruits had nutritional contents much higher than those of their parent lines. Additionally, X-1 (C. maxima) was found to act as a bridge species in interspecific hybridization. Collectively, in this study, the barriers to interspecific hybridization of C. moschata × C. maxima were addressed through crossing methods and choice of parents, and the obtained results are expected to provide novel support for interspecific hybrid breeding between C. moschata and C. maxima.

1. Introduction

Cucurbita (2x = 2n = 40) is a genus of important cucurbitaceous plants, widely grown as horticultural crops across the globe. There are more than 30 cultivated varieties and wild relatives [1]. Some resources have become important rootstocks or resistance resources for Cucurbitaceae crops, such as watermelon and cucumber. Cucurbita includes five cultivated species: Cucurbita maxima Duch., Cucurbita moschata Duch., Cucurbita pepo L., Cucurbita ficifolia Bouché, and Cucurbita mixta Pangalo [2]. C. pepo, C. maxima, and C. moschata are three commonly cultivated species of pumpkin with large cultivation areas [3]. Cucurbita species are known to have high nutritional value and health-protective properties; therefore, they have attracted increased interest in recent years.

Crop improvement is largely dependent on conventional methods of introduction, selection, or hybridization utilizing cultivated genotypes of a species. However, in a majority of crops, most cultivars are developed with relatively narrow genetic diversity. An estimated 75% of crop genetic diversity was lost in the twentieth century [4]. As is well known, interspecific hybridization can not only polymerize the nuclear genes of the parent, but also create new materials with different cytoplasmic genetic backgrounds, creating new paths for the breeding of breakthrough varieties. This method has been widely used in crops such as rice [5], cotton genus [6], brassica [7,8], and wheat [9,10]. The interspecific heterosis of Cucurbita crops is obvious [11], which has attracted the interest of breeders for a long time. Winter squash originated in the high dry areas of southern Peru, Bolivia, and northern Argentina in the Americas [12]. Under high-temperature conditions, virus disease and powdery mildew may cause it great harm. It is suitable for planting in cold and cool areas in the summer, which limits its widespread planting. However, pumpkin (C. moschata) originated in Mexico and Central and South America [12]. It also has a long history of cultivation in China, where it has formed many local varieties. It generally shows heat resistance, barren resistance, disease resistance, and wide adaptability, but its quality is greatly affected by cultivation conditions and its taste is considered far inferior to that of winter squash. Thus, improving the adaptability and stress resistance of winter squash by hybridization with pumpkin has attracted the attention of breeders.

Previous studies have shown hybrid incompatibility between pumpkin (C. moschata) and winter squash (C. maxima) [11,13]; however, interspecific crossing between different cultivated species has led to a diverse range of results [11,14,15]. Some interspecific Cucurbita maxima × Cucurbita moschata hybrid cultivars have been bred and widely used as a rootstock variety in vegetable production [16]. However, it is difficult to obtain F₂ generation interspecific hybrid inbred lines due to the hybrid sterility barrier of F₁ plants, which allows only immature hybrid embryos to be obtained from Cucurbita moschata × Cucurbita maxima [17].

Drude has pointed out that interspecific hybridization of Cucurbita maxima × Cucurbita moschata is subject to an incompatibility barrier [11]. Furthermore, Jiang Yu et al. observed pollen germination and pollen tube elongation in the process of positive and negative hybridization of C. moschata and C. maxima by aniline blue staining. They believed that distant hybridization is incompatible in the stage of interaction between pollen and stigma, and only a very small number of immature distant hybrid embryos could be obtained. It is difficult to obtain fertile F₁ embryos of hybrid combination with Cucurbita moschata as the female parent [18]. Back-crossing and add-crossing hybridization are important ways in which interspecific hybridization incompatibility can be solved [14,19,20]. Therefore, in this study, we used these methods for interspecific hybridization of Cucurbita moschata × Cucurbita maxima, and the characteristics of F₁ and F₂ plants were studied to provide a reference basis for the creation of new resistance resources.

