The glutamic acid-rich–long C-terminal extension of troponin T has a critical role in insect muscle functions

The troponin complex regulates the Ca2+ activation of myofilaments during striated muscle contraction and relaxation. Troponin genes emerged 500–700 million years ago during early animal evolution. Troponin T (TnT) is the thin-filament–anchoring subunit of troponin. Vertebrate and invertebrate TnTs have conserved core structures, reflecting conserved functions in regulating muscle contraction, and they also contain significantly diverged structures, reflecting muscle type- and species-specific adaptations. TnT in insects contains a highly-diverged structure consisting of a long glutamic acid–rich C-terminal extension of ∼70 residues with unknown function. We found here that C-terminally truncated Drosophila TnT (TpnT–CD70) retains binding of tropomyosin, troponin I, and troponin C, indicating a preserved core structure of TnT. However, the mutant TpnTCD70 gene residing on the X chromosome resulted in lethality in male flies. We demonstrate that this X-linked mutation produces dominant-negative phenotypes, including decreased flying and climbing abilities, in heterozygous female flies. Immunoblot quantification with a TpnT-specific mAb indicated expression of TpnT–CD70 in vivo and normal stoichiometry of total TnT in myofilaments of heterozygous female flies. Light and EM examinations revealed primarily normal sarcomere structures in female heterozygous animals, whereas Z-band streaming could be observed in the jump muscle of these flies. Although TpnT–CD70-expressing flies exhibited lower resistance to cardiac stress, their hearts were significantly more tolerant to Ca2+ overloading induced by high-frequency electrical pacing. Our findings suggest that the Glu-rich long C-terminal extension of insect TnT functions as a myofilament Ca2+ buffer/reservoir and is potentially critical to the high-frequency asynchronous contraction of flight muscles.


(ABSTRACT)
The troponin complex regulates the Ca 2+activation of myofilaments during striated muscle contraction and relaxation. Troponin genes emerged 500-700 million years ago during early animal evolution. Troponin T (TnT) is the thin filament-anchoring subunit of troponin. Vertebrate and invertebrate TnTs have conserved core structures, reflecting conserved functions in regulating muscle contraction, and also contain significantly diverged structures, reflecting muscle type-and species-specific adaptations. TnT in insects contains a highly diverged structure consisting of a long glutamic acid-rich C-terminal extension of ~70 residues with unknown function. We found here that C-terminally truncated Drosophila TnT (TpnT-CD70) retains binding of tropomyosin, troponin I, and troponin C, indicating a preserved core structure of TnT. However, the mutant TpnT CD70 gene residing on the X chromosome resulted in lethality in male flies. We demonstrate that this X-linked mutation produces dominant-negative phenotypes, including decreased flying and climbing abilities, in heterozygous female flies. Immunoblot quantification with a TpnTspecific monoclonal antibody indicated expression of TpnT-CD70 in vivo and normal stoichiometry of total TnT in myofilaments of heterozygous female flies. Light and electron microscopy examinations revealed primarily normal sarcomere structures in female heterozygous animals, whereas Z-band streaming could be observed in the jump muscle of these flies. Although TpnT-CD70expressing flies exhibited lower resistance to cardiac stress, their hearts were significantly more tolerant to Ca 2+ overloading induced by high-frequency electrical pacing. Our findings suggest that the Glu-rich long C-terminal extension of insect TnT functions as a myofilament Ca 2+ buffer/reservoir, potentially critical to the high-frequency asynchronous contraction of flight muscles.
Muscle contraction is a vital function of animals. Striated muscles represent a highly differentiated tissue type characterized by bundles of myofibrils consisting of tandem repeats of sarcomeres that make up the contractile units [1]. The sarcomeres are formed by overlapping myosin thick filaments and actin thin filaments. In vertebrate skeletal and cardiac muscles, contraction is powered by actin-activated myosin ATPase under the regulation of intracellular Ca 2+ via the troponin complex in the thin filament [1][2][3]. The myofilament protein contents of vertebrate and invertebrate striated muscle sarcomeres are conserved, both containing the thin filament regulatory proteins tropomyosin and troponin [4]. The troponin complex has three protein subunits: The calcium-binding/receptor subunit troponin C (TnC), the inhibitory subunit troponin I (TnI), and the tropomyosinbinding/thin filament-anchoring subunit troponin T (TnT) [3][4][5].
Troponin emerged during early animal evolution approximately 700 million years ago. Genes encoding the subunits of troponin are found in all invertebrate animals higher than Cnidaria (jellyfish and sponges) [6]. This phylogenetic chronicle corresponds to the emergence of central nervous system [7], implying an essential role of troponin in the more coordinated muscle contractions of higher animal species. To study the molecular evolution of troponin by characterizing its conservation and diversity in invertebrate and vertebrate muscles is a powerful approach to understand the structure-function relationship of troponin in determining muscle contractility in physiological and pathophysiological conditions.
By interaction with tropomyosin in the thin filament, TnT plays an organizer role in the center of the sarcomeric Ca 2+ regulatory system [8]. Vertebrates have evolved with three homologous genes encoding the cardiac, fast and slow skeletal muscle type isoforms of TnT, while most invertebrates such as insects have only a single TnT gene [6]. Gene structure and molecular evolution studies identified the fast skeletal muscle TnT as the ancestral form of the three vertebrate muscle type-specific isoforms [9]. The phylogenetic tree in Fig. 1A, constructed from protein sequences of representative invertebrate TnT and vertebrate fast TnT, shows their evolutionary lineage and overall divergence.
Previous studies have extensively demonstrated the critical role of troponin in the Ca 2+ -regulation of insect muscle contraction [10]. The Drosophila TpnT locus on chromosome X encodes several physiologic alternative splicing variants whereas aberrant splicing or missense mutations cause muscle abnormalities such as the upheld and indented thorax phenotypes [11], muscleblind myotonic dystrophy [12], or abnormal development of flight muscles [13].
Linear structure comparison between vertebrate and Drosophila TnT (Fig. 1B) shows a large portion of conserved structures, corresponding to their conserved core functions. Vertebrate and invertebrate TnTs both have an alternatively spliced N-terminal variable region [5,14] and a pair of mutually exclusively alternative spliced C-terminal coding exons, i.e., exons 16 and 17 of vertebrate fast skeletal muscle TnT [15,16] and exons 10A and 10B of Drosophila TnT [13]. After several decades of research, the functional significance of these variable structures of TnT remains not fully understood.
A diverged structure of invertebrate TnT found in flying insects is a long C-terminal extension with high contents of glutamic acid residues (Fig. 1B) [11]. The function of this highly diverged structure of insect TnT and its fitness value during natural selection are of significant interest. The N-terminal variable region of some vertebrate TnTs also has high Glu contents [5]. This similar trait is most striking in the fast skeletal muscle TnT specifically expressed in adult avian pectoral muscles [17]. This potentially convergent structural similarity can be clearly seen in the paired dot plot comparison between eagle pectoral muscle TnT and Drosophila TnT (Fig.  1C). Considering that avian leg muscle TnT has a much shorter N-terminal variable region [18] and fast TnT of emu, a flightless bird, has lost this structure (Fig. 1B), we propose that the long Glu-rich segments in insect TnT and avian pectoral muscle TnT may have both evolved from an analogous evolutionary selection for flight abilities.
Supporting this hypothesis, a previous study found that the Glu-rich N-terminal segment of avian pectoral muscle TnT has a physiologically relevant Ca 2+ -binding capacity [19], which may play a novel role in the Ca 2+regulation of muscle contraction. In the present study, we genetically engineered flies to demonstrate that deletion of the C-terminal 70 amino acids of Drosophila TnT (TpnT-CD70) significantly decreases muscle and heart functions, establishing a biological importance of the Glu-rich C-terminal extension of insect TnT. In the meantime, fly hearts expressing TpnT-CD70 exhibited significantly increased tolerance to Ca 2+ overloading during high frequency electrical pacing, supporting the potential role of the Glu-rich segment in TnT as a myofilament Ca 2+ buffer/reservoir.

