Evidence that weakened centromere cohesion is a leading cause of age-related aneuploidy in oocytes

Abstract

Aneuploidy arising early in development is the leading genetic cause of birth defects and developmental disabilities in humans. Most errors in chromosome number originate from the egg, and maternal age is well established as the key risk factor. Although the importance of this problem for reproductive health is widely recognized, the underlying molecular basis for age-related aneuploidy in female meiosis is unknown. Here we show that weakened chromosome cohesion is a leading cause of aneuploidy in oocytes in a natural aging mouse model. We find that sister kinetochores are farther apart at both metaphase I and II, indicating reduced centromere cohesion. Moreover, levels of the meiotic cohesin protein REC8 are severely reduced on chromosomes in oocytes from old mice. To test whether cohesion defects lead to the observed aneuploidies, we monitored chromosome segregation dynamics at anaphase I in live oocytes and counted chromosomes in the resulting metaphase II eggs. About 90% of age-related aneuploidies are best explained by weakened centromere cohesion. Together, these results demonstrate that the maternal age-associated increase in aneuploidy is often due to a failure to effectively replace cohesin proteins that are lost from chromosomes during aging.

Figure 1. Sister kinetochores are farther apart in old oocytes both at MII and MI

A-C, Oocytes from young (6–14 weeks of age) or old (16–19 months of age) mice were matured in vitro, treated with monastrol to disperse the chromosomes at MII, then fixed and stained for DNA (Sytox, green) and kinetochores (CREST, red). Representative images (A) show increased distance between sister kinetochores, with insets magnified to show sister kinetochore pairs. Inter-kinetochore distances were measured for all sister kinetochore pairs in an oocyte and averaged over each oocyte. The populations of young (n=24) and old (n=17) oocytes are represented by histograms (B) and by the mean (±sem) for each group (C). D-F, Oocytes were treated with monastrol at MI, then fixed and stained as above. Insets (D) show increased separation of sister kinetochores; each box includes all four kinetochores of a bivalent. Sister kinetochore configurations in MI were classified as indistinguishable, overlapping, or distinct (E, left to right). Outer kinetochore distances were measured from the outer edges of sister kinetochore pairs for young and old oocytes (n=200 kinetochore pairs from 10 oocytes in each group) (F). Images (A, D) are maximal intensity projections of confocal z-series, scale bars 5 μm. See also Figure S1.

Figure 2. REC8 protein levels on chromosomes are severely reduced in old oocytes while total REC8 is constant

A, B, Chromosome-associated REC8 was detected by immunocytochemistry in young and old oocytes at MI. Representative images (A) show DNA (Sytox, green) and REC8 (red), scale bar 5 μm. Similar results were obtained with multiple young (n=11) and old (n=9) oocytes, and the REC8 fluorescence intensity was quantified (B). C, D, Total REC8 levels in young and old oocytes at MI (n=80 oocytes in each lane) were determined by Western blot, with TUBB as a loading control. A representative blot is shown; similar results were obtained in two independent experiments, and band intensities were quantified (D). E, Levels of chromosome-associated REC8 were measured as in (A) with oocytes from mice at different ages (in months): 3 (n=21), 6 (n=21), 9 (n=19), 12 (n=21), 15 (n=7), 17 (n=3). F, G, MII eggs from mice at various ages were analyzed to determine inter-kinetochore distances (F) as in Fig. 1A-C, and the percent of eggs with chromosome segregation errors, including aneuploidies and separated sister chromatids, was determined (G). Mice were used at the following different ages (in months): 3 (n=39), 6 (n>28), 9 (n=20), 12 (n>24), 15 (n>14), 17 (n=16). See also Figure S2.

Figure 3. Proposed outcomes of weakened centromere cohesion in MI

Schematics show the predicted outcomes of normal (A) or weakened (B-E) centromere cohesion in MI. Red boxes follow the progression of a sister kinetochore pair. Cohesion status and sister kinetochore orientation on the MI spindle determine chromosome segregation dynamics in Anaphase I and the chromosome content of the MII egg. For prediction (E), sister kinetochore separation is shown at MI but could alternatively arise at a later time. Note that mouse chromosomes are telocentric.

Figure 4. Analysis of Anaphase I chromosome segregation dynamics and MII chromosome numbers in old oocytes

A, An old oocyte was micro-injected with H2b-gfp cRNA, then matured in vitro and imaged live: initially by DIC to determine the time of anaphase onset (t=0) and subsequently by fluorescence to detect lagging chromosomes in Anaphase I. GFP images are magnified from white boxes in the DIC images. No lagging chromosomes are detected in this example. B, The MII egg resulting from the oocyte shown in (A) was fixed and stained for DNA (Sytox, green) and kinetochores (CREST, red) to count chromosomes. The egg is euploid with 19 paired sister kinetochores (numbered 1–19) and two unpaired kinetochores (yellow and orange asterisks). Insets show the two unpaired kinetochores and an intact sister kinetochore pair. C, Another old oocyte was imaged live in Meiosis I as in (A). In this example a lagging chromosome (arrow) eventually segregates to the polar body. The final kinetochore count in the MII egg is 39, indicating loss of a single chromatid. Arrowheads mark the polar bodies; GFP images are maximal intensity projections of confocal z-series; scale bars 5 μm.