2. Materials and Methods

2.1. Pumpkin Lines and Breeding Methods

2.1.1. Laboratory Lines

Our study was conducted from autumn 2019 to autumn 2021 in the greenhouse at the Hunan University of Humanities, Science and Technology, PR China (27°42′55.00″ N 111°59′54.27″ E). Pumpkin seeds were sown into 72-cell plastic seedling plates (2.7 cm × 2.7 cm; 12 mL volume), which were filled with mixed substrate of peat: vermiculite: perlite = 1:1:1 (v:v:v). When seeds sprout and cotyledon appeared, these seedlings were planted into the soil in the greenhouse. The average day/night temperature was 25–28 °C/16–20 °C, relatively humidity was 60–70%, and CO₂ was ambient. Six pumpkin and winter squash inbred lines were obtained from the Department of Horticulture Research of Hunan University of Humanities, Science and Technology in 2019 (

Table 1. Characteristics of the six pumpkin (Cucurbita moschata Duch.) and winter squash (C. maxima Duch.) inbred lines.

Inbred Line Name	Pumpkin Taxon	Growth Potential	Powdery Mildew Resistance Level	Maturity	Fruit Quality Level	Fruit Shape and Color
J-1	C. moschata	Strong	High	Late maturity	Moderate	Round with ribbed
N-1	C. moschata	Weak	Susceptible	Early maturity	High	Short, rod-shaped, and smooth
X-1	C. moschata	Medium	Moderate	Medium maturity	Moderate	Highly spherical and ribbed
10-37	C. maxima	Strong	Susceptible	Late maturity	High	Oblate with grey white skin
10-04-3	C. maxima	Weak	Susceptible	Medium maturity	High	Oblate with green skin and dark green spots
10-05-2	C. maxima	Strong	Moderate	Late maturity	High	Oblate with light green skin and green spots

). N-1, J-1, and X-1 are pure pumpkin (C. moschata) lines. J-1 has strong growth potential, high resistance to powdery mildew, late ripening, medium quality, round fruit, and rib grooves; N-1 has weak growth, no resistance to powdery mildew, early maturity, good quality, and short rod-shaped and smooth fruit; and X-1 has medium growth potential, moderate resistance to powdery mildew, medium ripening period, and medium quality. The fruit is highly spherical and has rib grooves. 10-37, 10-04-3, and 10-05-2 are pure winter squash (C. maxima) lines, where 10-37 has the highest content of flavonoids, strong growth potential, no resistance to powdery mildew, late ripening, good quality, and flattened round gray fruit; 10-04-3 has weak growth potential, no resistance to powdery mildew, medium ripening, good quality, oblate fruit, green skin, and dark green spots; and 10-05-2 has strong growth potential, moderate resistance to powdery mildew, late ripening, good quality, flat and round fruit, light green skin, and green spots.

2.1.2. Crossing Methods

From autumn 2019 to autumn 2021, the three pumpkin (C. moschata) and three winter squash (C. maxima) pure lines were crossed. A small number of F₁ (C. moschata × C. maxima) seeds were harvested. Subsequently, back-crossing, add-crossing, three-way-crossing, and four-way-crossing were conducted (Figure 1). The experiment used three pumpkin (C. moschata) inbred lines and three winter squash (C. maxima) high-generation inbred lines as materials. Nine interspecific hybridizations were conducted with C. moschata as the female parent and C. maxima as the male parent. The trial was carried out in the autumn of 2019, in order to evaluate the hybrid F₁ seed acquisition rate. Next, the F₁ generation was planted in the spring of 2020. Self-crossing, back-crossing, add-crossing, and multi-parent crossing were carried out to obtain the fruit or seed, in order to record their characteristics. In the spring of 2021, the harvested three-crossing, four-crossing, and self-crossing seeds were cultivated, the field plant and fruit characteristics were investigated and recorded, and the seeds of interspecific hybridization offspring were harvested. In the autumn of 2021, the seeds of the interspecific hybrid F₂ generation were sown and cultivated, and the plant and fruit characteristics were investigated.

2.2. Characteristics of Interspecific Hybridization and Inbred Progeny Plants

The seed characteristics, seed acquisition rate, and number of plants were investigated in three plants, and the average values of the three tests were taken. The fertility of F₁ and F₂ male flowers, as well as the characteristics of F₁ plants, seeds, and fruits, were investigated using the following method:

F₁ seed acquisition rate (%) = number of full seeds per fruit/total number of seeds per fruit × 100%. F₁ plant acquisition rate (%) = number of F₁ plants/number of seeds sown × 100%. F₂ seed acquisition rate (%) = number of full seeds per fruit/total number of seeds per fruit × 100%.

