string: Biological Overview | Evolutionary Homologs | Regulation | Developmental Biology | Effects of Mutation and Overexpression | References

Gene name - string

Synonyms -

Cytological map position - 99A5-6

Function - protein tyrosine phosphatase

Keywords - cell cycle

Symbol - stg

FlyBase ID:FBgn0003525

Genetic map position - 3-[100]

Classification - cdc25 homolog

Cellular location - nuclear

NCBI links: Precomputed BLAST | Entrez Gene
Recent literature
Wang, P., Chen, Y., Li, C., Zhao, R., Wang, F., Lin, X., Cao, L., Li, S., Hu, L., Gao, Y., Li, Y. and Wu, S. (2015). Drosophila eye developmental defect caused by elevated Lmx1a activity is reliant on chip expression. Biochem Biophys Res Commun. PubMed ID: 26718403
The LIM-homeodomain (LIM-HD) family member Lmx1a has been successfully used to induce dopaminergic neurons from other cell types, thus showing significant implications in replacement therapies of Parkinson's disease, but the underlying mechanism remains elusive. This study used Drosophila eye as a model system to investigate how forced expression of dLmx1a (CG4328) and dLmx1a (CG32105), the fly homologs of human Lmx1a, alters cell identify. Ectopic expression of dLmx1a suppresses the formation of Drosophila eye tissue; the LIM and HD were found to be two essential domains. dLmx1a requires and physically binds to Chip, a well-known cofactor of LIM-HD proteins. Chip connects two dLmx1a proteins to form a functional tetrameric complex. In addition, evidence is provided showing that dLmx1a expression results in the suppression of two retina determination gene eyes absent (eya) and string (stg). Taken together, these findings identified Chip as a novel partner of dLmx1a to alter cell differentiation in Drosophila eye through repressing eya and stg expression, and provide an animal model for further understanding the molecular mechanism whereby Lmx1a determines cell fate.
Momen-Roknabadi, A., Di Talia, S. and Wieschaus, E. (2016). Transcriptional timers regulating mitosis in early Drosophila embryos. Cell Rep 16: 2793-2801. PubMed ID: 27626650
The development of an embryo requires precise spatiotemporal regulation of cellular processes. During Drosophila gastrulation, a precise temporal pattern of cell division is encoded through transcriptional regulation of cdc25string in 25 distinct mitotic domains. Using a genetic screen, it was demonstrated that the same transcription factors that regulate the spatial pattern of cdc25string transcription encode its temporal activation. buttonhead and empty spiracles were identified as the major activators of cdc25string expression in mitotic domain 2. The effect of these activators is balanced through repression by hairy, sloppy paired 1, and huckebein. Within the mitotic domain, temporal precision of mitosis is robust and unaffected by changing dosage of rate-limiting transcriptional factors. However, precision can be disrupted by altering the levels of the two activators or two repressors. It is proposed that the additive and balanced action of activators and repressors is a general strategy for precise temporal regulation of cellular transitions during development.

String protein regulates three cycles of cell division immediately following the formation of the cellular blastoderm: cell cycles 14, 15 and 16. Bursts of String transcription are both required and sufficient to trigger mitosis during these cycles. String activates mitosis by removing phosphate groups from cdc2, a cyclin dependent kinase that forms heterodimers with Cyclin A and Cyclin B. Cdc2 is held inactive in the phosphorylated state by phosphorylation of tyrosine 15, an invarient residue, one that is highly conserved in diverse organisms from yeasts to mammals. In other words, String is a phosphorylase, acting as a critical regulator of activity for the cyclin/ckc2 dimers responsible for driving cells into mitosis. How does String function to regulate postblastoderm mitosis?

The transcription of string, and hence the timing and pattern of mitosis in the postblastoderm embryo, is under complex developmental control. The postblastoderm embryo is divided into mitotic domains, each domain composed of a group of neighboring cells. Each such group follows the same within-group mitotic timing sequence, although the timing between groups varies. After the degradation of maternal String mRNA transcripts during early interphase 14, expression of string occurs in a sequence of brief pulses, timed differently in different regions of the embryo. The order of the appearance of String mRNA generally corresponds to the order of mitoses. What regulates the predictable, yet complex expression of string in postblastoderm mitotic cells?

