Tumbleweed/RacGAP50C: Biological Overview | Evolutionary Homologs | Regulation | Developmental Biology | Effects of Mutation | References
Gene name - Tumbleweed

Synonyms - DRacGAP, FlyBase name: tumbleweed

Cytological map position - 50C4--6

Function - signaling

Keywords - cytokinesis, dorsal closure, axonogenesis

Symbol - tum

FlyBase ID: FBgn0086356

Genetic map position - 2R

Classification - GAP domain in GAPs for Rho-family GTPases, diacylglycerol binding

Cellular location - nuclear and cytoplasmic

NCBI link: Entrez Gene
tum orthologs: Biolitmine
Recent literature
Nakamura, M., Verboon, J. M., Prentiss, C. L. and Parkhurst, S. M. (2020). The kinesin-like protein Pavarotti functions noncanonically to regulate actin dynamics. J Cell Biol 219(9). PubMed ID: 32673395
Pavarotti, the Drosophila MKLP1 orthologue, is a kinesin-like protein that works with Tumbleweed (MgcRacGAP) as the centralspindlin complex. This complex is essential for cytokinesis, where it helps to organize the contractile actomyosin ring at the equator of dividing cells by activating the RhoGEF Pebble. Actomyosin rings also function as the driving force during cell wound repair. Previous work has shown that Tumbleweed and Pebble are required for the cell wound repair process. This study shows that Pavarotti also functions during wound repair and confirm that while Pavarotti, Tumbleweed, and Pebble are all used during this cellular repair, each has a unique localization pattern and knockdown phenotype, demonstrating centralspindlin-independent functions. Surprisingly, it was found that the classically microtubule-associated Pavarotti binds directly to actin in vitro and in vivo and has a noncanonical role directly regulating actin dynamics. Finally, this actin regulation by Pavarotti was shown to not be specific to cellular wound repair but is also used in normal development.

Cytokinesis, the final step in cell division, involves the formation and constriction of an actomyosin-based contractile ring. The mechanism that positions the contractile ring is unknown, but derives from the spindle midzone. An interaction between Pebble [a Rho GTP exchange factor (GEF)], and the Rho family GTPase-activating protein, RacGAP50C, has been shown to connect the contractile ring to cortical microtubules at the site of furrowing in D. melanogaster cells. Pebble regulates actomyosin organization, while RacGAP50C and its binding partner, the Pavarotti kinesin-like protein, regulate microtubule bundling. All three factors are required for cytokinesis. As furrowing begins, these proteins colocalize to a cortical equatorial ring. It is proposed that RacGAP50C-Pavarotti complexes travel on cortical microtubules to the cell equator, where they associate with the Pebble RhoGEF to position contractile ring formation and coordinate F-actin and microtubule remodeling during cytokinesis (Somers, 2003).

RacGap50C has also been shown to negatively regulate the wingless pathway during Drosophila embryonic development. In the ventral epidermis of Drosophila embryos, Wg specifies cells to secrete a characteristic pattern of denticles and naked cuticle that decorate the larval cuticle at the end of embryonic development. The Drosophila ventral epidermis was used as an assay system in a series of genetic screens to identify new components involved in Wg signaling. Two mutant lines that modify wg-mediated epidermal patterning represent the first loss-of-function mutations in the RacGap50C gene. On their own, these mutations cause increased stabilization of Armadillo and cuticle pattern disruptions that include replacement of ventral denticles with naked cuticle; this suggests that the mutant embryos suffer from ectopic Wg pathway activation. In addition, RacGap50C mutations interact genetically with naked cuticle and Axin, known negative regulators of the Wg pathway. These phenotypes suggest that the RacGap50C gene product participates in the negative regulation of Wg pathway activity (Jones, 2005).

Finally, RacGAP50C corresponds to the tumbleweed (tum) gene previously identified based on its defects in dendrite development of sensory neurons (Gao, 1999). Using mushroom body neurogenesis and morphogenesis as a model, Tumbleweed (Tum), Pavarotti (Pav), and their association are shown to be required for neuroblast proliferation. Tum with a mutation predicted to disrupt the GTPase-activating protein (GAP) activity still largely retains its activity in regulating cell division but is impaired in its activity to limit axon growth. Tum and Pavarotti regulate the subcellular localization of one another in postmitotic neurons, and cytoplasmic accumulation of both proteins disrupts axon development in a GAP-dependent manner. Taken together with previous studies of RacGAP50C in regulating cytokinesis, it is proposed that Tum serves as a scaffolding protein in regulating cell division but acts as a GAP to limit axon growth in postmitotic neurons (Goldstein, 2005).

