The Effect of Self-Compatibility Factors on Interspecific Compatibility in Solanum Section Petota

The relationships of interspecific compatibility and incompatibility in Solanum section Petota are complex. Inquiry into these relationships in tomato and its wild relatives has elucidated the pleiotropic and redundant function of S-RNase and HT which tandemly and independently mediate both interspecific and intraspecific pollen rejection. Our findings presented here are consistent with previous work conducted in Solanum section Lycopersicon showing that S-RNase plays a central role in interspecific pollen rejection. Statistical analyses also demonstrated that HT-B alone is not a significant factor in these pollinations; demonstrating the overlap in gene function between HT-A and HT-B, as HT-A, was present and functional in all genotypes used. We were not able to replicate the general absence of prezygotic stylar barriers observable in S. verrucosum, which has been attributed to the lack of S-RNase, indicating that other non-S-RNase factors play a significant role. We also demonstrated that Sli played no significant role in these interspecific pollinations, directly conflicting with previous research. It is possible that S. chacoense as a pollen donor is better able to bypass stylar barriers in 1EBN species such as S. pinnatisectum. Consequently, S. chacoense may be a valuable resource in accessing these 1EBN species regardless of Sli status.


Introduction
The value of wild species in the genetic improvement of potato was first demonstrated over a century ago in response to late blight epidemics in the northern hemisphere [1]. Unfortunately, several barriers inhibit efficient introgression of traits from wild species. These barriers can be largely separated into three categories: differences in ploidy, prezygotic barriers, and post-zygotic barriers [2]. Here, we will examine prezygotic interspecific barriers associated with gametophytic self-incompatibility in potato.
The 107 wild species of potato represent a rich and diverse source of disease resistance and tuber quality traits for cultivated potato [2,3]. However, in the century-long history of potato breeding the enormous value of wild species traits has been realized only occasionally [1]. Among all the species characterized, the diploid 1 Endosperm Balance Number (1EBN) species are of particular interest, as they exhibit exceptional resistance to economically important pests such as early blight (Alternaria solani), Colorado potato beetle (Leptinotarsa decemlineata), and late blight (Phytophthera infestans) [4][5][6]. However, due to prezygotic stylar barriers and differences in effective ploidy between 1EBN species and cultivated germplasm, accessing the traits from these species is extraordinarily difficult [2,7,8].
Here, we seek to clarify the role of prezygotic barriers that prevent the use of 1EBN species in diploid potato breeding schemes.
Gametophytic self-incompatibility (GSI) is the default condition in diploid cultivated potato and most of its wild relatives [9][10][11]. GSI in Solanaceae is governed by a the multiallelic S-locus on chromosome 1, containing S-RNase tightly linked to multiple SLF genes [12][13][14]. This GSI system also serves the dual function of preventing interspecific

SC Factors Tested
Two different mechanisms are used as sources of SC in potato, S-RNase based, and Slocus inhibitor (Sli) based SC. S-RNase based SC aims to remove barriers by the introduction of dysfunctional S-RNase alleles from wild species or knocking out functional alleles via gene editing [12,25]. S-RNase based SC can also be supplemented by the addition of dysfunctional alleles of HT and possibly other factors [20,23,26]. Central to the GSI system, the function of S-RNase is to degrade the RNA of incompatible pollen tubes. In plants with non-functional S-RNase such as tomato, the central barrier to self-pollen is eliminated, resulting in SC and broad interspecific compatibility ( Figure 1) [15,20,23]. While S-RNase based SC aims to eliminate barriers associated with GSI, Sli-based SC hijacks the GSI system to achieve SC. In functional GSI systems a suite of 16-20 SLF proteins act as part of a collaborative non-self-recognition system to degrade non-like S-RNase proteins [13,27]. Sli is a non-S-locus F-box protein located on chromosome 12 capable of interacting with a wide range of S-RNase variants [27]. Thus, the presence of Sli in self-pollen initiates the degradation of S-RNase and confers SC even when the GSI system is still functional ( Figure 1) [27,28].
