Moore, DP; Page, AW; Tang, TT; Kerrebrock, AW; and Orr-Weaver, TL. (1998). The Cohesion Protein MEI-S332 Localizes to Condensed Meiotic and Mitotic Centromeres until Sister Chromatids Separate. J. Cell Biol. 140: 1003-1012.
ANALYSIS OF RESULTS
The authors identify three questions they seek to answer in this paper: First they want to determine the localization of MEI-S332 during meiosis in female Drosophila melanogaster. Having shown this protein's location during segregation, the authors then ask whether MEI-S332 is degraded or not. Then showing that it is not degraded in meiosis, they examine MEI-S332 behavior in embryonic meiosis and mitosis. Thus I will follow this logical progression in discussing the results presented in this paper.
To determine MEI-S332 localization in oocyte meiotic divisions, and specifically whether it localizes to meiotic centromeres, the authors isolated oocytes from females carrying four copies of a GFP-fused mei-S332 transgene (mei-S332+::GFP). Figure 1 displays flourographs of these fixed oocytes, stained for GFP (green) and chromatin (red), at seven stages of meiosis. Fig. 1 A shows an unactivated oocyte (arrested at metaphase I) with a circular nucleus and two "caps" of GFP at opposite ends of the nucleus. Citing that "the centromeric regions of chromosomes are positioned on opposite sides of the chromatin mass during the metaphase I arrest in Drosophila oocytes" (Dernburg et al., 1996), the authors claim that these MEI-S332-GFP caps most likely represent centromeric localization. Fig. 1 B shows an oocyte fixed at anaphase I. Seven MEI-S332 GFP dots are seen (as the authors note, the fourth chromosome containing the seventh and eighth centromere IS difficult to visualize). Although no polar orientation is shown on the image, the authors state that the dots are located on the leading edges of the chromosomes, one per pair of sister chromatids. Fig. 1 C shows a stained oocyte at late anaphase I, with two areas of separating chromatin, each with two MEI-S332-GFP dots visible at the leading edges. Between the first and second meiotic divisions, shown in Fig. 1 D, two clusters of chromatin "balls" are seen, each containing several MEI-S332-GFP dots. The authors state that each cluster is composed of the four pairs of sister chromatids. During metaphase II, shown in Fig. 1 E, when sister chromatids begin to separate, each ball of chromatin is seen separating, with at least two MEI-S332 GFP dots visible per chromatin ball. Once sister chromatids have separated at the end of anaphase II, Fig. 1 F, no MEI-S332-GFP dots are visible. As the chromosomes decondense during postmeiotic interphase, Fig 1 G, MEI-S332 GFP remains undetected.
These results greatly support the authors' hypothesis that MEI-S332 is localized to the centromeres. There are eight (seven distinct) MEI-S332-GFP dots seen during anaphase I (Fig. 1 B). As noted, this is exactly the number expected if the protein localizes to the centromeres, since the haploid chromosome number for Drosophila is four. In addition, the MEI-S332-GFP dots are seen along the leading edges of the separating chromosomes (Fig. 1 C), and at the centromeric locations in chromatid separation (Fig. 1 E). Then, once sister chromatid cohesion has been lost, MEI-S332-GFP is no longer visible (Fig. 1 F,G).
To further examine the temporal nature of MEI-S332 localization, the authors isolated stage 13 and 14 oocytes from the ovaries of females transgenic for mei S332-GFP, and visualized MEI-S332-GFP presence with respect to tubulin spindle formation. They report that although MEI-S332 is observed in spermatocyte cytoplasm during prophase I, it is necessary to observe localization in oocytes, as oocytes and spermatocytes differ in several significant respects during meiotic activity. Figure 2 shows this set of meiotic flourographs, stained for MEI-S332-GFP (green), tubulin (blue), and chromatin (red). Panel A shows the respective staining at prometaphase I, the first stage at which MEI-S332-GFP is observed. It shows two areas of MEI-S332-GFP at opposite ends of the chromatin (3rd column) and a general scattering of tubulin (4th column). Panels B and C show that as tubulin becomes increasingly elongated and bipolar (4th column), MEI-S332-GFP organizes more clearly into the caps at opposite ends of the chromatin mass (3rd column).
From these results, the authors conclude that the first MEI-S332 observation is coincident with spindle formation. It is relevant to note that nothing more than coincidence can be assigned in this simultaneous appearance of MEI-S332 and tubulin organization. It is likely that the two events are initiated by the same cellular control mechanisms, but from this experiment alone, no causal association can be made.
MEI-S332 Fate in Meiosis
Having observed that MEI-S332 is no longer visible after anaphase II (in either male or female meioses), the authors then seek to determine what happens to the protein after the sister chromatids separate. Citing the cyclin destruction machinery of the yeast and Xenopus mitoses which degrades a sister chromatid separation inhibitor, the authors speculate that MEI-S332 is degraded in a similar manner.
