Discussion
Expanding the genetic basis of sorghum to cope with parasitism byStriga , one of the biggest constraints to cereal production in SSA, could have far-reaching impact towards alleviating food insecurity. An efficient way to achieve Striga resistance is to harness advances in genomics and exploit the vast genetic diversity of sorghum. Building on previous successful identification of Strigaresistant sorghum from global collections (Kavuluko et al. 2021; Mallu et al. 2021, 2022), we sought to identify and characterize new sources of resistance. Our motivation for selecting the SAP collection was obtaining sorghum with increased resistance toStriga parasitism combined with additional traits that accrue benefits to smallholder farmers in Africa. The SAP is also extensively sequenced providing vital data for downstream resistance breeding. Based on our analysis and previous work (Boatwright et al. 2022), the SAP comprises all major botanical races of sorghum (caudatum, kafir, guinea, and durra) and a mixed population believed to have originated from bicolor – one of the earliest races to undergo domestication (Harlan & Stemler, 1976).
We adopted a simple methodology for discovery of Strigaresistance alleles in sorghum based on PCR screening using two sets of primers – making it applicable in basic laboratories without the need for sophisticated equipment and resources for advanced molecular screening. Our approach hinged on identifying sorghum genotypes harboring a chromosomal deletion in the sorghum LOW GERMINATION LOCI 1 (LGS1 ) region which makes some genotypes inefficient in stimulating the germination of the parasitic plant Striga (Gobenaet al. 2017). Mutants devoid of this region, termed lgs1genotypes are Striga resistant because they primarily exude the low potent SL, orobanchol, compared to their susceptible wild type counterparts (LGS1 genotypes) that exude 5-deoxystrigol which is more potent. We found 12 SAP lgs1 accessions; among them two had been previously described (Bellis et al . 2020). This represented 3.5 % of our screened population and importantly, 67 % of thelgs1 -like genotypes originated from Africa. This finding is important and significant because it points to increased prevalence ofStriga resistance in sorghum domesticated in Africa, which can be attributed to its adaptation to the parasite as both Striga and sorghum have their natural distribution ranges in SSA.
SAP lgs1 accessions bore the expected phenotype of the referencelgs1-like sorghum SRN39 whose hallmark is the production of high proportions of the SL orobanchol relative to 5-deoxystrigol (Gobenaet al. 2017). Germination assays further affirmed the expected low germination stimulation of lgs1 -like sorghum genotypes. Our results were consistent with previous work that have reported germination efficiencies of 38 % in lgs1 -like genotypes (Malluet al. 2021). Unexpectedly, PI656054, an LGS1 accession, showed notably low germination induction. Our hypothesis of involvement of ABA in mediating the low germination of this genotype was inconclusive as we did not find any correlation between germination and ABA content. One possibility is the production of another less potent SL that was not tested in this in this study. Determination of the actual mechanisms will be an interesting subject of further investigation.Striga seed germination stimulation assayed by agar gel method that measures MGD, the maximum germination distance which host root can stimulate Striga germination in vitro agar culture showed concurrence of our results with previous reports. For example, Gobenaet al . (2017) found the MGD of SRN39 to be 1mm. In pearl millet, the Striga resistant 29AW had a MGD of 7.96 mm (Dayou et al. 2021).
To further validate the resistance of SAP lgs1 accessions, we performed pot experiments where sorghum was planted inStriga- infested soil and the number of emerging Strigacounted. Our results showed that most SAP lgs1 had low emergence. We would like to point out PI655979, PI656040, and PI533576, that consistently showed low germination stimulation in the bioassay and low emergence in pot experiments. We would also like to point out PI 656054, the SAP LGS1 that we described earlier as having lowStriga germination induction. Because Striga emergence in pot experiments is a function of both pre- and post-attachment resistance, all these genotypes represent good candidates for further field evaluation.
The final aspect of our study was to evaluate SAP lgs1 lines under field infestations. Consistently, SAP lgs1 lines that showed low germination stimulation and emergence in the laboratory bioassays endured low Striga infestation under field conditions. The genotypes PI533976, PI656094 and PI655979, had Strigaemergence numbers comparable to the resistant check SRN39. A closer look at the number of Striga emergence in SAP lgs1 showed concurrence with other Striga resistant sorghum. For example, the number of emerging Striga in Framida and IS9830 were like those of SAP lgs1 PI533976, PI656094 and PI655979. Serendipitously discovering PI656054, an LGS1 sorghum with pre-attachment resistance that appears to be independent of the SL signaling pathway, opens new avenues for Striga resistance studies. Once validated phenotypically, the accession can be used as donors to incorporate new diversity into breeding lines.
Another important Striga management strategy is tolerance – ability of a crop to produce yield even under Striga infestation (Mwangangi et al. 2021). This trait is well exemplified by N13, a popular Striga resistant and tolerant durra sorghum from eastern Africa (Rodenburg, Bastiaans, Weltzien & Hess 2005). Identification of potentially tolerant SAP lgs1 has important implications because researchers are now advocating for a combination of resistance and tolerance in an integrated Striga management approach. We found that most SAP lines could produce yields even under reasonableStriga infestation. Particularly, SAP lgs1 accessions: PI656094 and 656096 are good candidates for deployment as tolerant accessions. Although the mechanism of Striga tolerance has not been fully explored, one can extrapolate a hypothesis based on the adverse effects of Striga parasitism on growth retardation and yield loss. ABA is a critical hormone for controlling plant responses to water limitation, inducing stomatal closure to limit water loss through transpiration (Mittelheuser & van Steveninck 1969). Striga are insensitive to ABA, resulting in higher transpiration rates than their hosts and the maintenance of a water potential gradient favoring nutrient and water transfer to parasites (Fujioka et al. 2019b a). One possibility is that host insensitivity to ABA could be extended to imply Striga tolerance if host plants control the water potential gradient under Striga parasitism via ABA signaling. Because there is no validation data for this hypothesis, for now it remains a subject for further investigation.
Field evaluation provided a further opportunity to study flowering time and yields of SAP lgs1 lines. Regarding flowering time, all the accessions were within the time considered as early maturity in sorghum. This is important because early maturity helps cope with Strigaby reducing infestation. This is also important as a drought coping mechanism, another major constraint for sorghum production in SSA.
To conclude, we: (i) describe a simple Striga resistance allele mining assay that could be adapted for allele discovery in other sorghum collections or populations, particularly in Africa where sorghum is prone to be enriched for Striga resistance, (ii) identify 12 newStriga resistant SAP lgs1 accessions that can be integrated in breeding programs in SSA, and (iii) demonstrate that identified sorghum accessions provide the additional advantage of early maturity and tolerance to Striga parasitism.