INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Muscle contraction is usually activated by a neurally stimulated intracellular release of Ca²⁺. In vertebrate striated muscle, released Ca²⁺ binds to the thin filament troponin-tropomyosin (Tn-Tm) complex, which consists of tropomyosin (Tm) and the troponins T, I and C (TnT, TnI, and TnC). Current models (see Farah and Reinach, 1995; Geeves and Lehrer, 1998; Squire and Morris, 1998) agree that Ca²⁺ binding to TnC changes the conformation of the Tn complex, releasing the inhibitory binding of TnI to F-actin and allowing Tm to move across the actin surface (Vibert et al., 1997). This movement cooperatively increases myosin accessibility to F-actin, activating the cross-bridge cycle.

Muscle protein interactions are generally well characterized, but the pathway from Ca²⁺ binding to contraction cannot yet be described in molecular detail. The amino acid sequences of vertebrate thin filament proteins (actin, Tm and Tn) are known and their general location in the Tn complex has been determined by low resolution structural studies (White et al., 1987; Al Khayat et al., 1995; Cabral-Lilly et al., 1997) and biochemical analyses. The atomic structure of part of the complex was recently determined (Vassylyev et al., 1998) and functional information derives from considerable biochemical research (see Farah and Reinach, 1995; Geeves and Lehrer, 1998).

Thin filament protein mutants with phenotypic effects are important in identifying residues with significant in vivo roles. Such mutants are known in humans through familial hypertrophic cardiomyopathies (HCM) (Redwood et al., 1999), in the nematode Caenorhabditis elegans (Kagawa et al., 1997; McArdle et al., 1998), and Drosophila melanogaster, where mutations affecting the indirect flight muscles (IFM) have been recovered (see Bernstein et al., 1993). The D. melanogaster IFMs are striated muscles and contain a Ca²⁺-regulated thin filament system but activation also requires applied strain (Peckham et al., 1990), as also found in vertebrate cardiac muscle.

We aim to study the function of the Drosophila Tm-Tn complex by isolating mutations that suppress the phenotypes of selected TnI mutants. held-up² (hdp²) a missense mutation of the TnI gene, produces a fully penetrant raised wing phenotype. Six dominant suppressors of hdp², including D53, were recovered (Prado et al., 1995). Here we show that D53 is a missense mutation, S185F, of the Tm2 tropomyosin gene and characterize the functional and structural effects of the suppression. We discuss these results with respect to known interactions within the vertebrate Tn-Tm complex. We observed that S185F occurs within a region of the Drosophila Tm which is homologous to that containing two human cardiac -tropomyosin mutations which cause HCM and compare the effects of D53 with these mutations.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Stocks and Fly Strains

Flies were maintained at 25°C on a yeast-sugar-agar medium. Unless otherwise stated, strains were obtained from the European Stock Center (Umeå) or the MidAmerica Drosophila Stock Center (Bowling Green). All chromosome and gene symbols unless specifically described are as in Flybase on http://flybase.bio.indiana.edu/. For the two Tm genes, the symbols Tm1 (was TmII - Bautch et al., 1982) and Tm2 (was TmI) are used as proposed in Karlik et al. (1984). The hdp and up alleles are in genes which encode respectively TnI, wupA (wings-up A) (Beall and Fyrberg, 1991; Barbas et al., 1991) and TnT, up (upheld) (Fyrberg et al., 1990). The wupA gene is the sole TnI-encoding gene in Drosophila (Barbas et al., 1991). The hdp² mutation is in constitutive exon 5 (Beall and Fyrberg, 1991; Prado et al., 1995) where it causes a substitution of alanine116 by valine, A116V (using the numbering system of Prado et al., 1995 for Drosophila TnI residues which includes the IFM-specific exon 3 which introduces a 61-residue N-terminal extension).

Microscopy Procedures

For polarised light photomicroscopy thoraces were prepared and mounted as described in Nongthomba and Ramachandra (1999). Fly half-thoraces were prepared for transmission electron microscopy (TEM) as described in Kronert et al. (1995).