The ability of Sli to overcome interspecific reproductive barriers (IRBs) was first reported by Sanetomo et al. [29], and the role of HT and S-RNase in IRBs has been demonstrated in Solanum section Lycopersicon by Tovar-Mendez et al. [23]. In potato, S. verrucosum is well recognized for its SC and lack of general IRBs [2,19,30]. The strong SC and lack of IRBs observable in S. verrucosum is likely rooted in its lack of a functional S-RNase protein [25,31]. While it is known that there is no detectable S-RNase on the protein level, it is unknown if the alleles for S-RNase in S. verrucosum are missing, non-functional, or inhibited in some way, or if HT or other factors play a role in these interactions [31].
The two objectives in this study were to: First, characterize the role of S-RNase and HT in the IRBs between cultivated S. tuberosum as the female and the wild species S. bulbocastanum, S. commersonii, S. jamesii, and S. pinnatisectum as pollen donors. CRISPR-Cas9 gene knockouts of S-RNase and HT-B in the S. tuberosum clone DRH195 were used to examine each of these factors independently and together. These results were compared with S. verrucosum as the female to see if knocking out S-RNase and HT-B replicated the phenotype observed in S. verrucosum. Second, in order to better capture the value of Sli, a set of crosses was designed to confirm the findings of Sanetomo et al. [29] that Sli contributes to interspecific compatibility and determine if these findings can be applied to other species. The ability of Sli to overcome IRBs could be an important factor in accessing the sexually isolated 1EBN species [29]. Additionally, the widespread use of Sli based jamesii, and S. pinnatisectum were used as pollen donors. In experiment 1, wild-type DRH195 (S. tuberosum) was used as the self-incompatible (SI) control and compared with CRISPR-Cas9 knockouts of S-RNase and HT-B in DRH195. Experiment 2 evaluated interspecific reproductive barriers in S. verrucosum, where the self-compatibility (SC) factors are not fully understood. In experiment 3, the S. chacoense clones M6 and USDA8380-1 were used as pollen donors since the SC factor being tested (Sli) is pollen expressed.

The Role of S-RNase and HT-B in Interspecific Pollen Rejection
In order to determine the contributions of S-RNase and HT-B in interspecific pollinations, we utilized independent and dual CRISPR-Cas9 knockouts (KO) of S-RNase and HT-B of the S. tuberosum clone DRH195 as females in this experiment ( Table 1). The wild species S. bulbocastanum, S. commersonii, S. jamesii, and S. pinnatisectum were used as pollen donors to test the strength of the IRBs in the female lines used. Pollinations were made using fresh pollen on newly opened flowers, and styles were collected for pollen tube measurements 48 h post pollination.

