Self-Incompatibility in S-RNase-Based Systems: Are there Differences in the Pollen Rejection Mechanisms Among Species?

SI is a selective mechanism that discriminates between selfand non-self-pollen, preventing self-fertilization by rejecting self-pollen (incompatible cross). Pollen rejection occurs as the result of a complex pollen-pistil interaction network, that is, as the starting point, genetically regulated by the single multiallelic S locus [5,6], except for some grasses (Poaceae), in which two loci (S and Z) control self-pollen rejection [7,8].

The SSI is a more restrictive pollen rejection mechanism because pollen will carry on its surface two S-proteins synthesized by the anther tapetum. For example, an S 1 S 2 plant will deposit the S 1 and S 2 -proteins on the pollen grain exine. If this pollen lands on an S 1 S 2 stigma, it will be rejected. Even more, if pollen lands on S 1 S 3 or S 2 S 4 stigmas, it will also be rejected, but no for S 3 S 4 stigmas. By contrast, GSI is slightly more flexible. For instance, in a plant S 1 S 2 , 50% of the pollen grains will display S 1 haplotype and 50% S 2 . The pistil will reject either pollen grains ( Figure 1). However, pollen may come from a close-related individual. For example, in an S 1 S 2 X S 1 S 3 cross, only S 1 pollen will be rejected but not S 3 (Figure 1), which does not occur in SSI (where pollen would display an S 1 S 3 phenotype). Therefore, while in SSI, there are only compatible and incompatible crosses, semi-compatible crosses may occur in GSI.

Figure 1:
S-specific pollen rejection response in the gametophytic self-incompatibility system (Nicotiana alata). An S 1 S 2 plant expresses both S alleles in the pistil and produces pollen with S 1 and S 2 haplotypes (50% each). When the pollen grains (asterisks) land on the stigmatic surface, they germinate, and the pollen tubes (arrows) start to grow through the extracellular matrix of the stylar transmitting tissue. The pistil recognizes and rejects pollen with either S 1 or S 2 haplotype (incompatible; left side) at the upper part of the style, preventing homozygotes in the progeny. By contrast, if the pollen portrays a different haplotype (e. g. S 3 ), it will be allowed to grow and reach the ovary to exert fertilization (compatible; right side), ensuring heterozygotes in the progeny. For these images, N. alata pistils were hand-pollinated with incompatible or compatible pollen. After 36h, pistils were squashed in a blue aniline solution to stain the pollen tube wall and observed under fluorescence microscopy.
SSI has been deeply studied in Brassicaceae. The research on this has identified both S determinants, how they interact, and determined the resultant signaling cascade triggered in an incompatible cross that prevents pollen germination on the stigmatic surface [rev. in 5,10,12,13]. On the other hand, two types of GSI control have been described: the Papaver SI system, only known and studied in Papaver rhoeas, and the S-RNase-based SI system, investigated in Plantaginaceae, Rosaceae, Rubiaceae, Rutaceae, and Solanaceae [9].

