Recombination patterns in maize reveal limits to crossover homeostasis

Abstract

During meiotic recombination, double-strand breaks (DSBs) are formed in chromosomal DNA and then repaired as either crossovers (COs) or non-crossovers (NCOs). In most taxa, the number of DSBs vastly exceeds the number of COs. COs are required for generating genetic diversity in the progeny, as well as proper chromosome segregation. Their formation is tightly controlled so that there is at least one CO per pair of homologous chromosomes whereas the maximum number of COs per chromosome pair is fairly limited. One of the main mechanisms controlling the number of recombination events per meiosis is CO homeostasis, which maintains a stable CO number even when the DSB number is dramatically altered. The existence of CO homeostasis has been reported in several species, including mouse, yeast, and Caenorhabditis elegans. However, it is not known whether homeostasis exists in the same form in all species. In addition, the studies of homeostasis have been conducted using mutants and/or transgenic lines exhibiting fairly severe meiotic phenotypes, and it is unclear how important homeostasis is under normal physiological conditions. We found that, in maize, CO control is robust only to ensure one CO per chromosome pair. However, once this limit is reached, the CO number is linearly related to the DSB number. We propose that CO control is a multifaceted process whose different aspects have a varying degree of importance in different species.

Keywords: crossing-over; crossover homeostasis; double-strand breaks; meiosis; recombination.

Fig. S1.

Quantification of chiasmata in maize. Chiasmata were identified using a combination of chromosome spreading and 3D image reconstruction. (A) Typical morphology of a rod bivalent with a single chiasma at a chromosome end. (B) A ring bivalent with two chiasmata, one at each end. (C) A ring bivalent with three chiasmata, two terminal and one interstitial. (D) A bivalent with two terminal chiasmata and a chromosome twist. (E) A bivalent with two chiasmata located close to each other. (F) A bivalent with a poorly visible terminal chiasma that could make the bivalent be confused for two univalents. Chiasmata are indicated by green arrows. A chromosome twist is indicated by a purple arrow.

Fig. 1.

Variation in the mean number of chiasmata per cell in male meiocytes in a set of diverse maize inbreds. Each blue circle represents a chiasma count from a single meiocyte. Bars = ±SE. The coefficient of variation (C.V.) for chiasma data is indicated for each inbred line below the inbred name on the x axis.

Fig. 2.

Maize inbreds exhibiting substantial differences in chiasma number have similar synaptonemal complex (SC) length. (A–C) Immunolocalization of the ZYP1 protein to measure the SC length in maize. (A) Green, ZYP1 immunolocalization; red, DAPI-stained chromatin. (B) Green, ZYP1 immunolocalization. (C) Measurement of the SC length. The total length of synaptonemal complex in the nucleus was determined by tracing the extent of the ZYP1 protein (white) using the Distance Measurement tool in the softWoRx software. (Scale bar, 5 μm.) (D) Comparison of the SC length and chiasma number in the A344 and CML228 inbreds. Bars = ±SD.

Fig. 3.

CO numbers vary coordinately with DSB number in maize. (A and B) Immunolocalization of RAD51 in zygotene meiocytes in B73 (A) and CML228 (B). Green, RAD51 immunolocalization; red, DAPI-stained chromatin. Images are flat projections of three consecutive Z sections out of a total of 80–90 Z sections through the entire nucleus. (Scale bar: 5 μm.) (C) Variation in the mean number of RAD51 foci per cell in male meiocytes at midzygotene. Each blue circle represents a RAD51 focus count from a single meiocyte. Darker color represents multiple cells that have similar RAD51 foci numbers. Bars = ±SE. C.V., coefficient of variation. (D) Comparison of chiasma numbers and the numbers of RAD51 foci in six maize inbred lines. Bars = ±SD.

Fig. S2.

Distribution of RAD51 focus numbers per meiocyte in B73, CML228, and the B73 × CML228 hybrid at midzygotene. Each circle represents a single nucleus. The color gradation from lighter to darker represents a higher number of cells with the same focus number. Bars = ±SD.

Fig. S3.

H3K4me3 localization is not a good predictor of meiotic recombination sites in maize. (A) Dynamics of H3K4me3 during meiosis in maize. H3K4me3 foci first appear during leptotene and persist throughout meiosis I. Green, H3K4me3 immunolocalization; red, DAPI-stained chromatin. (Scale bar: 5 μm.) (B) Colocalization of H3K4me3 and RAD51 at zygotene in the B73 inbred. Green, RAD51 immunolocalization; red, H3K4me3 immunolocalization; blue, DAPI-stained chromatin. (Scale bar: 5 μm.)

Fig. 4.

Maize inbreds exhibiting significant differences in CO number also differ in the strength of CO interference. (A) Bivalents in B97, Mo18w, and CML228 exhibiting differences in chiasma locations. (Scale bar: 5 μm.) (B) Correlation between the strength of CO interference and the mean chiasma number in maize inbreds. Bars = ±SD.

Fig. 5.

Model explaining how CO homeostasis functions in lines with low and high DSB number. In inbreds with low DSB numbers (Left), DSB distribution in some cells may not be sufficient to ensure proper chromosome pairing, which, in maize, is tightly linked to recombination (19). In these meiocytes, more DSBs are then generated until complete pairing is achieved, resulting in high cell-to-cell variability. In inbreds with high DSB numbers (Right), the distribution of the initially formed DSBs is already sufficient for proper pairing.