In some mutants, such as twist, snail and buttonhead, string expression is completely deleted in the specific domain that corresponds to the normal spatial and temporal expression of the mutant gene. This is consistent with the notion of direct regulation. In other mutants, such as bicoid, hunchback and Krüppel, string expression patterns are not deleted, but are globally distorted. This suggests indirect, combinatorial, or concentration-dependent regulation. For example, the pair-rule periodicity of string expression in the lateral epidermis is not significantly affected in pair-rule mutants, but is altered in gap gene mutants. Expression in the dorsal ectoderm is affected by mutations in gap genes as well, in a similar fashion to pair rule gene regulation. Expression of string in mesectoderm is precisely coincident with expression of single-minded. However, string expression is unaffected by sim mutation. Presumably string and sim are regulated independently, in parallel, by a similar mechanism, that is by combinations of broadly distributed dorsoventral pattern gene products. There are also examples of indirect affects. dpp regulates the spatial patterns of twist and zerknüllt, two transcription factors shown to alter string expression.

Activation of string expression is independent of cell cycle progression. Arrest of cell cycle progression achieved by various cyclin mutations causes few perturbations in the dynamics of string transcription following arrest. Thus, like a number of DNA synthesis genes expressed at the G1 to S transition, string does not behave like a true 'cell cycle-regulated' gene in vivo. However, string activity, or its consequence (cdc2 activation and mitosis), contributes to the shut-off of string transcription at the close of cycle 14, but such effects are not uniformly essential. For example, in embryos arrested in G2 of cell cycle 16 by Cyclin A mutants, the shut off of string expression after arrest is essentially normal, as is continued expression in the brain and CNS. Thus cell cycle is not an obligatory factor in string regulation.

Mitosis in most Drosophila cells is triggered by brief bursts of transcription of string (stg), a Cdc25-type phosphatase that activates the mitotic kinase, Cdk1 (Cdc2). Promoter analyse defines four string position specific elements that drive transcription in distinct sets of cells: one drives mesoderm expression, a second drives early expression in ventral neuroectoderm, a third contains elements that act in a number of cell types in the head, the nervous system and the trachea and a fourth is inferred to drive expression in lateral epidermis, mesectoderm, tail, head and the ventral neurogenic region. Thus string is subject to position-specific regulation in much the same way that achaete and even-skipped are regulated (Edgar, 1994 and Reed, 1995).

To understand how string transcription is regulated, the expression of string-lacZ reporter genes covering ~40 kb of the string locus were examined. Protein coding fragments of the string locus of 6 kb to 31.6 kb were tested for their ability to complement loss of string function in embryos and imaginal discs. A plethora of cis-acting elements spread over >30 kb control string transcription in different cells and tissue types. Regulatory elements specific to subsets of epidermal cells, mesoderm, trachea and nurse cells were identified, but the majority of the string locus appears to be devoted to controlling cell proliferation during neurogenesis. Consistent with this, compact promotor-proximal sequences are sufficient for string function during imaginal disc growth, but additional distal elements are required for the development of neural structures in the eye, wing, leg and notum. It is suggested that, during evolution, cell-type-specific control elements were acquired by a simple growth-regulated promoter as a means of coordinating cell division with developmental processes, particularly neurogenesis (Lehman, 1999).

DNA fragments from the transcription start sites (from 0 kb to -26.4 kb upstream) drive lacZ transcription in distinct subsets of string expressing cells, and thus these sequences are referred to as position-specific elements (PSEs). Many of these PSEs activate string expression in specific mitotic domains (MDs) in the embryo (Foe, 1989). For example, a 4.9 kb fragment (in pstgbeta-E4.9 centered on -4 kb) drives expression in cycle 14 domains, including the mesoderm (MD 10), the mesectoderm (MD 14), the ventral neurectoderm (MD 21, N), and the ventral epidermis (MD M). Another PSE, the 6.4 kb fragment (in pstgbeta-E6.4, centered on -10 kb) drives expression in a different set of cycle 14 domains (MD 1, 2, 15, 18). For most of the PSE fragments tested, lacZ expression occurs in spatial and temporal patterns that mimic a subset of the normal string expression pattern. This fine correlation indicates that the PSEs can function independent of one another and that their spacing relative to the string promotor is not critical. Most of the string PSEs activate transcription in multiple cell types and at several developmental stages, suggesting that they are composites of smaller more specific PSEs. This possibility was confirmed in several instances when a large PSE was bisected to give smaller PSEs with separate activities. Many PSEs also drive expression within a particular cell lineage during consecutive cell cycles. For example, the 6.4 kb PSE (in pstgbeta-E6.4 centered on -10kb) drive expression in cells of mitotic domains 1 and 2 during embryonic cycles 14, 15 and 16. Similarly, the PSEs that drive expression in cycle 14 (MDs 10, 14, 15, 21, N and M) also promoted expression in the analogous MDs during cycle 15 and in some cases during cycle 16. However, many cycle 14 domains are subdivided during cycles 15 and 16 (Foe, 1989), and several instances have been found in which a particular PSE drives expression in some subdomains and not in others. It is concluded that the string PSEs function in a cell type-specific fashion, rather than as developmental timers. Their activities most likely depend upon the expression of position-specific transactivators that are expressed over times spanning several cell cycles within a given cell lineage (Lehman, 1999).

