Hawley Lab R. Scott Hawley, Ph.D. Investigator rsh@stowers-institute.org

     We focus on elucidating the mechanisms by which chromosomes pair, the mechanism and control of recombination, and the mechanisms by which chiasmata ensure segregation. All of our studies begin with genetic screens to identify genes and chromosomal sites required for the meiotic process. Indeed, a reviewer of an American Cancer Society grant proposal once stated that we "have raised the genetic analysis of meiosis in Drosophila to an art form". The work in our laboratory during the last eight years is divided into several basic areas, each of which is described briefly below.

1. The analysis of homologous chromosome pairing in Drosophila in both mitotic and meiotic cells (Justin Blumenstiel, American Cancer Society Post-doctoral Fellow)

     To ensure the proper segregation of chromosomes during meiosis, it is essential that chromosomes become paired and aligned with one another. In many organisms, this pairing is restricted to meiosis and occurs in early meiotic prophase. However, in Drosophila, and in other dipterans, somatic pairing is ubiquitous. Recent work in the lab using the LacI-GFP system that allows direct visualization of chromosome dynamics has indicated that chromosome pairing is established prior to meiosis in the mitotically dividing germline (Gong, McKim and Hawley, 2005). This suggests the possibility that meiotic chromosome pairing in Drosophila is established in the early differentiating germline during embryogenesis. It is therefore possible that the mechanism that establishes somatic chromosome pairing during embryogenesis may also establish chromosome pairing in the germline and that this pairing is maintained up until meiosis. Alternatively, we are also exploring the possibility that the re-establishment of pairing during G1 might be a general property of Drosophila cells, both meiotic and mitotic.

2. The mechanisms of synapsis and synaptonemal complex (SC) formation and function during meiosis (Justin Blumenstiel and Jennifer Jeffress)

     The SC forms between homologous chromosomes during meiotic prophase and functions to regulate meiotic recombination and maintain chromosome pairing. The Drosophila C(3)G protein is structurally and functionally similar to ZIP1 and SCP1 of yeast and mammals, respectively, and is a component of the transverse filaments of the SC (Page et al., 2001; Anderson et al., 2005). The secondary structure prediction for C(3)G shows a central coiled coil-rich (CC) domain flanked by N- and C-terminal globular domains. To identify functional domains of C(3)G, Jennifer has designed a full length (FL) wild type C(3)G expression construct and a series of six constructs that express C(3)G with an in-frame deletion of either of the globular domains, or portions of the CC domain. Deletion of the N-terminus, or the N-terminal end of the CC domain ablates the ability for C(3)G to promote synapsis and recombination. C(3)G lacking the C-terminal domain failed to localize along chromosomes but rather concentrated within a polycomplex-like structure that we have further analyzed by electron microscopy. These data demonstrate that the N- and C-termini of C(3)G play important, but distinct roles in C(3)G function. The C-terminus may be necessary for connection with the lateral elements of the SC. The N-terminus and, in particular, a CC region adjacent to the N-terminus could be important for the dimerization of C(3)G or connecting C(3)G homodimers across the SC central element.

     Perhaps the most important question we can ask about the SC concerns its function. We propose that the SC serves to communicate to two homologous centromeres that they are connected by a crossover and thus facilitate their co-orientation. Evidence that crossingover does indeed ensure proper centromere pairings has already been provided in yeast by the Dawson and Roeder labs. Youbin Xiang studies of the process of secondary nondisjunction in Drosophila, in which two nonexchange X chromosomes can segregate from a Y chromosome while exchange chromosomes segregate from each other (ignoring the Y), suggest that exchanges alter the spatial relationships between centromeric regions long before nuclear envelope breakdown, and in doing so set up centromere co-orientation. Therefore, we propose that the primary function of the SC is to communicate distal exchanges to the centromeres (Xiang and Hawley, 2006).