Cloned cDNAs encoding two predominantly expressed and one novel RNA splicing variants of Drosophila TpnT
cDNAs encoding three N-terminal alternative splicing variants of Drosophila TpnT were cloned via RT-PCR from total RNA extracted from whole adult flies. The corresponding protein variants are compared in the linear structural maps in Fig. 2. A novel splice form of Drosophila TnT with the exon 5-encoded segment in the N-terminal variable region excluded and exon 10A in the Cterminal mutually exclusive splicing region was identified among the cloned cDNAs. The sequence has been deposited to GenBank with accession number MK227440.
The other two cDNA variants cloned encode the high and low molecular weight TnT splice forms that are predominantly expressed in normal adult Drosophila muscles (Fig. 2).

Production of C-terminal truncated Drosophila TnT in E. coli
Intact Drosophila TnT proteins exhibited significant toxicity to host E. coli cells and did not yield a preparative level of expression (data not shown). In sharp contrast, Drosophila TpnT-CD70 exon 10A and TpnT-CD70 exon 10B proteins with the C-terminal Glu-rich segment removed express in E. coli at high levels ( Fig. 3). Both exon 10A and exon 10B variants of TpnT-CD70 proteins were purified using conventional biochemical methods ( Fig.  3) for functional characterization.

C-terminal truncated Drosophila TnT retains the ability to bind tropomyosin, TnI and TnC
Since the C-terminal Glu-rich extension is an evolutionarily diverged structure in invertebrates such as insects, we anticipated its deletion would not destroy the core functions of Drosophila TnT. To validate this hypothesis, protein binding studies showed that both exon 10A and exon 10B variants of Drosophila TpnT-CD70 have high affinity binding to rabbit a-tropomyosin (Fig. 4A), bovine cardiac TnI (Fig. 4B), and chicken fast skeletal muscle TnC in the presence of 0.1 mM CaCl2 (Fig. 4C) or 0.1 mM EGTA (Fig. 4D), comparable with the control of chicken fast skeletal muscle TnT that is a native binding partner of the vertebrate tropomyosin, TnI and TnC used in the present study to overcome the lack of purified insect myofilament proteins. The biochemical assessment demonstrates that deletion of the Cterminal Glu-rich segment does not abolish the core functions of Drosophila TnT.
Within physiological range, Drosophila TpnT-CD70 proteins showed lower affinity for vertebrate tropomyosin but higher affinity for vertebrate TnI (Fig. 4 A and 4B). The binding of TpnT-CD70 proteins to TnC is Ca 2+dependent, similar to that of chicken fast TnT (Fig. 4D) and consistent with previous studies of vertebrate troponin [20]. The preserved biochemical activities of TnT-CD70 proteins demonstrate their preserved core functions in muscle thin filament regulation.
The presence of the mutually exclusive alternative exon 10A or 10B produced different binding affinities for tropomyosin and TnC (Fig. 4), indicating functional impacts of this alternatively spliced segment of TnT, which are under investigation together with vertebrate counterparts, exon 16 and exon 17, in fast TnT [5,6].

TpnT-CD70 mutant flies
With the evidence that deletion of the Cterminal Glu-rich segment does not abolish the core function in Drosophila TpnT-CD70 protein (Fig. 4), TpnT-CD70 mutant flies were developed. Illustrated in Fig. 5, the mutant allele was constructed by inserting a stop codon in the exon 11 of TpnT gene on chromosome X using CRISPR/Cas9 genomic editing to truncate the last 70 amino acids of the polypeptide chain and delete the C-terminal Glu-rich segment. The RFP selection marker was subsequently removed by crossing with a Cre line to establish a line of RFP -TpnT CD70 mutant fly. After confirmation by DNA sequencing of the targeted region, phenotypic studies were performed using both RFP + and RFPmutant flies.
Although the CD70 truncation of Drosophila TnT does not abolish its biochemical functions (Fig. 4), the TpnT CD70 allele produced embryonic lethality in vivo. Shown in the segregation pattern during breeding (Table 1), the X-linked TpnT CD70 mutation causes 100% death of mutant males before hatching. Therefore, TpnT CD70 homozygous females would be lethal as well although test breeding could not be done due to the lack of mutant male parents.
With the presence of one wild type X chromosome, heterozygous TpnT CD70/+ females are viable and fertile, appearing in the expected Mendelian ratio (Table 1). Therefore, TpnT CD70/+ females were used as subjects for phenotypic characterization.