2.3. Methods for Assessing Nutritional Components of the Fruits

The nutritional components of the fruit pulp were determined using the following methods. The polysaccharide content was determined by phenol sulfuric acid colorimetry, the starch content was determined by anthrone colorimetry, the cellulose content was determined by furfural derivative colorimetry, the bioflavonoid content was determined by aluminum ion colorimetry, and the protein content was determined by protein Cu⁺ colorimetry. All of the above determinations were carried out using a 100T/96S kit (Jiangsu Keming Biotechnology Co., Ltd., Suzhou, China), and microdetermination was carried out using an enzyme labeling instrument with a 96-well plate.

2.4. Statistical Analysis

Excel 2007 was used for data processing and mapping, and SAS 8.1 software (SAS Institute, Cary, NC, USA) was used for the analysis of variance and correlation.

3. Results

3.1. Analysis of Plant Characteristics of Interspecific F₁ Hybrids

The trial schedule is shown in Figure 1b. Three pumpkin and three winter squash self-crossing lines, as shown in Figure 1a, were used to breed nine interspecific hybrids, the characteristics of which are provided in Table 1. Crossing was carried out according to the nine combination schemes shown in Figure 1b. As detailed in

Table 2. The seed and plant characteristics of interspecific hybrids of C. moschata × C. maxima.

Hybridization No.	Hybridization	Average Seed Acquisition Rate of F1 (%) ± SD	F1 Plant Acquisition Rate (%) ± SD	F2 Seed Acquisition Rate (%)	F1 Male Flower Fertility	F1 Plant Characteristics	F1 Seed Characteristics	F1 Fruit Characteristics
S-1	N-1 × 10-04-3	22.1 ± 0.5 d	8.4 ± 0.7 d	0	Sterile	Similar to female parent	Shape is similar to the male parent and the color is similar to the female parent	Fruit shape is a combined type, the color is similar to male parent and the flesh quality is similar to male parent
S-2	N-1 × 10-37	0	-	-	-	-	-	-
S-3	N-1 × 10-05-2	0	-	-	-	-	-	-
S-4	J-1 × 10-04-3	0	-	-	-	-	-	-
S-5	J-1 × 10-37	0	-	-	-	-	-	-
S-6	J-1 × 10-05-2	59.7 ± 1.9 c	67.7 ± 2.1 b	0	Fertile	Similar to female parent	Shape and color are similar to the female parent	Fruit shape, color, and flesh quality are of combined type
S-7	X-1 × 10-04-3	56.4 ± 0.7 c	29.9 ± 1.0 c	0	Partly fertile	Similar to female parent	Shape and color are similar to the female parent	Fruit shape, color, and flesh quality are similar to female parent
S-8	X-1 × 10-37	79.8 ± 0.8 a	72.2 ± 1.0 a	0	Sterile	Similar to female parent	Shape and color are similar to the male parent	Fruit shape and flesh are similar to male parent, color is of combined type
S-9	X-1 × 10-05-2	65.2 ± 2.7 b	62.6 ± 2.4 b	0	Fertile	Similar to female parent	Shape and color are similar to the female parent	Fruit shape is similar to female parent, color and flesh are of combined type

Values are means ± SEM (n = 3), with different lowercase letters above the table representing significant differences (p < 0.05).

, F₁ generation seeds were harvested from interspecific crosses S-1, S-6, S-7, S-8, and S-9, where the seed acquisition rates of the F₁ generation were 22.1%, 59.7%, 56.4%, 79.8%, and 65.2%, respectively, while the plant acquisition rates of the F₁ generation were 8.4%, 67.7%, 29.9%, 72.2%, and 62.6%, respectively. However, F₁ seeds were not obtained from S-2, S-3, S-4, and S-5. Furthermore, the F₁ generation interspecific crosses S-1, S-6, S-7, S-8, and S-9 were incompatible, and F₂ seeds could not be obtained. It should be noted that pumpkin X-1, as a female line, yielded F₁ seeds after distant crossing with winter squash 10-04-3, 10-37, and 10-05-2 (Table 2).

