Figure 1.
Proposed mechanism for age-dependent loss of chiasma maintenance in human oocytes.
Different colors are used to represent the homologous chromosomes, each composed of identical sister chromatids. Meiotic cohesion is represented by solid black circles linking the sister chromatids. Centromeres are depicted as white bi-lobed structures. Sister chromatid cohesion distal to the site of a crossover is necessary to hold recombinant homologues together and this linkage ensures their accurate segregation during meiosis I. During meiotic prophase I, cohesion-dependent association of recombinant homologues is manifest as chiasmata that occur at crossover sites. Because meiotic recombination in human oocytes occurs during fetal development and is followed by a prolonged dictyate arrest, chiasma maintenance requires that sister cohesion remain intact for decades. One explanation for why meiotic segregation errors are more prevalent in older women is that cohesion between sister chromatids deteriorates with age and renders recombinant chromosomes susceptible to missegregation.
Figure 2.
Scheme for aging Drosophila oocytes and measuring nondisjunction.
(A) Each Drosophila ovary is composed of several ovarioles that contain a linear array of oocytes at progressive stages of development. Meiosis is initiated at the anterior tip of the ovariole, in the germarium (left). After exit from the germarium, the oocytes move posteriorly through the ovariole (stages 2 through 14). Not all oocyte stages are present at any given time in a single ovariole. This schematic is adapted from Robinson et al. [45]. (B) For the aging regimen, virgin females of the same genotype are divided into two groups. Control females (left) are incubated with males and lay eggs continuously. Experimental females (right) are deprived of males and egg-laying is suppressed. In these females, developmental progression of oogenesis is halted and oocytes “age” within the female. After four days of this aging regimen, control and experimental females are placed in vials with males and the flies are transferred at 24-hour intervals to generate “24 hour broods” of progeny for NDJ analysis. Each brood corresponds to oocytes that were halted at specific stages of development during the aging regimen (see panel A). Meiotic NDJ is measured in all three broods of both aged and non-aged oocytes (see below). (C) To measure NDJ, we made use of an X chromosome dominant marker, Bar (B−), that affects eye shape. Bar+ experimental females are crossed to Bar− males that contain an attached X∧Y chromosome. The sperm generated by these males will contain either X∧Y with the Bar− marker or no sex chromosome (designated “0”). If meiotic segregation is normal in the female, oocytes will be produced that contain one X chromosome. However, segregation errors during female meiosis will result in exceptional gametes that contain either two X chromosomes (Diplo) or no X chromosomes (Nullo). The adult progeny resulting from normal and exceptional gametes can be distinguished based on their sex and eye shape (due to Bar−). In this test, all of the normal progeny survive, but only half of the exceptional progeny are viable. Therefore, the number of exceptional progeny is doubled and the total number of progeny is adjusted when calculating the %NDJ.
Figure 3.
Chiasma formation is not severely disrupted when SMC1 is reduced.
(A) Whole ovary extracts from wild-type and smc1+/− females were analyzed by immunoblot analysis using SMC1 antiserum [11]. α-tubulin was used as a loading control. Note that the relative intensity of SMC1 signal is lower in the smc1+/− extract than in wild type. A visual comparison suggests that SMC1 is reduced by approximately two-fold (SMC1 signal intensity is similar in the lanes containing 20 µg of smc1+/− extract and 10 µg of wild-type extract). (B) Meiotic recombination in wild-type and smc1+/− females was measured in four intervals along the X chromosome: sc-cv, cv-v, v-f, f-car. Crossovers were reduced in the distal intervals (sc-cv, cv-v) in smc1+/− oocytes. The tetrad exchange rank was computed from the recombination data and the percentage of bivalents for each rank is shown. Note that E0 tetrads increase in the mutant females at the expense of E1, E2, E3 tetrads.
Figure 4.
Scheme to assay loss of chiasmata with age in Drosophila oocytes.
In the absence of crossovers, accurate segregation of bivalents in Drosophila oocytes is governed by the achiasmate pathway which relies on pericentric heterochromatin. In the presence of a functional achiasmate system, bivalents that lose their chiasmata will still segregate properly. Therefore, the achiasmate pathway must be compromised to determine whether chiasmata are lost as Drosophila oocytes age. Top panel: A functional achiasmate segregation system (depicted by green stars between homologues) will ensure accurate segregation of recombinant bivalents that have lost chiasmata during the aging process. Bottom panel: When the achiasmate pathway is compromised (depicted by the absence of green stars), recombinant bivalents that have lost chiasmata should missegregate.
Figure 5.
Both chiasmate and achiasmate homologues exhibit age-dependent NDJ in smc1+/− mtrm+/− oocytes.