The Role of S-RNase and HT-B in Interspecific Pollen Rejection
In order to determine the contributions of S-RNase and HT-B in interspecific pollinations, we utilized independent and dual CRISPR-Cas9 knockouts (KO) of S-RNase and HT-B of the S. tuberosum clone DRH195 as females in this experiment ( Table 1). The wild species S. bulbocastanum, S. commersonii, S. jamesii, and S. pinnatisectum were used as pollen donors to test the strength of the IRBs in the female lines used. Pollinations were made using fresh pollen on newly opened flowers, and styles were collected for pollen tube measurements 48 h post pollination. Stylar squash assays were employed to visualize the growth and inhibition of pollen tubes within the style. Results from statistical analysis of pollen tube growth are consistent with previous findings in tomato showing that S-RNase plays a central role in interspecific pollen rejection [23]. This is readily observable in the growth of pollen tubes from S. bulbocastanum and S. pinnatisectum. Pollen tubes from these two species were able to penetrate significantly further down the styles in the S-RNase KO of DRH195 compared to the wild-type. Additionally, when S. bulbocastanum and S. pinnatisectum were the pollen donors, the longest pollen tubes were nearly able to traverse the entire length of the style in S-RNase knockouts. With S. jamesii and S. commersonii as the pollen donors the effect of S-RNase was not statistically significant, although the inhibitive effect on pollen tube growth of S-RNase was still evident for S. jamesii ( Figure 2). It should be pointed out that in pollinations with S. pinnatisectum, the difference between pollen tube growth in the S-RNase KO vs the S-RNase/HT-B KO was statistically significant, but not practically significant. Pollen tube penetrance in these pollinations was highly variable, and although a few pollen tubes from S. pinnatisectum were able to reach the ovary, no fruit development was observed. Stylar squash assays were employed to visualize the growth and inhibition of pollen tubes within the style. Results from statistical analysis of pollen tube growth are consistent with previous findings in tomato showing that S-RNase plays a central role in interspecific pollen rejection [23]. This is readily observable in the growth of pollen tubes from S. bulbocastanum and S. pinnatisectum. Pollen tubes from these two species were able to penetrate significantly further down the styles in the S-RNase KO of DRH195 compared to the wild-type. Additionally, when S. bulbocastanum and S. pinnatisectum were the pollen donors, the longest pollen tubes were nearly able to traverse the entire length of the style in S-RNase knockouts. With S. jamesii and S. commersonii as the pollen donors the effect of S-RNase was not statistically significant, although the inhibitive effect on pollen tube growth of S-RNase was still evident for S. jamesii ( Figure 2). It should be pointed out that in pollinations with S. pinnatisectum, the difference between pollen tube growth in the S-RNase KO vs the S-RNase/HT-B KO was statistically significant, but not practically significant. Pollen tube penetrance in these pollinations was highly variable, and although a few pollen tubes from S. pinnatisectum were able to reach the ovary, no fruit development was observed. Results from the same analyses demonstrate that HT-B alone is not a significant factor in the inhibition of pollen tubes from these species. No significant difference was observed between the ability of pollen tubes to penetrate the styles of wild-type S. tuberosum compared to the HT-B KO. Likewise, the differences between the S-RNase KO and the S-RNase/HT-B KO were not significant; showing that knocking out both S-RNase and HT-B compared to S-RNase alone offered no advantage to pollen tubes attempting to traverse the style. For example, pollen tubes from S. jamesii were generally arrested closer to the stigma in the HT-B KO compared to the wild type S. tuberosum ( Figure 2). These results do not contradict the findings from Tovar-Méndez et al., [20,23] where HT proteins did play a significant role in interspecific pollen rejection. Due to the vicissitudes of gene editing, KOs of HT-A were not available during this study. Since HT-A remains functional and active, with its expression in DRH195 proven by RT-PCR and cDNA sequencing in Results from the same analyses demonstrate that HT-B alone is not a significant factor in the inhibition of pollen tubes from these species. No significant difference was observed between the ability of pollen tubes to penetrate the styles of wild-type S. tuberosum compared to the HT-B KO. Likewise, the differences between the S-RNase KO and the S-RNase/HT-B KO were not significant; showing that knocking out both S-RNase and HT-B compared to S-RNase alone offered no advantage to pollen tubes attempting to traverse the style. For example, pollen tubes from S. jamesii were generally arrested closer to the stigma in the HT-B KO compared to the wild type S. tuberosum (Figure 2). These results do not contradict the findings from Tovar-Méndez et al. [20,23] where HT proteins did play a significant role in interspecific pollen rejection. Due to the vicissitudes of gene editing, KOs of HT-A were not available during this study. Since HT-A remains functional and active, with its expression in DRH195 proven by RT-PCR and cDNA sequencing in all of the female genotypes used, it can be concluded that either: (1) HT-A plays a greater role in interspecific pollen rejection compared to HT-B; (2) there is a significant overlap in gene function between HT-A and HT-B so the loss of HT-B is not significant; or (3) HT in general is not a significant component in interspecific pollen rejection in these combinations.