A high affinity mAb specifically recognizes all alternative splice forms of Drosophila TnT and the CD70 truncation
From immunization using Drosophila TpnT-CD70 protein, a mouse hybridoma cell line secreting anti-Drosophila TnT mAb (2C12) was established after multiple rounds of subcloning. Characterization of hybridoma cell culture supernatant using Western blotting showed that mAb 2C12 specifically recognizes TnT bands in total muscle homogenates of adult fly and total protein extracts of whole larvae without detectable cross reaction to any other proteins (Fig. 6A). The Western blot showed that mAb 2C12 does not cross react with cardiac or skeletal muscle TnT in multiple vertebrate species including the ancestral species hagfish (Fig. 6A), indicating its recognition of a fly TnT-specific epitope.
Shown in the high sensitivity ELISA titration of 2C12 hybridoma culture supernatant in Fig. 6B, mAb 2C12 has equally high affinities against the exon 10A and exon 10B variants of Drosophila TpnT-CD70 with no cross reaction to chicken breast muscle fast TnT and human cardiac TnT, confirming its specificity to fly TnT under nondenaturing conditions. This new mAb provides a critical tool to examine the expression of TpnT-CD70 proteins in mutant fly muscle and heart in order to establishing the causal effect on functional phenotypes.
The Western blot data in Fig. 6A further show that there are two major TnT splice forms expressed in adult Drosophila muscles, of which the low molecular weight form is predominant in the indirect flight muscle (IFM) whereas the adult jump muscles express both splice forms at similar levels. In contrast, the total protein extracts from Drosophila larvae predominantly express the high molecular weight splice form (Fig. 6A).
With protein controls expressed in E. coli from cloned cDNAs (Fig. 7), the high and low molecular weight splice forms of TnT predominantly expressed in adult Drosophila muscles correspond to the two previously identified N-terminal alternative splicing variants shown in Fig. 2. An intermediate size TnT splice form was detected in Drosophila larva (Fig. 6A), of which the exon inclusion pattern and developmental and functional significance remain to be investigated.

Expression and myofilament integration of TpnT-CD70 protein in mutant fly muscles
Age-matched (5 and 14 days old as noted for each experiment) female TpnT CD70/+ heterozygote and w 1118 control flies were studied for muscle and heart phenotypes. We first examined the expression of TpnT-CD70 protein in mutant fly muscles in vivo to validate this new experimental model. mAb 2C12 Western blots on total protein extracted from flight and jump muscles isolated from TpnT CD70/+ heterozygote female flies detected significant amounts of two variants of truncated TnT proteins (Fig. 7). To confirm them as the C-terminal truncated version of the two N-terminal alternative splice forms ( Fig. 2) that are predominantly expressed in adult fly muscles (Fig. 6A), we showed their SDS-gel mobility identical to that of the corresponding CD-70 protein size controls expressed in E. coli from cloned cDNA (Fig. 7).
The levels of TpnT-CD70 proteins relative to their intact counterparts in TpnT CD70/+ heterozygote female fly muscles were calculated from Western blot densitometry (Fig.  7). Removal of the RFP cassette significantly increased the expression of TpnT-CD70 protein in both in IFM and jump muscle (Fig.  7), indicating a negative effect of the RFP cassette on gene transcription and/or mRNA stability.
The ratios of high and low molecular weight splice forms in jump muscles were similar for the intact and CD70 forms (Fig. 7B), indicating that the insertion of stop codon in exon 11 did not affect the alternative splicing pattern of the N-terminal coding exons (Fig. 2).
Densitometry quantification of total TnT detected in Western blots vs the level of actin determined in parallel SDS-gels showed similar stoichiometry in wild type and transgenic fly muscles (Fig. 7), excluding haploid insufficiency from possibly causing the lethality of TpnT CD70 male embryos or the phenotypes of heterozygote females.
Previous studies have determined that the level of troponin protein detected in striated muscle cells reflects the actual amount integrated in the myofilaments [21] and nonmyofilament-integrated TnT is rapidly degraded in muscle cells [22]. Therefore, the protein stoichiometry study demonstrated that TpnT-CD70 is effectively integrated into the myofilaments in Drosophila muscle in vivo.
The in vivo protein expression and stoichiometry data established that dominant negative effects of the C-terminal truncation of Drosophila TnT causes embryonic lethality in males and the phenotypes of heterozygote females, justifying the TpnT-CD70 fly model for use as an informative system to investigate the functional significance of the Glu-rich long C-terminal extension of insect TnT.

Decreased muscle functions in TpnT CD70/+ flies
To investigate the functional effect of deleting the insect-specific C-terminal Glu-rich extension of TnT on muscle functions in vivo, acute flight ability was assessed in heterozygote mutant (TpnT CD70/+ , without RFP) and wild type (w 1118 ) flies. The results in Fig.  8A show that the average landing height of the mutant flies was decreased, indicating a negative impact of TnT-CD70 on the function of flight muscles.
Climbing velocity of the mutant flies is also significantly decreased compared to wild type control (Fig. 8B), consistent with decreased muscle functions.
Reduced function is also apparent when endurance is tested by repetitive induction of climbing (Fig. 8C), indicating dysfunction in both acute and chronic muscle activities.

