INTRODUCTION

The black node disease is a severe fungal alteration that affects beans in regions with cold to moderate temperatures (15-25°C), continuous rainfall, and high relative humidity (80-100%). In the Americas, the Black node disease is commonly found in cold climate regions at more than 1,500 m a.s.l. Infestation can be very severe, causing 100% yield loss in bush beans in Ecuador, Colombia and Peru, among other countries in the Andean region of South America, and is also significant in high elevation areas of Guatemala in Central America (^{Garzón et al., 2011}; ^{Miranda et al., 2021}).

Sustainable production strategies in current agri-food systems are important for reducing unfavorable impacts on the environment and generating products that alleviate malnutrition for millions of people (^{Powers and Thavarajah, 2019}). The common bean is of interest when improving food systems because of its great diversity and use of genetic recombination that transfers helpful genes for resistance to biotic and abiotic factors (^{Montejo et al., 2022}). Recombination is achieved from artificial crosses between genotypes of interest, either at the intraspecific or interspecific level (^{Singh et al., 1991}; ^{Montejo et al., 2022}). This technique is widely used in bean breeding programs in search of resistant materials and increased dry grain yield (^{Barrios et al., 2011}). However, research on this topic has been limited by the lack of continuity and the scarce publication of studies on interspecific breeding in beans. In addition, there are no reports on the mechanisms of inheritance of resistance to some abiotic factors and limiting diseases for bean production, as is the case for black node (Boeremia noackiana [Allesch.] Aveskamp, Gruyter & Verkley).

According to ^{Beebe and Corrales (1991)}, intraspecific crosses do not usually show problems as opposed to interspecific crosses that frequently display incompatibility between materials. Intraspecific crosses are used when it is necessary to introgress genes of interest and are widely applied to transfer genes with resistance to pathogens (^{Silue et al., 2014}).

^{Hanson et al. (1993)} reported that there are no known Phaseolus vulgaris L. materials with resistance to the black node disease caused by the pathogen Boeremia diversispora (Bubák) Aveskamp, Gruyter & Verkley. However, several studies have reported genotypes with an intermediate reaction (^{Garzón et al., 2011}). Thus, the ASC population from the Phaseolus polyanthusGrenm. resistant genotype G35575 developed by the International Center for Tropical Agriculture (CIAT) was explored. This species has a close evolutionary relationship with P. vulgaris (^{Chávez, 2017}).

Segregating populations of these interspecific crosses are unstable, especially given their reversal to P. vulgaris when this species is used as the female parent. However, reciprocal crosses using P. polyanthus as a female parent attenuate this reversal to P. vulgaris. Thus, the recombination frequency for tolerance to B. noackiana that is typical of P. polyanthus is increased, while maintaining the fertility and grain quality of P. vulgaris (^{Camarena and Baudoin, 1987}).

The objective of this study was to develop populations of filial generations with possible genetic resistance to the black node disease (B. noackiana) in susceptible common bean cultivars with commercial grain characteristics using the introgression of desirable P. polyanthus genes.

MATERIALS AND METHODS

This study was carried out under greenhouse conditions in the facilities of the Faculty of Agricultural Sciences at the Universidad Nacional de Colombia, Bogotá campus during 2017-2018. Permanent cultivation was established under tropical climate conditions, at an altitude of 2,650 m a.s.l., with an average annual temperature of 18.6°C, maximum and minimum temperatures of 40 and 6 °C, respectively, and average relative humidity of 70%.

The plant materials were ASC 160, ASC 162, and NCB 226 provided by CIAT, within the framework of the project “Analysis of the distribution and genetic and pathogenic diversity of ascochyta (Boeremia spp.) in beans as a tool for the development of improved cultivars and small agriculture in Colombia”. The first two materials were obtained from the progeny of the resistant parent G35575 (P. polyanthus) and CAL 96 (P. vulgaris), with 16 successive generations of individual and mass selections, from CIAT’s genebank. These materials show high resistance with disease reaction values of 1 for leaves and 2 for pods according to the CIAT’s scale for the pathogen Boeremia spp. (^{Schoonhoven and Pastor-Corrales, 1987}). The materials were selected for their resistance and genetic closeness to P. vulgaris. Phaseolus polyanthus is believed to be a natural interspecific cross between cultivated forms of P. vulgaris and Phaseolus coccineus, which could facilitate the development of populations of filial generations (^{Silue et al., 2014}). The material NCB 226 (P. vulgaris) was used as the susceptible control.

