A Surveillance System Ensures Crossover Formation in C. elegans

Abstract

Crossover (CO) recombination creates a physical connection between homologs that promotes their proper segregation at meiosis I (MI). Failure to realize an obligate CO causes homologs to attach independently to the MI spindle and separate randomly, leading to nondisjunction. However, mechanisms that determine whether homolog pairs have received crossovers remain mysterious. Here we describe a surveillance system in C. elegans that monitors recombination intermediates and couples their formation to meiotic progression. Recombination intermediates are required to activate the system, which then delays further processing if crossover precursors are lacking on even one chromosome. The synaptonemal complex, a specialized, proteinaceous structure connecting homologous chromosomes, is stabilized in cis on chromosomes that receive a crossover and is destabilized on those lacking crossovers, a process that is dependent on the function of the polo-like kinase PLK-2. These results reveal a new layer of communication between crossover-committed intermediates and the synaptonemal complex that functions as a cis-acting, obligate, crossover-counting mechanism.

Keywords: C. elegans; SUN-1; checkpoint; cosa-1; germline; him-5; meiosis; plk-2; synapsis; synaptonemal complex.

Fig. 1. Framework for a crossover surveillance system

(A) Chromatin (green) morphology and SUN-1 Ser12Pi (magenta) distribution change as meiotic nuclei progress through prophase of meiosis I. Leptotene/zygotene is referred to as transition zone (TZ) in C. elegans; early pachytene (EP); pachytene (P). (B) Proposed model for the functioning of a CO surveillance system. The system will be inactive prior to either the onset of meiotic DSBs or the formation of the first CO intermediates. Once a single chromosome has received this intermediate, the system will be activated and all chromosomes will be subject to monitoring (grey nuclear environment). The system will remain active until a CO is received on all chromosomes, at which time chromosomes will disperse from their clustered morphology. Blue cross is a CO intermediate. Magenta chromosomes are the X chromosomes.

Fig. 2. Activation of the surveillance system alters SUN-1 phosphorylation

(A–F) Dissected germ lines were stained with DAPI (top) and anti-SUN-1 Ser12Pi antibody (bottom). Transition zone (TZ) is demarcated with hashed vertical lines. Scale bars: 20μm. (A) Wild type. (B) him-5(ok1896) shows a “dead zone” of pSUN-1 staining in early pachytene (white bar). (C) The wild-type staining pattern is restored by exposing him-5 to 10 Gy IR. (D) him-17(e2806) has a dead zone (white bar). (E) spo-11(me44) shows extended perinuclear pSUN-1 and rare nuclei with pSUN-1 foci. (F) Upon exposure to 2 Gy IR, spo-11 germ lines phenocopy him-5. (Dead zone, white bar). (G, H) Phosphorylated SUN-1 assembles in distinct patches in the irradiated spo-11 germ cells. (G) spo-11 (me44). (H) spo-11 exposed to 2 Gy IR. Scale bars: 5μm. See also Figures S1, S2 and Movie S1.

Fig. 3. Sub-threshold crossovers elicit desynapsis

(A–D, F–I) Mid-late pachytene nuclei stained with anti-SYP-1 (magenta), anti-HTZ-1 (yellow), and DAPI (green) to label the SC, autosomes, and DNA, respectively. (A) Wild type. X chromosomes desynapsis (white arrowheads) is seen in (B) xnd-1(ok709), (C) him-5(ok1896), and (D) him-17(e2806). (E) Desynapsis is seen as a progressive loss of SYP proteins as nuclei progress from mid-late pachytene (left to right) as show by live imaging of nuclei with mCherry::HIS-58 to mark DNA and SYP-1::GFP to visualize the SC. (F) spo-11 exposed to 3 Gy IR. (G) dsb-2(me96) on day 1 of adulthood. Arrowheads mark desynapsed X chromosomes; arrows, desynapsed autosomes. (H, I) Desynapsis is not observed in DSB-defective mutants: (H) mre-11(ok179), (I) dsb-1(me11). Scale bars: 5 μm. (J) Frequency of late pachytene nuclei exhibiting desynapsis. Counts represent percent of nuclei in the last 10 rows prior to diplotene for each genotype. Note that the full synapsis category includes nuclei with crossovers on every chromosome, as in wild type, or no crossovers, as in spo-11. See also Figure S3.