Decreased cardiac function in TpnT CD70/+ flies
Assessed for functional capacity by pacing at 6 Hz for 30 sec, the hearts of TpnT CD70/+ mutant flies were significantly more sensitive to the stress pacing compared to wild type control ( Fig. 9A), demonstrating that the Cterminal Glu-rich extension of TnT is required for normal cardiac functional durability.

Increased tolerance of TpnT CD70/+ fly hearts to short-duration high frequency pacing
To test the hypothesis that the Glu-rich Cterminal extension of insect TnT functions as a myofilament Ca 2+ buffer/reservoir, as previously suggested for the N-terminal Glurich segment of avian pectoral muscle TnT [19], mutant flies were tested in short durations of high frequency cardiac pacing. In contrast to the decreased tolerance to overall cardiac stress shown in Fig. 9A, TpnT CD70/+ hearts were more resistant to 5 sec episodes of high frequency pacing, with fewer fibrillation (Fig. 9B) and arrest ( Fig.9C) events than wild-type control. This result suggests that the mutant fly hearts are more resistant to acute Ca 2+ overloading produced by the electrical pacing-induced membrane depolarization of cardiomyocytes. This novel finding supports the notion that removal of the long Glu-rich C-terminal extension of Drosophila TnT diminishes a Ca 2+ buffer in the myofilaments, resulting in a lower baseline content of Ca 2+ , which allows cardiomyocytes to better tolerate acute Ca 2+ overloading.

Preserved myofibril and sarcomere structure in TpnT CD70/+ flies
Whole-mount imaging of heart and muscle tissues of age-matched TpnT CD70/+ and control flies showed similar striation patterns without notable overall or regional myofibril disorganization (Fig. 10A). Quantitative analysis showed that the resting sarcomeres are longer in the heart (Fig. 10B) and jump muscle (Fig. 10C) of TpnT CD70/+ flies than that in the controls. In contrast, it is not different in IFMs of TpnT CD70/+ and control flies (Fig. 10D), likely a phenotypic difference between synchronous and asynchronous muscles.
The ultrastructure of TpnT CD70/+ mutant fly muscles was examined using transmission electron microscopy (EM) at 14 days of age. In comparison with w 1118 controls, primarily normal sarcomeric structure are seen in both IFM and jump muscle (Fig. 11). TpnT CD70/+ jump muscle but not IFM showed signs of Zband streaming (Fig. 11), an evidence of the negative impact of deleting the long Glu-rich C-terminal extension of TnT.

DISCUSSION
By interactions with TnC, TnI and tropomyosin, TnT sits at a central position in the thin filament regulatory system of striated muscles. It has been extensively studied as an indicator to understand the evolution of vertebrate muscle types. The structural divergence among TnT isoforms and during evolution provides valuable information to understand the structure-function relationship of troponin.
TnT is encoded by a single gene in insects [6]. A significantly diverged structure of insect TnT is the Glu-rich long C-terminal extension [11]. Intrigued by the presence of similarly Glu-rich segment in the N-terminal hypervariable region of vertebrate TnT (Fig. 1), especially the fast TnT in adult avian pectoral muscles [18], we tested the hypothesis that such Glu-rich segments in TnT may have an analogous function that has been convergently selected during the evolution of flight muscles.
Our study found that deletion of the Glurich C-terminal extension of Drosophila TnT results in dominantly negative phenotypes in muscle and heart. In the meantime, TpnT CD70/+ fly hearts exhibited significantly increased tolerance to Ca 2+ overloading, supporting a novel role of the Glu-rich segment in TnT as a myofilament Ca 2+ buffer/reservoir. The following findings provide new insights into key aspects of troponin function in striated muscle myofilament regulation.

Removal of the C-terminal Glu-rich segment from Drosophila TnT does not abolish core biochemical activities
Previous studies in vertebrates have located two tropomyosin binding sites of TnT, one in the middle conserved region and the other in the beginning of the C-terminal T2 region where the TnI and TnC binding sites are also located [23] (Fig. 1B). Amino acid sequence alignment showed that the regions containing binding sites for tropomyosin, TnI and TnC are conserved between vertebrates and Drosophila TnTs ( Fig. 1B and 1C). Drosophila TnT lacking the C-terminal 70 amino acids retains physiological binding to tropomyosin, TnI, and TnC (Fig. 4).
The ELISA protein binding test involves multiple high stringency washes, and thus reflects high binding affinities of TpnT-CD70 protein for its three known physiological partners in the myofilaments. Although we did not have an intact Drosophila TnT protein as control due to the toxicity in bacterial expression, the binding affinities of TpnT-CD70 proteins to vertebrate tropomyosin, TnI and TnC are comparable with that of chicken fast TnT, a native binding partner of vertebrate tropomyosin, TnI and TnC used in the assay (Fig. 4), demonstrating preserved core functions of TpnT-CD70 proteins.
The preserved tropomyosin, TnI and TnC binding sites in TpnT-CD70 protein also supports that the C-terminal extension of insect TnT is an evolutionarily added structure similar to the N-terminal variable region of vertebrate TnT [9]. Since the C-terminal extension is common among flying insects, it may serve an evolutionarily selected function with a fitness value in insect muscle specific to high efficiency flight activities.

mAb 2C12 provides a common tool for Drosophila muscle and heart research
The 2C12 mAb that we developed in the present study shows high affinity and specificity to Drosophila TnT in Western blotting and ELISA (Fig. 6). It is able to detect intact and CD70 truncated Drosophila TnT as well as spliced variants containing various combinations of alternative exons in the Nterminal variable region (exons 3, 4 and 5) and the C-terminal region (exons 10A and 10B) ( Fig. 6 and Fig. 7A). Therefore, mAb 2C12 must recognize an epitope encoded by constitutively spliced exon(s) in the middle region of Drosophila TnT. The fact that mAb 2C12 does not recognize any vertebrate TnT (Fig. 6A) reflects a significantly diverged structure in Drosophila TnT, which is not directly involved in the binding sites for other myofilament proteins. The data show that mAb 2C12 provides a useful tool for the detection of Drosophila TnT in muscle and heart studies.