The contrast between S. commersonii and the other pollen donors highlights the diversity in interspecific pollen rejection systems. S. commersonii was unique as a pollen donor as both S-RNase and HT-B did not appear to affect the rejection of pollen in the styles of DRH195. As discussed previously, since HT-A is still functional, HT-A alone may mediate the S-RNase/HT-B independent rejection of S. commersonii pollen in these pollinations. If this is the case, this would be consistent with the findings of Tovar-Méndez et al. [20]. The rejection or allowance of interspecific pollen is entirely dependent on the specific combination of species used as parents [15]. This is also the case with the role of certain genes in interspecific pollen rejection. HT-A, HT-B, S-RNase, and other factors such as the 120 kDa arabinogalactan protein or farnesyl pyrophosphate synthase (FPS2) may play significant or inconsequential roles during the rejection of interspecific pollen, and these roles are entirely dependent on the specific species and genotypes present in these interactions [15,16,33].

Prezygotic Interspecific Barriers in S. verrucosum
S. verrucosum was used as a female in this experiment, with S. bulbocastanum, S. commersonii, S. jamesii, and S. pinnatisectum as pollen donors. The same pollination protocols were used for all experiments and are detailed in the materials and methods. S. verrucosum is well recognized for its SC and lack of IRBs which can partially be explained by the lack of a functional S-RNase protein [2,19,25,30,31]. Because of the lack of IRBs, the results from this experiment were considered as the benchmark for compatible interspecific pollinations with these pollen donors.
In S. verrucosum, pollen tubes from all pollen donors, except for S. bulbocastanum, were able to easily travel down the full length of the style (Figure 3). Pollen tubes from S. jamesii and S. pinnatisectum were always present in the ovary at 48 h post pollination. The majority of pollen tubes from S. bulbocastanum did not penetrate the full length of the style. Despite this difference, fruit formation was observed for all four pollen donor species and hybrid progeny between S. verrucosum and the pollen donors S. bulbocastanum and S. commersonii were recovered. Overall, no significant stylar barriers were observed and pollen tubes from all four pollen donors were observed in the vicinity of the ovules. The contrast between S. commersonii and the other pollen donors highlights the diversity in interspecific pollen rejection systems. S. commersonii was unique as a pollen donor as both S-RNase and HT-B did not appear to affect the rejection of pollen in the styles of DRH195. As discussed previously, since HT-A is still functional, HT-A alone may mediate the S-RNase/HT-B independent rejection of S. commersonii pollen in these pollinations. If this is the case, this would be consistent with the findings of Tovar-Méndez et al. [20]. The rejection or allowance of interspecific pollen is entirely dependent on the specific combination of species used as parents [15]. This is also the case with the role of certain genes in interspecific pollen rejection. HT-A, HT-B, S-RNase, and other factors such as the 120 kDa arabinogalactan protein or farnesyl pyrophosphate synthase (FPS2) may play significant or inconsequential roles during the rejection of interspecific pollen, and these roles are entirely dependent on the specific species and genotypes present in these interactions [15,16,33].

Prezygotic Interspecific Barriers in S. verrucosum
S. verrucosum was used as a female in this experiment, with S. bulbocastanum, S. commersonii, S. jamesii, and S. pinnatisectum as pollen donors. The same pollination protocols were used for all experiments and are detailed in the materials and methods. S. verrucosum is well recognized for its SC and lack of IRBs which can partially be explained by the lack of a functional S-RNase protein [2,19,25,30,31]. Because of the lack of IRBs, the results from this experiment were considered as the benchmark for compatible interspecific pollinations with these pollen donors.