P. vulgaris commercial cultivars requiring resistance to this pathogen were selected for their traits of grain size, color, shape, and consumer acceptance. The cultivars were: Sutagao, an improved climbing cultivar with red seeds, and Bacata, an improved shrubby cultivar with red seeds. The regional varieties Simijaca and Cabrera, both with red seeds, were used as controls (Tab. 1).

Two sowing systems were used: first, seeds were sown directly in the soil; six 20 m long rows were used at distances of 1.1 m between rows and 0.2 m between plants, for an area of 132 m². The parental plant was sown at 1.5 linear m for shrubby beans and 2 linear m for climbing beans. The sowing was carried out in complete rows, staggered for all genotypes. The second sowing system was in 3 kg pots with a mixture of soil and rice husk plus 20 g of the commercial product 15-15-15 (NPK) for edaphic fertilization, as well as foliar fertilization with amino acids and microelements at 30, 40, and 60 days after sowing (DAS). Additionally, adequate management was carried out to control typical insect pests in common beans, such as whitefly (Trialeurodes vaporariorum Westw.), mites (Tetranychus sp.), thrips (Frankliniella sp.), and cutworms (Agrotis ipsilon Hufn.). Fungicides were also applied to manage powdery mildew (Erysiphe polygoni). The two sowing systems were complementary, for better use of the availability of flowers for crosses since the genotypes had very dissimilar life cycles up to flowering (Tab. 1). Therefore, the sowing methods were not quantified.

The hybridization technique with emasculation was used for the crosses. This method ensures that the resulting seed is the progeny of the cross between two different parents. This procedure was suggested by ^{Pimentel et al. (2021)}, which consisted of taking a flower bud (representing the mother plant) and opening the petals with fine-tipped tweezers; then, the standard was opened by the suture, accessing the two keel petals to remove the anthers. The male parent was a completely open flower; as in the previous step, the standard and the keel were pulled in opposite directions with tweezers. The stigma that was covered with viable pollen from the mature anthers was extracted. In the final step, the stigmas were rubbed together and left intertwined on the mother's flower bud to provide more time for pollination.

^{Blair et al. (2011)} and ^{Garzón et al. (2011)} determined the days to anthesis for some cultivars when working on resistance to Boeremia spp. This information, along with the growth habit of the cultivars, was considered for the sowing. Shrubby materials flower between 30 and 65 DAS, while variable materials flower on average at 95 DAS (Tab. 1).

To obtain F1 seeds, the homozygous lines ASC 160 and ASC 162 (resistant to the disease) were crossed with five cultivars, four of them of commercial grains adapted to high tropics, and the non-commercial cultivar NBC 226 (Tab. 1). The direct cross was susceptible parent 1 × resistant parent 2, and the reciprocal cross was resistant parent 1 × susceptible parent 2 (Tab. 2 and 3). A cross was viable when it produces a pod with at least one seed, the efficiency of the cross is the percentage of viable crosses out of the total number of crosses made for each pair of parents, and the number of seeds per pod results from the average number of seeds of viable crosses once deformed, fractured and very small seeds are discarded.

For the second cycle of crosses, the best parent of P. vulgaris was taken in the interspecific crosses because of greater efficiency in the number of viable crosses and seed quantity. The parents and F1 were sown in a staggered manner, where F2 populations were obtained with self-pollination in some hybrid seeds. Backcrosses (BC) were carried out from the crosses between F1. BC1 was the cross between F1 and its female parent, and BC2 was the cross between F1 and its male parent (Tab. 4).

No experiment design was used to obtain the bean progeny; the treatments corresponded to the families. The variables recorded per plant were: Number of total crosses for each resistant parent, number of viable crosses, date of crossing, and seed quantity. Differences between treatments were determined using t-tests (P<0.05) with SAS^® .

RESULTS AND DISCUSSION

The broad crossing plan in this study consisted of three genotypes susceptible to B. noackiana and three genotypes with characteristics of commercial grain for Colombia with an unknown reaction to B. noackiana, which were crossed with three resistant genotypes from CIAT (Tab. 1). According to ^{Muñoz and Hidalgo (1986)}, the use of a wide range of genotypes with different origins, seed types, and growth habits makes it possible to significantly increase interspecific hybrids. Consequently, to increase the success of crosses, cultivars adapted to the Andean zone were used, and a high number of pollinations were carried out. ^{Toussaint et al. (2004)} carried out studies at the Faculty of Agricultural Sciences of the University of Gembloux with the genotypes: G 21245 wild form (WLD) and NI 637 cultivated form (CV) in P. vulgaris, NI 1015 (CV) and G 35348 (CV) in P. Polyanthus, CIAT origin, and reported abnormalities observed during embryo development, which depended to a great extent on the compatibility between the crossed genotypes. This showed the difficulties caused by interspecific crosses and the slow progress when this type of cross is carried out in a breeding program.