To determine MEI-S332 levels, they raised polyclonal rabbit antibodies against a COOH-terminal fragment of the MEI-S332 protein (binding epitope indicated by arrow in Fig. 3 A). Fig. 3 B shows the Western blot generated using the affinity purified a-MEI-S332 probe. This was performed to be certain that the antibody does indeed recognize the MEI-S332 protein. Lane 1 is a positive control using extracts from transgenic ovaries that had four extra copies of mei-S332, giving a total of 6 copies of the transgene. Lanes 2 and 3 show negative controls from both oocyte (lane 2) and ovary (lane 3) extracts made with a mutated MEI-S332, mei-S3327. This mutation creates a nonsense codon which prematurely truncates the MEI-S332 so that it does not contain the epitope for the rabbit antibody (stop region indicated in Fig. 3 A). Lanes 4-5 show the experimental lanes, consisting of extracts from immature ovaries (lane 4), ovaries (lane 5). Lanes 6-7 relate to a later section of the paper (see below: MEI-S332 Fate in Mitosis). The same size band (~55kD) appears in each of lanes 1 and 4-5, but is absent from lanes 2-3 (Fig. 3 B). The conclusion that this band is MEI-S332 is further supported by the expected result that the positive control band is significantly more intense (as these oocytes contains at least three times as many genes).
In order to actually determine MEI-S332 levels during the metaphase II-anaphase II transition, further oocyte extracts were analyzed, using the affinity purified a MEI-S332 antibody (Fig. 3 C). Lane 1 of the Western blot shows a positive control of wild-type oocytes. Lane 2 shows a negative control using the mei S3327 mutated transgene. Lanes 3-4 are the experimental lanes, containing unactivated oocytes (lane 3) and oocytes fixed 60 min. after in-vitro activation (lane 4). The authors claim that activated eggs can be selected so that 95-99% have completed meiosis. In light of the positive and negative controls, the authors seem justified in their conclusion that the top band (present in all lanes except the negative control) is the MEI-S332 protein, and the lower band (present in all lanes) is an experimental artifact. These results (repeated several times according to the authors) lead them to conclude that MEI-S332 is not degraded after chromatid separation, but rather is present in similar levels before and after separation (although it is difficult to objectively quantify protein levels simply from a Western blot). However, to further substantiate the conclusion that MEI S332 is not degraded, the authors repeat the experiment using a translational inhibitor. They feared that new translation of MEI-S332 could mask degradation levels, so they analyzed unactivated oocytes (lane 5) and oocytes activated in the presence of cycloheximide (lane 6), which reportedly inhibits translation to 5% of wild-type levels, yet allows for 95% of oocytes to still complete meiosis. The same two bands appeared in both lanes (Fig 3 C) , confirming the conclusion that MEI-S332 is present at similar levels both before and after sister chromatid separation.
MEI-S332 Fate in Mitosis
Citing that the phenotype of mei-S332 mutants is only known to be only meiotic and the findings that MEI-S332 is not degraded after anaphase II, the authors then asked whether the protein persisted in the developing embryo. They determined protein level as before, probing 0-2 hour embryos (Fig. 3B, lane 6) and 2-4 hour embryos (Fig. 3 B, lane 7) with rabbit a-MEI-S332 antibody. The results are displayed in Fig 3 B. Again, from the positive and negative controls, the authors seem justified in their conclusion that, like the ovary extracts, the embryos also contained MEI-S332. The authors also observe that the protein level seemed to increase between the 0-2 and 2-4 hour embryo. Although somewhat speculative, the bands do seem to support such a conclusion.
The authors then ask if MEI-S332-GFP localizes to embryonic chromosomes in both mitosis and meiosis. They analyze embryonic localization using the same mei-S332+ transgene immunofluorescence method as was used in Fig. 1. Fig. 4 A shows centrally located MEI-S332-GFP dots in meiotic polar body rosettes. These rosettes form from chromosomes which have replicated and condensed into a metaphase like-state. Citing that the centromeres are expected to be located centrally in these rosettes, they conclude that MEI is localized to the centromeres. Fig. 4 B simply shows a close-up of Fig. 4 A. The authors claim that in Fig. 4 B, 24 chromosome foci are distinguishable (which is the expected number after chromosome replication), although I find it difficult to distinguish all 24. Nonetheless, the authors observe that similar to oocyte meiosis, embryonic polar bodies also show a localization of MEI-S332 to the centromeric regions of replicated sister chromatids. Turning to mitosis, Fig. 4 C shows embryonic chromosomes condensed in metaphase and MEI-S332-GFP localized in discrete dots on these chromosomes. As before, the authors claim that these dots are specifically localized to the centromeric regions. Additionally, MEI-S332-GFP is observed in diffuse clouds around the mitotic nuclei. Fig. 4 D shows embryonic cells in interphase, but no discrete MEI-S332-GFP dots are observed, only the diffuse areas surrounding the nuclei. Fig. 4 E simply shows a close up of Fig. 4 D, better illustrating the absence of discrete MEI-S332-GFP dots and only the presence of diffuse MEI-S332-GFP. Although not certain what the areas of diffuse MEI-S332-GFP are, the authors speculate that they may correspond to energids, regions of yolk-free cytoplasm that have been observed in the early cycles of Drosohpila embryos. Nonetheless, this figure suggests that in both embryonic mitosis and meiosis, MEI-S332 localizes to centromeric regions on the chromosomes.