3. The role of the replication protein Mcm5 in meiotic recombination (Cathy Lake)

     In a genetic screen engineered to identify mutations in essential genes Cathy recovered a meiosis-specific allele of mcm5. This mutant which is homozygous viable and fertile shows high levels of chiasmate X nondisjunction. Several lines of evidence indicate that this new meiotic mutant is the first allele of the Drosophila mcm5 gene. First, sequence analysis of the mcm5 gene revealed a single A-T transversion when compared to the target chromosome. This mutation is in a conserved residue in the C terminal region of the protein. Mcm5 is known to be involved in the initiation of replication, transcription and recombination. The allele isolated by Cathy shows normal localization of the synaptonemal complex protein C(3)G and double strand break formation appears normal. However, there is a significant reduction in the frequency of recombination. This allele of mcm5 encodes a protein that is not required for viability, fertility, or the repair of DNA damage in mitotic cells. Cathy is now exploring the possibility that this allele of mcm5 defines a meiosis-specific domain or residue of the Mcm5 protein. In addition she is collaborating with Professor Susan Forsburg at the University of Southern California to create the same mutation in yeast and study its role in meiotic recombination. Finally, like other known mutants that show defects in recombination, this allele of mcm5 is defective in the ability to arrest at metaphase I.

4. The role of heterochromatin in ensuring achiasmate segregation (Bill Gilliland, American Cancer Society Post-doctoral Fellow)

     Meiotic segregation can be viewed as the events that mediate the stable co-orientation of the homologous centromeres toward opposite poles of the meiotic spindle. Establishing stable co-orientation requires a balancing of plateward and poleward forces acting on the kinetochores. This is usually achieved by chiasmata, the physical consequence of recombination. The key to this process is that chiasmata physically connect homologous chromosomes, thus balancing the poleward forces on the two oppositely oriented centromeres. However, like many organisms, Drosophila females possess a mechanism that ensures the regular segregation of achiasmate homologs. The central focus of my lab's research over the last decade has been to elucidate the mechanisms that allow proper meiotic segregation in the absence of exchange. The principal result of these studies is a delightfully simple model of achiasmate segregation in which the pattern of achiasmate segregation is determined by heterochromatic pairing, and in which the non-exchange DNA-DNA linkages within the heterochromatin serve as the functional equivalent of chiasmata in terms of holding homologs together. Evidence for these assertions is described briefly below.

     Achiasmate segregations are the result of heterochromatic pairings. The first evidence that heterochromatic homology is crucial to the segregation of achiasmate homologs came from our studies of the effects of homologous duplications on the segregation of two normal 4th chromosomes (Hawley et al., 1993a). Based on these data, we proposed that achiasmate segregations are facilitated by the pairing of heterochromatic regions surrounding the centromeres. In collaboration with Abby Dernburg and John Sedat at the University of California, San Francisco, we used three-dimensional fluorescent hybridization to demonstrate that in Drosophila meiosis, heterochromatic pairings persist beyond the dissolution of the synaptonemal complex at pachytene and facilitate centromere co-orientation at prometaphase (Dernburg et al., 1996) (we note that the development of this model, and its verification, occurred in parallel with work in Gary Karpen's lab at Salk, and reflects collaborations of many kinds between the Hawley, Karpen, and Sedat labs).

     During Bill’s cytological studies of oocytes in females defective for ald/mps1 (see below), he noticed threads of DAPI-staining material between chromosomes, even between the normally non-crossover 4th chromosomes. We realized that the threads could be due to DNA-DNA linkages (which could have been established by the process of repairing stalled replication forks, a normal process in replicating highly repetitive DNA) holding paired chromosomes together, which had not had time to be resolved by topoisomerase-like enzymes before normal chromosome separation. Bill has demonstrated that the threads hybridize to a heterochromatin-specific FISH probe, that the threads can be observed even in the absence of recombination (by use of a null allele for mei-W68, the spo11 homolog), and that chemical inhibition of topoisomerase II can result in elevated nondisjunction. The analysis of these DNA linkages and their role in facilitating achiasmate segregation is a major focus of our laboratory.

     Although these heterochromatic pairings are sufficient to hold pairs of achiasmate homologs together, they are not sufficient to allow the immediate co-orientation of achiasmate homologs (see next section), this requires additional time provided by the spindle assembly checkpoint. Once those chromosomes have co-oriented at the metaphase plate, the achiasmate homologs begin a slow separation and progression to opposite poles. By metaphase arrest, when the achiasmate chromosomes are still balanced on the metaphase plate, achiasmate chromosomes are often symmetrically positioned between the plate and the poles along the same arc of the metaphase spindle. This movement and position requires the functioning of the Nod kinesin-like protein (see Section seven below).