As shown in Table 2, the plant and seed characteristics of S-1, S-6, S-7, S-8, and S-9 were compared and analyzed. It was found that S-1 was male-sterile and S-7 was partially male-sterile. The plant characteristics of S-1, S-6, S-7, S-8, and S-9 were similar to those of the female parent. The shape of S-1 seeds was similar to that of the male parent, while the color was similar to that of the female parent. The shape and color of S-8 seeds were similar to those of the male parent. The fruit of S-1 was of combined type with respect to the male and female parents, and the color and fruit quality were partially similar to those of the male parent. The S-6 fruit was of combined type. The S-7 fruit shape, color, and quality were partially similar to those of the female parent. The fruit type and quality of S-8 were partially similar to the male parent, while the fruit pulp color was of combined type. The S-9 fruit shape was partially similar to that of the female parent, while the color was of combined type.

3.2. Analysis of Seed and Plant Characteristics of Back-Cross F₁

The back-crossing F₁ seed and plant characteristics are provided in

Table 3. The characteristics of back-cross combination BC₁F₁ and BC₁F₂ seeds and plants.

Hybridization No.	Back-Crossing	BC1F1 Seed Acquisition Rate (%) ± SD	BC1F1 Plant Acquisition Rate (%) ± SD	BC1F2 Seed Acquisition Rate (%) ± SD	BC1F1 Male Fertility	BC1F1 Plant Characteristics	BC1F1 Seed Characteristics	BC1F1 Fruit Characteristics
T-1	N-1 × 10-04-3 × 10-04-3	2.2 ± 0.2 d	0.0	-	-	-	-	-
T-2	N-1 × 10-04-3 × N-1	0.0	-	-	-	-	-	-
T-3	J-1 × 10-05-2 × 10-05-2	69.7 ± 0.2 b	61.8 ± 2.9 c	0.0	Sterile	Similar to C. moschata, strong growth potential	Light brown color, not smooth with ribbed edge	Black skinskin, long pendant shape
T-4	J-1 × 10-05-2 × J-1	0.0 ± 0.0	-	-	-	-	-	-
T-5	X-1 × 10-04-3 × 10-04-3	43.1 ± 1.6 c	0.0	-	-	-	-	-
T-6	X-1 × 10-04-3 × X-1	0.0 ± 0.0	-	-	-	-	-	-
T-7	X-1 × 10-37 × 10-37	81.8 ± 0.6 a	75.1 ± 3.0 a	0.0	Fertile	Similar to C. moschata, medium growth potential	Light brown color, smooth	Oblate shape, green or grey skin
T-8	X-1 × 10-37 × X-1	0.0	-	-	-	-	-	-
T-9	X-1 × 10-05-2 × 10-05-2	78.2 ± 3.5 a	67.6 ± 5.7 b	0.0	Fertile	Similar to C. moschata, medium growth potential	White color, smooth	Oblate shape, dark yellow with brown spots
T-10	X-1 × 10-05-2 × X-1	0.0 ± 0.0	-	-	-	-	-	-

Values are means ± SEM (n = 3), with different lowercase letters above the table representing significant differences (p < 0.05).

. In terms of the interspecific F₁ hybrids that could be obtained (S-1, S-6, S-7, S-8, and S-9), BC₁F₁ seeds could not be obtained when the recurrent parent was pumpkin (C. moschata). Meanwhile, BC₁F₁ seeds could be harvested when the recurrent parent was winter squash (C. maxima), and some BC₁F₁ generation plants were obtained. Specifically, T-1 (N-1 × 10-04-3 × 10-04-3) and T-5 (X-1 × 10-04-3 × 10-04-3) BC₁F₁ plants were not obtained, while T-3 (J-1 × 10-05-2 × 10-05-2), T-7 (X-1 × 10-37 × 10-37), and T-9 (X-1 × 10-05-2 × 10-05-2) yielded BC₁F₁ plants with plant acquisition rates of 61.8%, 75.1%, and 67.6%, respectively. The BC₁F₁ plant acquisition rate of T-7 was the highest, and T-3 was male-sterile. Therefore, seeds or BC₁F₁ plants can be obtained if winter squash (C. maxima) is selected as the recurrent parent for the interspecific hybridization of pumpkin (C. moschata) × winter squash (C. maxima).