In S. verrucosum, pollen tubes from all pollen donors, except for S. bulbocastanum, were able to easily travel down the full length of the style (Figure 3). Pollen tubes from S. jamesii and S. pinnatisectum were always present in the ovary at 48 h post pollination. The majority of pollen tubes from S. bulbocastanum did not penetrate the full length of the style. Despite this difference, fruit formation was observed for all four pollen donor species and hybrid progeny between S. verrucosum and the pollen donors S. bulbocastanum and S. commersonii were recovered. Overall, no significant stylar barriers were observed and pollen tubes from all four pollen donors were observed in the vicinity of the ovules.

The Role of Sli in Interspecific Pollen Rejection
Sli operates differently than the SC mechanisms in the previous experiments. Knocking out pistil expressed S-RNase and HT-B renders the GSI system non-functional, effectively removing the barrier to self-pollen [12,20,23,26]. Sli is expressed in the pollen and initiates the degradation of S-RNase to confer SC in functional GSI systems [27,28]. Because of this, the direction of the crosses in this experiment are reversed from the previous two. The wild species S. bulbocastanum, S. commersonii, S. jamesii, and S. pinnatisectum were used as females, and the two S. chacoense clones M6 and USDA8380-1 (80-1) were used as pollen donors. M6 (Sli +/+) is a homozygous, well characterized, and widely used source of Sli-based SC [32,34]. In order to reduce confounding genetic variables, clone 80-1 (Sli −/−) was selected as the non-Sli control as it is largely homozygous, and also the same species as M6 (S. chacoense) [35].
A statistically significant difference was observed between Sli +/+ and Sli −/− pollen donors in the styles of S. bulbocastanum with pollen tubes of the Sli +/+ clone M6 traveling further down the style (Figure 4). However, the same analysis failed to find any statistical significance between pollen donors in the styles of S. commersonii, S. jamesii, or S. pinnatisectum, which is inconsistent with the findings in previous research [29]. Pollen tubes from M6 generally traveled further down the style in S. bulbocastanum, S. commersonii, and S. pinnatisectum compared to pollen tubes from the SI (Sli −/−) clone 80-1 ( Figure 4). However, even the statistically significant differences between pollen donors failed to have any practical significance, and did not change the outcome of these interspecific pollinations. For example, in the styles of S. bulbocastanum pollen tube growth from 80-1 was arrested closer to the stigma, while the majority of pollen tubes from M6 penetrated much further down the style. But, since all pollen tubes from both pollen donors were arrested in the first third of the S. bulbocastanum styles, this difference lacked any practical significance. Sli presence or absence did not have a significant impact on interspecific pollen-tube inhibition in the styles of S. commersonii, S. jamesii, and S. pinnatisectum. Pollen tubes from M6 generally grew further down the style in S. pinnatisectum with the opposite being true for S. jamesii, but these differences between pollen donors lacked significance. Unlike the other species combinations analyzed, there was a substantial amount of variability in the growth of M6 pollen tubes in S. jamesii styles. It isn't clear why this is the case, but it may be attributable to unknown environmental factors. No pollen tubes from either pollen donor entered the ovary in S. jamesii, while most pollen tubes from both pollen donors traversed the entire style in S. commersonii and S. pinnatisectum. Pollinations of S. pinnatisectum by both M6 and 80-1 also initiated fruit set, although all the seeds were aborted likely due to differences in effective ploidy.