5. The co-orientation of achiasmate chromosomes requires additional time during meiotic prometaphase — time that is provided by Ald/Mps1 and the Spindle Assembly Checkpoint (Bill Gilliland, American Cancer Society Post-doctoral Fellow)

     Our recent ability to visualize meiosis in living oocytes has revealed that the achiasmate chromosome segregation system is a slower system in terms of positioning the bivalents on the metaphase plate during meiosis I. Chiasmate bivalents appear to co-orient immediately while achiasmate bivalents require nearly 30-40 minutes to stabilize at the metaphase plate. Again this is explained by our hypothesis that exchanges function, via the SC, to predispose centromeres to co-orient long before nuclear envelope breakdown. In the absence of such proper exchanges achiasmate ‘bivalents’ fail to co-orient and thus must undergo cycles of mal-orientation and release prior to finally establishing proper co-orientation.

     Obviously, this extended time required for proper co-orientation requires delaying the events of metaphase (see discussion of Nod below) until proper achiasmate co-orientation can be achieved. We have recently shown that this delay requires the spindle assembly checkpoint protein Mps1. Bill’s first project in the lab was to clone a meiotic mutant, ald, which is a single nucleotide change in the Drosophila homolog of mps1. Mps1 is a component of the spindle checkpoint, and loss of this protein results in early separation of sister chromatid cohesion and precocious entry into anaphase. In the absence of an extended prometaphase, achiasmate chromosomes fail to properly co-orient and non-exchange X chromosomes often segregate at random. We are now exploring the role of other checkpoint proteins in the process of prometaphase extension and achiasmate segregation.

6. The mtrm gene encodes a protein that interacts with Polo and acts to ensure achiasmate centromere co-orientation in Meiosis I (Youbin Xiang)

     Mutants in the matrimony (mtrm) gene exhibit dosage-sensitive effects on achiasmate segregation in Drosophila oocytes, as evidenced by high levels of achiasmate nondisjunction in mtrm heterozygotes. Fluorescent in-situ hybridization showed that achiasmate nondisjunction in mtrm heterozygotes results from a defect in the ability to ensure proper centromere co-orientation at metaphase. While the deleterious effects of reducing the dose of Mtrm protein by 50% appear to be restricted to impairing achiasmate segregation, complete loss of Mtrm protein results in defects in both the control of pre-meiotic nuclear division and in proper spindle assembly during meiotic prometaphase. Consistent with multiple roles in the control of meiotic and mitotic cell cycles, Mtrm protein interacts with Polo kinase. The polo alleles strongly suppress the meiotic defects in mtrm/+ heterozygotes. Over-expression of Polo in mtrm heterozygotes leads to female sterility. Mtrm is co-immunoprecipitated with Polo, indicating that a physical interaction exists between them. A mutation in one of the two Polo Box Domain binding sites of Mtrm disrupts this physical interaction and ablates Mtrm function. Exploring the function of Mtrm and its interaction with Polo is the major goal of Youbin’s research efforts.

7. The role of the Nod protein is to position achiasmate chromosomes during late prometaphase and metaphase (Li-Jun Huo, Kim Collins)

     Following co-orientation of achiasmate bivalents, their proper separation and movement to opposite poles of the spindle requires the Nod protein. We have recently demonstrated by both genetic and cytological means that the Nod protein plays a crucial role in balancing the poleward forces acting on achiasmate centromeres (Matthies et al., 1999). In collaboration with Lisa Sproul and Susan Gilbert at the University of Pittsburgh, Wei Cui and Heiner Matthies, former post-docs in the lab, showed that Nod is not a motor. Rather Nod, which is localized along the arms of meiotic chromosomes, binds to the plus ends of microtubules and stimulates microtubule polymerization. In doing so, Nod serves to “push” chromosome arms back towards the metaphase plate and countering the pole forces acting at the centromeres (Cui et al., 2005a and 2005b)