3.3. Seed and Plant Characteristics of Triple-Cross or Four-Way-Cross F₁ and F₂

Triple-cross combinations ZJ-1 (X-1 × 10-37 × N-1), ZJ-6 (X-1 × 10-37 × J-1), and ZJ-13 (X-1 × 10-05-2 × J-1) yielded F₁ seeds and plants. Notably, ZJ-13 also yielded F₂ seeds and plants. This demonstrates that the distant hybridization incompatibility barrier can be overcome by adding hybridization, and F₂ plants can be obtained by self-crossing in ZJ-13 (

Table 4. The characteristics of triple-cross and four-way-cross F₁ seeds and plants.

Hybridization No.	Method of Hybridization	Triple- and Four-Way-Cross F1 Seed Acquisition Rate (%) ± SD	Triple- and Four-Way-Cross F1 Plant Acquisition Rate (%) ± SD	Triple- and Four-Way-Cross F2 Seed Acquisition Rate (%) ± SD	Triple- and Four-Way-Cross F1 Male Fertility	Triple- and Four-Way-Cross F1 Plant Characteristics	Triple- and Four-Way-Cross F1 Seed Characteristics	Triple- and Four-Way-Cross F1 Fruit Characteristics
ZJ-1	X-1 × 10-37 × N-1	41.8 ± 4.7 f	94.9 ± 5.3 b	0.0	Fertile	Similar to C. moschata	Light yellow color and smooth skin	Round shape with green skin with spots
ZJ-2	X-1 × 10-37 × N-1 × 10-37	67.7 ± 0.9 d	97.3 ± 4.7 a	0.0	Fertile	Similar to C. moschata	Light yellow color and smooth skin	Pendant shape, yellow smooth skin
ZJ-3	(J-1 × 10-05-2 × 10-05-2) × 10-05-2	90.9 ± 1.7 b	97.0 ± 1.4 a	0.0	Sterile	Similar to C. moschata, strong growth potential	Light brown and not smooth with ribbed edge	Long pendant with black skin
ZJ-4	X-1 × 10-37 × J-1 × 10-37	67.8 ± 2.2 d	95.7 ± 2.0 b	0.0	Fertile	Similar to C. moschata	Brown color and smooth skin	Round shape with green spotted skin
ZJ-5	X-1 × 10-37 × J-1 × 10-05-2	0.0	-	-	-	-	-	-
ZJ-6	X-1 × 10-37 × J-1	47.1 ± 1.2 e	98.4 ± 2.7 a	-	Fertile	Similar to C. moschata	Light brown color and smooth skin	Round shape with green spotted skin
ZJ-7	X-1 × 10-37 × J-1 × JG-1	20.8 ± 1.6 g	0.0	-	-	-	-	-
ZJ-8	X-1 × 10-37 × 10-37 × 10-37	92.1 ± 2.5 b	98.3 ± 2.9 a	92.3 ± 8.4 b	Fertile	Similar to C. moschata	Light brown color and smooth skin	Round shape with green spotted skin
ZJ-10	(X-1 × 10-37 × 10-37) × (X-1 × 10-05-2 × J-1)	0.0	-	-	-	-	-	-
ZJ-11	(X-1 × 10-37 × 10-37) × 10-04-3	86.3 ± 2.3 c	97.7 ± 2.0 a	97.2 ± 2.6 a	Fertile	Similar to C. moschata	Black brown color and smooth skin	Oblate shape with green spotted skin
ZJ-12	X-1 × 10-05-2 × J-1 × JG-1	0.0	-	-	-	-	-	-
ZJ-13	X-1 × 10-05-2 × J-1	98.0 ± 0.5 a	99.0 ± 1.7 a	97.2 ± 1.5 a	Fertile	Similar to C. moschata, resistance to powdery mildew and strong growth potential	Grayish white color and smooth skin	Pendant shape with brown smooth spotted skin
ZJ-14	X-1 × 10-05-2 × J-1 × 10-05-2	86.0 ± 2.3 c	97.3 ± 1.0 a	0.0	Sterile	Similar to C. moschata, strong growth potential	Light brown color and smooth skin	Round shape with dark green spotted skin

Values are means ± SEM (n = 3), with different lowercase letters above the table representing significant differences (p < 0.05).

