Disabling of ARC1 through CRISPR–Cas9 leads to a complete breakdown of self-incompatibility responses in Brassica napus

Self-incompatibility (SI) is a genetic mechanism utilized by many flowering plants to promote outcrossing by recognizing and rejecting self-pollen. While the molecular mechanisms controlling SI can differ in various families, plants in the Brassicaceae family have been found to regulate SI through a S-haplotype-specific receptor-ligand interaction. The recognition of self-pollen leads to the activation of ARMADILLO-REPEAT-CONTAINING1 (ARC1), an E3 ligase which is thought to degrade the compatibility factors required for pollen germination, thus resulting in SI. However, the role of ARC1 in SI in the Brassicaceae family has been disputed within the scientific community. While various studies demonstrate that the manipulation of ARC1 and its downstream targets can break-down the SI response, others show that the SI phenotype can be imparted to self-compatible cultivars without the ARC1 gene. This study investigated the involvement of ARC1 in SI through the generation of loss-of-function ARC1 mutants using CRISPR-Cas9 in B. napus. Loss of ARC1 resulted in complete breakdown of SI in two different S-haplotypes providing strong evidence for the essential nature of ARC1 for mounting a successful SI response.

Despite the partial breakdown of SI shown in two different systems, the role of ARC1 in the Brassicaceae SI response has been questioned for the past several decades. Stable coexpression of Brassica SLG, SRK, and ARC1 was insufficient to confer the SI phenotype in compatible Arabidopsis thaliana (Bi et al., 2000). In another report, when A. thaliana plants were transformed to express SRK b and SCR b genes, a strong SI phenotype was observed in the absence of ARC1 (Kitashiba et al., 2011). A. thaliana plants that exhibit the SI phenotype through transformation of SRK and SCR alone show trends in heritability, developmental regulation, and intensity of SI response similar to those of self-incompatible Brassica cultivars (Nasrallah and Nasrallah, 2014). Although, evolutionarily, Arabidopsis that lacked a bona fide ARC1 ortholog could have re-purposed other E3 ligases to assume the role of ARC1, the question of how complete loss of ARC1 could influence SI in Brassica sp. remained unresolved.
To unequivocally examine the role of ARC1 during SI, we created ARC1 loss-of-function B. napus using the CRISPR-Cas9 platform. In order to efficiently design specific targets for the CRISPR constructs, we retrieved available gene copies of ARC1 or similar genes present in B. napus (allotetraploid from the A and C genomes of B. rapa and B. oleracea) (http://cbi. hzau.edu.cn/bnapus/index.php) (Song et al., 2020). After validation of the sequences across the published NCBI database, TOPO cloning was performed with BnARC1-specific cDNA amplification products obtained from an RNA pool of fully mature B. napus stigmas. Sanger sequencing of the clones revealed that BnARC1 is a single-copy gene derived from Brassica rapa (A genome of rapa), whereas its homolog from the C genome (oleracea) was identified at low frequency and was most similar to PUB17 family genes ( Figure 1A). When the expression profiles of these two genes were assessed at various stages of stigma maturity, ARC1 had significantly higher expression than PUB17 and peaked at stigma maturity ( Figure 1B).
The Westar cultivar is self-compatible due to an insertion in the promoter region of SP11, leading to a lack of expression of SP11, the pollen ligand required for SI response (Okamoto et al., 2007). However, it retains all downstream SI signaling components (SRK and ARC1), as it readily rejects pollen of S47 haplotype origin with a functional SP11 (Okamoto et al., 2007). The T2 generations of two independently edited wes-arc1 plants with biallelic edits were used for pollination assays with compatible Westar or S8 pollen and incompatible S47 pollen. When B. rapa S47 haplotype pollen was applied to control Westar stigmas, a robust SI was observed in the Westar stigmas, which readily rejected pollen of the S47 haplotype ( Figure 1D-1F). By contrast, the wes-arc1-1 and wes-arc1-2 plants showed a complete breakdown of the SI response when pollinated with S47 pollen ( Figure 1D-1F). When either Westar pollen or B. rapa pollen from an S8 haplotype (SRK8 is absent in Westar) was used as a positive control for stigma receptivity, full acceptance was observed in both Westar controls and wes-arc1 plants, indicating that stigma receptivity was not modified in the edited line ( Figure 1D).
We next sought to test whether nullifying ARC1 in the well characterized self-incompatible W1 background would result in (legend continued on next page) complete breakdown of SI. The isogenic, self-incompatible W1 line was created by introgression of the dominant 910 B. rapa haplotype (SRK 910/SP11-910); it displays a very strong SI phenotype when self-pollinated but is compatible when crossed with Westar. We predicted that regardless of the upstream receptor/ligand complex, abolishing ARC1 function should lead to complete breakdown of SI.
Because W1 plants are incompatible and difficult to propagate/ regenerate through conventional Agrobacterium-based tissue culture approaches, W1 plants were crossed with the wes-arc1 mutant transgenic lines harboring the ARC1 editing system. In the F1 generation, we identified a W1-arc1-1 line that displayed strong breakdown of SI when self-pollinated. Sequencing of ARC1 revealed edits at both the T 3/4 sites ( Figure 1G, only T 3/4 shown). When pollen from W1-arc1-1 was tested on W1 stigmas, W1 stigmas rejected W1-arc1-1 pollen, indicating that the SP11-910 haplotype was not altered or deleted. On the other hand, ARC1-edited W1-arc1-1 plants readily accepted pollen from self-incompatible W1 plants ( Figure 1H). This complete breakdown resulted in full seed set ( Figure 1I) comparable to that observed when stigmas were pollinated with compatible Westar pollen ( Figure 1J and 1K). This compromise in SI was further confirmed in at least four successive generations of W1-arc1-1 plants that harbored an intact SRK910. Flowers from these lines were also used to pollinate W1 stigmas and confirm the SI reaction, demonstrating that the S-haplotype was unmodified in these lines.
We have convincingly demonstrated that elimination of ARC1 results in complete breakdown of SI in two different S-haplotypes (S47 and SRK910), confirming the essential nature of ARC1 for SI response in Brassica. In both situations, the closely related PUB17 ortholog was unaltered, indicating the exclusive nature of ARC1 for mediating the SI response. Although this study eliminates any doubt as to whether ARC1 is required for SI in Brassica, the fact that ARC1 was shown to be dispensable in certain cases in A. thaliana suggests that there could be an evolutionary significance to this observation. In a self-incompatible A. lyrata ecotype, ARC1 is present and required for SI response (Indriolo et al., 2012;, whereas in selfcompatible Arabidopsis species, ARC1 is often found to have been deleted (Indriolo et al., 2012  (J and K) Bar graphs representing the average pod length (J) and seed set (K) in W1-arc1-1 pollinated with W1 or self-pollinated compared with selfpollinated Westar. Error bars represent ±SEM.