Rejection of S-heteroallelic pollen by a dual-specific s-RNase in Solanum chacoense predicts a multimeric SI pollen component.

S-heteroallelic pollen (HAP) grains are usually diploid and contain two different S-alleles. Curiously, HAP produced by tetraploids derived from self-incompatible diploids are typically self-compatible. The two different hypotheses previously advanced to explain the compatibility of HAP are the lack of pollen-S expression and the "competition effect" between two pollen-S gene products expressed in a single pollen grain. To distinguish between these two possibilities, we used a previously described dual-specific S(11/13)-RNase, termed HVapb-RNase, which can reject two phenotypically distinct pollen (P(11) and P(13)). Since the HVapb-RNase does not distinguish between the two pollen types (it recognizes both), P(11)P(13) HAP should be incompatible with the HVapb-RNase in spite of the competition effect. We show here that P(11)P(13) HAP is accepted by S(11)S(13) styles, but is rejected by the S(11/13)-RNase, which demonstrates that the pollen-S genes must be expressed in HAP. A model involving tetrameric pollen-S is proposed to explain both the compatibility of P(11)P(13) HAP on S(11)S(13)-containing styles and the incompatibility of P(11)P(13) HAP on styles containing the HVapb-RNase.