In testing the vectors used to make the various string reporter genes, several interesting properties of the string promotor were observed. The promoter contains sequences that allow it to respond specifically to distant PSEs. Such promotor/enhancer specificity has been noted in studies of other Drosophila loci, and may be a common mechanism by which enhancers like the string PSEs activate only the relevant gene within a chromosomal region. Other experiments suggest that some interactions between the PSEs and the string promotor are repressive. Specifically, the pstgHZ and pstgb vectors, which contain only promotor-proximal sequences, drive ectopic expression patterns that differ both spatially and temporally from normal string expression. These consist of abnormal expression throughout the head at the cellular blastoderm stage and in the mesoderm, anterior midgut (AMG) and posterior midgut (PMG) during gastrulation. Interestingly, the ectopic expression in the head and mesoderm is lost when certain PSEs are added to pstgbeta (as in pstgbeta-E6.4), and the ectopic AMG and PMG expression is lost in constructs containing sequences 3' to the promotor, such as pstgbeta-3.2 and pSTG6.0. A similar relationship was discovered in the developing optic lobe of the larval nervous system: the pstgbeta vector is expressed throughout a region known as the outer proliferative center (OPC), but parts of this expression are lost when various PSEs are added to pstgbeta. This suggests that, in addition to positive regulatory elements, the string locus contains negative elements that restrict the activity of the promotor (Lehman, 1999).

Embryonic neuroblasts delaminate from the neurectoderm in five waves, S1-S5, followed by string expression and then cell division. Greater than 15 kb of the string regulatory region is dedicated primarily to expression in neuroblasts. Within this region, the expression patterns promoted by four separable and contiguous PSEs were analyzed. The 6.4, 2.6, 5.3 and 6.7 kb PSEs (centered on -10kb, -14kb, -18kb and -25kb respectively) all drive expression in overlapping subsets of neuroblasts throughout embryogenesis. The 6.4 kb PSE is a strong activator for all early S1 neuroblasts except one cell-type: MP2. In contrast, the 2.6, 5.3, and 6.7 kb PSEs express in smaller subsets of S1 neuroblasts. Mitosis in embryonic neuroblasts is immediately followed by S-phase, and therefore BrdU pulse-labeling was used to track the division pattern in these cells. This analysis indicates that the neuroblasts of the lateral row (NBs 2-5, 3-5, 5-6, 7-4) plus NB 5-2 and 5-3 divide first, followed by the division of NB 7-1 and 1-1, and subsequently NB 3-2 and MP2. Interestingly, three or four PSEs activate transcription in those neuroblasts that divide earliest. In contrast, fewer PSEs drive expression in the later dividing S1 neuroblasts. This suggests that the timing of neuroblast divisions may depend on rates of string accumulation driven by the additive effect of multiple PSEs. During larval neurogenesis, String mRNA is expressed in neuroblasts of the ventral nerve cord (VNC) and the central brain (CB), and in complex patterns in the developing optic lobe, including the inner and outer proliferation centers (IPC and OPC) and the lamina. Patterns of beta-gal protein expression driven by the string PSEs were analyzed in the CNS of second and third instar larvae. Those PSEs that activated expression in embryonic neuroblasts also function in larval neuroblasts. The 0.7 kb promoter in pstgbeta, which is active in a few CNS neuroblasts late in embryogenesis, is expressed in many larval neuroblasts. All transgenes containing this 0.7 kb promotor show expression in neuroblasts of the CB and the thoracic VNC during the second and third larval instars. In addition, distinct, PSE-specific expression patterns were observed in the developing optic lobe. For example, in second instar larvae, the 4.9 kb PSE drives expression in the IPC and OPC, while a different PSE, the 2.6 kb, does not. In third instar larvae, the 4.9 kb PSE drives expression in the entire OPC while the 2.6 kb PSE drives expression in the IPC and only the posterior portion of the OPC. Yet another PSE, the 6.4 kb, drives expression in a different subset of cells in IPC and OPC regions that lie under the surface of the brain. This pattern may correspond to the progeny of the optic lobe neuroblasts going through additional divisions after budding interior to the proliferation centers. Finally, the 5.3 kb PSE drives expression in cells of the developing lamina. These results indicate an important role for the multiple neuroblast PSEs in regulating the complex proliferation patterns of optic lobe development (Lehman, 1999).