Identification of the two major N-terminal alternative splice forms expressed in adult fly muscles
In addition to identification and cloning of a novel splice form of Drosophila TnT from whole body RNA extract (Fig. 2), we have cloned cDNAs that encode the major high and low molecular weight TnT protein variants expressed in normal adult Drosophila muscles. By expressing the cloned cDNAs in E. coli for use as protein gel mobility controls, we identified the major splice form of TnT expressed in thoracic muscles (Fig. 7A) as the low molecular weight variant with exons 3, 4 and 5 excluded (Fig. 2), consistent with a previous report [13]. In jump muscles, two major splice forms of TnT are present (Fig. 7B), of which the low molecular weight variant is the same as the major form in thoracic muscles whereas the high molecular weight variant has the alternative exons 3, 4 and 5 all included (Fig. 2).
Expression of cDNAs encoding CD70truncated Drosophila TnT with known Nterminal variations in bacterial culture provided control proteins for use as SDS-gel mobility markers to identify C-terminal truncated TnT variants in the muscles of TpnT CD70/+ mutant flies (Fig. 7). The results verified that the insertion of CD70 stop codon in exon 11 of TpnT gene did not alter the native N-terminal alternative splicing pattern of TpnT RNA, further validating the TpnT CD70/+ mutant fly for use in focused studies on the function of the C-terminal Glu-rich extension of TnT.

Embryonic lethality of TpnT CD70 mutation
The embryonic lethal phenotype of TpnT CD70 mutation causing 100% male lethality (Table 1) required validation since the C-terminal extension truncated TnT retains core functions (Fig. 4), effectively expresses in myocytes and integrates into myofilaments (Fig. 7). The mutant flies were generated by CRISPR/Cas9 gene editing that may cause offtarget mutations which could be X-linked and lethal. A more likely explanation is that the effect of deleting the Glu-rich C-terminal extension of Drosophila TnT on decreasing muscle contractility as seen in heterozygote female flies may be critical to the contractility of early embryonic muscles in generating sufficient force for successful hatching. Detailed developmental studies are merited to identify the primary mechanism for the embryonic lethality of TpnT CD70 mutation.

Dominantly negative phenotypes of TpnT-CD70 in muscle and heart
Decreased functions are observed in both flight and leg muscles, as well as in cardiac muscle of TpnT CD70/+ flies, indicating the loss of a commonly important function. As no other muscle function-related genes are present near the TpnT locus, the significantly decreased muscle and heart functions of TpnT CD70/+ flies are attributed to dominantly negative effects of the mutation.
The phenotypes of TpnT CD70/+ flies are distinct from that reported previously for splicing or single amino acid missense mutations such as upheld and indented thorax [11], muscleblind myotonic dystrophy [12], and abnormal muscle development [13]. In contrast to destructions of conserved core structure or a complete loss of function reported the previous studies, TpnT-CD70 mutant produces unique dominant effects on muscle and heart functions ( Fig. 8 and Fig. 9A) with various proportions of the truncated protein (Fig. 7) to indicate a critical role of the Glu-rich C-terminal extension of Drosophila TnT.
Previous studies have documented an inverse correlation between baseline free Ca 2+ and sarcomere length [24]. Previous studies also established that decreased intracellular calcium weakens muscle contraction [25]. An increase of sarcomere length in cardiac muscle at diastole is an adaptation to utilize Frank-Starling mechanism for compensating weakened contractility [26], which is also a fundamental property of skeletal muscles [27]. Although the overall development, growth and structure of TpnT CD70/+ fly muscles are apparently normal, the increased resting length of sarcomeres in jump muscle and heart but not in IFM (synchronous vs asynchronous muscles) (Fig. 10) may correspond to muscle weakness as well as to indicate the lower intracellular Ca 2+ suggested by the cardiac pacing experiment (Figs. 9B and 9C).
EM images detected the presence of Z-band streaming in TpnT CD70/+ fly jump muscle (Fig.  11), which is often a sign of workload and/or muscle weakness-caused injury [28]. This mechanical load and stretch-induced phenotype was not detected in IFM of the same flies. Therefore, the potentially synchronous muscle-specific impact of TpnT-CD70 is worth further investigation. To study the contractility of isolated muscle preparations from TpnT CD70/+ flies should provide mechanistic insights into the physiological function of the Glu-rich C-terminal extension of insect TnT.

