Requirement of the calcineurin subunit gene canB2 for indirect flight muscle formation in Drosophila

Abstract

Calcineurin is a calcium-activated protein phosphatase involved in multiple aspects of cardiac and skeletal muscle development and disease. Genes encoding calcineurin subunit proteins are highly conserved among animal species. Toward the goal of identifying new calcineurin-interacting loci that function in myogenic processes, we expressed an activated form of mouse calcineurin A in Drosophila and screened for suppressors of the phosphatase-induced lethality. Here, we demonstrate that a mutation in the canB2 gene, which encodes a regulatory subunit of Drosophila calcineurin, can suppress a pupal developmental arrest phenotype to adult viability. As canB2 is an essential gene and rare homozygous escapers are flightless, we further analyzed canB2 expression and function in pupae and adults. The gene is expressed in the forming indirect flight muscles and central nervous system during pupal development. A canA gene is comparably expressed in these tissues. Consistent with the observed muscle expression, canB2 mutants exhibit severe defects in the organization of their indirect flight muscles, a phenotype that is likely caused by muscle hypercontractility. Together, these findings demonstrate a vital role for the phosphatase in this specific facet of Drosophila myogenesis and show conserved fly and vertebrate calcineurin genes contribute prominently to fundamental processes of muscle formation and function.

Calcineurin is a calcium-activated protein phosphatase composed of a 60- to 63-kDa catalytic subunit (calcineurin A), a 16- to 18-kDa regulatory subunit (calcineurin B), and the calcium-binding protein calmodulin (1). The serine/threonine-specific phosphatase functions within a signaling pathway that regulates biological responses in a variety of cell types (2, 3). Calcineurin was initially discovered as a signal transduction molecule used during the activation of T cell lymphocytes in response to foreign antigens (4). In neurons, calcineurin activity is important for axonal guidance and reinforcing synaptic connections crucial for aspects of learning and memory (5). The role of calcineurin in these diverse cellular processes includes the dephosphorylation of nuclear factor of activated T cells (NF-AT) transcription factors and their cytoplasm to nucleus translocation, culminating in altered gene expression profiles.

The calcium/calcineurin/NF-AT signaling cassette is also integral to the control of cardiac and skeletal muscle development and disease. Null mutations in the murine NF-ATc1 gene lead to an absence of heart valve formation and cardiac septum defects, suggesting a role for calcineurin in regulating NF-AT nuclear translocation in endocardial cells adjacent to the cardiac jelly (6, 7). Forced expression of an activated form of calcineurin in cardiomyocytes during mouse heart development leads to a pronounced cardiac hypertrophy (8); enzyme-inhibitory drugs, such as cyclosporin A and FK-506, can block this hypertrophic response (9). In skeletal muscle, calcineurin activity may be essential for normal myogenesis, as cyclosporin A blocks muscle differentiation in vitro (10). Calcineurin gene expression is up-regulated during the skeletal muscle hypertrophy induced by insulin-like growth factor 1 (IGF1), and both the phosphatase and IGF1 can induce the GATA2 transcription factor, which associates with NF-ATc1 in the regulation of muscle gene transcription (11, 12). Calcineurin also activates the MEF2 transcription factor in the reprogramming of muscle fiber-specific gene expression during the physiological adaptation of exercised skeletal muscles (13).

Genes encoding the calcineurin subunit proteins are highly conserved among animal species. As in vertebrate organisms, the Drosophila genome contains three canA and two canB genes that are 75% and 88% homologous to their vertebrate relatives, respectively (14). Recently, Sullivan and Rubin (14) expressed an activated form of a fly canA gene (canA^act) in the Drosophila eye and carried out a dominant modifier screen to identify potential regulators of calcineurin function. These studies uncovered nine complementation groups that either enhanced or suppressed the eye phenotype induced by canA^act expression and demonstrated that calcineurin negatively regulates Egf receptor signaling during Drosophila development. One of the suppressor groups corresponds to the canB2 gene, which encodes a regulatory subunit of the enzyme. canB2 was shown to be an essential gene, with mutant animals dying at a late larval/early pupal stage. However, the cause of this lethality was not elucidated.