     Kim Collins is continuing our structure function analysis of Nod with a specific focus on identifying the domains required to activate and de-activate this protein in meiosis. She will also pursue a proteomics-based approach towards identifying proteins that interact with Nod as well as searching for those DNA-sequences which facilitate Nod binding. Using a genetic approach to identify Nod-interactors, Li-Jun Huo has begun a clever screen for mutants that enhance or suppress the dominant nod allele nod^DTW when it is over-expressed in the eye. GAL4-dependent over-expression of nod^DTW in the developing eye creates a strong reduced and rough eye phenotype, and deficiencies lines from the Bloomington Stock Center and the Exelexis Deficiency Collection are being screened for their capacity to dominantly enhance or suppress the eye defects caused by over-expression of nod^DTW. Candidate P insertions can also easily be identified in a simple visual F1 screen and then tested quickly for relevance to the meiotic phenotype. Because the enhancing and suppressing mutants were created by P insertion, it will be straightforward to identify the disrupted genes. The synergy between the molecular approaches employed by Kim and Li-Jun’s genetic screens seem likely to quickly identify the proteins that regulate this novel and important meiotic component.

8. Axs and the sheath-like structure that surrounds the meiotic spindle (Stacie Hughes)

     A dominant mutation in Aberrant X segregation (Axs^D) was isolated which increased achiasmate chromosome nondisjunction during female fly meiosis without affecting the incidence of nondisjunction of chiasmate chromosomes. The Axs gene was found to encode a member of a new family of transmembrane proteins that includes members in all major kingdoms except Monera (Kramer et al., 2003). The 8 transmembrane domain protein encoded by the Axs gene localizes to the endoplasmic reticulum until germinal vesicle breakdown. Axs then localizes to a novel membranous sheath encasing the meiotic spindle and chromosomes. The Axs^D mutation also causes defects in meiotic chromosome positioning, spindle structure and increases the incidence of anaphase. These phenotypes suggest that the membranous sheath is important for holding the chromosomes in place during meiosis, for focusing the microtubules on the developing spindle, and for controlling the meiotic cell cycle.

     Future research questions to be pursued include where does the sheath surrounding the meiotic spindle come from and what are other components of this sheath? While Axs is recruited from the endoplasmic reticulum the only other protein found to colocalize with Axs to this sheath is Myt-1, a regulator of mitotic and meiotic entry, that has a Golgi localization signal. Biochemical and candidate approaches should yield additional components of the sheath and suggest from where the components are recruited. The signaling and meiotic proteins Axs functionally interacts with during meiosis may be elucidated through a screen looking for genes that suppress the AxsD phenotypes when over-expressed

9. The very essence of our lab — screens for new meiotic mutants (Susan Flynn, Jenny Griffiths, Rachel Nielsen, Kathy Teeter, Diana Hiebert, and, when necessary, everybody else).

     A large number of mutants that affect meiosis have been discovered through genetic screens in Drosophila. Unfortunately, nearly all of those screens exclude mutants in essential genes because of the requirement for testing viable and fertile homozygotes. We have performed a genetic screen designed to recover mutants whose protein products are involved in female meiosis, including mutants in essential genes. Our approach utilized the FLP-DFS system to generate and select for homozygous mutant germline clones that were assayed for meiotic nondisjunction. We screened over 20,000 EMS treated chromosome arms for 2L and over 25,000 EMS treated chromosome arms for 3R. Twelve mutants were recovered; two on 2L and 10 on 3R. The two mutants on 2L were viable and found to be alleles of c(2)M, a previously characterized meiotic mutant. Of the 10 mutants on 3R, one is lethal, one is semi-lethal and the remaining mutants are homozygous viable. All but three mutants have been mapped to a specific gene (alleles of mcm5, corona, and ald which were recovered in this screen are described above). In addition, we have identified a cold-sensitive recombination mutant that we have named trade embargo (trem), which is an allele of what appears to be an essential gene, CG4413. This mutant is defective both for DSB formation and for crossover production. Surprisingly, of the nine mutants that have been mapped to a specific gene on either 2L or 3R, five are the result of a Doc retrotransposon insertion into or near the gene of interest, while only three are the result of a point mutation. We are now extending this screen to 2R and 3L.

Academic Appointments: Professor of Molecular and Integrative Physiology, School of Medicine, University of Kansas Medical Center; Adjunct Professor of Biological Sciences at the University of Missouri Kansas City; Adjunct Professor of Undergraduate Program in Biology, The University of Kansas