The presence or absence of Sli did not change the outcome of any of the pollinations made. Additionally, no significant difference was observed in pollinations made with S. pinnatisectum, which is inconsistent with the findings of Sanetomo et al. [29] despite the use of the same S. pinnatisectum PI (275232). The lack of consistent findings between these experiments and those of Sanetomo et al. [29] is likely attributable to some unknown factors in pollen donors besides Sli. The pollinations conducted in this study demonstrated that the clones M6 and 80-1 were comparable in their ability to bypass stylar barriers, or in their uniform inhibition. Unlike the pollen donors used by Sanetomo et al. [29] these pollen donors were of one species, S. chacoense, while the pollen donors used by Sanetomo et al. [29] came from interspecific crosses between S. chacoense and S. phureja. The difference between pollen donors is the likely source of inconsistencies; the introduction of unknown genetic variables from S. phureja may have introduced confounding genetic variables. The ability to form viable seeds cannot be assumed to be attributable to the presence or absence of Sli either, as postulated by Sanetomo et al. [29]. As observed by Hosaka and Hanneman [36,37] Sli inhibits S-RNase function in the pollen tube, thus overcoming the prezygotic GSI barrier. Inability to produce viable seed after fertilization due to endosperm failure is a post-zygotic barrier, resulting from differences in effective ploidy [38,39]. These two interspecific barriers are independent even though the outcome is the same. The presence or absence of Sli did not change the outcome of any of the pollinations made. Additionally, no significant difference was observed in pollinations made with S. pinnatisectum, which is inconsistent with the findings of Sanetomo et al. [29] despite the use of the same S. pinnatisectum PI (275232). The lack of consistent findings between these experiments and those of Sanetomo et al. [29] is likely attributable to some unknown factors in pollen donors besides Sli. The pollinations conducted in this study demonstrated that the clones M6 and 80-1 were comparable in their ability to bypass stylar barriers, or in their uniform inhibition. Unlike the pollen donors used by Sanetomo et al., [29] these pollen donors were of one species, S. chacoense, while the pollen donors used by Sanetomo et al. [29] came from interspecific crosses between S. chacoense and S. phureja. The difference between pollen donors is the likely source of inconsistencies; the introduction of unknown genetic variables from S. phureja may have introduced confounding genetic variables. The ability to form viable seeds cannot be assumed to be attributable to the presence or absence of Sli either, as postulated by Sanetomo et al. [29]. As observed by Hosaka and Hanneman, [36,37] Sli inhibits S-RNase function in the pollen tube, thus overcoming the prezygotic GSI barrier. Inability to produce viable seed after fertilization due to endosperm failure is a post-zygotic barrier, resulting from differences in effective ploidy [38,39]. These two interspecific barriers are independent even though the outcome is the same.

Discussion
The relationships of interspecific compatibility and incompatibility in Solanum section Petota are complex. Inquiry into these relationships has elucidated the pleiotropic and redundant function of S-RNase and HT which tandemly and independently mediate both interspecific and intraspecific pollen rejection [15,20,23]. The outcome of interspecific pollinations is specific to the individual species involved and direction of the cross due to the direct and indirect involvement of multiple factors such as HT and S-RNase, 120 kDa, Cul-lin1 (CUL1), and farnesyl pyrophosphate synthase (FPS2) [16,20,40,41].
The findings presented here are consistent with previous work conducted in Solanum section Lycopersicon showing that S-RNase plays a central role in interspecific pollen rejection [15,23]. Statistical analyses also demonstrated that HT-B alone is not a significant factor in these pollinations. These results do not contradict the previous findings from tomato, where HT proteins played a significant role in interspecific pollen rejection [20,23].

Discussion
The relationships of interspecific compatibility and incompatibility in Solanum section Petota are complex. Inquiry into these relationships has elucidated the pleiotropic and redundant function of S-RNase and HT which tandemly and independently mediate both interspecific and intraspecific pollen rejection [15,20,23]. The outcome of interspecific pollinations is specific to the individual species involved and direction of the cross due to the direct and indirect involvement of multiple factors such as HT and S-RNase, 120 kDa, Cullin1 (CUL1), and farnesyl pyrophosphate synthase (FPS2) [16,20,40,41].
The findings presented here are consistent with previous work conducted in Solanum section Lycopersicon showing that S-RNase plays a central role in interspecific pollen rejection [15,23]. Statistical analyses also demonstrated that HT-B alone is not a significant factor in these pollinations. These results do not contradict the previous findings from tomato, where HT proteins played a significant role in interspecific pollen rejection [20,23]. Since HT-A remained functional in all of the female genotypes used, these findings demonstrate the overlap in gene function between HT-A and HT-B, the greater importance of HT-A, or that HT is not a significant factor in these combinations, implicating other mechanisms [15,33].