Potential role of the Glu-rich C-terminal extension of insect TnT as a myofilament Ca 2+ reservoir
While the overall cardiac function is decreased (Fig. 9A), consistent with the skeletal muscle phenotypes in Fig. 8, the hearts of TpnT CD70/+ flies showed a significantly higher tolerance to Ca 2+ overloading with less cardiac arrest/fibrillation during high frequency pacing (Fig. 9B and Fig. 9C). This finding implicates a lower baseline Ca 2+ content in the cardiomyocytes.
The insect muscle-specific Glu-rich Cterminal extension of TnT shows a striking similarity in amino acid sequence to the Nterminal variable region of avian pectoral muscle TnT (Fig. 1C). We previously found that the also Glu-rich segment of avian pectoral muscle TnT has an intriguing capacity to bind Ca 2+ with an affinity of ~14 µM, just below that of the regulatory Ca 2+ -binding sites of TnC [19]. It is interesting to note that this segment is absent in the TnT of emu, a flightless bird (Fig.  1B). Therefore, this unique structure may function as a myofilament Ca 2+ reservoir residing closely to TnC as a local Ca 2+ buffer that might reduce the amount of Ca 2+ to be cycled between cytoplasm and sarcoplasmic reticulum (and the extracellular compartment) during muscle contraction and relaxation. This mechanism would reduce the energy expenditure of Ca 2+ pumps and increase the energetic efficiency of muscle, which is important for sustaining the highly demanding muscle work during flight, especially the high frequency and asynchronous contraction of insect IFM.
In this hypothesis, the deletion of the Glurich C-terminal 70 amino acids of Drosophila TnT removes a potential myofilament Ca 2+ reservoir/buffer to decrease the level of free intracellular Ca 2+ in resting myocytes. This effect may chronically lower the total calcium contents in the myocytes to decrease contractility [25], as shown by the overall muscle weakness in TpnT CD70/+ flies. The loss of the Ca 2+ reservoir function of TpnT-CD70 protein to reduce the total calcium content of myocyte is supported by the increased tolerance of the mutant fly heart to pacinginduced calcium overloading ( Fig. 9B and Fig.  9C).
This notion suggests a novel mechanism in the calcium-regulation of straited muscle contraction. Further studies by imaging Ca 2+ dynamics in beating TpnT CD70/+ fly hearts may validate this potentially underlying mechanism. To overcome the limitation that whole fly body electrical pacing is not cardiac muscle specific, the use of an optogenetic reagent, channelrhodopsin-2, for heart-specific and noninvasive pacing as reported in a recent study [29] may be considered.
In conclusion, the present study suggests that the Glu-rich segments independently evolved in TnT of insect and bird muscles may provide a novel mechanism to facilitate muscle contraction and relaxation as well as to reduce the energy expenditure for calcium cycling in these flying species. To further establish this function of TnT may lead to the development of a new treatment of muscle weaknesses and heart failure.

EXPERIMENTAL PROCEDURES
This research is approved by the Institutional Animal Care and Use Committee of Wayne State University.

Cloning of cDNAs encoding Drosophila TnT and construction of C-terminal deletion mutants
Total RNA was extracted from whole bodies of adult w 1118 flies using the TRIZOL reagent. Two µg of total RNA was used as template for reverse transcription (RT) of poly-A mRNA using an anchored oligo dT primer (TV20). PCR amplification of cDNA corresponding to the entire coding region of Drosophila TpnT mRNA was performed using a pair of primers, one specific to the translation initiation codon site in exon 2, the other specific to the translation termination codon site in exon 11. The amplified cDNA with expected size was digested at a 5' NdeI site and a 3' EcoRI site introduced by the PCR primers for insertion into NdeI and EcoRI-cut pAED4 plasmid. Colonies of JM109 E. coli transformed with the recombinant plasmids were screened using PCR.
To identify cDNA clones containing the mutually exclusive exon 10A or exon 10B, which cannot be distinguished by size, PCR using exon 10A-and exon 10B-specific primers was employed. A Drosophila TpnT cDNA encoding an N-terminal alternatively spliced low molecular weight variant expressed in flight muscles containing exon 10A in the Cterminal region was reconstructed by exchanging a restriction fragment from an exon 10B cDNA clone to generate a comparable pair of exon 10A/10B variants. cDNAs encoding Cterminal 70 amino acids truncated Drosophila TnT (TpnT-CD70) were then constructed using PCR with a mutagenesis reverse primer to introduce a premature stop codon.
All cloned Drosophila TpnT cDNAs and mutants were confirmed by DNA sequencing.

Expression of Drosophila TpnT-CD70 in E. coli and purification
C-terminal truncated Drosophila TpnT-CD70-exon 10A and TpnT-CD70-exon 10B proteins were expressed in transformed BL21(DE3)pLysS E. coli cultured in 2X TY media containing 100 mg/L ampicillin and 25 mg/L chloramphenicol. After induction with 0.4 mM isopropyl β-D-1 thiogalactopyranosie at mid-log phase in a 37°C shaking incubator for 3 hours, the bacterial cells were harvested by centrifugation.
The purification steps were carried out at 4°C. The cell pellets were resuspended in lysis buffer (50 mM sodium phosphate buffer, pH 8.0, 2 mM EDTA, 1 mM PMSF, 15 mM bmercaptoethanol) and lysed using a French cell press. The bacterial lysate was centrifuged to remove any insoluble materials and fractionated by ammonium sulfate precipitation. 30-50% saturation fraction enriched with TpnT-CD70 protein was dialyzed against deionized water, added solid urea to 6 M, EDTA to 0.1 mM, bmercaptoethanol to 6 mM, PMSF to 0.5 mM, adjusted to pH 5.6 with 20 mM sodium acetate buffer, and centrifuged to remove any precipitates before loading on a CM52 cation exchange column equilibrated in 6 M urea, 20 mM sodium acetate, pH 5.6. The column was eluted with a linear gradient of 0-500 mM KCl and the A280nm peaks were analyzed by SDSpolyacrylamide gel electrophoresis (PAGE) to identify the fractions containing TpnT-CD70 protein. The TpnT-CD70 fractions were pooled, dialyzed against 0.1 mM EDTA and lyophilized.

Development of a specific monoclonal antibody against Drosophila TnT
To develop an anti-Drosophila TnT monoclonal antibody (mAb) for use as detection tool, a short-term immunization procedure was used to immunize a 5 weeks old female Balb/c mouse. As previously described [30], the initial immunization was done by intraperitoneal and intramuscular injections of a total of 100 µg purified TpnT-CD70 protein (1:1 mix of the exon 10A and exon 10B variants) in 100 µL PBS mixed with an equal amount of Freud's complete adjuvant. On day 11 and day 12 after the primary immunization, the mouse was boosted intraperitoneally with 100 µg TpnT-CD70 protein in 150 µL PBS without adjuvant. On day 14, the mouse was euthanized and spleen cells were harvested to fuse with SP2/mIL-6 mouse myeloma cells (ATCC CRL-2016) using 50% polyethaglycol 3500 containing 7% dimethyl sulfoxide. Hybridoma colonies were selected with HAT (0.1 mM hypoxanthine, 0.4 µM aminopterin, and 16 µM thymidine) media containing 20% fetal bovine serum and screened using indirect enzyme-linked immunosorbent assay (ELISA) against purified TpnT-CD70 protein using horse radish peroxidase (HRP)-conjugated anti-mouse immunoglobulin secondary antibody (Santa Cruz). Positive hybridoma clones secreting anti-TpnT mAb were subcloned three or four times using limiting dilution method. Multiple vials of the final generation of positive hybridoma subclones were stored in liquid nitrogen. Exhausted hybridoma cell culture supernatant of mAb 2C12 was collected for characterization and lyophilized in aliquots for uses in the detection of Drosophila TnT.