). In addition, back-combination ZJ-3 (J-1 × 10-05-2 × 10-05-2 × 10-05-2) yielded C₂F₁, the plants of which were male-sterile, so BC₂F₂ seeds and plants could not be obtained. However, ZJ-8 (X-1 × 10-37 × 10-37 × 10-37) yielded BC₂F₂ seeds and plants (Table 4), demonstrating again that the barrier of distant hybridization incompatibility can be overcome by back-crossing hybridization, as F₂ plants could be obtained by self-crossing in ZJ-8.

Four-way-cross combinations ZJ-5 (X-1 × 10-37 × J-1 × 10-05-2), ZJ-10 ((X-1 × 10-37 × 10-37) × (X-1 × 10-05-2 × J-1)), and ZJ-12 (X-1 × 10-05-2 × J-1 × JG-1) did not yield F₁ seeds or plants. However, ZJ-4 (X-1 × 10-37 × J-1 × 10-37), ZJ-7 (X-1 × 10-37 × J-1 × JG-1), and ZJ-11 (X-1 × 10-37 × 10-37 × 10-04-3) did. It is vital to note that ZJ-14 (X-1 × 10-05-2 × J-1 × 10-05-2) yielded F₁ seeds and plants which showed 100% male sterility, and ZJ-11 (X-1 × 10-37 × 10-37 × 10-04-3) yielded F₂ seeds and plants (Table 4). These results indicate that distant hybridization can be used as a way to obtain male-sterile materials and add-crossing hybridization can overcome the obstacle of distant hybridization to obtain F₂ seeds and plants.

In the 13 different combinations (see Table 4), X-1 participated in the interspecific hybridization combinations of ZJ-13, ZJ-8, and ZJ-11, from which F₂ seeds and plants could be obtained, indicating that X-1 can be used as a bridge germplasm resource for interspecific hybridization between pumpkin and winter squash. Our results also show that the selection of bridge materials in interspecific hybridization is crucial to overcoming the incompatibility obstacle.

3.4. Comparative Analysis of Fruit Quality of Interspecific Hybridization

As shown in Figure 2, the bioflavonoid contents of the interspecific hybrid ZJ-13 and winter squash 10-37 were 3.19 mg/g FW and 3.40 mg/g FW, respectively, with that of ZJ-13 being significantly lower than that of 10-37. The starch contents of interspecific hybrids ZJ-7, ZJ-8, and ZJ-4 were 57.62 mg/g FW, 56.55 mg/g FW, and 48.28 mg/g FW, respectively, significantly higher than those of the six inbred lines; notably, there was no significant difference between ZJ-7 and ZJ-8. The polysaccharide contents of ZJ-7 and ZJ-8 were 30.98 mg/g FW and 22.61 mg/g FW, respectively, significantly higher than those of the six inbred lines. ZJ-1, ZJ-2, ZJ-4, and X-1 had moderate polysaccharide contents, with no significant difference among them. The cellulose contents of ZJ-7, ZJ-2, ZJ-1, ZJ-4, and ZJ-8 were 30.07 mg/g FW, 28.88 mg/g FW, 25.33 mg/g FW, 25.27 mg/g FW, and 24.06 mg/g FW, respectively, significantly higher than those of the six inbred lines. ZJ-1, ZJ-4, and ZJ-8 showed no significant difference among them. The soluble protein contents of ZJ-7 and ZJ-6 were significantly higher than those of the six inbred lines, with their contents being 62.36 mg/g FW and 53.71 mg/g FW, respectively. The dry matter contents of ZJ-8 and ZJ-7 were 20.22% and 19.83%, respectively, with no significant difference between the two. The dry matter contents of ZJ-1, ZJ-2, and ZJ-4 were significantly higher than those of the six inbred lines.

The fruit quality of the interspecific hybrids is shown in Figure 2. Among the six inbred lines, 10-37 had the highest content of bioflavonoids, up to 3.40 mg/g FW. The bioflavonoid content in the back-crossing hybrid BC₂F₁ fruit was 1.72 mg/g FW; thus, the bioflavonoid content could not be significantly increased by recurrent hybridization when 10-37 was used as a recurrent parent with high bioflavonoid content. However, the bioflavonoid content in the fruit of an interspecific F₁ hybrid was higher than that of either of its parents, each having moderate bioflavonoid content (Figure 2a).