Although NF-AT transcription factors are defined targets of calcineurin dephosphorylation, these calcium-responsive proteins are found only in vertebrate systems and not in invertebrates such as Drosophila (15). Considering the demonstrated involvement of calcineurin in several myogenic events in vertebrates, and the potential to use genetic means to identify new calcineurin interacting loci, we expressed an activated form of mouse calcineurin A (mcanA^act) in Drosophila muscles. In this report, we show the lethality induced by this expression can be suppressed by seven different deficiencies that delete specific intervals of chromosomes 2 and 3. One of these regions contains the canB2 gene, and a canB2 mutation alone can prevent the mcanA^act-induced lethality. We analyzed phenotypes resultant from canB2 mutations and demonstrated the gene is essential for normal indirect flight muscle (IFM) formation during the pupal stage. In contrast, calcineurin seems to have no function in earlier myogenic events occurring within the embryo. Such findings provide some of the first critical insights into the selective requirement of calcineurin activity in Drosophila and further demonstrate conserved calcineurin genes function in a variety of essential myogenic processes during development.

Materials and Methods

Fly Strains.

24B-Gal4, Sco/CyO GFP-Actin, and chromosome deficiency kits were obtained from the Bloomington Stock Center (Indiana University, Bloomington); the EP(2)0774/CyO strain was provided by the Szeged Stock Centre (University of Szeged, Szeged, Hungary). The insertion site of the EP vector within canB2 was determined by generating a genomic PCR product by using primers from canB2 intron 1 and the EP DNA, followed by cloning of the 258-bp fragment into the TOPO pCRII vector (Invitrogen) and sequencing. canB2¹⁸⁰/CyO-GFP^arm (14) was obtained from K. Sullivan (University of California, Berkeley), and MHC-GFP stocks were established as described in Chen and Olson (16). To generate UAS-mcanA^act lines, a murine canAα cDNA encoding amino acids 1–398 was cloned into the pUAST P element vector. This cDNA produces a truncated form of calcineurin A, with its autoinhibitory and calmodulin-binding domains deleted (8). Multiple transgenic strains expressing UAS-mcanA^act were established by using standard procedures (17). The EP(2)0774, MHC-GFP chromosome 2 and 24B-Gal4, MHC-GFP chromosome 3 were generated by standard recombination crosses. To facilitate phenotypic analyses, EP(2)0774, MHC-GFP was balanced over CyO, GFP-Actin. This combination allows for the easy identification of homozygous EP(2)0774, MHC-GFP larvae because the MHC-GFP transgene expresses strongly in body wall muscles (16), whereas GFP-Actin generates a fluorescent signal in portions of the gut.

Suppressor Screen Using Drosophila Chromosome 2 and 3 Deficiencies.

Females homozygous for UAS-mcanA^act inserted on the X chromosome were crossed to males carrying individual chromosome 2 and 3 deficiencies and marked balancer chromosomes. All F₁ male progeny inherit UAS-mcanA^act and either the deficiency or balancer chromosome. UAS-mcanA^act/Y; Df/+ males were then mated to homozygous 24B-Gal4 females; it was expected that all F₂ male progeny should survive, as they lack the UAS-mcanA^act chromosome. All F₂ females inherit both UAS-mcanA^act and 24B-Gal4, which is a late larval/early pupal lethal combination. Half of these females also receive the deficiency chromosome, and when such deficiencies delete a gene needed for activated calcineurin function, female adults may be observed. Based on this strategy, a suppression value of 100% would occur when the number of eclosed females corresponded to one half the number of eclosed males.

canB2 Mutant Phenotype Analyses.