SDS-PAGE and Western blotting
Protein samples were prepared in SDS-PAGE sample buffer, heated at 80°C for 5 min, and clarified by centrifugation. The samples were electrophorized in 14% SDS-PAGE gel with an acrylamide:bisacrylamide ratio of 180:1 as previously described [31]. The resulting gels were stained with Coomassie Blue R250 and de-stained in 10% acetic acid to visualize the protein bands.
Duplicated SDS-PAGE gels were blotted on nitrocellulose membrane using a semi-dry electrotransfer apparatus from Bio-Rad. The blotted membranes were blocked with 1% bovine serum albumin (BSA) in Tris-buffered saline (TBS, 136.9 mM NaCl, 2.7 mM KCl, 25 mM Tris-HCl, pH 7.4) at room temperature for 45 min and incubated with anti-TpnT mAb 2C12 diluted in TBS containing 0.1% BSA at room temperature for 2 hours. Following washes with TBS containing 0.5% Triton X-100 and 0.05% SDS, the membranes were incubated with alkaline phosphatase-labeled anti-mouse IgG secondary antibody (Santa Cruz), washed as above, and developed in 5bromo-4-chloro-3-indolylphosphate and nitro blue tetrazolium substrate solution as previously described [32].

Microtiter plate ELISA protein binding assay
ELISA solid phase protein binding experiments were performed to assess the binding of C-terminal truncated Drosophila TnT with TnI, TnC and tropomyosin. As described previously [33], purified TpnT-CD70 exon 10A, TpnT-CD70 exon 10B and chicken fast TnT control [17] proteins were coated on 96-well microtiter plates at 2 µg/mL in Buffer A (0.1 M KCl, 3 mM MgCl2, 10 mM PIPES, pH 7.0), 100 µL/well at 4°C overnight. The plates were washed with Buffer T (0.05% Tween-20 in Buffer A) and blocked with Buffer B (Buffer T plus 1% BSA) at room temperature for 1 hour. Serial dilutions of tissue-purified bovine cardiac TnI, rabbit atropomyosin and recombinant mouse fast TnC were added to triplicate wells in Buffer D (Buffer T plus 0.1% BSA) to incubate at room temperature for 2 hours. The TnC binding assay was done in the presence of 0.1 mM CaCl2 or 0.1 mM EGTA.
The core function of TpnT-CD70 in binding these conserved myofilament proteins was examined via anti-TnI mAb TnI-1 [34], anti-a-tropomyosin mAb CH1 [23] and antifast TnC mAb 2C3 [35], respectively, diluted in Buffer D. After incubation with the primary antibodies at room temperature for 1 hour and washes with Buffer T, HRP-conjugated antimouse immunoglobulin secondary antibody was added to the plates and incubated at room temperature for 45 min. After final washes with Buffer T, H2O2-2,2'-azinobis-(3ethylbenzthiazolinesulfonic acid) substrate was added for color development at room temperature. Absorbance at 420 nm was read using an automated microplate reader at 5 min intervals for 30 min. Protein binding curves were derived from data collected at a time point prior to the end of the linear phase of color development.

Generation and maintenance of TpnT-CD70 flies
Drosophila TpnT-CD70 mutation was constructed in w 1118 strain with CRISPR/Cas9 gene editing at a commercial service facility (Well Genetics, Taiwan, ROC). To express Cterminal 70 amino acids-deleted TnT, a stop codon was inserted into the exon 11 sequence of TpnT gene. An RFP selection marker cassette was introduced after the stop codon, which was removed with Cre-loxP recombination after successful establishment of the mutant fly line. The mutant TpnT-CD70 allele was identified using PCR, and both RFP + and RFPlines were separately maintained as heterozygotes in a stable line balanced with FM7a (TpnT-CD70/FM7a, referred to as TpnT CD70/+ in the present study).
The flies were maintained at 25°C and 50% humidity on a 12-hour light/dark cycle with a standard 10% sucrose and 10% yeast diet. Flies were housed at constant density for at least 2 generations prior to phenotype studies. Gravid female TpnT CD70/+ or w 1118 flies were bred in aerated 6 oz bottles capped with grape juice agar plates spread with a small amount of yeast paste [36] for 48 hours, at which time adults were removed and total embryo counted. Embryo were transferred to 6 oz. bottles containing standard 10% sucrose/10% yeast media and allowed to develop at 25°C. Viable adults were phenotypically scored and counted. Experiments were performed in at least duplicate.
The expression of TpnT-CD70 protein in Drosophila muscles was examined using Western blot. Indirect flight muscle and jump muscle were isolated from adult flies anesthetized with Flynap (triethylamine) as described [37]. SDS-PAGE of total muscle protein extract and Western blotting using mAb 2C12 were done as above. Muscles of agematched w 1118 flies and total protein extracted from w 1118 larvae were used as controls.

