LETTER TO THE EDITOR

Polarity of Meiotic Recombination in the bronze Locus of Maize

Meiotic recombination hotspots in plants are of considerable interest because their occurrence has important implications for the mechanism of meiotic recombination and the development of plant breeding strategies. We therefore enjoyed reading the detailed investigations of recombination events in the bronze (bz) locus of maize by Dooner and Martinez-Ferez 1997 . Although these authors conclude that recombination occurs uniformly within the bz gene, we would like to point out the clear evidence in their data for a 5' to 3' polarity in meiotic gene conversion at this locus.

To obtain their data, which they interpret within the context of the dou-ble-strand DNA break (DSB) model for the initiation of meiotic recombina-tion, Dooner and Martinez-Ferez 1997 mapped the recombination junctions in a large number of intragenic recombinants (IGRs) at the bz locus. For convenience, we will also adhere to the DSB model in this letter, although it should be noted that meiotic DSBs have not been demonstrated in organisms other than budding yeast (reviewed by Lichten and Goldman 1995 ).

According to the DSB model, the recombination junctions mapped by Dooner and Martinez-Ferez 1997 can be interpreted as representing the sites at which a Holliday junction was resolved or the initiating DSB occurred (see their Figure 8). Because the authors did not observe a gradient in the frequency of recombination (i.e., recombination junctions) in bz, they concluded that recombination occurs uniformly across this locus and that there are no preferred sites for the initiation of meiotic recombination in bz.

However, polarity is usually not detected as a gradient in the frequency of recombination (or of recombination junctions), but as a gradient of gene conversion. Thus, a sensitive method for detecting recombination polarity is to compare the relative frequency of conversion of the two involved markers in heteroallelic two-point crosses (reviewed by Whitehouse 1982 ). The data presented in Tables 2 and 3 of Dooner and Martinez-Ferez 1997 allow such a comparison, and we show the results in Figure 1. In all but two of the marker pairs, the 5'-most marker is converted more frequently than is the 3' marker, thereby defining a 5' to 3' polarity of gene conversion across the bz locus (Figure 1). Moreover, for the combined data in Figure 1, the difference between the conversion frequencies of the 5' markers and of the 3' markers is significant (P < 0.05; Wilcoxon's matched pairs test). However, because of the small numbers of recombinants, we were not able to analyze this variable for subsets of marker pairs and/or crosses.

View larger version (14K): [in this window] [in a new window]   Figure 1. Polarity of Gene Conversion in the bz Locus of Maize. The horizontal axis shows the physical locations of the E and M mutations. Each line represents the result from a cross between strains carrying the mutations corresponding to the end points of the line; the number to the left of each line denotes the total number of recombinants per cross. The vertical axis shows the percentage of recombinants with conversion of either the 5' mutation (left end of each line) or the 3' mutation (right end of each line) in each cross (i.e., the P1 and P2 categories in Tables 2 and 3 of Dooner and Martinez-Ferez 1997 ).

What is special about the two exceptions? Both of these marker pairs include the mutation E9, and so it is possible that a marker-specific effect, for instance the efficiency by which E9 is repaired in heteroduplexes, accounts for the apparent reversal in the polarity of gene conversion exhibited in these two crosses.

The patterns of polymorphisms in the IGRs analyzed by Dooner and Martínez- Férez (1997) (see their Figures 4 and 5) are also compatible with a 5' to 3' polarity of gene conversion across the bz locus. With a single exception, all of the recombination junctions are located between the E and M mutations used in a cross (Figure 1), which is in agreement with conversion of the 5' E or M mutation, subsequent coconversion of adjacent mutations, and crossover resolution.

Although a clear gradient of recombination junctions in the region between these two mutations is indeed debatable, the polarity of coconversion is not. Mutations that occur upstream of the 5' E or M mutation are (almost) invariably coconverted, mutations that occur between the two E or M mutations sometimes are, and mutations downstream of the 3' E or M mutation are never coconverted. This is a strong indication that there is a 5' to 3' polarity of gene conversion in the bz gene.

We propose that a preferred recombination initiation site 5' to the bz locus would explain the gene conversion polarity that our reevaluation of these data has uncovered. However, this is not the only possibility. Alternative explanations are that the molecular nature of individual markers may cause a bias in the recovery of certain alleles or may affect the site(s) at which recombination is initiated.

For example, the two crosses that exhibit the most obvious polarity (Figure 1; Dooner and Martinez-Ferez 1997 ) involve two heteroallelic Ds insertions, termed m1 and m2, which do not transpose in the maize lines used by Dooner and Martinez-Ferez 1997 . However, we do not think that these observations reflect a bias against conversion to the insertional allele during maize meiosis because both the 5' marker (m1) and the 3' marker (m2) in these crosses are insertions. The other possible explanation for these observations, that m1 (but not m2) causes recombination polarity by acting as a preferred site for recombination initiation, also seems unlikely. This is because the polarity of gene conversion in crosses that involve m1 and a point mutation is weaker than it is in those crosses that involve both m1 and m2.

The crosses that involve both m1 and m2 not only show a stronger polarity of gene conversion than do other crosses, but also a lower frequency of recombination and a lower frequency of crossovers among recombinants (see Tables 2 and 3 in Dooner and Martinez-Ferez 1997 ). These observations support our contention that there is a recombination hotspot 5' of bz because they suggest that branch migration of heteroduplex intermediates into the bz locus may be hampered by the presence of the two Ds insertions. If, as we propose, recombination is preferentially initiated upstream from the bz gene, a block to expansion of heteroduplex regions in recombination intermediates would account for the lower frequency of recombination between m1 and m2 (because of frequent resolution 5' of the m1 insertion) and the stronger polarity of gene conversion (because of infrequent resolution 3' of the m2 insertion). Finally, the low frequency of crossovers among recombinants from m1/m2 crosses could be explained by postulating that the heteroallelic Ds insertions in these crosses provoke a preferential resolution of recombination intermediates into noncrossovers.

REFERENCES

Dooner, H.K., and Martínez-Férez, I.M. (1997) Recombination occurs uniformly within the bronze gene, a meiotic recombination hotspot in the maize genome. Plant Cell 9:1633-1646[Abstract].

Lichten, M., and Goldman, A.S.H. (1995) Meiotic recombination hotspots. Annu. Rev. Genet. 29:423-444[CrossRef][ISI][Medline].

Whitehouse, H.L.K. (1982). Genetic Recombination: Understanding the Mechanisms. (Chichester, UK: John Wiley and Sons).

Dooner, H.K., English, J., Ralston, E., and Weck, E. (1986) A single genetic unit specifies two transposition functions in the maize element Activator.. Science 234:210-211[Abstract/Free Full Text].

Martínez-Férez, I.M., and Dooner, H.K. (1997) Sesqui-Ds, the chromosome-breaking insertion at bz-m1, links double Ds to the original Ds element. Mol. Gen. Genet. 255:580-586[CrossRef][Medline].

This article has been cited by other articles:

				H. K. Dooner Extensive Interallelic Polymorphisms Drive Meiotic Recombination into a Crossover Pathway PLANT CELL, May 1, 2002; 14(5): 1173 - 1183. [Abstract] [Full Text] [PDF]

This Article Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via ISI Web of Science (1) Citing Articles via Google Scholar Google Scholar Articles by Thijs, H. Articles by Heyting, C. Search for Related Content PubMed PubMed Citation Articles by Thijs, H. Articles by Heyting, C. Agricola Articles by Thijs, H. Articles by Heyting, C.