To determine the status of IFM formation, thoraces from pupal or adult animals were embedded and processed as described (18). Images of hematoxylin- and eosin-stained sections were captured with a Zeiss Axioplan 2 microscope and digital camera using AXIOVISION V.3.1 software. For developmental time courses monitoring MHC-GFP expression and muscle formation, wandering third instar larvae of known genotypes were transferred to grape agar plates and periodically checked for puparium formation. Early white prepupae were collected and timed as 0 h after puparium formation. Once pupae had reached stage P5, they were dissected from their pupal cases and returned to plates to continue development. GFP expression was monitored with a Leica MZFLIII stereomicroscope, with photographs taken intermittently by using IMAGEPRO PLUS and MEYER IN-FOCUS software.

Calcineurin Expression Analyses.

cDNAs were provided by T. Nguyen (M. D. Anderson Cancer Center, Houston). Embedding of pupae and preparation of pupal sections for in situ hybridization were as described in Lyons et al. (19). DIG-labeled cRNA probes were generated by using the DIG RNA Labeling Kit (Roche Molecular Biochemicals); in situ hybridizations were carried out according to Ingham and Jowett (20) by using Histoclear (National Diagnostics) as a replacement for xylene. Stained sections were fixed in 4% (vol/vol) paraformaldehyde and mounted in 50% glycerol. RNA in situ hybridization of embryos was performed as described (17). Images were obtained with a Zeiss Axioplan 2 microscope using AXIOVISION V.3.1 software.

Results

Expressing mcanA^act in Drosophila Muscles Results in Lethality.

Forced expression of a constitutively active form of calcineurin A in cardiomyocytes during murine postnatal development results in cardiac hypertrophy (8). By using the Gal4/UAS transcription system (21), we expressed the same mcanA^act cDNA in developing Drosophila muscles with the 24B-Gal4 driver. This genetic combination is fully lethal as no adults are observed. To determine the effective lethal phase, we followed the progression of wild-type and mcanA^act; 24B-Gal4 animals by examination of external structural characteristics and MHC-GFP muscle expression in late larval and pupal stages. During pupal day 1, wild-type animals evert their cephalic complexes, undergo hypodermis formation, and evaginate their wing and leg discs, to note a few landmarks (Fig. 1A). Also, the MHC-GFP marker is expressed strongly in a larval pattern of abdominal muscles before their histolysis (Fig. 1B). By day 4, normal pupae undergo extensive eye and cuticle pigmentation along with patterned macrochaete and microchaete formation and pigmentation (Fig. 1C). In parallel, the MHC-GFP transgene becomes strongly expressed in forming IFM of the thorax and weakly expressed in a subset of adult abdominal muscles (Fig. 1D).

Figure 1.

A canB2 mutation suppresses the arrested development phenotype induced by activated calcineurin. Pupal development was followed in animals of the indicated genotypes over 5 days. Animals were examined for imaginal disk evertion, eye and body pigmentation, and bristle formation (A, C, E, G, I, and K) or expression of the GFP marker in thoracic and abdominal muscles (B, D, F, H, J, and L). Animals are shown at days 1 and 4. After day 5, normal eclosion is observed with the 24B-Gal4, MHC-GFP/+, and UAS-mcanA^act/+; EP(2)0774/+; 24B-Gal4, MHC-GFP/+ suppressed animals.

In contrast, mcanA^act; 24B-Gal4 animals showed an arrested development phenotype around the puparia to pupa transitional period. About 25% of animals failed to completely evert their anterior spiracles and retained an elongated and sometimes curved larval shape, with abdominal contraction away from the pupal case. Most other animals initiated prepupal or pupal development, with the latter everting their head sacs and evaginating wing and leg discs during day 1 (Fig. 1E). However, these animals often showed a cleft thorax phenotype because of an incomplete fusion of the hypodermis. Externally, the mcanA^act; 24B-Gal4 animals failed to progress in developmental terms by day 4 or later as no eye, cuticle, or bristle pigmentation was observed (Fig. 1G). Additionally, the pattern of MHC-GFP expression at day 4 was identical to that observed at day 1 (Fig. 1F). That is, the GFP marker was expressed in the array of larval abdominal muscles and not the expected adult pattern of indirect flight and abdominal wall muscles (Fig. 1H). Hypertrophy of the abnormally persistent larval muscles was not detected. Thus the mcanA^act; 24B-Gal4 combination seems to trigger an arrest in development sometime during pupal day 1.