Having demonstrated the likelihood of centromeric MEI localization in embryonic mitosis and meiosis, the authors finally seek to confirm the protein's mitotic localization during the metaphase-anaphase transition. They examine the mitotic divisions in embyros during the later syncytial divisions in which the nuclei migrate to the surface and are easily visible during the subsequent waves of mitosis. These waves of mitosis allow for observations of different cell cycle stages within the same organism. Embryos from mothers hemizygous for the mei-S3327 mutation (which produces the truncated protein form) and carrying two copies of the mei-S332+::GFP transgene were used to simulate the mei-S332 gene dosages of oocytes. Fig. 5 A is a fluorograph of these embryonic cells in a wave of mitotic activity, with metaphase cells at the top, anaphase cells in the middle, and late anaphase cells at the bottom. DNA appears as red and MEI-S332 appears as green, with column 2 staining only for DNA and column 3 only for MEI-S332. As is evident, in columns 1 and 3, MEI S332-GFP dots are observed in most of the metaphase cells at the top of the figure (Fig. 5 A, arrow), but ceases to be observed in the anaphase or late anaphase cells. Fig. 5B is a close-up of a two cells--one in early anaphase (left) and the other in late anaphase (right). MEI-S332-GFP dots are also observable at the leading edge of chromosomes (Fig. 5 B, arrowhead) in the early anaphase cell, but was not observed anywhere in the late anaphase cell. These observations confirm that, similar to oocyte and embryonic meiosis, MEI-S332 is localized to centromeric regions in embryonic mitosis during the during the metaphase anaphase transition, but is not detectable in later stages of anaphase, once chromosomes separate.
PROPOSED AVENUES OF FURTHER RESEARCH
Although several important questions are raised by this paper--such as 1) if the authors are correct in their claim that embryonic MEI-S332 shows a different mobility form (Fig. 3 B), then does this second, embryonic form differ functionally from the oocyte form? 2) what role does the diffuse cloud of MEI S332 observed around the nucleus (Fig. 4 D, E) in embryonic interphase play? and 3) are there truly no phenotypes of mitotic mei-S332 mutants?--I would like to focus on the most pertinent question: what happens to MEI-S332 after chromatid separation? In the follow few paragraphs, I will propose several experimental procedures which seek to answer this question.
Having observed that MEI-S332 localizes to the centromeric regions of condensed chromosomes in oocyte meiosis and embryonic mitosis and meiosis, the authors write, Òprecisely what happens to MEI-S332 when sister chromatids separate at anaphase is a question of great interest.Ó In their discussion they propose three conceivable possibilities: 1) the protein is degraded at the metaphase-anaphase transition; 2) the protein dissociates at this transition, only to be later degraded; or 3) MEI-S332 is inactivated before dissociation, permitting anaphase movement and subsequent separation. The authors answer part of this question in their experiment: they confirm that MEI-S332 is not degraded, but rather is observed in similar levels before and after the metaphase II-anaphase II transition (Fig. 3). This result seems to implicate one of the latter two possibilities.
Several methods could be employed to confirm these speculations. The authors write that MEI-S332 Òmay be regulated at some level by phosphorylation,Ó observing that the protein has over 30 recognizable phosphorylation sites. Since this is the case, it might be possible to isolate MEI-S332, attach it to beads, and run an affinity chromatography in an attempt to isolate a regulatory protein. If one (or even multiple) regulators were identified, and its expressing gene were identified through expression libraries, it would be possible to create knock-out oocytes which were deficient for the regulatory protein. Visualization experiments similar to those performed in this paper could then confirm if cells deficient for this MEI-S332 regulatory protein were able to complete meiotic division. Even more significant, it would be possible to identify at which meiotic stage the cell arrested. If meiotic activity was halted during meiosis I, then it would seem that a similar regulatory mechanism would be present in mitosis; however, if meiotic activity was halted only during meiosis II, then it would seem that the regulatory mechanism was specific to meiosis. Thus experiments using knock-out cells deficient in the supposed regulatory protein could confirm such a regulatory mechanism for MEI-S332 in chromatid separation.
Another possible experiment could involve controlled levels of MEI-S332. The authors observe that decreased amounts of MEI-S332 are detected on chromosomes in early anaphase (Fig. 5 B), suggesting that there might be a wave of inactivation/dissociation. It would seem plausible then, that there might be a threshold level of MEI-S332 which is responsible for chromosome separation. In an attempt to identify such a threshold level, an experiment could be performed which creates several levels of MEI-S332 expression through the translational inhibitor cycloheximide. Through this experiment it would be possible to determine if there is a level of MEI-S332 expression required for meiotic completion. Such an experiment might be able to distinguish the dissociation model from the inactivation-dissociation model. Thus, each of these experiments seeks to determine what happens to MEI-S332 after chromotid separation. In turn, such experiment coupled will provide further insight into the mechanisms of both meiosis and mitosis.
Please send your comments, suggestions to email@example.com.
Back to GSN's homepage.
Link to Davidson College's Biology Homepage.