GSI S determinants and other participants
In P. rhoeas, the female S determinant encodes a stigma-specific secreted small protein called PrsS [14], which presumably interacts with the pollen S determinant, PrpS, a putative membrane protein specifically expressed in pollen [15]. In an incompatible cross, the S-specific interaction between the specificity determinants leads to a rapid increase of intracellular calcium concentration that triggers a series of biochemical and cellular changes, including cytoskeleton rearrangements [16], that end in the pollen programmed cell death
All the hallmarks of pollen rejection response in P. rhoeas are observed when PrsS and PrpS are heterologous expressed in Arabidopsis thaliana [18,19]. It suggests that the Papaver SI system uses common proteins for the signal cascade and that the only SIspecific proteins are the S-determinants themselves.
The S-RNase-based SI system has been studied in Solanaceae, particularly in Nicotiana and Petunia. Both self-(incompatible) and non-self (compatible) pollen can germinate on the stigmatic surface ( Figure 1), and the PTs begin their journey towards the ovary through the extracellular matrix (EM) of the stylar transmitting tissue (STT). However, when the pollen matches the S-haplotype with one of the two S alleles in the pistil, the PT growth is inhibited within the first third of the style, preventing self-fertilization ( Figure 1).
The female S-determinant is a glycoprotein with ribonuclease activity, called S-RNase [20,21]. The S-RNase is specifically expressed in STT cells with a final localization onto the EM [20]. During PT growth through the EM of the STT, S-RNases are taken up by both self-and non-self-PTs [22,23]. S-RNases function as cytotoxins degrading the PT RNA in an incompatible cross [24].
The S-pollen determinant comprises a suite of proteins called SLF (S-Locus F-box proteins), and pollen grain expressed [25][26][27][28]. SLF proteins contain an F-box domain at their N-termini and work as E3 ligase proteins within the SCF (Skp1-Cullin-F-box) complex, which is involved in ubiquitylation of target proteins, leading to their degradation through the proteasome 26S [29].
The pollen rejection specificity relies on S-specific interaction between the S-RNase and some of the SLF proteins [30]. However, genetic, biochemical, and cellular evidence indicate that other genes unlinked to the S locus, known as modifier genes (MGs), are required to reject the pollen in an S-specific manner successfully. So far, nine MGs have been identified. Four of them are pistil-expressed genes: HT-B, 120K, NaStEP, and NaTrxh [31][32][33][34]; and five from the pollen side: PhUBC1, SSK1, MdABCF, PiCUL1-G, and NaSIPP [29,[35][36][37][38]. Their function in pollen rejection was proved by gain and loss of function assays in transgenic plants.

GSI S-RNase-based SI, an intricated system
The S-RNase-based SI system appears to be quite complicated to dissect. First, it has not been possible to establish a clear in vitro or semi-in vitro assay, like in P. rhoeas, that allow elucidating in a clear and more accessible manner the complex interactions during SI response in different families such as Solanaceae and Rosaceae.
The first identified MG, HT-B, encodes a small asparaginerich pistil-expressed protein that, when suppressed, the SI phenotype is lost [31], which means that it is directly involved in S-specific pollen rejection. Although its biochemical role is unknown, it is incorporated into the PT, whether it is a compatible or an incompatible cross [23]. However, its stability within PT depends on the type of cross. While in an incompatible PT, HT-B is degraded, it remains stable when it is a compatible one [23,33], in agreement with the results obtained by genetic suppression. If HT-B is hydrolyzed, the PT will reach the ovary. A similar pattern is observed with the arabino galactoprotein 120K, encoded by other pistil MG [32], but in this case, 120K is stabilized in compatible PTs and disappears in incompatible ones [23].
The degradation/stabilization of HT-B and 120K indicates that proteolytic activity is essential during SI. Likewise, NaStEP, another pistil-expressed MG, encodes a proteinase inhibitor [39,40]. When NaStEP is suppressed in Nicotiana transgenic plants, the ability to reject self-pollen is abolished. In the absence of NaStEP, the HT-B protein is degraded in PT of both compatible and incompatible crosses, suggesting that NaStEP positively regulates HT-B stability in PT [33], inhibiting probably to a putative pollen protease [33,40]. However, direct evidence of this is lacking.
NaTrxh encodes a thioredoxin (Trx) type h with expression in several plant organs, including the pistil. This Trx localizes onto the EM of the STT [41]. NaTrxh interacts and reduces S-RNase in vitro and it has been proved that this reduction results in a seven-fold increase on its ribonuclease activity [34,41,42]. When Nicotiana transgenic plants co-express a non-functional NaTrxh variant with its active site mutated, the SI response is disrupted, indicating that its redox activity is essential to pollen rejection in an S-specific manner [34].