Screen for Chromosome Intervals That Suppress the mcanA^act-Induced Lethality.

Based on the phenotype caused by mcanA^act expression, we designed an F₂ deficiency screen to search for chromosome intervals that harbor genetic suppressors of the induced lethality. Conceptually, the objective was to assay flies with a 50% reduction in the dose of one or more genes for their ability to prevent the detrimental effect of mcanA^act expressed under the control of 24B-Gal4. Deficiency kits that contained deletion strains encompassing most of chromosomes 2 and 3 were obtained, with individual chromosomes tested for their suppression ability based on the crosses outlined in Fig. 2. A total of 62 chromosome 2 and 63 chromosome 3 strains were assayed and scored in a stringent manner. That is, whereas all mcanA^act; 24B-Gal4 females die unless the deficiency overcomes the developmental arrest phenotype, only those crosses that generated eclosed adult females were scored as positive. No females were observed in 72 (58%) of the assays, whereas a single or a few females managed to eclose in 46 (36%) of the tests. However, this latter group seemed insignificant because the females represented <5% of all progeny. Seven deficiency chromosomes resulted in a high degree of suppression; the locations of the specific intervals along chromosomes 2 or 3 are indicated in Fig. 3. Based on the scoring of duplicate assays, the suppression values for the intervals were 41.6% (1), 66.0% (2), 23.9% (3), 35.9% (4), 13.6% (5), 29.7% (6), and 18.8% (7).

Figure 2.

Genetic crosses used to screen Drosophila chromosome deficiencies for their ability to suppress the lethality induced by UAS-mcanA^act expression. The diagram depicts the use of chromosome 2-deficiency strains; the screen of chromosome 3-deficiencies was essentially the same. Based on the F₁ mating, all F₂ females inherit the UAS-mcanA^act and 24B-Gal4 chromosomes, which is a lethal combination. All females die unless a specific deficiency suppresses the lethal phenotype. All males survive as they fail to receive the UAS-mcanA^act chromosome. Based on this scheme, 100% suppression would occur when 50 adult females appear for every 100 adult males.

Figure 3.

Deletion intervals and canB2 mutation that suppress the UAS-mcanA^act; 24B-Gal4-induced lethality. (A) The map locations of seven suppressing regions are indicated above the schematic of Drosophila chromosomes 2 and 3. The coordinates of the intervals are: 1, 21A01–21B08; 2, 42B03–43E18; 3, 46A–46C; 4, 59A1–59D4; 5, 60E2–60E12; 6, 66F5–66F5; and 7, 76B1–76B5. Only those deficiencies that showed >10% suppression in duplicate experiments are indicated. (B) Exon/intron organization of the canB2 gene, which maps within deficiency interval 2 at position 43E16. In the EP(2)0774 allele of canB2, the P element insertion has occurred at nucleotide 29 of the first intron.

A canB2 Mutation Functions as a Strong Suppressor of mcanA^act.

The strongest suppressing deficiency we observed corresponded to a deletion of the 42B03–43E18 region of chromosome 2. To determine whether a single gene might be responsible for this suppression, we surveyed the FlyBase database (available at http://flybase.bio.indiana.edu/) for candidate genes that mapped within interval 2. Intriguingly, the canB2 gene was located at position 43E16. We obtained the EP(2)0774 mutant allele of canB2 and verified the presence of the EP element inserted at nucleotide 29 of the first intron (Fig. 3B). Next, we tested the ability of this mutation to prevent the developmental arrest phenotype and lethality observed in the mcanA^act; 24B-Gal4 flies. The presence of a single EP(2)0774 chromosome in this genetic background resulted in a 34.3% suppression value and the appearance of multiple mcanA^act; EP(2)0774; 24B-Gal4 adult females. Such animals showed normal development in terms of eye and cuticle pigmentation and bristle formation (Fig. 1K). Additionally, they showed the normal switch in MHC-GFP expression from the larval abdominal muscles during day 1 to the adult pattern of indirect flight and abdominal muscles by day 4 (Fig. 1L). Finally, these animals were able to eclose and fly. Thus, this suppression of phenotype demonstrates a normal dosage and function of this calcineurin regulatory gene is needed for the adverse effects of mcanA^act expression on animal development and viability.