Of all the possible direct or reciprocal crosses between the five susceptible genotypes of P. vulgaris and the two resistant genotypes of the ASC population, the direct crosses where P. vulgaris was the female parent showed the highest efficiency. On the other hand, when the resistant materials were used as the female parent, the efficiency dropped drastically, reaching values of zero. This difference was significant according to the Student's t-test (P≤0.05). Tables 2 and 3 summarize the number of crosses, as well as the variables related to the percentage of crossing efficiency.

The detachment of immature pods and the low generation and viability of seeds stood out among the factors that affected the efficiency between crosses in this study, whether direct or reciprocal. These problems coincide with those described by ^{Gepts (1981)}, specifically for interspecific crosses of P. vulgaris with P. coccineus and P. polyanthus. However, the crosses described in tables 2 and 3 provide a viable option to prioritize crosses based on greater efficiency and seed production.

According to ^{Hadley and Openshaw (1980)}, one of the possible causes of pod abscission is low embryo viability, which results in the loss of F1 hybrid zygotes or developing embryos generated by successful pollination. ^{Ibrahim and Coyne (1975)} were able to maintain a high percentage of developing pods from interspecific crosses in Phaseolus for 30 d with the help of hormones, as compared to pod abortion or abscission in the control treatment 15 d after pollination. Several studies have achieved normal plant growth and development, obtaining hybrid plants through in vitro culture of immature embryos (^{Geerts et al., 2011}). There are different options to expand the genetic base of bean cultivars through interspecific hybridization; however, because of the low costs, conventional breeding with crosses in low-resource programs may be the most viable option.

Another important factor in the formation of interspecific hybrids was the lower efficiency in reciprocal crosses, as compared to direct crosses when the ASC population was used as the female parent. This result agrees with the studies cited by ^{Gepts (1981)}. ^{Camarena and Baudoin (1987)} reported that the direction of crossing results from the nucleus-cytoplasm interaction, which does not occur when the female parent is P. vulgaris. ^{Gepts (1981)} suggested that the combination of parents is of interest in the success rate. Thus, if the female parent is P. vulgaris, there is an average of 1.2 seeds per pollination; whereas, if the cross is reciprocal (i.e., P. coccineus is the female parent), only 0.004 seeds are obtained per pollination. This behavior was observed in this study; even when the success rate in the cross Bacata × ASC 160 had an efficiency of 12% and a mean number of seeds per pod of 0.6, its reciprocal cross had an efficiency of 6% and a mean number of seeds per pod of 0.8, which were higher than in the other tested combinations. Cultivars Bacata and Simijaca may have lower specific genetic barriers that allow greater interspecific hybridization efficiency. In addition, efficiency varies according to the source of resistance. This can be attributed mainly to the genetic differences of the parents, as suggested by ^{Gepts (1981)}, and to the adaptation of the cultivars to the environmental conditions of the experiment when making the crosses (^{Berke, 2000}).

According to ^{Ibrahim and Coyne (1975)} and ^{Silue et al. (2014)}, the transfer of a trait from one species to another is always accompanied by a sterility effect. The hybrid is easier to obtain when P. vulgaris is used as the female parent, but the degree of sterility in F1 is higher than in reciprocal hybrids. Similar results were obtained in this study since the mean crossing efficiency was 8% when the female parent was P. vulgaris; whereas, for the reciprocal cross, the mean crossing efficiency was 5%. However, the mean number of seeds per pod was higher for the reciprocal cross with P. polyanthus as the female parent (Tab. 2 and 3). ^{Manshardt and Waines (1983)} stated that the gametes of the zygotes had a high proportion of P. vulgaris alleles from the cytoplasm of the same species, displaying a higher degree of selectivity. ^{Chenq et al. (1981)} studied F1 lines from bean interspecific hybrids and reported that meiosis was abnormal, possibly because of an interaction between the genome and the cytoplasm or interactions between chromosomes. In the case of F1 hybrids of P. vulgaris crossed with P. coccineus, a close relative of P. polyanthus (^{Chávez, 2017}), abnormal meiosis was reported, where at least two pairs of chromosomes were differentiated by inversions. Studies on anaphase I and II pollen mother cells have indicated that inversions are located at the end of differentiated chromosomes and are responsible for up to 79% of pollen abortion in interspecific hybrids (^{Cheng et al., 1981}).