PCR and Sequencing

A 3334 bp fragment including the complete Act88F coding sequence was PCR amplified with Taq DNA polymerase from genomic DNA of hdp² and hdp²; D53 flies using the 5'TGTAGGTGGAGCTAACCGTGTGC (sense) and 5'GCTGCCTTTGAAGAGCTTTCGG (antisense) primers. The amplified product was gel-purified (Geneclean, BIO101), ligated into the pGEM-T vector system (Promega), and transformed into TG1 rec^o cells. DNA from three separate clones was extracted, purified, sequenced with internal oligonucleotide primers using an ABI Prism Dye Cycle Sequencing Kit containing AmpliTaq DNA polymerase (Perkin Elmer-Cetus), and analyzed using an in-house automated sequencer.

RT-PCR was used to amplify the Tm1 gene exons that are spliced to form the `mTm' (muscle tropomyosin) mRNA. This mRNA is derived by splicing together exons 1-3, 5, 7, 8, 10, 11, 13 and 17 (Karlik and Fyrberg, 1986; Hanke and Storti, 1988). Total RNA was isolated from newly emerged hdp² and hdp²; D53 flies and reverse transcribed using random nucleotide hexamers and Moloney murine leukemia virus reverse transcriptase (Life Technologies). The cDNA was used in PCR reactions with Pfu DNA polymerase (Stratagene) using 5'GTTCAAGTCGCGGATAACTCCGAATAAAAGTT (sense) and 5'CAGTGCGCCTACGATTATGC (antisense) primers to amplify specifically the coding region. The PCR products were gel purified, ligated into pGEM-T, and transformed into TG1 rec^o cells. Three clones from two separate RT-PCR reactions were sequenced as above.

DNA extracted from Oregon-R, hdp², and D53 flies was used for PCR amplifications with Pfu DNA polymerase (either native or cloned) of overlapping Tm2 genomic fragments containing exons 2, 3 and the coding sequence of exon 4. The primers used were 5'AGGATTCAGTTATTCAGCATAC, 5'TCCTCAATCTGTTGCACCTT, 5'TTGGTATCGGCATCCTCAGC, 5'GGAGGAGGAGCTGAAGGTGG, 5'GGGAGTTGCCGACAACCTAGT, 5'GGGTGGTCAAGGGCATTGTGTG, and 5'CAAACTGAACGAGGAGGTGC. Amplified products from a minimum of two PCR reactions for each fragment were purified and sequenced as described above.

Adult and Larval Muscle Performance Tests

Flight testing was done as previously described (Drummond et al., 1991). Individual flies were scored for flight upwards (U), horizontal (H), downwards (D), or not at all (N). The mean flight score, [U + H]/Total flies, was arcsine transformed to give a Normal distribution. Five samples, > 50 flies, were scored. Wing beat records were obtained and analyzed as described in Schmitz et al. (2000) using 6 flies per genotype, each sampled 3 times.

Adult walking ability was measured in a vertical 100-ml glass measuring cylinder, internal diameter 30 mm, with a line marked 80 mm above the base. For each test 8-30 flies were introduced into the cylinder. Flies were knocked to the cylinder bottom and the time taken for 50% of the flies to walk past the marked line was scored.

Late third instar larval crawling was measured on 1.5% (wt/vol) agar-sucrose medium in plates marked with a 0.5-cm grid, as total gridlines crossed in 5 min. Larval feeding movements were quantified by counting nonlocomotory, pharyngeal movements in a 2-min period on agar plates thinly coated with yeast. Larvae which did not feed continuously were discarded.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Suppressor D53 Is a Tropomyosin 2 Mutation

In hdp² flies the IFMs hypercontract (Figure 1C) so that myofibrillar material remains (arrows) only near the muscle attachment sites. This effect is completely suppressed in hdp²; D53/+ (Figure 1D) and hdp²; D53/53 flies (not shown), which are indistinguishable from wild-type (Figure 1A). Even in pupae at 78 h APF (after puparium formation), ~ 20 h before adult emergence, the hdp² muscles (Figure 1B) begin to hypercontract. IFMs cannot be imaged by birefringence before 78 h, but using a lacZ gene promoter-reporter construct and -galactosidase staining we have shown that developing hdp² IFMs have a normal structure at 75 h APF (Nongthomba and Sparrow, in preparation). D53 suppression mapped to a region on chromosome 3 (Prado et al., 1995) that contains the IFM-specific actin gene, Act88F, and two tropomyosin genes, Tm1 and Tm2. Due to the very tight linkage of these genes D53 was located by sequencing the relevant coding regions of all three genes. The Act88F sequence was identical to the GenBank sequence (M18830) except for seven silent codon changes.