Within the ~50 kb region under study, PSEs responsible for only a subset of all proliferating cells were identified. One explanation for the failure to detect PSEs for all cell types is that expression in certain regions requires synergistic interactions between multiple PSEs. To test this, a 31.6 kb genomic DNA fragment was isolated covering the string transcription unit and 24 kb of intact upstream sequence (STG31.6). The function of this fragment was tested in two string mutants that completely block postblastoderm cell divisions. As expected, String mRNA and BrdU incorporation (a measure of cell cycle progression) are detected in transduced embryos in all the mitotic domains where lacZ expression is driven by the individual PSEs. Interestingly, STG31.6 also drives string expression and mitosis in a few domains that are not detected using the stg-lacZ reporter lines. These included parts of cycle 14 MD 11 and MD 23 and cycle 15 MD 3, MD 6 and MD 19. Thus the PSEs may interact additively or synergistically to drive portions of string’s expression pattern. Despite these findings, the division patterns driven by STG31.6 still represent only a subset of the wild-type division pattern. Consistent with this, transduced embryos die with mild cuticular defects that can be attributed to partial loss of cell division in MD11 (the dorsolateral epidermis). Studies of the stg-lacZ reporter-genes, and also tests of genomic string transgenes, indicate that additional control elements do not reside in the 16 kb 3' to string. 5' to -28 kb, two additional PSEs have been detected, but these promote expression patterns unlike those of the normal string gene, suggesting that they might not normally regulate string. PSEs controlling string expression in MDs 4, 5, 9, 12 and 20 have yet to be identified, and results pertinent to MDs 7, 8, 11, 16, 17, 22, 24 and 25 remain ambiguous. These missing regulatory elements may be revealed by analysis of sequences beyond -35 kb (Lehman, 1999).

Imaginal discs are epithelial primordia that undergo growth and cell proliferation in the larva. They differentiate structures such as wings, legs and eyes in the adult. string is required and rate-limiting for G2/M progression in the discs. During the initial 6-8 cycles of disc growth, String mRNA is expressed in periodic, spatially random patterns that may reflect oscillation during the cell cycle, and during the final 2-3 divisions, as disc cell cycles become synchronized with the onset of cell differentiation, string displays position-specific expression patterns (Milan et al., 1996; Thomas et al., 1994; Johnston and Edgar, 1998). To identify the control regions required for string expression in imaginal discs, clones of sting mutant cells were generated in the presence of rescuing string transgenes possessing different amounts of flanking regulatory sequence. Imaginal disc cells homozygous for mutant stg divide only once, giving 2-celled clones that are eliminated by cell competition. In contrast, stg mutant cells carrying particular transgenes divide many times and give large clones of cells. Mutant clones rescued by the largest string transgene are equal in size to their wild-type sister clones (‘twin spots’) and thus appear to grow normally. Mutant clones rescued by the other string transgenes are smaller than their twin-spots, and also show increased cell size and Cyclin A accumulation. This suggests that cells rescued by the shorter string transgenes have a slower cell cycle with a lengthened G2 phase. All of the string transgenes are able to rescue cell division in all regions of the wing, leg and eye imaginal discs. This suggests that region-specific PSEs are not used during imaginal disc growth. Very large clones of mutant cells rescued by any of the string transgenes could be generated using the Minute technique. These clones often encompass the majority of the disc tissue, and discs containing them grow to full size and differentiate normally sized adult structures. This confirms that even the smallest string transgene is sufficient to support cell cycle proliferation in all regions of the imaginal disc cells. It is concluded that an imaginal disc PSE resides between -1 kb and +5 kb. In performing these rescue experiments, it was noted that adult flies carrying clones of string mutant cells rescued by any of the string transgenes have defects in differentiated cuticular structures. These include fused facets and missing bristles in the eye, and missing sensory bristles (macrochaetae and microchaetae) in the wing, leg, and notum. These deletions appear to be specific to neural cell types since, in most cases, sensilla are lost without deletions of the underlying epidermis. Losses of epidermal tissue are rare; only the scutellum is frequently affected. It is inferred that sequences not encompassed by the longest transgene, STG31.6, are required specifically for cell cycle control in the neural cell lineages that generate sensilla and ommatidia in the adult cuticle (Lehman, 1999).