The dry matter content of ZJ-8 (X-1 × 10-37 × 10-37 × 10-37) was 20.22%, significantly higher than that of the recurrent parent 10-37 and the nonrecurrent parent X-1 (Figure 2f).

ZJ-7 had the highest polysaccharide content (30.98 mg/g FW), followed by ZJ-8 (22.61 mg/g FW). These contents were both significantly higher than those of the parent lines X-1, 10-37, and J-1, with polysaccharide content of 4.87 mg/g FW, 18.46 mg/g FW, and 5.42 mg/g FW, respectively, indicating that interspecific hybridization can increase the content of fruit polysaccharides through add-crossing or back-crossing (Figure 2c).

As shown in Figure 2b,d,e, the starch, protein, and cellulose contents in the fruits of the add-crossing hybrid ZJ-7 were 57.62 mg/g FW, 62.36 mg/g FW, and 30.07 mg/g FW, respectively, while those in the fruits of back-crossing hybrid ZJ-8 were 56.55 mg/g FW, 30.60 mg/g FW, and 24.06 mg/g FW, respectively. Therefore, the content of starch, protein, and/or cellulose can also be increased through add-crossing or back-crossing.

4. Discussion

Breeding between individuals from different species belonging to the same genus (interspecific hybridization) or two different genera of same family (intergeneric hybridization) is called distant hybridization, and such crosses are known as distant crosses, interspecific crosses, or wide crosses [19]. Generally speaking, wide hybridization has been widely used to transfer specific gene(s) in vegetable crops.

Repeated back-crossing of wide hybrids to their parental species has also contributed to the evolution and speciation of some species by gene introgression, i.e., the chromosome fragments from one species into another through repeated back-crossing of wide hybrids to their parental species [21], thereby resulting in changes in the genotypes and phenotypes of the progenies. The early generations that were obtained from all the multiple-parent populations by sib- or self-pollinations segregated enormously in all traits containing plant and fruit types, disease resistance, sex expression, fertility, and self- and sib-compatibility. During the development of the interspecific families and lines, it was reported that the frequencies of some characteristics were skewed in favor of C. moschata and C. maxima in all populations, especially fruit and plant type [20]. We also found that single-, triple-, or four-way-cross F₁ plant characteristics were similar to C. moschata, with the seed and F₁ fruit characteristics being similar to C. moschata and C. maxima in very few instances, except for the biased hybrids.

The difficulties encountered during the production of such hybrids are mainly due to incompatibilities. The incompatibilities or barriers in interspecific hybridization can be broadly classified as pre- or postzygotic barriers. Not all of the interspecific and most of the intergeneric hybridizations confer success in achieving their targets. Crossability relations determine the potential gene exchange between crop plants and their wild relatives under either natural conditions or by experimental techniques. Techniques for overcoming these barriers include chromosome doubling [22], embryo rescue [23], or back-crossing [20]. In previously published case studies, full F₁ seeds have been obtained in the interspecific hybridization of C. maxima × C. moschata [16], but the F₁ hybrids indicated a sterility barrier that prevented F₂ generation hybrids. Otherwise, C. moschata × C. maxima crosses suffer from hybrid incompatibility and hybrid sterility [17], and most of them yield only immature hybrid embryos. Here, we found that recurrent parental selection was important in interspecific hybridization and using C. moschata as a recurrent parent to be crossed with interspecific hybrid F₁ plants did not result in BC₁F₁ seeds, while using C. maxima as the recurrent parent allowed for BC₁F₁ seeds and some plants to be obtained (Table 3).

When direct crosses between two species with the same or different ploidy levels are difficult or impossible to accomplish, a third species (called a bridging species) can be used to produce such crosses [19,24]. Lin D. [11] and the Food and Agriculture Organization of the United Nations (FAO) stated, in the global report Genetic Resources of Cucurbitaceae Plants in 1983, that C. moschata occupies the central position of interspecific hybridization, while W. Whitaker believed that hybridization between C. moschata and C. maxima is impossible. In previous studies, the interspecific inbred lines derived from the three-species base populations had a broader compatibility than the lines resulting from the two-species base populations, although successive selection pronouncedly increased the compatibility of all the lines [20]. A wild species [25,26], interspecific F1 [27,28], amphidiploidy, or induced polyploidy [29,30] as a genetic bridge played an important role in overcoming species barriers and the male sterility of interspecific F1 for gene transfer, none of these genetic bridges can solve the male sterile, incompatible, and infertile problems in the later generations [31,32]. Notably, we found that X-1, as one of the parents (C. moschata), participated in the interspecific hybridization combinations of ZJ-13, ZJ-8, and ZJ-11, all of which yielded F₂ generation seeds and plants by self-crossing (Table 4). These results indicated that X-1 could be used as a bridge species for interspecific hybridization between C. moschata and C. maxima, consistent with the study of Li B. et al. [15].