In Drosophila melanogaster embryonic epithelial cells, constriction occurs during anaphase B and telophase to generate two daughter cells, each containing one set of the recently separated sister chromatids. Constriction of the cleavage furrow proceeds through the activity of the myosin II motor protein acting on an F-actin network. Members of the Rho subfamily of small G proteins are potent regulators of the actin cytoskeleton in a variety of contexts. Like all small G proteins, Rho1 is active when GTP is bound and inactive when GDP is bound. Activation is mediated by guanine nucleotide exchange factors (GEFs) that catalyze the displacement of GDP and the uptake of GTP, whereas inactivation is regulated by GTPase-activating proteins (GAPs) that stimulate the intrinsic GTPase activity of the G protein (Somers, 2003).

Molecular and genetic studies have shown that the D. melanogaster RhoGEF, Pebble (PBL), and its mammalian ortholog, the protooncogene ECT2, are required for cytokinesis. pbl mutant embryos proceed normally through the first 13 syncytial mitotic cycles following fertilization and cellularize normally during G2 phase of cycle 14, but they fail to undergo cytokinesis during the fourteenth and subsequent division cycles. Pbl binds to Rho1, but not Rac1 or Cdc42, and sensitized pbl mutant alleles show strong genetic interactions with Rho1 but not Rac1 or Cdc42 alleles. During cytokinesis in epithelial cells of the embryo, Pbl accumulates in the contractile ring during furrowing, where it appears to stimulate Rho1-mediated organization and activity of the actomyosin contractile ring (Somers, 2003).

Reorganization of the actomyosin contractile apparatus occurs coincident with reorganization of the microtubule network. During anaphase, the mitotic spindle is remodeled to form a midzone bundled microtubule structure referred to as the central spindle, which is further compacted into a late cytokinetic structure termed the midbody. Curiously, another regulator of Rho family G protein activity, the Caenorhabditis elegans CYK-4 GAP, is required for microtubule bundling, because microtubule reorganization fails in cyk-4 mutant embryos. It also fails in embryos mutant for the zen-4/CeMKLP1 gene, which encodes a kinesin-like protein that forms a complex with CYK-4. This complex has been shown to bundle microtubules in vitro. The CYK-4 and ZEN-4 proteins and their respective mammalian orthologs localize to the central spindle and are all essential for cytokinesis, as is Pavarotti (Pav), the D. melanogaster ortholog of ZEN-4 (Somers, 2003).

It is not known how remodeling of the microtubule and F-actin networks is coordinated during cytokinesis. Although initial studies focused on their role in F-actin remodeling, recent studies have now linked Rho family members to microtubule organization. For example, depolymerization of microtubules results in an increase in the amount of active RhoA and the formation of contractile actin bundles, while microtubule polymerization results in an increase in the amount of active Rac1 and the formation of lamellipodia. RhoA can also mediate selective microtubule stabilization, while the Rac1/Cdc42 effector PAK is capable of activating the microtubule destabilizer Stathmin. It is possible, therefore, that Rho family members play roles in both F-actin and microtubule organization during dynamic processes such as cytokinesis. An important but poorly understood aspect of the relationship between the microtubule and F-actin networks is the nature of the signal that positions the contractile ring and initiates furrowing. It is now generally accepted that the signal originates from the midzone of the anaphase microtubule network, although the nature of the stimulus is unknown (Somers, 2003).

A complex has been identified between two Rho family regulators, the RhoGEF Pbl and RacGAP50C, the D. melanogaster ortholog of the CYK-4 Rho family GAP. A ring of RacGAP50C and Pav, associated with cortical microtubules, colocalizes with Pbl in dividing embryonic epithelial cells, forming a link between the actomyosin and microtubule networks. These observations suggest a molecular model for contractile ring positioning and function whereby RacGAP50C-Pav complexes, positioned at the equatorial cortex of the cell by their association with microtubules, interact with cortical Pbl to activate Rho1, initiate formation of the contractile ring, and coordinate F-actin and microtubule dynamics during furrowing (Somers, 2003).