Analysis of an IFM Phenotype in canB2 Mutants.

canB2 was recently reported to be an essential Drosophila gene, with mutants presenting lethality in the late larval/early pupal period (14). However, in our analysis of EP(2)0774 flies, we observed a low level of homozygous adults that were flightless, with wings positioned at abnormal angles. As this phenotype was suggestive of irregularities among certain thoracic muscle groups, we analyzed sections of canB2 mutants for the organization of their IFM. In wild-type adults, the thorax is filled with six central pairs of dorsal longitudinal (DLM) and seven lateral pairs of dorsal ventral (DVM) IFM (Fig. 4 A and B). Also, certain sections identify the paired trochanter muscles at mediolateral positions. This intricate muscle pattern is observed in both anterior and posterior thoracic domains.

Figure 4.

Abnormal DLM organization in eclosed canB2 mutant flies. Hematoxylin and eosin-stained paraffin sections of adult thoraces from animals of the indicated genotypes. (A, C, E, and G) Anterior thoracic sections. (B, D, F, and H) Posterior sections. Arrowheads point to DLM.

canB2 mutants, in contrast, exhibited abnormal organizations of their thoracic muscles. In EP(2)0774 homozygous (Fig. 4C), EP(2)0774/canB2¹⁸⁰ transheterozygous (Fig. 4E), and EP(2)0774/Df(2R)1888 hemizygous (Fig. 4G) animals, large gaps or deletions were observed among the centrally located DLM pairs. In most cases, one or more dorsal muscles were present, and on rare occasions, one or more ventral muscles were alternatively observed. Additionally, although the pattern of DLM was severely affected, the dorsal ventral and trochanter muscle pairs seemed intact and normally arranged. Although muscle gaps were observed in anterior sections of canB2 mutants of the three genotypes, the analysis of posterior sections revealed significantly more complex muscle patterns. That is, the posterior thoraces of EP(2)0774/EP(2)0774 (Fig. 4D), EP(2)0774/canB2¹⁸⁰ (Fig. 4F), and EP(2)0774/Df(2R)1888 (Fig. 4H) animals contained large muscle masses that were poorly defined. This occurrence was in direct contrast to the precisely arrayed IFM present in posterior sections of wild-type animals (Fig. 4B).

The DLM develop from three pairs of persistent larval oblique muscles that fail to histolyze during metamorphosis and split into six pairs that serve as templates for DLM growth (22). To determine whether the absence of most DLM pairs from anterior domains of canB2 mutant thoraces was caused by a lack of larval muscle templates, we used the MHC-GFP marker to assay for their existence in wild-type and mutant animals. During pupal day 1, we observed the presence of the persistent larval oblique muscles in flattened thoracic preparations of both EP(2)0774, MHC-GFP/CyO (Fig. 5A) and EP(2)0774, MHC-GFP/EP(2)0774, MHC-GFP (Fig. 5B) animals. Additionally, the staining of thoracic sections from animals of the same genotypes during pupal day 2 revealed a normal splitting of the templates and formation of six DLM pairs in canB2 mutants (Fig. 5D), as compared with wild-type pupae (Fig. 5C). Thus, the prominent loss of DLM from anterior thoracic regions of canB2 mutants is unlikely to be caused by defects in the early architectural events of DLM formation.

Figure 5.