In comparison to the other species used as pollen donors, S. commersonii was unique as the absence of S-RNase did not appear to affect the rejection of pollen in S. tuberosum. Since the rejection or acceptance of interspecific pollen is entirely dependent on the specific combination of species used as parents, HT, S-RNase, and other factors may play significant or inconsequential roles during the rejection of interspecific pollen [15,16]. In this instance HT-A alone may mediate the S-RNase independent rejection of S. commersonii pollen in DRH195.124-001, where both HT-B and S-RNase are non-functional. If this is the case, this would be consistent with the findings of Tovar-Méndez et al. [20].
Prezygotic stylar barriers were not observable in these pollinations using S. verrucosum as the female parent. The lack of prezygotic stylar barriers in S. verrucosum may be comparable to S. lycopersicum [15]. This lack of IRBs is a very valuable phenotype which allows S. lycopersicum to not only be SC, thus allowing the development of inbreds and F1 hybrids, but also serves as a parent for wide interspecific crosses allowing the introgression of valuable traits from genetically distant wild relatives [15,42,43]. The lack of SC and interspecific incompatibility in diploid S. tuberosum has been a significant barrier to the effective use of wild species, and the ability to replicate the phenotype observed in S. verrucosum and S. lycopersicum within S. tuberosum would be a significant advancement in potato breeding [7,12]. Understanding the causative genetic mechanisms behind SC and interspecific compatibility in S. verrucosum is necessary to replicate this phenotype. Additionally, stylar barriers are not the only significant factors barring the use of wild potato species in potato breeding. Therefore, further inquiry into the genetic mechanisms underlying effective ploidy and endosperm balance number will be necessary in the future.
Although Sli played no significant role in these pollinations, it is possible that S. chacoense as a pollen donor is better able to bypass stylar barriers in EBN1 species such as S. pinnatisectum compared to S. megistacrolobum, S. demissum, and S. phureja (S. tuberosum group Phureja), as reported by Sanetomo et al. [29]. Consequently, S. chacoense may be a valuable resource in accessing these EBN1 species regardless of Sli status. Currently, other methods such as bridge crossing with S. verrucosum are more reliable and feasible in their ability to access 1EBN species [2]. Unlike the methods proposed by Sanetomo et al. [29] and Carputo et al. [44], the use of S. verrucosum bridge crosses is also more accessible for most plant breeders, as it does not require ploidy manipulation, interploidy crosses, or the use of advanced techniques [2].

Plant Material
The wild species S. bulbocastanum, S. commersonii, S. jamesii, and S. pinnatisectum were used as both pollen donors and females for this study. Multiple plant introduction populations for each species were acquired from the USDA-ARS Potato Germplasm Introduction Station (Sturgeon Bay, WI, USA), and individual genotypes were selected from their respective field-grown plant introduction populations based on their fertility and overall vigor (Table 1). In this study, S-RNase and/or HT-B mediated interspecific pollen rejection was investigated using CRISPR-Cas9 gene knockouts of S. tuberosum clone DRH195 and untransformed controls (Table 1) [12,26]. Interspecific compatibility in S. verrucosum was investigated using the clone SV607845.02. This S. verrucosum clone was selected as a female based on its previous ability to create interspecific hybrid offspring with S. bulbocastanum. The S. chacoense clone USDA8380-1 and the S. chacoense self-compatible inbred line M6 were selected as pollen donors to evaluate Sli [34]. The M6 clone was selected as it is a homozygous, well characterized, and widely used source of Sli-based SC [32,34]. In order to reduce confounding genetic variables, clone USDA8380-1 was selected as the non-Sli control as it is largely homozygous, and also the same species as M6 [35]. Each genotype was maintained in tissue culture on Murashige and Skoog Basal Medium with Vitamins and Sucrose (PhytoTech Labs, Lenexa, KS, USA) and Phyto Agar (Research Products International, Mount Prospect, IL, USA) prepared with DI water and balanced with 1M HCl and 8N NaOH to a pH of 5.8 and cultured in growth chambers with 16-h-light/8-h-dark photoperiod at 22 • C and average light intensity of 200 µmoles m −2 s −1 .