Thus, in Drosophila embryonic epithelial cells at the onset of cytokinesis, the two Rho family regulators are part of a cortical double-ring structure at the site of cleavage furrowing. The RacGAP50C ring is associated with cortical microtubules, presumably through its interaction with the Pavarotti kinesin-like protein. Pav colocalizes with RacGAP50C, and coimmunoprecipitation experiments have shown that they form a complex in vivo. The RacGAP50C-Pav ring appears to abut or overlap the Pbl-containing contractile ring. The Pav kinesin-like protein, RacGAP50C and Pbl RhoGEF form a trimolecular complex that simultaneously associates with, and has the capacity to control, both the actin and microtubule cytoskeletons as they are remodeled during cytokinesis. Furthermore, this complex appears to be a conserved feature of animal cytokinesis, since the mammalian Pbl and RacGAP50C orthologs, the protooncogene ECT2 and MgcRacGAP, bind to each other in a yeast two-hybrid assay and colocalize during mitosis (Somers, 2003).

The interaction between Pbl and RacGAP50C occurs through an extended BRCT domain of Pbl and an N-terminal coiled-coil domain of RacGAP50C. RacGAP50C binds Pav through sequences adjacent to the Pbl-interacting domain, indicating the presence, in Drosophila, of the so-called centralspindlin complex (Mishima, 2002) first identified from analysis of CYK-4 and ZEN-4, the C. elegans RacGAP50C and Pav orthologs (Somers, 2003).

RacGAP50C-Pav complexes were found to be cytoplasmic at prophase, associated with mitotic spindles during metaphase, concentrated in the spindle midzone during anaphase, and localized to the midbody at cytokinesis and to the nucleus during interphase. During late anaphase and early telophase in Drosophila epithelial cells, RacGAP50C-Pav complexes not only localize to the overlapping microtubules of the centrally located anaphase spindle, but also to distinct cortical microtubules. Cortical microtubules have been reported in dividing Drosophila neuroblasts (Savoian, 2002), and they can be seen in all D. melanogaster anaphase cells examined. Localization of the RacGAP50C-Pav complexes to the microtubule midzone is independent of its interaction with Pbl, since RacGAP50C is found to localize appropriately in Pbl mutant cells. However, localization of RacGAP50C is dependent on the Pav kinesin-like protein. The affinity of the RacGAP-KLP complex for microtubules, the cortical localization of the microtubules, and the plus end-directed nature of the Pav kinesin-like motor protein appears sufficient to account for localization of the complex to an equatorial cortical ring (Somers, 2003).

The molecular signal that positions the contractile ring and initiates furrowing remains to be elucidated. A number of studies have shown that the signal derives from the overlapping midzone microtubules that form during anaphase. One of the most striking aspects of the formation of the Pbl-RacGAP50C ring is that it is present in the earliest examples of furrowing that were observed. The existence of this ring at the onset of cytokinesis suggests a molecular model for the positioning and regulation of the contractile ring. Specifically, it is proposed that the initiation signal corresponds to the microtubule-mediated arrival of the RacGAP50C-Pav kinesin-like protein complex at its equatorial ring and establishment of the interaction with the Pbl RhoGEF. It is proposed that this interaction results in activation of RhoGEF activity. Rho1 would then be activated to initiate contractile ring formation and furrowing through activation of factors such as Diaphanous and myosin. This model accounts for the role of microtubules in positioning the contractile ring, because microtubules deliver the RacGAP50C-Pav complexes to their interaction site with Pbl. It also accounts for the conclusion, made by Gatti and colleagues from their studies of cytokinesis, that there is a requirement for both the central spindle and a cortical Pbl-containing apparatus for the onset of cytokinesis (see Somma, 2002; Somers, 2003).