Analysis of patterns expressed by the stg-lacZ reporters in imaginal discs uncovered several phenomena that are consistent with this scenario. For instance, in the eye disc, the pstgbeta vector (centered at +1.0 kb) is expressed at moderate levels anterior to the morphogenetic furrow (MF); it is depressed in the furrow and expressed at lower levels posterior to the furrow. These patterns are a subset of the normal string expression pattern in the eye (Thomas, 1994). However, none of the stg-lacZ reporter genes drive the strong stripe of expression exhibited by string just anterior to the MF. This stripe is thought to synchronize cells in G1 prior to the onset of differentiation, and loss of cell cycle synchronization in the MF results in roughening of the eye (Thomas, 1994). Loss of string-mediated cell cycle synchronization and consequent defects in the patterning of cell differentiation may explain the patterning defects in eyes composed of sting mutant tissue rescued by the STG31.6 and other specific transgenes. Interestingly, a viable string allele, stgHWY, fails to express string in the stripe anterior to the MF, and causes roughening of the eye and loss of macrochaetae (Lehman, 1999). These defects in stgHWY cannot be rescued by the STG31.6 transgene (H. Stocker and E. Hafen, personal communication to Lehman, 1999).

Short-term integration of Cdc25 dynamics controls mitotic entry during Drosophila gastrulation

Cells commit to mitosis by abruptly activating the mitotic cyclin-Cdk complexes. During Drosophila gastrulation, mitosis is associated with the transcriptional activation of cdc25(string), a phosphatase that activates Cdk1. This study demonstrated that the switch-like entry into mitosis observed in the Drosophila embryo during the 14th mitotic cycle is timed by the dynamics of Cdc25(String) accumulation. The switch operates as a short-term integrator, a property that can improve the reliable control of timing of mitosis. The switch is independent of the positive feedback between Cdk1 and Cdc25(String) and of the double negative feedback between Cdk1 and Wee1. It is proposed that the properties of the mitotic switch are established by the out-of-equilibrium properties of the covalent modification cycle controlling Cdk1 activity. Such covalent modification cycles, triggered by transcriptional expression of the activating enzymes, might be a widespread strategy to obtain reliable and switch-like control of cell decisions (Di Talia, 2012).

During Drosophila gastrulation, transcriptional activation of string is associated with entry into mitosis. This study investigated how String accumulation results in abrupt switch-like activation of Cdk1 and entry into mitosis. The time interval between String transcriptional activation and entry into mitosis is controlled by the rate of string expression. The concentration of String integrated over 2 min is the quantity that best correlates with the decision of entering mitosis and such integration time might be determined by the response time of Cdk1 activity to changes in String concentration (Di Talia, 2012).

Two ultrasensitive steps (activation of Cdk1 by String and entry into mitosis by Cdk1) control mitosis. Such a cascade can provide high ultrasensitivity from two moderately ultrasensitive steps. Positive feedback does not play an important role in the control of entry into mitosis in the Drosophila gastrula and it is proposed that the ultrasensitivity is rather due to the out-of-equilibrium properties of the covalent modification cycle controlling Cdk1. Feedback mechanisms are conserved in Drosophila: it can be shown that mutants (string9A and Wee19A) that disable feedbacks have the expected effects on cell cycle control when overexpressed (Di Talia, 2012).