Defective IFM development in canB2-mutant pupae. Animals of the indicated genotypes were analyzed. Pupae were examined for the presence of persistent larval dorsal oblique muscle templates (asterisks in A and B), splitting of the templates and initial formation of the six pairs of DLM (C and D), and late-stage development of IFM (E–H). Arrowheads point to forming DLM; arrow indicates retracted IFM.

To analyze further the muscle phenotype presented by the canB2 mutants, we followed IFM formation in living pupae by using the MHC-GFP marker. During pupal day 2, muscle formation appeared comparable in wild-type (Fig. 5E) and mutant (Fig. 5F) pupae, with both DLM and DVM developing and elongating at their normal positions. However, by the end of pupal day 4, a striking abnormality was observed in canB2 animals, as all DLM and DVM became located at a posterior thoracic position (Fig. 5H). To understand better the genesis of this defect, we followed the formation of the canB2 IFM over time to determine whether the posterior presence was caused by a rapid detachment and withdrawal of muscles or a more gradual movement and contraction of the complex. By using the MHC-GFP marker, we monitored the status of the IFM in canB2 mutants during pupal day 3 at 30-min intervals and determined there was a slow, progressive retraction of the muscles to the posterior of the thorax after they had initiated observable muscle contractions.

Expression of Calcineurin Subunit Genes in Developing IFM.

Based on the muscle phenotypes observed in canB2-mutant pupae and adults, we sought to verify that the gene was expressed in IFM during normal development. We isolated and sectioned pupae at day 3 and performed in situ hybridizations with a canB2 anti-sense probe. Such studies showed that canB2 is prominently expressed in forming IFM and developing brain and nerve cord (Fig. 6 A and B). We also demonstrated the calcineurin A gene Pp2B-14D is expressed in these same tissues (Fig. 6D), whereas the second calcineurin regulatory subunit gene canB is not expressed in IFM (Fig. 6C). Thus, a canA gene and canB2 are expressed at a place and time during pupal development where they could assume a critical role in the formation of IFM. Further analysis of the calcineurin subunit genes demonstrated transcription of all three in the CNS, but not in body wall muscles or heart, of developing embryos (Fig. 6 E–G).

Figure 6.

Expression of calcineurin subunit genes. (A) Peripheral sagittal section of a day-3 pupa showing canB2 RNA in DLM (dlm) and a brain hemisphere (bh). (B) Central sagittal section of a slightly older pupa detecting canB2 RNA in DLM, brain, and ventral nerve cord (nc). (C) canB RNA in a sagittal section of a day-3 pupa, showing expression in brain but not DLM (*). (D) Central sagittal section of a day-3 pupa demonstrating expression of the calcineurin A gene Pp2B-14D in DLM, brain, and nerve cord. (E–G) Lateral views of stage 13 embryos hybridized with canB2, canB, and Pp2B-14D anti-sense probes. All show strong expression in the CNS.

Discussion

In this paper, we demonstrate that expression of an activated form of mouse calcineurin A in Drosophila muscles results in adverse developmental effects and lethality. Although the precise nature of the mcanA^act-induced phenotype remains unknown, it is noteworthy that the pupal arrest observed in mcanA^act; 24B-Gal4 animals is reminiscent of certain loss-of-function phenotypes seen with the ecdysone-inducible E74 gene (23). E74 encodes two transcription factor isoforms of the ets proto-oncogene family, both with proven functions in larval muscles during pupariation and pupation. A recent report also shows a global block to development can occur with the forced expression of dominant–negative versions of the ecdysone receptor, even when this expression is restricted locally to specific cell or tissue types (24). The generation of comparable phenotypes by mcanA^act expression may be indicative of a role for the phosphatase in the normal reprogramming of gene expression during early metamorphosis.