Greenhouse Pollination Assays
In October of 2020, 2 individuals from each genotype were transferred directly from tissue culture to 14 L (3.8 gallon) plastic pots filled with Suremix Perlite peat and perlite soilless medium (Michigan Grower Products Inc., Galesburg, MI, USA). For the duration of the experiments greenhouse conditions were maintained at 20 • C with a 16-h photoperiod under Philips GreenPower light-emitting diode (LED) DR/W-MB lights (Philips Lighting Holding B.V., Eindhoven, The Netherlands). To validate the male fertility of the pollen donors, pollen was collected from 5-6 anthers directly onto a glass slide and immediately stained with acetocarmine-glycerol as described by Ordoñez [45], covered with a cover slip, and sealed using clear nail polish. These slides were stored at room temperature in the dark and visualized the same day using a Leica DM750 binocular microscope (Leica Microsystems, Wetzlar, Germany) and the associated Leica imaging software at 10× and 40× magnification. A minimum of 100 pollen grains were then used to calculate the percentage of viable pollen. Stained, turgid pollen was classified as viable, while any grains that were shriveled, unstained, or unusually large or small were classified as unviable.

Stylar Squash Assays
For the evaluation of stylar barriers in interspecific crosses, 6-8 newly opened flowers (within 24 h after anthesis) from each of the female genotypes were carefully emasculated and pollinated with fresh pollen collected on a glass slide directly from the pollen donor. Styles were collected 48 h post-pollination, by removing the remaining petals and sepals and storing the remaining intact style, ovary, and receptacle in a 1.5 mL microcentrifuge tube containing a 3:1 ethanol/acetic acid fixation solution. These styles were then kept in the dark at room temperature for at least 24 h. Styles and ovaries were then softened using an 8N NaOH solution at 60 • C for 1 h. Samples were then rinsed 3 times with distilled water and stained with 0.1% aniline blue in 0.1N K 3 PO 4 keeping them in dark conditions with light shaking. Styles with attached ovaries were then placed on glass slides, gently squashed under a coverslip, and sealed with nail polish for subsequent visualization. Samples were visualized using a Nikon Eclipse Ni-U upright microscope (Nikon Instruments Inc., Melville, NY, USA) with a SOLA light engine (Lumencor, Beaverton, OR, USA), and photographed with the attached ANDOR Zyla sCMOS camera (Oxford Instruments, Abingdon, UK) and NIS-Elements BR 5.02 software. Images of each stylar sample were stitched together using Image Composite Editor 2.0 software. Pollen tube growth measurements were made using ImageJ 1.53e software [46]. Measurements of pollen tube growth were calculated as a proportion of the total length of the style due to the variation of total style length with a single genotype. Pollinations were made at 3 discrete time points, separated by 30 days, with the first time point being approximately 6 weeks after plants were transferred from tissue culture to the greenhouse.

Data Collection and Analysis
Measurements of the pollen tube front, or the point where the majority of the pollen tubes stop, the longest pollen tube, and the total length of the style were collected using ImageJ software ( Figure 5). From these measurements, distance from the surface of the stigma to the pollen tube front and length of the longest pollen tube were calculated based on the total length of the style in which they were measured (Supplemental File S1). Significant differences between means (α = 0.05) from three replicated measurements were calculated using analysis of variance (ANOVA) and Tukey's honestly significant difference (HSD) using "R" software version 4.0.4.