View larger version (120K): [in this window] [in a new window]   Figure 1.   Polarized light micrographs of dissected thoraces to show the progressive hypercontraction of hdp2 IFMs and suppression of this phenotype by D53. Each thorax is shown with the anterior-posterior axis running right to left. (A) Wild-type fly dorso-longitudinal IFMs (asterisk). (B) hdp2 pupa at 78 h APF to show the initiation of hypercontraction (arrowhead). (C) hdp2 fly just after emergence showing IFM remnants (arrowheads) after hypercontraction and (D) a hdp2; D53/+ fly showing complete suppression of hdp2 hypercontraction. Bar = 0.120 mm.

The Tm1 gene produces four alternatively spliced products (Hanke and Storti, 1988). Only one, the mTm mRNA, has an expression pattern coincident with the muscles in which hdp² phenotypes are suppressed by D53 (see below). The Tm1 coding sequences from hdp² and hdp²; D53 flies, obtained by RT-PCR, and the wild-type mTm isoform sequence (Swissprot ID:TPM1_DROME) generated by splicing appropriate exons from EMBL sequences L00355-L00363 (Hanke and Storti, 1986, 1988) were identical.

Larval muscles express a Tm2 mRNA, containing exons 1-4; IFM and leg muscles express a Tm2 mRNA consisting of exons 1-3 and 5 (Basi et al., 1986; Karlik and Fyrberg, 1986). As D53 suppresses hdp² effects in IFMs, legs, and larval muscles (see below), the common coding exons (2 and 3) were sequenced by PCR of genomic DNA from hdp² and hdp²; D53 flies. In hdp² flies Tm2 exons 2 and 3 were identical to the wild-type sequence (EMBL AC: K03277; Swissprot TPM2_DROME; Basi and Storti, 1986). However, exon 2 in hdp²; D53 flies contained a TCC to TTC mutation in codon 185, which will replace serine by phenylalanine (S185F); exon 3 was wild-type.

Suppression of IFM Defects

Flies hemi- or homozygous for hdp² hold their wings vertically, but when D53 was present in one (+/D53) or two (D53/D53) gene doses, their wings were in the resting position. Young (2-3 days) hdp²; +/D53 or hdp²; D53/D53 flies flew, but less well than wild-type (Table 1A). By 6-7 days old the flight ability of hdp²; D53/+ flies, but not hdp²; D53/D53, had been dramatically reduced. These data show that D53 suppression of the hdp² flightless phenotype is dominant and almost complete. The age effect suggests that incomplete suppression in hdp²; D53/+ heterozygotes causes a progressive deterioration in muscle function. Does this correlate with structural changes to the myofibrils?

Electron micrographs of IFMs from 2-3 d-old hdp² flies showed (Figure 2B) an almost complete absence of wild-type myofibrillar structure (Figure 2A). The fibers contained disorganized thick and thin filaments with only small islands of myofibrillar lattice (asterisk) remaining. IFMs of 2-3 d-old hdp²; D53/+ flies (Figure 2C) showed an almost complete restoration of wild-type myofibrillar structure. However, some myofibrils showed small areas of disorganized filament lattice at their centers which in longitudinal sections (Figure 2F) were visible as Z-disk dislocations. By 6-7 days, all the myofibrils of hdp²; D53/+ flies (Figure 2D) contain extensive disorganized centers, an effect never seen in hdp²; D53 (Figure 2E) or D53 homozygotes (not shown). These structural changes correlate with the reduced flight ability of hdp²; D53/+ flies with increasing age (Table 1). Since neither occur in D53 homozygotes, complete D53 suppression is a recessive character.