There is strong support for this model beyond the evidence described here. Importantly, consistent with the observation that DRacGAP50C RNAi-treated S2 cells show no furrowing, pav and pbl mutant cells fail to form a contractile ring and do not undergo furrow ingression. However, contradictory evidence has come from C. elegans, where cyk-4 and zen-4 mutant cells initiate but fail to complete furrowing. Two possible explanations are suggested for these contradictory observations. The first is that D. melanogaster epithelial cells may use a different cytokinesis mechanism than that used in the early C. elegans embryo. In support of this, the cell types are very different in size and exhibit differences in microtubule organization during anaphase and telophase. Alternatively, it is possible that the C. elegans phenotypes do not represent the true null phenotype. The cyk-4 allele used to determine the phenotype is a temperature-sensitive allele, which may not abolish all activity at the restrictive temperature. The zen-4 allele used to generate germline mutant clones is a premature truncation that would eliminate all function. However, to observe the phenotype, germline clones were generated, perhaps requiring zen-4 activity to undergo the previous division. Some of the product may therefore have persisted to produce the partial furrowing observed in the mutant embryos. It therefore remains to be seen whether the model proposed here is applicable to cytokinesis in all animal cells (Somers, 2003 and references therein).

Midzone microtubule bundles have been shown to be required continuously for cytokinesis in cultured cells. The cortical Pbl-RacGAP50C-Pav ring, which persists and narrows as cytokinesis proceeds, is ideally positioned to coordinate actomyosin contraction and the bundling of microtubules. Actin filament activity is regulated by Pbl, which is required for establishment and/or maintenance of the contractile ring through activation of the Rho1 GTPase. Microtubule bundling activity has been demonstrated for CYK-4 and ZEN-4. It is likely, therefore, that the complex between the Pbl RhoGEF, RacGAP50C and the Pav kinesin-like protein functions to coordinate F-actin and microtubule remodeling during contractile ring constriction (Somers, 2003).

While actomyosin regulation and microtubule bundling may constitute the primary regulatory roles of these factors, there are additional ways that the Pbl RhoGEF and RacGAP50C could influence both the actin and microtubule-based cytoskeleton. Rho downstream effectors have been shown to regulate both cytoskeletal systems. For example, the Rho1 target, Diaphanous, mediates actin reorganization but also affects the stability of microtubules (Somers, 2003).

The CYK-4 and ZEN-4 microtubule bundling activity does not require the presence of any of the small G proteins, but the site-directed mutant analysis described in this study suggests a requirement for the GTPase-activating domain of RacGAP50C. Consistent with this, a GAP domain-defective form of MgcRacGAP appears to act as a dominant-negative protein, inducing cytokinetic defects. If such a target of RacGAP50C GAP activity exists, it has still not been identified. However, the evidence is inconsistent with RacGAP50C acting as the Rho1 GAP that opposes Pbl, based on the synergistic nature of pbl and RacGAP50C genetic interactions and on the absence of genetic interactions between RacGAP50C and Rho1. Consistent with this, in vitro assays show that the CYK-4 and MgcRacGAP homologs target Rac and Cdc42 with far greater efficiency than Rho1 (Somers, 2003 and references therein).

Ths study has identified complexes between the RhoGEF Pbl and the Rho family GAP, RacGAP50C, and between RacGAP50C and the kinesin-like protein, Pav, that connect the contractile ring to cortical microtubules during cytokinesis. During late stages in anaphase and during telophase, these proteins localize to a cortical ring where furrowing is initiated, constricting as furrowing proceeds. These observations suggest a model for the molecular control of cytokinesis in animal cells, whereby microtubule-dependent cortical equatorial localization of RacGAP50C-Pav kinesin-like protein complexes is the positioning signal generated by the central spindle microtubules, and formation of complexes with the Pbl RhoGEF allows coordination of F-actin and microtubule remodeling (Somers, 2003).

Peripheral astral microtubules ensure asymmetric furrow positioning in neural stem cells

Neuroblast division is characterized by asymmetric positioning of the cleavage furrow, resulting in a large difference in size between the future daughter cells. In animal cells, furrow placement and assembly are governed by centralspindlin (Pavarotti and Tumbleweed) that accumulates at the equatorial cell cortex of the future cleavage site and at the spindle midzone. In neuroblasts, these two centralspindlin populations are spatially and temporally separated. A leading pool is located at the basal cleavage site and a second pool accumulates at the midzone before traveling to the cleavage site. The cortical centralspindlin population requires peripheral astral microtubules and the chromosome passenger complex for efficient recruitment. Loss of this pool does not prevent cytokinesis but enhances centralspindlin signaling at the midzone, leading to equatorial furrow repositioning and decreased size asymmetry. These data show that basal furrow positioning in neuroblasts results from a competition between different centralspindlin pools in which the cortical pool is dominant (Thomas, 2021).

