SUMMARYIn plants, double fertilization requires successful sperm cell delivery into the female gametophyte followed by migration, recognition and fusion of the two sperm cells with two female gametes. We isolated a null allele (lre-5) of LORELEI, which encodes a putative glycosylphosphatidylinositol (GPI)-anchored protein implicated in reception of the pollen tube by the female gametophyte. Although most lre-5 female gametophytes do not allow pollen tube reception, in those that do, early seed development is delayed. A fraction of lre-5/lre-5 seeds underwent abortion due to defect(s) in the female gametophyte. The aborted seeds contained endosperm but no zygote/embryo, reminiscent of autonomous endosperm development in the pollen tube reception mutants scylla and sirene. However, unpollinated lre-5/lre-5 ovules did not initiate autonomous endosperm development and endosperm development in aborted seeds began after central cell fertilization. Thus, the egg cell probably remained unfertilized in aborted lre-5/lre-5 seeds. The lre-5/lre-5 ovules that remain undeveloped due to defective pollen tube reception did not induce synergid degeneration and repulsion of supernumerary pollen tubes. In ovules, LORELEI is expressed during pollen tube reception, double fertilization and early seed development. Null mutants of LORELEI-like-GPI-anchored protein 1 (LLG1), the closest relative of LORELEI among three Arabidopsis LLG genes, are fully fertile and did not enhance reproductive defects in lre-5/lre-5 pistils, suggesting that LLG1 function is not redundant with that of LORELEI in the female gametophyte. Our results show that, besides pollen tube reception, LORELEI also functions during double fertilization and early seed development.

No abstract

SUMMARY Polarized cell elongation is triggered by small molecule cues during development of diverse organisms. During plant reproduction, pollen interactions with the stigma result in the polar outgrowth of a pollen tube, which delivers sperm cells to the female gametophyte to effect double fertilization. In many plants, pistils stimulate pollen germination. However, in Arabidopsis, the effect of pistils on pollen germination and the pistil factors that stimulate pollen germination remain poorly characterized. Here, we demonstrate that stigma, style, and ovules in Arabidopsis pistils stimulate pollen germination. We isolated an Arabidopsis pistil extract fraction that stimulates Arabidopsis pollen germination, and employed ultrahigh resolution ESI FT-ICR and MS/MS techniques to accurately determine the mass (202.126 daltons) of a compound that is specifically present in this pistil extract fraction. Using the molecular formula (C10H19NOS) and tandem mass spectral fragmentation patterns of the m/z (mass to charge ratio) 202.126 ion, we postulated chemical structures, devised protocols, synthesized N-Methanesulfinyl 1- and 2-azadecalins that are close structural mimics of the m/z 202.126 ion, and showed that they are sufficient to stimulate Arabidopsis pollen germination in vitro (30 µM stimulated ~50% germination) and elicit accession-specific response. Although N-Methanesulfinyl 2-azadecalin stimulated pollen germination in three species of Lineage I of Brassicaceae, it did not induce a germination response in Sisymbrium irio (Lineage II of Brassicaceae) and tobacco, indicating that activity of the compound is not random. Our results show that Arabidopsis pistils promote germination by producing azadecalin-like molecules to ensure rapid fertilization by the appropriate pollen.

We present a general model for tracking smooth trajectories of multiple targets in complex data sets, where tracks potentially cross each other many times. As the number of overlapping trajectories grows, exploiting smoothness becomes increasingly important to disambiguate the association of successive points. However, in many important problems an effective parametric model for the trajectories does not exist. Hence we propose modeling trajectories as independent realizations of Gaussian processes with kernel functions which allow for arbitrary smooth motion. Our generative statistical model accounts for the data as coming from an unknown number of such processes, together with expectations for noise points and the probability that points are missing.For inference we compare two methods: A modified version of the Markov chain Monte Carlo data association (MCMCDA) method, and a Gibbs sampling method which is much simpler and faster, and gives better results by being able to search the solution space more efficiently. In both cases, we compare our results against the smoothing provided by linear dynamical systems (LDS).We test our approach on videos of birds and fish, and on 82 image sequences of pollen tubes growing in a petri dish, each with up to 60 tubes with multiple crossings. We achieve 93% accuracy on image sequences with up to ten trajectories (35 sequences) and 88% accuracy when there are more than ten (42 sequences). This performance surpasses that of using an LDS motion model, and far exceeds a simple heuristic tracker.