View larger version (198K): [in this window] [in a new window]   Figure 2.   Transmission electron micrographs of hdp2; D53/+ IFMs to show age effects. (A-C) Transverse sections of IFM myofibrils (Myo) from 2- to 3-day-old flies: (A) Wild-type. (B) hdp2 showing the almost complete disruption of the myofibrils as a result of hypercontraction and (C) hdp2; D53/+ showing suppression but with some myofibrillar lattice disorder at the centers. (D, E) Transverse sections of myofibrils from 6- to 7-day-old flies: (D) hdp2; D53/+ flies demonstrating the increased central disruption of all myofibrils and (E) hdp2; D53/D53 myofibrils showing the complete suppression of hdp2 myofibrillar defects. (F) Longitudinal section of 6- to 7-day-old hdp2; D53/+ myofibrils, with disruptions at some of the Z-disk centers. Bar = 1 µm.

Suppression in Adult, Non-IFM Muscles

Since the hdp² mutation occurs in a constitutive exon of the sole Drosophila TnI gene (Beall and Fyrberg, 1991), its dysfunction in non-IFM muscles should affect behavior. hdp² flies cannot jump (Deak et al., 1982) but hdp²; D53/+ and hdp²; D53/D53 flies jumped readily. Since jumping is powered by the tergal depressor of trochanter (TDT) muscles, D53 must be expressed there. The walking speed of hdp² flies was reduced even in young flies (Figure 3A) compared with wild-type, as reported previously (Deak, 1977) and was further reduced as flies aged (Figure 3A). Older flies appeared crippled, continually fell over, or showed a propensity to fall off vertical surfaces. One or two gene doses of D53 restored walking ability of hdp² flies to almost wild-type levels, indicating that D53 is expressed and suppresses the hdp² phenotype in the muscle groups required for walking. Neither hdp²; D53/D53 nor hdp²; D53/+ flies showed a reduced walking ability with age (Figure 3A). Leg muscles from 12-d wild-type flies (Figure 3B) had a normal fibrillar structure, but in hdp² flies of similar age (Figure 3C), although a few muscles showed a normal muscle structure (arrow) most muscles (asterisk and arrowhead) lacked myofilaments and exhibited large intracellular vacuoles (asterisk) and electron dense staining of cells and nuclei (arrowhead). One muscle (arrowhead) has detached from one of its apodemes. These effects correlate with the behavioral studies. Aged hdp²; D53/+ flies showed (Figure 3D) an intermediate phenotype in which some muscles were clearly damaged (arrowhead) but most appeared normal (arrow). Although the behavioral studies suggested a complete dominant suppression of hdp² by D53, these structural data indicate that dominant suppression is incomplete.

View larger version (131K): [in this window] [in a new window]   Figure 3.   The effects of age on wild-type and hdp2 leg muscles. (A) Walking ability of hdp2 (), Oregon-R (), hdp2; D53/+ (×) and hdp2; D53/D53 (). `Seconds' is mean time (10 tests/sample) for 50% of flies to walk upwards > 80 mm. Error bars ± 1 SD. (B-D) Electron micrographs of the proximal apodeme (ap) of femur from 12-day-old (B) wild-type, (C) hdp2, and (D) hdp2; D53/+ males. In hdp2 flies a few muscles were structurally normal (compare arrows in B and C), but some were detached (arrowhead) and with most of the others showed degenerative changes with intracellular vacuoles (*) and electron dense staining of cells and nuclei (arrowhead); in hdp2; D53/+ flies (D) although some muscles were collapsed (arrowhead), others were attached and had normal structure (compare arrows B and D). Bar = 2 µm.

Suppression of Larval Muscle Defects

As wupA must be expressed in larval muscles (see above), hdp² should affect larval behaviors. Crawling of hdp² larvae was very different from wild-type (Table 2). Many appeared paralyzed in their posterior segments; frequently the posterior was raised, reminiscent of the `hook' phenotype of kinesin or kinesin-related mutants (Hurd and Saxton, 1996). As hdp<sup>2</sup> larvae crawl, they `roll'. Locomotory and feeding behaviors of hdp²; D53 larvae were not significantly different from those of wild-type confirming a) that the Tm2 gene is expressed in larval muscles and, b) that D53 suppresses the effects of hdp² (Table 2) in these muscles.