Asymmetric cell division is a robust process that ensures that two daughter cells inherit different fates and sizes. The Drosophila NB is a powerful and widely used model system to study this specialized form of division. Rhis study challenges asymmetric cell division by modifying MT growth dynamics. It was possible to increase mitotic spindle length using overexpression of MT-polymerizing MAPs (Msps and Ensconsin), as well as by RNAi-mediated depletion of Klp67A, a member of the kinesin-8 family of MT-depolymerizing kinesins. Despite the presence of long and bent mitotic spindles under these conditions, the NB cell size ratio remained unchanged relative to control NBs. This reveals that asymmetric cell division and asymmetric positioning of the cleavage furrow are resistant to an excess of abnormally long and stable MTs during cell division. By contrast, decreasing MT stability and shortening the mitotic spindle produced more symmetric cell divisions. This change was due to an apical shift of the cleavage furrow during its ingression, following apical and basal cortex expansion. This phenotype was not MAP dependent and was observed following overexpression of either Klp10A (kinesin-13) or Klp67A (kinesin-8) MT depolymerases and in ensc mutants. Rather, the data suggest that spindle size or interference with MT dynamics is responsible for the phenotype. In agreement with this, spindle size is restored in Klp10A-OE telophase cells, which display the shortest spindles at metaphase. Interestingly, sas-4s2214 mutants, which are reported to lack functional centrosomes and thus astral MTs, yielded reduced levels of cell size asymmetry despite harboring longer metaphase spindles. This suggests that MT-asters and not spindle length are the key determinant factor for size asymmetry in NBs. Consistent with this, loss of either apical or basal MT-asters, through targeted laser irradiation and ablation prior to anaphase onset, also reduced sibling cell size asymmetry. Together, these results strongly suggest that astral MTs are required to maintain a cleavage site, which normally favors a basal position in the fly NB. It is proposed that a specific population of these astral MTs, called peripheral MTs, is positioned in direct contact with the division furrow and plays a determining role in maintaining its stable position during anaphase until cytokinesis. For technical reasons, it was not possible to quantify peripheral MT bundles in live dividing cells. However, fixed-cell analyses, despite a possible bias in the determination of late-anaphase substages, support this and revealed a significant decrease in peripheral MTs in ensc, Klp10A-OE, and Klp67A-OE NBs. The results are in accord with reports indicating that a subpopulation of these stable astral MTs plays a key role in the initiation of furrowing in symmetrically dividing cells and that, in some systems, furrowing can occur without the presence of a stable central spindle. However, in contrast to previous studies, the data reveal that in asymmetrically dividing control NBs, the astral MT-furrowing pathway dominates over the midzone pathway. Prior investigations indicated that NBs have two genetically separable pathways to drive cytokinesis. The first, the polarity-dependent pathway, triggers the clearing of apical myosin, resulting in apical cortical expansion. Interference with this pathway leads to simultaneous apical and basal clearing, symmetrical cortex expansion, equatorial furrow positioning, and symmetric division. The second, the spindle pathway, is proposed to rely on the spindle midzone and the CPC. This triggers the subsequent basal myosin clearing and basal cortical expansion. Several of the results presented in this study have led to the proposal of another mechanism for furrow positioning that would rely on peripheral astral MTs with a minor contribution from the spindle midzone. This is supported by several observations. (1) Live-cell imaging and analyses utilizing GFP-Pav-klp as a marker of centralspindlin position revealed that this master controller of cytokinesis accumulated at the basal cortex throughout the entire furrow ingression process. (2) Centralspindlin levels were low at the midzone during furrow placement and ingression compared to the cortex. (3) It was consistently found that the midzone, as defined independently using both GFP-Pav-klp and Feo-GFP, was spatially independent from the furrowing site. Moreover, inhibition of midzone formation through Feo depletion did not impair furrow positioning but did interfere with the late stages of cytokinesis. (4) The midzone consistently relocated from an initial location to a final position that was coincident with the furrow. The converse was never observed, confirming previous observations made in embryonic NBs. (5) Finally, genetic or photo-based interference with centrosomes precluded astral MT formation and interaction with the cortex. Accordingly, the cortical centralspindlin pool was diminished and NBs exhibited a size asymmetry defect (Thomas, 2021).