For the second cycle of crosses from the five susceptible parents, cultivar Bacata was selected as one of the more efficient in both direct and reciprocal crosses with both ASC sources of resistance in the first cycle of crosses (Tab. 2 and 3). 'Bacata'’s precocious life cycle facilitated the formation of populations of subsequent BC generations and F2. Table 4 shows that the two F1s, both 'Bacata' with 'ASC160' and 'Bacata' with 'ASC162', exhibited a crossing efficiency equal to or greater than 10%. When this progeny was crossed again with its parents, the crossing efficiency of the new BC was increased with values that ranged between 31 and 67%. This difference was significant according to the Student’s t-test (P≤0.05). For the sources, 'ASC 162' showed the highest crossing efficiency without differences with 'ASC 160' in the mean number of seeds per pod and seed weight. The results are summarized in table 4.

In the populations generated from the parent 'Bacata' (F1, BC, and F2), the mean number of seeds obtained per cross was 2.13. This value was higher than that reported by ^{Gepts (1981)}, who obtained a mean of 1.3 seeds. The better performance in the 'Bacata' populations may have been due to the fact that the ASC population was used. This population comes from a simple cross between P. vulgaris and P. polyanthus, which has been increased during several cycles and has greater adaptation than F1 populations from recent interspecific crosses. Therefore, the ASC lines may work as ‘genetic bridges’, being compatible with the selected 'Bacata' parent; whereas, the values reported by ^{Gepts (1981)} were the result of a simple cross between P. vulgaris × P. polyanthus parentals.

Despite the low rate of seed production per crossing, the efficiency of the crosses varied significantly between the populations of single F1 crosses, backcrosses (BC), and F2 of 'Bacata'. This behavior was similar to that observed by ^{Gepts (1981)}, who indicated that efficiency depends on the combination of the parents and that subsequent generations have a progressive restoration of the viability and fertility of the populations. In this study, the backcross (Bacata × ASC 162) × ASC 162 reached an efficiency of 67%.

The fact that the backcrosses with the sources of resistance, 'ACS 160' and 'ACS 162', have shown crossing efficiency values between 31 and 67%, a valuable contribution to broadening the useful genetic base. The ability to exploit it for all traits of agronomic value that can be transferred from the P. polyanthus source provides a great opportunity for breeding for resistance. According to ^{Camarena and Baudoin (1987)}, early generations of P. vulgaris × P. polyanthus and P.coccineus crosses are characterized by a reversal to P. vulgaris, with a loss of 50% of the traits of the sources of resistance in each generation because of the interaction between the cytoplasm and the genome. This was avoided in the present study since the ASC sources were used as the female parent of the F1 of Bacata × ASC 160 or Bacata × ASC 162 from backcrossings. Additionally, these crosses obtained 30 and 78 seeds, respectively, which means that using backcrosses may see a rapid reversal of the traits of interest to the recurrent parent in subsequent generations. In the case of the 'ASC 162' source, the seed weight was 0.82 g, a large size. This is an important factor for grain selection processes for commercial characteristics in high Andean bean production systems.

Consequently, it was possible to obtain enough seeds to continue developing populations of advanced generations for selection processes of breeding lines with varietal resistance. These results suggested that it is possible to perform introgression of genes for resistance to the black node disease (B. noackiana) in common beans (P. vulgaris) through interspecific crosses with P. polyanthus using conventional procedures. To continue the studies on greater efficiency in the use of genetic variability in P. vulgaris, research that determines the deficiency in endosperm development and failures in the transport of nutrients from the female parent tissue to the zygote can complement the information reported by ^{Geerts et al. (2002)} and ^{Nguema et al. (2007)}.

CONCLUSIONS

The genotypes (CIAT) ASC 160 and ASC 162, which are resistant to the black knot disease (Boeremia noackiana), showed greater genetic compatibility with P. vulgaris when used as male parents, with 8% cross viability. The viability in the reciprocal crossings was 5%. These values are higher than those reported in previous studies on the formation of interspecific hybrids in bean breeding programs in the Andes. The shrub-type Bacata and the climbing Simijaca were more compatible with the source genotypes of resistance. Thus, the Bacata × ASC 160 and Bacata × ASC 162 populations and the Simijaca parent, when used as a male parent, can constitute the basis for an interspecific bean improvement program for varietal resistance since they produce viable seeds with grain characteristics close to commercial ones.