Muscle Effects due to D53 Mutation Per Se

Often genetic suppressors have their own mutant phenotype. IFMs from D53 homozygotes have a wild-type structure (Figure 2E). To look for more subtle effects the flight ability of D53/+ and D53/D53 flies was measured (Table 1). Both genotypes showed a small, significant reduction in flight ability that was age-independent. Oregon-R flies gave a wing beat frequency of 240.6 ± 23.4 Hz (n = 6 flies) while D53 homozygotes produced 214.7 ± 18.6 Hz, a modest but significant reduction (Student's t test p = 0.041). The walking speed of D53 and wild-type flies was very similar and did not change with age (data not shown). Thus, D53 has a mild IFM phenotype, which is not detected in leg muscles.

Specificity of the D53 Suppression

Suppression could result from altered stoichiometry or from specific amino acid alterations affecting protein interactions. Gene dosage studies should reproduce the former type of suppression, whereas the latter should be allele-specific. Df(3R)ea^5022rxl lacks both the Tm1 and Tm2 genes and Tm2^C10 lacks a functional Tm2 copy (Kreuz et al., 1996). Males that had hdp²/Y; Df(3R)ea^5022rxl/+ or hdp²/Y; Tm2^C10/+ genotypes showed partial suppression of IFM hypercontraction but did not fly; hdp²/Y; Df(3R)ea^5022rxl/+ flies had the wings-up phenotype (Table 3). We define partial IFM suppression as thoraces in which muscle fibers often span the complete thorax, but where some do not show continuous birefringence (cf. Figure 4B) from one attachment site to the other. The hdp²/Y; Df(3R)ea^5022rxl/+ and Tm2^C10/+ results suggest that the reduction in Tm1 gene copy number did not contribute to suppression. This was confirmed by a lack of hdp² dominant suppression (data not shown) by recessive lethal P-element insertion Tm1 alleles (gift of Dr. C. O'Kane). Other recessive lethal Tm2 mutations partially suppressed IFM hypercontraction, but all the flies were flightless (Table 3).

View larger version (52K): [in this window] [in a new window]   Figure 4.   Suppression of up101 by D53. (A-C) Polarized light microscopy of IFMs. Each thorax is shown with anterior-posterior axis running left to right. Bar = 0.125 mm. (A) In up101/Y birefringence is seen only at the muscle ends. The dorso-longitudinal (DLM) flight muscles are most readily seen, but the opposing dorso-ventral (DVM) muscles are visible. (B) The more complete DLMs typical of partial suppression in up101/Y; D53/+ males. (C) Age effects on walking ability of up101 (), up101; D53/+ (), and wild-type () flies.

Flies hemi- or homozygous for the hdp³ and hdp⁵ mutations of wupA also show the wings-up phenotype but in these cases the IFMs fail to develop. These IFM-specific mutations affect alternative transcript splicing of IFM mRNA and produce no TnI protein in the IFMs (Barbas et al., 1993). D53 did not suppress the muscle phenotype of either mutation suggesting that suppression requires the presence of functional TnI protein. The myosin heavy chain "rod" domain mutation Mhc¹³, which causes a recessive hypercontraction of IFMs (Kronert et al., 1995), is also not suppressed by D53. Thus D53 is not a general suppressor of IFM hypercontraction. Its lack of effect on hdp³ and hdp⁵ and its more effective suppression of hdp² than the Tm2 null mutations suggest that it is an allele-specific suppressor.

up¹⁰¹ is a missense mutation of the troponin T gene (Fyrberg et al., 1990). It produces a recessive wings-up phenotype (~ 90% penetrant) and IFM hypercontraction (100% penetrant) indistinguishable from that of hdp² (compare Figures 4A and 1C). Homozygous D53 completely suppressed both phenotypes but up¹⁰¹/Y; D53/+ flies showed a range of partially (Figure 4B) or completely suppressed IFM phenotypes. The wing position phenotype was also partially suppressed by D53/+; 39% of the flies had wings in the normal position compared with 10% of control up¹⁰¹/Y flies. No up¹⁰¹; D53/+ or up¹⁰¹; D53/D53 flies could fly.

