In Arabidopsis the quartet qrt mutant enables the
In Arabidopsis the quartet1 (qrt1) mutant enables the same type of analysis of gametophytes (pollen) in sets of four (tetrads) derived from individual meioses 61., 62.. Pollen tetrads from one accession (Columbia-0) are used to pollinate a second polymorphic accessions (Landsberg erecta) resulting in four viable F1 seeds, which can be resequenced to detect recombination events. Using this method, 18 COs and 4 NCOs were detected. The maximum lengths of meiotic recombination related gene conversion were estimated to range from 306 to 3,288bp, with a median of 1,115bp. About half of COCTs were less than 1kb, and one fifth were larger than 2kb . These results were supported by the subsequent study of 13 Arabidopsis tetrads, which estimated the maximum COCT lengths ranged from 127 to 12,127bp, with a 889bp median . A recent study in maize measured 5 fined-mapped COs, and estimated COCT lengths from 220 to 1,875bp . Together, these results suggest an average COCT length of 558bp in plants. This value corresponds closely to one derived from an independent method in Arabidopsis which assayed gene conversion at 7 transgenic test loci in 1.05 million tetrads to estimate an average gene conversion tract length of 605bp . It is notable that these estimates are considerably shorter than maximum median COCT length of 2,643bp in AM251 yeast  and also suggests that COCT lengths are vary both within and between species.
A proposed new model of interference sensitive meiotic recombination pathway In current models of meiotic recombination (Fig. 2), RAD51 and DMC1 facilitate 3′ single end invasion following DSB formation by SPO11. Using the invading 3′-end as a primer and one strand of a non-sister chromatid as a template, DNA synthesis extends the D-loop enabling second end capture and resulting in the formation of a dHJ, which is resolved to yield a CO or, at least theoretically, a NCO. Evidence from S. cerevisiae suggests that the primary products of the DSBR pathway are COs . The mechanistic origin of NCOs has not been determined in plants, but current models assume that, as in yeast, they primarily arise from the SDSA pathway. Based on this assumption, most DSBs are thought be repaired to yield NCOs via SDSA pathway, while the rest are processed into interference sensitive and insensitive COs depending on ZMMs and MUS81 proteins, respectively (Fig. 2). As described above, molecular genetic evidence supports that both DNA leading and lagging strand synthesis factors are required for the formation of meiotic type I crossovers. Since the primary function of those genes are essential for DNA replication, we incorporated both lagging strand synthesis and leading strand elongation in a revised DSBR model, mainly focusing on the type I pathway (Fig. 3). In the proposed new model, we differentiate type I COs into two sub-types by the length of COCT. Using an average length of Okazaki fragment within 100–200bp for lagging strand synthesis , a threshold of 200bp was selected to distinguish long COCTs (>200bp) from short COCTs (<200bp). Since both leading and lagging strand synthesis factors are required for the formation of type I COs, it appears that both long and short COCTs depend on coordination of leading and lagging strand synthesis. The fact that long COCTs require more leading strand elongation is consistent with the processivity Polε. However, aside from Polδ’s function in elongating Okazaki fragments during lagging strand synthesis, it can also replace Polε’s function under certain conditions 47., 68., 69., 70., 71., such as elongation of stretches shorter than Okazaki fragments. Based on this, short COCTs may also require both Polε and Polδ, with Polδ elongating short fragment in the absence of Polε. This model is consistent with the observation that Polε mutants only influence some type I COs while rfc1 has more severe meiotic defects compared to pol2a.
Perspectives Comparison of the Arabidopsis male meiocyte transcriptome with those of somatic tissues revealed over 16,000 expressed genes in both mitotic cells and meiocytes, including DNA replication genes . This supports the idea that mitotic DNA replication and meiotic recombination share components of the DNA synthesis machinery. These genes are often difficult to study in meiosis because loss of their function causes embryonic lethality. A combination of approaches has made Arabidopsis an excellent system for overcoming these difficulties. Several genes have been identified as being involved in these processes, with the exception of the DNA synthesis factor following single end invasion. The use of gene knockdown with the meiosis-specific DMC1 gene promoter provides a feasible approach to investigate the meiotic functions of essential genes, such as POL2A and RFC1 described in this review. There is now strong genetic evidence to support a role for lagging strand synthesis and leading strand elongation in meiotic recombination, but the underlying molecular mechanisms remain to be fully elaborated.