(A) An odds ratio is used to represent the relative level of NDJ in aged versus non-aged oocytes for 24-hour broods. An odds ratio >1 indicates that NDJ is more likely in the aged oocytes than in the non-aged oocytes. The error bars correspond to ±1 standard error. Two independent experiments are shown. In both experiments, NDJ in aged oocytes is significantly higher than in non-aged oocytes for broods 1 and 2 (**P = 0.0010, *P = 0.0062; ••P = 0.0017, •P = 0.0439) but age-dependent NDJ is not observed for Brood 3 (P = 0.4751, P = 0.1416 for Experiments I and II respectively). Although the absolute levels of NDJ vary, the same pattern is observed for both experiments. For each experiment, at least 1000 progeny were scored for each brood. The raw data for the two independent experiments are shown in Tables S3A and S3B. (B) Diplo-X females from brood 1 of Experiment II were genotyped (see Figure 6). R denotes Diplo-X females containing at least one recombinant X chromosome. NR denotes Diplo-X females in which both X chromosomes were non-recombinant (as judged by the markers scored). N is the number of total progeny scored in the NDJ assay. The frequency at which R and NR chromosomes missegregate also is shown (R/N and NR/N). Missegregation of both recombinant and non-recombinant chromosomes is more prevalent in aged oocytes.
Figure 6.
Scheme to determine the recombinational history of X chromosomes that missegregate in the NDJ test.
A hypothetical example is presented in the two panels to illustrate how we determine the genotype of the two missegregating X chromosomes that give rise to a Diplo-X female. Top panel: In this scenario, a single crossover between cv-v on the X chromosome in the oocyte generates a bivalent with two recombinant and two non-recombinant chromatids. Missegregation during meiosis I results in an egg with two X chromosomes that is fertilized and develops into a “Diplo-X” adult female. In this schematic, the Diplo-X female inherited one recombinant and one non-recombinant X chromosome from the oocyte. Bottom panel: The Diplo-X female is crossed to a normal male and her male progeny are scored for the different markers on the X chromosome. Genotyping is complicated by the fact that crossovers can occur between the two X chromosomes in the oocytes of the Diplo-X females. However, the genotype of the X chromosomes in the Diplo-X female (right, top panel) can be assigned because in any recombination analysis, the parental chromosomes are the most predominant. In this example, the parental chromosomes are shown in red. Note that it is possible for a Diplo-X female to inherit two non-recombinant chromatids from a recombinant tetrad. In addition, a double crossover between two adjacent markers will be invisible. For these reasons, the frequency at which recombinant chromosomes nondisjoin is underestimated in our analyses.
Figure 7.
Quantification of oocyte stages present in non-aged and aged ovaries at different time-points following the aging regimen.
The number of oocytes at each of the indicated stages was tabulated for one ovary from three different smc1+/− mtrm+/− females and the average is presented. Immediately after the aging regimen (time-point 0), the number of mature oocytes (stage 13/14) is ∼2.5 fold higher in the aged ovaries than in non-aged ovaries. This contrasts sharply with the decline in stage 10 oocytes at this same time point. The number of stage 8 oocytes is similar in aged and non-aged ovarioles immediately following aging regimen. These results are consistent with a previous report [23] that when egg-laying is inhibited, stages 1–8 halt in development and “age” while stages 9–13 continue to develop and only arrest once they reach stage 14. After egg-laying resumes, the relative abundance of each of the different stages is similar in aged and non-aged ovaries by 16 hours although fewer oocytes are present in the aged ovaries.
Figure 8.
Diplotene-like oocytes are most vulnerable to age-dependent nondisjunction.
(A) For each of the different 8-hour sub-broods indicated, the relative level of NDJ in aged versus non-aged smc1+/− mtrm+/− oocytes is represented as an odds ratio. The error bars correspond to ±1 standard error. Oocytes laid between 16–32 hours after the aging regimen (sub-broods 3 and 4) exhibit significant age-dependent NDJ (**P = 0.0009 and *P = 0.0035). The data shown represent pooled data from two independent experiments in which at least 1200 progeny were scored for each sub-brood. The combined raw data are shown in Table S6. (B) For aged oocytes, the frequency at which recombinant chromosomes missegregate (R/N) in sub-brood 3 is significantly different from that for both sub-broods 1 and 6 (P = 0.0037, for sub-brood 3 versus 1, P = 0.022 for sub-brood 3 versus 6). In contrast, the frequency of R Diplos is similar in all four sub-broods analyzed from non-aged oocytes (0.73<P<0.92 for pair-wise comparisons). (C) Schematic of an ovariole following the aging regimen reflects the accumulation of metaphase I arrested oocytes (stage 14) that occurs at the expense of stages 9–13 (compare with Figure 2A). Unlike stages 9–13 which continue to progress, oocytes at stages 1–8 halt in development during the aging regimen. However, only oocytes that undergo aging during diplotene (stages 7–8) exhibit significant levels of age-dependent NDJ. In contrast, oocytes that remain arrested at metaphase I are not susceptible to age effects. Schematic is adapted from Robinson et al. [45].