The localization studies suggest that, under normal conditions, midzone-associated centralspindlin does not perform a key role in positioning of the cleavage site and that this function is served by the more abundant centralspindlin pool associated with the cortex at the cleavage site. When peripheral astral MTs were impaired, centralspindlin enrichment at the furrow was often diminished, and in some cases was accompanied by an increase in the midzone-associated pool, leading to a decreased midzone/furrow centralspindlin ratio and a reset of the furrowing toward the equatorial midzone. This indicates that the two populations of centralspindlin are competent to signal furrowing but that the cortical pool delivered by astral MTs may be dominant. Thus, the spatial localization and the cortical/midzone ratio of centralspindlin are the pivotal determinants of final furrow position in the Drosophila NB. Interestingly, a recent study has shown that a similar competition between centralspindlin pools also occurs in human cells, revealing an evolutionary conservation of the mechanism (Adriaans, 2019). As with human cells, It was found that CPC activity seems essential in this regulatory event (Thomas, 2021).

In contrast to a recent study in the symmetrically dividing S2 cells, this study do not observe GFP-Pav-klp labeling at the plus ends of astral MTs even when studied by enhanced-resolution imaging methods. Instead, it was consistently found that centralspindlin coats the entire length of astral MTs emanating from both centrosomes, suggesting that the plus-end-directed motor activity of Pav-klp is used to bring centralspindlin to the furrow in Drosophila NBs, similar to findings in early embryos. It is therefore likely that centralspindlin, depending on the cell type, utilizes preferentially EB1-mediated MT plus ends or the motor activity of Pav-Klp to reach the cleavage site (Thomas, 2021).

In total, these data suggest a model in which competition between different centralspindlin populations is a key determinant of asymmetric division in Drosophila NBs. The consecutive actions of the polarity-dependent cleavage-furrow-positioning pathway and the MTs emanating from the asters serving as centralspindlin delivery arrays are essential in the whole process. In this system, it is proposed that the ability of the spindle midzone to define furrow and cleavage location may only become engaged during late telophase or after subcortical astral MTs are compromised. Despite their clear role in governing size asymmetry, it was not possible to induce complete daughter cell size equality through any of a host of MT-perturbing treatments. It is possible that the few MTs that remain after the perturbations are sufficient to target enough cortical centralspindlin to provide some degree of asymmetry. However, additional mechanisms, such as MT initial asymmetric midzone position and displacement toward the cleavage site, also appear important to secure a minimal level of asymmetry in these cells. Elucidating these systems and their advantages for asymmetrically dividing stem cells will be important directions for future investigations on tissue homeostasis (Thomas, 2021).

This study reveals a critical role for astral MTs in maintaining the asymmetric position of the cleavage site during cytokinesis in neural stem cells. This is essential to preserve an appropriate NB/GMC size ratio. The perturbation of peripheral astral MTs and the subsequent loss of asymmetry occur in the presence of a timely apical myosin clearing, an event regulated by the polarity-dependent pathway that controls apical cortical expansion during anaphase. However, it is not possible to fully rule out that the observations are completely polarity independent. Other additional yet uncharacterized polarity-dependent mechanisms could be involved to contribute to the findings. Indeed, it remains possible that polarity proteins participate in the regulation of other components including centrosomal or MT-associated proteins, to control the dynamics of these peripheral astral MTs and ultimately basal furrow positioning (Thomas, 2021).


cDNA clone length - 2317

Bases in 5' UTR - 159

Exons - 4

Bases in 3' UTR - 281


Amino Acids - 625

Structural Domains

DRacGAP encodes a protein of 625 amino acids whose sequence is closely related to RacGAPs from different organisms. This protein contains a domain (residues 395-537) including the three conserved amino acid blocks and the arginine finger that define the GAP domain in GAPs for Rho-family GTPases. It also presents two motifs usually found in these proteins -- a proline rich region, putative recognition motif for SH3-domain containing proteins, and a cysteine-rich domain, similar to the diacylglycerol-binding domain of protein kinase C. The putative GAP domain is most similar to those of Drosophila RnRacGAP, a mouse RacGAP and human n-chimaerin, which is a GAP specific for Rac GTPase (Sotillos, 2000).

RacGAP50C: Evolutionary Homologs | Regulation | Developmental Biology | Effects of Mutation | References

date revised: 3 November 2005

Home page: The Interactive Fly © 2006 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.