Fig. 4. Desynapsis occurs only on chromosomes lacking CO intermediates

(A–E) Germline nuclei carrying GFP::COSA-1 transgene (green foci) were immunostained with anti-SYP-1 (magenta) and anti-HTP-3 (yellow) to visualize the SC and chromosome axes, respectively. (A) Wild type. (B) him-5(ok1896) (C) dsb-2(me96) (D) spo-11(me44) exposed to 2 Gy IR showing extensive desynapsis. (E) spo-11(me44) with a nucleus containing a single GFP::COSA-1 focus and desynapsis of 5 of the 6 homolog pairs. Nuclei with complete synapsis and 6 GFP::COSA-1 foci are encircled in red, 1–5 foci in cyan, no foci in yellow. Scale bars: 2μm. (F) Desynapsis precedes loss of DSB competency and GFP::COSA-1 recruitment. The graph represents the region from transition zone (leptotene) through the pachytene/ diplotene border. Blue bars show the distance that SUN-1 Ser8 persists in the germline with the dead zone demarcated by brackets. Green bars depict when GFP::COSA-1 is recruited to and localized at CO sites. Purple bars depict when desynapsis occurs in the relevant genotypes. (G) Rare nucleus showing a GFP::COSA-1 focus associated with a thin/near absent tract of SC at 4 hours post-IR. (DNA, yellow; SYP-1, magenta; GFP::COSA-1, green). See also Movies S2, S3.

Fig. 5. The CO surveillance system monitors strand exchange intermediates independently of known meiotic checkpoints and alters SC dynamics in a plk-2 dependent fashion

(A,B) Desynapsis of mid-late pachytene nuclei was quantified for nuclei in the last 10 rows prior to diplotene (see experimental procedures). Representative images of each genotype are shown in Fig. S4 and 5C,D. (A) Eliminating early steps in strand exchange and joint molecule stabilization through mutation of rad-51 and msh-4 or cosa-1, respectively, prevents desynapsis in him-5 and/or dsb-2 backgrounds. RNAi knockdown plk-2 suppresses desynapsis in him-5. (B) Desynapsis still occurs when germline checkpoint functions are abrogated. The lack of DNA damage checkpoint signaling in cep-1;him-5 and cep-1;him-17 mutants does not prevent the desynapsis, nor does the loss of pch-2. Mutations that impair signaling through the DNA damage dependent protein kinases, ATM-1 and ATL-1/ATR, also do not suppress desynapsis in the presence of sub-threshold DSBs, compare spo-11 vs atm-1;spo-11;atl-1 both exposed to 2 Gy IR. ***, p<0.0001, NS= not significant, χ squared, n>100 nuclei from at least 3 germ lines for all genotypes. (C) cosa-1 is epistatic to him-5 and no desynapsis is seen in the double mutants containing mCherry::HIS-58 to mark DNA (magenta) and SYP-1::GFP to visualize the SC (green). Scale bars: 5 μm. (D, E) FRAP shows changes in the SC dynamics associated with the formation of DSBs. (D) Comparison of mid-pachytene nuclei from wild type (red) and spo-11 (blue), and (E) him-5 X chromosome (green) and autosome (magenta). Times at 50% recovery are demarcated with dashed lines. Representative images from movies are found in Fig S5A. (F) Control RNAi does not affect desynapsis; plk-2(RNAi) suppresses desynapsis as shown by live imaging of worms carrying SYP-1::GFP and mCherry::H2B transgenes. See also Figures S4, S5.