Models pretending to explain the molecular basis of pollen rejection
The collaborative and compartmentalization models pretend to explain the molecular basis of pollen rejection response in S-RNase-based systems. The collaborative one states that the pollen rejection response only depends on the S-nonspecific interaction between SLF and S-RNase in PT, which leads to the ubiquitylation and degradation of non-self-S-RNase in a compatible cross [30,43]. Here, a suite of 16 to 20 SLF proteins collaborate to identified nonself-S-RNases, which will be ubiquitylated and degraded through the 26S proteasome pathway. In contrast, in an incompatible cross, self-S-RNase will not be recognized by any of the collection of SLFs in the PT and will escape proteasome degradation, remaining active to hydrolyze the PT RNA and, therefore, inhibiting PT growth towards the ovary [30]. Under this model, no MG is considered at any step of the pollen rejection leading cascade.
In contrast, the compartmentalization model states that S-RNases are sequestered within a vacuole when incorporated into the PT cytosol. If the cross is compatible, this vacuole will remain intact, preventing the S-RNase cytotoxic activity, and no RNA degradation will occur. By contrast, when the cross is incompatible, the vacuole will be broken down, releasing S-RNase into the cytosol, where they will degrade RNA [23]. The breaking down of the containing S-RNases vacuole will depend on the S-specific interaction between SLF and S-RNase.

MCDA.000684. 8(2).2021
Moreover, the pistil proteins HT-B and NaStEP and the pollen protein NaSIPP play a crucial role in the biochemical mechanism that results in self-pollen recognition [31,33,38]. Functional assays provide evidence that HT-B is an essential protein in the breakdown of the S-RNase-containing vacuole because, in its absence, the integrity of this vacuole persists [23]. Besides, the stability of HT-B in incompatible PTs depends on the presence of NaStEP, likely inhibiting the protease activity that degrades HT-B in compatible PT through its proteinase inhibitor activity [33,40].
NaStEP seems to be a multifunctional protein because a fraction of it interacts with NaSIPP, a mitochondrial phosphate carrier. When NaSIPP is suppressed in transgenic Nicotiana, the S-specific pollen recognition is disrupted [38]. Furthermore, NaStEP also has activity as a voltage-dependent channel blocker [40]. Taken together, we propose that PCD might be involved in the pollen rejection response in Nicotiana. We hypothesize that the PCD mechanism in an incompatible cross is initiated by the S-specific interaction between SLF and S-RNase, through NaStEP, which with its inhibitor proteinase function, will inhibit the protease that hydrolyzes HT-B [33,40], a necessary step to breakdown the S-RNase containing vacuole. Likewise, another fraction of NaStEP will impair the mitochondria through its interaction with NaSIPP [38] and some voltage-dependent channel [40], which will trigger the PCD that concludes with the disruption of the S-RNase-containing vacuole, probably through HT-B participation. Finally, once in the cytosol, S-RNases will be further activated by the reduction of one of their disulfide bridges by NaTrxh [34,44]. While the collaborative model evidence has mainly raised from research on Petunia, the compartmentalization model evidence is from Nicotiana. Both species belong to the Solanaceae family but, according to several phylogenetic analysis, are grouped in two different subfamilies [45,46].

Conclusion
Both collaborative and compartmentalization models seem to be excluding each other. While the collaborative model proposes that the S-nonspecific interaction between SLF and S-RNase protects compatible pollen; the compartmentalization model claims that the S-specific interaction between the specificity determinants, which triggers self-pollen rejection through a signal cascade downstream this interaction where the MG products are involved.
It calls attention that Petunia and Nicotiana SI mechanisms be so different because both belong to different subfamilies within Solanaceae [45,46]. One possible scenario is that, indeed, it does. Still, another option is that the collaborative model is limited to the S-RNase degradation because S-RNase compartmentalization and MG function, such as HT-B, NaStEP, NaSIPP, and NaTrxh, have not yet been evaluated in Petunia. Therefore, the question remains open, and more forth research in this direction is needed.