Adult up¹⁰¹ flies walked more slowly than wild-type (Student's t, P 0.001), a difference that increased with age (Figure 4C). This effect was smaller than in hdp² flies and became extreme only in much older flies (cf. Figure 3A). D53 suppression of up¹⁰¹ was apparently complete in walking muscles because up¹⁰¹/Y; D53/+ flies of all ages walked at wild-type velocities (Figure 4C).

Thus, the suppressing effects of D53 are clearly not gene-specific, but they are weaker in the IFMs with up¹⁰¹ compared with hdp² (compare Figures 1D and 4B); in walking behavior D53 acts equally well in suppressing both mutations.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The D53 suppressor is a missense mutation, S185F, of the Tm2 gene. It suppresses all the effects of hdp² on muscle function, consistent with D53 being a mutation in constitutively expressed exon 2, and Tm2 expression occurring in the IFMs, TDT, leg muscles and supercontractile larval muscles (Basi et al., 1984; Basi and Storti, 1986; Karlik and Fyrberg, 1986).

The dominant suppression of hdp² flightlessness in D53/+ heterozygotes, but not D53 homozygotes, deteriorates with age. Clearly a mixture of mutant and wild-type Tm2 proteins suppresses in young flies but incompletely so as progressive degeneration occurs, unusually, from the myofibril centers outwards. A similar disruption pattern occurs in the Drosophila Act88F^G245D mutant allele (Sakai et al., 1990) and the intragenic D3 suppressor of hdp² (Prado et al., 1995). These observations suggest that the myofibrillar center and periphery either have different protein constitutions or experience different physiological conditions. Myofibrillogenesis begins at the center (Reedy and Beall, 1993) and different isoforms could be assembled in the myofibril center compared with the periphery but there is no evidence for this. Calcium diffusion into and out from the myofibril upon activation and relaxation will produce temporal differences of Ca²⁺ concentration across the myofibril. If hdp² and D53 affect the response to Ca²⁺ a trans-myofibrillar calcium gradient could generate damaging forces.

Tm2 null mutant heterozygotes produce myofibrils with a normal central lattice but considerable peripheral disruption (Kreuz et al., 1996). Why do they partially suppress hdp²? Fiber destruction in the hdp² hypercontraction likely involves actomyosin generated forces. Kreuz et al. (1996) showed that Tm2^C10/+ fibers produced significantly less work than wild-type. This would explain the partial suppression by extreme Tm2 alleles and why flight is not restored. A similar argument may explain the effects of the myosin heavy chain gene suppressors of hdp² (Kronert et al., 1995) though other explanations including direct interactions between the myosin head and the Tn-Tm complex cannot be excluded.

Structural Implications

The vertebrate inhibitory, actin-binding TnI and the calcium-binding TnC form a globular domain (Herzberg and James, 1985) which sits on TnT, an elongated molecule (see Figure 6A). TnT is the major structural link between TnI + TnC and Tm binding (see Farah and Reinach, 1995) and extends along the carboxy-terminal third of the Tm coiled-coil to overlap with the N-terminus of the adjacent Tm dimer. Two regions of Tm bind TnT one of which, the Ca²⁺-sensitive binding region, includes residues 170-190.

Residue 185, phenylalanine in D53, is indicated in the alignment (Figure 5) of Tm sequences. The Drosophila Tm2 IFM isoform (TPM0_DROME) has considerable sequence homology with other Tm sequences in this region. TPM0_DROME and TPM1_DROME (the standard muscle Tm encoded by the Tm1 gene) show many residue differences. Vertebrate Tm sequences are more similar to each other than to invertebrate sequences and residue type conservation occurs at many positions (Figure 5). In part, this reflects the -helical coiled-coil heptad repeat requirement for hydrophobic residues at most `a' and `d' positions. Residue 185 is predicted to be a `c' residue, implying that the phenylalanine side chain of D53 will point away from the dimer axis and not affect coiled-coil stability. More probably it affects interactions between Tm and TnT or F-actin.