Based on our goal of identifying genes that interact with calcineurin, we used this phenotype to screen for chromosome intervals that harbor genes essential for calcineurin function. Seven distinct regions of chromosomes 2 and 3 were shown to suppress the mcanA^act-induced lethality to adult viability. The efficacy of the screen has been proven thus far by our demonstration of canB2 as a strong suppressor locus mapping within interval 2. Additionally, preliminary tests using D-mef2 gene mutations suggest this muscle differentiation factor contributes to the suppression activity of interval 3. Further experiments are necessary to investigate this genetic interaction, which would be consistent with the known activation of a vertebrate MEF2 protein by calcineurin in exercised skeletal muscles (13). Likewise, the characterization of other suppressor genes within the remaining chromosome intervals should provide additional insights into calcineurin function during Drosophila myogenesis.

The results of a screen for dominant modifiers of an activated calcineurin phenotype induced in the eye have been reported recently, and these studies identified four enhancing and five suppressing complementation groups (14). A major conclusion of this work is that calcineurin cooperates with other genes to negatively regulate Egf receptor/Ras signaling in cells of the eye imaginal disk via the receptor tyrosine kinase antagonist sprouty and the GTPase-activating protein Gap1. Interestingly, canB2 was identified as one of the suppressor groups, demonstrating the shared importance of this gene for calcineurin activity in the eye and pupal developmental assays. The other suppressor groups identified in the modifier screen map to locations different than the suppressing chromosome intervals identified in this study, suggesting calcineurin activity requires other cell-specific factors in addition to canB2.

canB2 has been classified as an essential gene, with homozygous animals dying at a late larval/early pupal stage (14). Because we noticed the occasional eclosion of homozygous EP(2)0774 adults and their inability to fly, a detailed analysis was undertaken on the structure of IFM in canB2 mutant pupae and adults, with two reproducible phenotypes observed. First, in animals that were able to eclose, abnormalities were observed in the pattern of the thoracic DLM, with most muscle pairs absent from anterior sections and disorganized muscle masses present in posterior sections. We were able to conclude this phenotype was not a result of the lack of larval muscle templates or initial DLM formation but was likely due to the displacement of most of the DLM to the posterior of the thorax. A second, more severe phenotype was also observed as we followed IFM development in living pupae. In cases where animals progressed only to the pharate adult stage, we documented the complete retraction of the IFM to a posterior thoracic position. One interpretation of these findings is that EP(2)0774 represents a canB2 hypomorphic allele and homozygous, transheterozygous, or hemizygous adults can eclose if they maintain a sufficient IFM integrity so as to be competent in their movement out of the pupal case. When a complete displacement of the IFM occurs, the animals may be unable to eclose because of thoracic compression and/or insufficient IFM contraction, resulting in lethality as pharate adults.

The slow, progressive retraction of IFM in canB2 mutants is reminiscent of the hypercontracted IFM phenotype observed in myosin heavy chain mutants (25) and the troponin I mutant heldup² (26). Muscle contraction is known to be a result of numerous protein–protein interactions and conformational changes that allow for the relative movement of thin and thick filaments, leading to sarcomere shortening. It is plausible that mutations in a signal-transducing molecule such as calcineurin B2 could have detrimental effects on the intricate transcriptional regulation of sarcomeric protein genes and subsequent myofiber growth and contraction. As various mutations exist in genes encoding thick and thin filament proteins, interaction studies among canB2 and contractile protein mutants should allow for mechanistic insights into the role of calcineurin in IFM formation and function. Genetic and molecular analyses should also identify those muscle transcription factors that serve as direct targets of calcineurin phosphatase activity.

In summary, our studies have led to the discovery of canB2 as a strong suppressor of an activated calcineurin-induced phenotype and a vital component of the regulatory network needed for the competence of developing flight muscles. The demonstrated expression and function of canB2 in the IFM provides definitive proof of the requirement of calcineurin in a Drosophila developmental process. These canB2 properties contrast with the apparent lack of function of the canB regulatory gene in IFM. Given also the absence of canB2, canB, and Pp2B-14D transcripts from embryonic muscles, a clear specificity exists in calcineurin's myogenic function during fly development.

Abbreviations

NF-AT

nuclear factor of activated T cells

IFM

indirect flight muscles

DLM

dorsal longitudinal IFM