This raises the question of why feedbacks do not play a role in WT cells. It is proposed that activation of mitosis in Drosophila is too rapid for feedback to make a significant contribution to Cdk1 activation. In Xenopus egg extract, positive feedback introduces a 10 min delay between the accumulation of cyclin to a critical threshold concentration and the activation of Cdk1. This delay has been interpreted as the time required to activate the feedback mechanism. Controlling entry into mitosis through rapid accumulation of String is, therefore, likely to make the contribution of feedback irrelevant. When string9A and Wee19A (Wee1) are overexpressed, the time between String activation and mitosis can become significantly longer providing enough time for feedback to contribute to activation of Cdk1. It is speculated that the molecular network controlling entry into mitosis is highly flexible and can operate as cyclin-driven switch dependent on feedback or as Cdc25-driven switch with the properties described in this article. These two different strategies to control the cell cycle might reflect different selective pressures on the control of mitosis and might be utilized at different stages during development (Di Talia, 2012).

The control of cell division during Drosophila gastrulation provides an extraordinary example of the temporal precision with which cell behaviors can be timed during embryonic development. This precision might be required to avoid incompatible cell behaviors that might easily arise due to the fast timescales of Drosophila embryonic development. It is suggested that controlling entry into mitosis by transcriptional expression of string rather than accumulation of cyclins avoids the possible delay due to activation of the positive feedback. Such a delay might be incompatible with the precise control of cell divisions observed during Drosophila gastrulation. Positive feedback also has the potential of amplifying noise in the expression of string and could in principle deteriorate the precision of cdc25string transcriptional control. It is speculated that String-driven switches similar to the one that this study has described might be a preferred solution for the precise temporal control of mitosis. Such switches, when operating as short-term integrators, have the ability to filter out the probably unavoidable fast fluctuations in the expression of string (Di Talia, 2012).

Covalent modification cycles are widespread signaling modules that can generate ultrasensitivity when operating in the zero-order regime. Theoretical work shows that transcriptionally driven covalent modification cycle can effectively generate an ultrasensitive response when they operate in the first-order regime as long as they are not close to steady state. In this regime, these cycles also display interesting signaling properties. They act as low-pass filters dampening fluctuations that happen on time scales faster than the response time. Effectively, they resemble short-term integrators with an integration time that is determined by the response time (Di Talia, 2012).

Because the response time depends on the concentration of the two opposing enzymes, the filtering properties of the cycle can be easily tuned to the desired frequency. It is proposed that by driving the cycle with string expression, Drosophila is able to achieve switch-like behavior while maintaining the robust filtering properties of covalent modification cycles. Positive feedback driven circuits that have similar integration properties and therefore achieve both ultrasensitive and precise control of cell decision might be much harder to design. It is speculated that covalent modification cycles, triggered by transcriptional expression of the activating enzymes, might be a widespread strategy to obtain reliable and switch-like control of cell decisions (Di Talia, 2012).


>cDNA clone length - 2308bp

Bases in 5' UTR -391

Bases in 3' UTR - 749


Amino Acids - 479

Structural Domains

STG protein contains a C-terminal region that is 34% identical to the C-terminal portion of the cdc25 protein of the yeast S. pombe. In a wild-type background, cdc25 is required for progression from G2 to mitosis; when overexpressed it causes premature initiation of mitosis (Edgar, 1989).

Cdc25 phosphatases activate the cell division kinases throughout the cell cycle. The 2.3 A structure of the human Cdc25A catalytic domain reveals a small alpha/beta domain with a fold unlike previously described phosphatase structures but identical to rhodanese, a sulfur-transfer protein. Only the active-site loop, containing the Cys-(X)5-Arg motif, shows similarity to the tyrosine phosphatases. In some crystals, the catalytic Cys-430 forms a disulfide bond with the invariant Cys-384, suggesting that Cdc25 may be self-inhibited during oxidative stress. Asp-383, previously proposed to be the general acid, instead serves a structural role, forming a conserved buried salt-bridge. It is proposed that Glu-431 may act as a general acid. Structure-based alignments suggest that the noncatalytic domain of the MAP kinase phosphatases will share this topology, as will ACR2, a eukaryotic arsenical resistance protein (Fauman, 1998).

string: Evolutionary Homologs | Regulation | Developmental Biology | Effects of Mutation and Overexpression | References

date revised: 1 April 98

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

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