View larger version (42K): [in this window] [in a new window]   Figure 5.   CLUSTALX multiple sequence alignment of muscle tropomyosin sequences corresponding to the TnT-2 binding site to show (a) position of S185F substitution in the protein sequence of the Drosophila Tm2 product expressed in the IFMs (TPM0_DROME); (b) consensus sequence (4 = KR, 6 = LIVM); (c) U = residues unique to the Tm2 product; (d) Heptad repeat `abcdefg ' prediction generated with MacStripe (Dr. A. E. Knight, http://motility.york.ac.uk:85/) using the algorithm of Lupas et al. (1991). SwissProt Sequences: TPM0_DROME, Drosophila Tm2 gene product (P09491); TPM1 DROME, Drosophila Tm1 gene product (P06754); TPM1_HUMAN (P09493) and TPM3_HUMAN (P06753), human skeletal muscle -chains; TPMA_BRARE (P13104), TPMA_COTJA (P18442), and TPMA_RABIT (P46902), zebrafish, quail, and rabbit skeletal/cardiac muscle -chains; TPMA_RANTE (P13105), TPMA_RAT (P04692), TPMA_XENLA (Q01173), European common frog, rat, and South African claw toed frog skeletal muscle -chains; TPMB_CHICK (P19352), TPMB_HUMAN (P07951), and TPMB_RABIT (P02560), chicken, human, and rabbit skeletal muscle -chains; TPMC PIG (P42639), pig cardiac muscle -chain; TPMM LOCMI (P31816) and TPMM_TRICO (P12324), locust and nematode muscle tropomyosins. Dots indicate residues identical to those in the TPM0 DROME sequence.

Our interpretations assume that current models for the vertebrate complex are relevant. This seems likely as Drosophila thin filament proteins show substantial sequence homology to their vertebrate counterparts although TnI (Barbas et al., 1991; Beall and Fyrberg, 1991), TnT (Fyrberg et al., 1990), and two IFM-specific Tm isoforms (Karlik et al., 1984; Hanke and Storti, 1988) have N- or C-terminal polypeptide additions. Drosophila and Lethocerus (waterbug) IFMs have a very similar physiology (Peckham et al., 1990) but in Lethocerus IFMs TnI is not detectable and is replaced by a unique 80 kDa molecule called TnH (Wendt et al., 1997). EM studies of Lethocerus Tn-Tm complexes (Wendt et al., 1997) suggest that the globular portion of the Tn complex lies not along the Tm as in vertebrates (Figure 6A) but at the Tm dimer overlap region (Figure 6B). While sequence homology strongly implies structural homology of the Drosophila Tn-Tm complex to that of vertebrates, common physiology argues for a structural similarity to Lethocerus. However, the orientation of the Lethocerus Tn complex with respect to the Tm N- and C-termini could not be determined from electron micrographs (Wendt et al., 1997). When their figure (Figure 6B) is inverted (Figure 6C) it shows that the Tm, TnT and F-actin in both insects and vertebrates could maintain the same relative position though the remainder of the troponin complex is located differently.

View larger version (51K): [in this window] [in a new window]   Figure 6.   ._Cartoon of tropomyosin-troponin complex arrangements redrawn from Wendt et al., (1997) to illustrate that an inversion of their diagram for Lethocerus would lead to a similar orientation between the Tm and TnT in vertebrate and invertebrate troponin-tropomyosin- complexes. (A) vertebrate, (B) Lethocerus redrawn from Wendt et al. (1997 and (C) proposed reorientation. The N- and C- termini labeling of Tm and TnT in (C) is based solely on the implied structural homology to the vertebrate structure.

Functional Implications

Current models (Geeves and Lehrer, 1998; Squire and Morris, 1998) suggest that in the absence of Ca²⁺, TnI binds actin maintaining the `blocked' filament state (Geeves and Lehrer, 1998) with Tm occluding much of the myosin binding site on actin. Ca<sup>2+</sup> binding to TnC changes TnC/TnI conformation, releasing the TnI inhibitory domain from actin, producing the `closed' state, and Tm with its attached Tn complex, can now move across the F-actin surface. This exposes the complete myosin binding surface on actin (`open' state), activating the actomyosin cross-bridge cycle.