Several techniques, such as embryo rescue, in vitro embryo culture, ovary/ovule culture, reciprocal crossing, grafting, back-crossing, and chromosome manipulation, have been widely adopted in annual crop plants. However, embryo rescue, back-crossing, and chromosome improvement are the most common methods applied for fruit crop improvement [21]. In our study, back-crossing and add-crossing enabled us to obtain F₁ (pumpkin × winter squash) seeds and plants, and some of these plants could be self-crossed to obtain F₂ seeds and plants. In previous studies, during the transfer of important characteristics through bridges, unfavorable species-specific traits are frequently carried along to subsequent populations by initial interspecific hybridization [33]. Nevertheless, these disadvantages may be mitigated by intervarietal hybridization and selection [31,34]. In this study, we found that the content of starch, protein, and/or cellulose can also be increased through add-crossing or back-crossing. The starch, polysaccharide, cellulose, soluble protein, and dry matter contents of interspecific hybrids ZJ-7 and ZJ-8 were significantly higher than those of the six inbred lines. Notably, the starch, polysaccharide, cellulose, soluble protein, and/or dry matter of ZJ-7 were 22.80 mg/g FW, 12.52 mg/g FW, 10.71 mg/g FW, 36.93 mg/g FW, 5.59%, 42.94 mg/g FW, 26.12 mg/g FW, 24.43 mg/g FW, 19.85 mg/g FW, 11.43%, 42.91 mg/g FW, 25.56 mg/g FW, 24.33mg/g FW, 31.75 mg/g FW, and 12.53% higher than that of parent X-1, 10-37 and J-1, respectively; and those in the fruits of the back-crossing hybrid ZJ-8 were 21.73 mg/g FW, 4.15 mg/g FW, 4.69 mg/g FW, 5.17 mg/g FW, 5.98%, 41.87 mg/g FW, 17.74 mg/g FW, 18.42 mg/g FW, and 11.82% higher than that of parent X-1 and recurrent parent 10-37, respectively, while the soluble protein content of ZJ-8 was 11.91 mg/g FW lower than that of 10-37. Furthermore, the bioflavonoid contents in the fruits of the interspecific hybrid ZJ-13 and winter squash 10-37 were 3.19 mg/g FW, and 3.40 mg/g FW with no significant difference and those of ZJ-13 were 1.12 mg/g FW, 1.43 mg/g FW, and 0.95 mg/g FW, significantly higher than that of parents X-1, 10-05-2, and J-1. This implies that interspecific hybridization has the advantage of being able to create high-quality interspecific hybrids.

5. Conclusions

In summary, the characteristics of interspecific hybridization between inbred progeny of pumpkin (Cucurbita moschata) and winter squash (Cucurbita maxima) were described in our study. The results strongly imply that interspecific hybrid sterility barriers can be solved through the use of back-crossing, add-crossing, and recombinant inbred lines. X-1 appears to act as an important bridge species in the interspecific hybridization of pumpkin and winter squash. Additionally, the interspecific hybridization combinations were ZJ-13, ZJ-8, and ZJ-11, from which normal F₂ seeds and fertile plants could be obtained. Thus, back-crossing, add-crossing, and this bridge species can be used to overcome the obstacle of distant hybrid infertility. Furthermore, the starch, polysaccharide, cellulose, soluble protein, and dry matter contents after interspecific hybridization were heterogenous. Our results provide new insights into a potential method to resolve the barriers of distant hybrid incompatibility, which are expected to enrich studies of pumpkin interspecific hybridization. Additionally, this work is beneficial for implementing the advantages of interspecific hybridization in order to create high-quality interspecific hybrids, for example in high-quality disease-resistance breeding.