Examination of Drosophila muscle functions in vivo
Functional studies were performed in 5 and 14 day old flies as previously described [38][39][40][41]. All assays were performed in triplicate with age-matched TpnT CD70/+ and w 1118 (Well Genetics) background controls.
Acute flight ability was assessed by gravity dropping 160 flies per genotype into a cylindrical acrylic tube containing a polycarbonate sheet coated with Tangle-Trap glue. The sheet was removed, placed against a white background, and photographed. Landing height of the flies was assessed using ImageJ software (NIH, Bethesda, MD).
Climbing velocity was assessed by measuring the distance that flies climbed in two seconds. Five vials of twenty flies each were secured in a vial rack and placed in front of a white background. The vial rack was manually lifted and dropped to knock the flies to the bottom of the vial. Flies' progress was photographed two seconds after the drop, with four repetitions. The four pictures were analyzed using ImageJ software. The height of the vials was divided into four quadrants and the total number of flies in each quadrant were used to generate a climbing index as described [39,40].
To assess fatigue tolerance, climbing endurance was measured as described [39]. Eight vials containing twenty flies each per experimental group were placed on a negative geotaxis machine (Power Tower) which induces flies to repetitively climb vials until fatigue. A vial of flies was scored as fatigued when 80% of the flies in the vial stopped initiating the climbing response for three consecutive stimuli.

Cardiac pacing
External electrical cardiac pacing was performed as described previously [41]. Using a modified microscope slide, wire leads were connected to a square-wave electric stimulator that allowed experimental control of heart rate in anesthetized live flies.
To test cardiac stress resistance, flies were paced at 40 volts and 6 Hz for 30 seconds. Immediately after pacing, fly hearts were visually scored through the transparent abdominal cuticle for either arrest or fibrillation. The percentage of flies that experienced either arrest or fibrillation was reported as the failure rate.
To assess the response of fly cardiomyocytes to acute Ca 2+ overloading, a high frequency cardiac pacing protocol was developed. Anesthetized flies were placed on the pacing apparatus as above and paced at progressively increasing frequencies at constant voltage (40V) for 5s episodes with 5s rest intervals. Flies were paced until hearts exhibited initial failure after a pacing episode. Cardiac fibrillation and arrest were scored and plotted separately as mortality curves.

Fluorescence confocal microscopy
Heart and thoracic muscle from 14-day-old kettin GFP (control, DGRC stock no 110855 [42]) or TnT CD70/+ ;;kettin GFP flies were fixed and stained using protocols described previously [43] [44] with the following modifications. Adult control and experimental flies were rapidly dissected ventral side up in relaxing solution (RS) containing 20 mM sodium phosphate buffer, pH 7.0, 5 mM MgCl2, and 5 mM EGTA. Flies were pre-fixed in RS plus 4% formaldehyde (FA) for 30 minutes, washed with relaxing buffer, and thoraces and abdomens separated into wells of a 96-well plate for a second 15 minute incubation in RS and final 15 minute RS+FA fixation. Hearts and thoraces were separately blocked in PBS, 0.05% Tween-20 (PBS-T) plus 10% normal goat serum (NGS) and stained with chicken anti-GFP antibody (Invitrogen) in PBS-T plus 10% NGS and goat anti-chicken and goat antimouse secondary antibodies. Alexafluor 488 goat anti-chicken secondary antibody (Invitrogen) was applied in the same buffer supplemented with Alexafluor 594-Phalloidin (Invitrogen) to co-stain F-actin.
Imaging was performed using a Leica DMI 6000 scope outfitted with a Photometrics Prime 95B CMOS camera and X-light spinning disc. Images were taken using 40x dry and 100x oil immersion lens, and sarcomere length measurement was performed on the 100x lens images using ImageJ software. Ten images were acquired for each sample, and experiments were performed in duplicate.

Transmission electron microscopy
14 days old TpnT CD70/+ and w 1118 flies were euthanized using Flynap and fixed in 1% paraformaldehyde and 1% electron microscopy grade glutaraldehyde in PBS. The specimens were sent for transmission electron microscopic imaging of muscle ultrastructure at the Central Microscopy Research Facility at the University of Iowa.

Data analysis
SDS-gels and Western blots were scanned at 600 psi for data documentation and densitometry quantification. Statistical analyses were done using paired Student's ttest for mAb affinity, protein binding and stoichiometry analysis, one-way ANOVA for fly muscle functional studies, chi-square test for chronic cardiac pacing, and a log-rank test for fatigue resistance and high frequency cardiac pacing studies.

by grants from the National Institutes of Health HL127691 and HL138007 (to JPJ) and AG059683 (to RJW).
Author Conflict of Interest -The authors declare that they have no conflicts of interest with the contents of this article.

FIGURE 2. Splicing variants of Drosophila
TnT cloned in the present study. In addition to cDNAs encoding the predominantly expressed high and low molecular weight (MW) N-terminal splice forms including or excluding the segment encoded by exons 3, 4 and 5, a cDNA encoding a novel alternative splicing variant excluding exon 5 and containing exon 10A has been cloned. The CD70 truncation site engineered to delete the insect-specific Glu-rich C-terminal extension in the two major splice forms is indicated.        TpnT CD70/+ flies showed lower resistance to cardiac stress-pacing at 6 Hz for 30 seconds than the control of w 1118 flies (***P=0.001 in Chi squared test; n=100 per group). (B) When paced for 5 seconds with 5 seconds intervals at progressively increasing frequency to test the effects of Ca 2+ overloading, TpnT CD70/+ fly hearts trend toward lower fibrillation rate at higher frequencies compared to w 1118 controls (P=0.0868 in log-rank test, n≥30). (C) When scored for cardiac arrest, TpnT CD70/+ flies exhibited significantly greater tolerance to high frequency pacing-induced Ca 2+ overloading than age-matched w 1118 controls (****P<0.0001 in log-rank test, n≥43). The experiments were performed in duplicate. Representative whole-mount fluorescence images of heart, jump muscle and IFM tissues from control (upper panels) and TpnT CD70/+ (lower panels) with kettin::GFP to indicate sarcomere Zdiscs (green) and Alexafluor 594-phalloidin for F-actin (red) show similar striation patterns without myofibril disorganization. Quantitative analysis showed that the resting sarcomeres are longer in the heart (B) and jump muscle (C) of TpnT CD70/+ flies but are not different in IFM (D) from age-matched controls, likely reflecting a phenotypic difference between synchronous and asynchronous muscles.