TnI sequence alignments (not shown) show that Drosophila TnI residue A116, (A116V in hdp²) coincides with conserved residue A25 of vertebrate skeletal muscle isoforms. A25 contributes to a contact with the `E' -helix of TnC (Vassylyev et al., 1998). The hdp<sup>2</sup> A116V substitution increases residue size which must affect TnC-TnI binding by changing TnI -helix and TnC `E' -helix interactions. These could result in hdp² either lowering the threshold for calcium activation or affecting the ability of the Tn-Tm complex to return to the relaxed state. We cannot yet distinguish which explanation is correct for hdp² but the latter one was previously proposed for a hypercontracting TnT mutant in Caenorhabditis elegans (McArdle et al., 1998) and would readily explain hypercontraction of hdp² IFMs. The less extreme effects on nonflight muscles argue strongly that hdp² can produce near-normal thin filament regulation.

The S185F substitution introduces a more bulky, hydrophobic residue. Its heptad `c' position means that the phenylalanine ring will point outwards from the Tm axis. It could suppress hdp² either by altering Tn-Tm complex movement across the F-actin surface or by changing the Tn complex orientation on Tm through the Ca²⁺ sensitive TnT-Tm binding site. The former suggestion is supported by observations (Bing et al., 1998) that the Act88F gene E93K mutation reduces Tm movement across the F-actin surface in vitro and suppresses hdp² IFM hypercontraction in vivo (Nongthomba and Sparrow, unpublished). In the latter suggestion S185F would reverse the conformational effects of hdp² on the TnI/TnC interaction by altering structural relationships between Tm and the Tn complex. Reversal of the TnI/TnC conformational effects of hdp² could require very specific changes such as found with the intragenic D3 suppressor, a TnI mutation (L188F) which suppresses hdp² (Prado et al., 1995) perhaps by changes in actin binding (L188 is close to an actin -binding domain). However, alterations in Tm mobility could affect any Tn mutant which causes slight regulatory changes. The partial suppression of up¹⁰¹ by D53 might thus seem to favor this latter explanation, except for the observation that up¹⁰¹, as a TnT mutation, is also located in the Ca²⁺-sensitive TnT-Tm interaction site.

Clinical Implications

Human hypertrophic cardiomyopathies (HCM) include two dominant mutations, D175N and E180G, in highly conserved residues (Figure 5) of the TPM1 gene (Thierfelder et al., 1994). As with D53, these mutations map within the tropomyosin region which interacts with TnT in a Ca²⁺-sensitive manner. In vitro motility experiments using reconstituted cardiac thin filaments showed that these HCM mutations affect the Ca²⁺ sensitivity of the fraction of filaments moving compared with wild-type, although in the absence of structural information on the Ca²⁺-sensitive Tm binding site for TnT, it is not clear how these mutations could cause this (Bing et al., 2000). There is evidence that HCM TnT mutations, I79N and R92Q, located within the same TnT-Tm interaction (residues 70-180) may affect Ca²⁺ sensitivity in tension versus force measurements (see Tobacman et al., 1999 for discussion). Further biochemical studies of all these human and Drosophila Tm and TnT mutants are required to ascertain the role of this Ca²⁺-sensitive contact between Tm and TnT in thin filament regulation.

Typically HCM symptoms do not manifest before adolescence, but penetrance and expressivity of HCM mutations are highly variable (Coonar and McKenna, 1997). This suggests that HCM mutations have relatively mild effects, but the functional deficits can induce cardiac changes, such as hypertrophy, leading to premature death. The variable expression of HCM mutations must be partly due to genetic background. Modifier genes may reduce or enhance HCM severity by affecting either the contractile machinery or the response to cardiac dysfunction. D53 is similar to HCM mutations in that reduced IFM function is detectable only in older organisms. The D53 suppression of the hdp² and up¹⁰¹ progressive myopathies suggests that Drosophila muscles may provide a useful genetic model with which to study HCM genes and their genetic modifiers.