A method to measure myocardial calcium handling in adult Drosophila

Abstract

Rationale

Normal cardiac physiology requires highly regulated cytosolic Ca²⁺ concentrations and abnormalities in Ca²⁺ handling are associated with heart failure. The majority of approaches to identify the components that regulate intracellular Ca²⁺ dynamics rely on cells in culture, mouse models, and human samples. However, a genetically robust system for unbiased screens of mutations that affect Ca²⁺ handling remains a challenge.

Objective

We sought to develop a new method to measure myocardial Ca²⁺ cycling in adult Drosophila and determine whether cardiomyopathic fly hearts recapitulate aspects of diseased mammalian myocardium.

Methods and Results

Using engineered transgenic Drosophila that have cardiac-specific expression of Ca²⁺ sensing fluorescent protein, GCaMP2, we developed methods to measure parameters associated with myocardial Ca²⁺ handling. The following key observations were identified: (1) Control w¹¹¹⁸ Drosophila hearts have readily measureable Ca²⁺-dependent fluorescent signals that are dependent on L-type Ca²⁺ channels and SR Ca²⁺ stores and originate from rostral and caudal pacemakers; (2) A fly mutant, held-up² (hdp²) that has a point mutation in Troponin I and has a dilated cardiomyopathic phenotype demonstrates abnormalities in myocardial Ca²⁺ handling that include increases in the duration of the 50% rise in intensity to peak intensity, the half-time of fluorescence decline from peak, the full duration at half maximal intensity (FDHM) and decreases in the linear slope of decay from 80% to 20% intensity decay; and (3) Hearts from hdp² mutants had reductions in caffeine-induced Ca²⁺ increases and reductions in ryanodine receptor (RyR) without changes in L-type Ca²⁺ channel transcripts compared to w¹¹¹⁸.

Conclusions

Our results show that the cardiac-specific expression of GCaMP2 provides a means to characterize propagating Ca²⁺ transients in adult fly hearts. Moreover, the adult fruit fly heart recapitulates several aspects of Ca²⁺ regulation observed in mammalian myocardium. A mutation in Drosophila that causes an enlarged cardiac chamber and impaired contractile function is associated with abnormalities in the cytosolic Ca²⁺ transient as well as changes in transcript levels of proteins associated with Ca²⁺ handling. This new methodology has the potential to permit an examination of evolutionarily conserved myocardial Ca²⁺ handing mechanisms by applying the vast resources available in the fly genomics community to conduct genetic screens to identify new genes involved in generated Ca²⁺ transients and arrhythmias.

Introduction

Cardiomyocytes require highly regulated Ca²⁺ handling to maintain normal contractile function and intracellular signaling.^{1, 2} Abnormalities in Ca²⁺ handling are associated with cardiac hypertrophy, cardiomyopathies, and heart failure.^3–6 Moreover, altered Ca²⁺ regulation can lead to loss of myocytes, dysregulated contractile function, and arrhythmias that lead to sudden cardiac death.^{1, 7–9}

While the major components involved in myocardial Ca²⁺ regulation are known, the ability to identify new molecules in Ca²⁺ cycling is limited to the availability of genetic models systems that allow for genetic screens. The fruit fly, Drosophila melanogaster, is an excellent genetic model that has vast resources to facilitate the identification and mapping of new genes that cause disease.^{10, 11} Previously, we and others have demonstrated that the fly heart can recapitulate aspects associated with mammalian cardiac contractile dysfunction.^{1, 4, 5, 12–19} Moreover, since many of the genes and signaling pathways are evolutionarily conserved among flies and mammals, studies of the fly cardiac system can provide insight into human heart disease.^{17, 20–25} For example, receptor tyrosine kinase signaling pathways, Notch pathways, and many transcription factors have been identified in the fly, thus providing insights into how corresponding orthologs function in mammals.^{13, 19, 23, 26–28} Therefore, we sought to develop a method to measure Ca²⁺ handling in the adult fly heart. This approach establishes the means to potentially identify new mutants previously not known to be involved in myocardial Ca²⁺ signaling.

The fly circulatory system consists of a single layer of tinC expressing myocytes arranged as an open linear tube with the main cardiac chamber located directly underneath the dorsal cuticle in the proximal abdominal region.^29–31 Additionally, a strap of non-tinC expressing striated muscle cells, referred to as the ventral longitudinal muscle or dorsal diaphragm, is tightly associated with the ventral aspect of the abdominal circulatory system.³² Despite a simple circulatory system, many genes that are critical for cardiac function in the fly are conserved among mammals, including humans. For example, transcription factors such as NKX2.5, structural proteins such as delta-sarcoglycan, dystrophin, myosins, and troponins, and receptor-mediated signaling pathways are required for normal cardiac function in the fly heart.^{13, 15, 18, 19, 33–40}

Based on the concept that genes and pathways necessary for Drosophila heart function are conserved among species, we developed a method to measure Ca²⁺ handling in the intact adult fly heart. Transgenic flies in a w¹¹¹⁸ genetic background that expressed the Ca²⁺-dependent fluorescent reporter GCaMP2, under the direct control of the 304 base pair cardiac-specific tinC genomic element, allowed us to measure myocardial propagating Ca²⁺ transients with high fidelity in adult Drosophila. We characterized myocardial Ca²⁺ transients in control w¹¹¹⁸ flies and in a fly mutant, hdp² that has an enlarged cardiac chamber and impaired contractile function similar to the functional abnormalities observed in mammalian dilated cardiomyopathy.¹⁷ Our results demonstrate that the adult fly heart has similar Ca²⁺ handling properties as compared to mammalian hearts and establish new methods to potentially identify fly mutants that have abnormalities in genes affecting the regulation of myocardial Ca²⁺.

Methods

Materials

w¹¹¹⁸ flies were obtained from Bloomington Stock Center. The hdp² was provided by Dr. James Vigoreaux. GCaMP2 vector was obtained from Addgene. pCasper5 and pGreenHPelican plasmids were obtained from the Drosophila Genome Resource Center. Rhod2-AM and Fluo-4 Ca²⁺ sensitive dyes were obtained from Invitrogen. Cytochalasin D and cadmium chloride were purchased from Sigma. Diltiazem, caffeine, thapsigargin, and tetrodotoxin (TTX) were purchased from EMD chemicals. Probes used for qPCR were obtained from Applied Biosystems, Inc.

Transgenic fly engineering

Transgenic flies harboring cardiac-specific GCaMP2 were generated in a w¹¹¹⁸ genetic background as described in the Online Data Supplement. To examine myocardial Ca²⁺ handling in hdp², the transgenic GCaMP2 line was introduced into the hdp² line that was previously backcrossed into the w¹¹¹⁸ genetic background. Therefore, the differences in fluorescence were attributed to the troponin I mutation since the w¹¹¹⁸ genetic background was used for all experiments.

Optical Coherence Tomography

The cardiac chamber sizes in awake, adult Drosophila were measured using optical coherence tomography as previously described.¹⁷ Adult female flies were collected between five and seven days after eclosion, briefly anesthetized using carbon dioxide exposure, gently placed on soft gel plates, and allowed to awaken before imaging using a 1310 nm OCT microscopy system (Bioptigen, Inc.). M-modes obtained from transverse oriented B-mode through the A1 segment of the Drosophila heart were used to calculate end-diastolic dimensions (EDD), end-systolic dimension (ESD), and heart rate. Fractional shortening was calculated as (EDD-ESD)/EDD × 100.

Heart dissections

Female flies were collected 48 hours after eclosion for myocardial Ca²⁺ fluorescence measurements. Hearts were prepared according to previously described methods.^{41, 42} Detailed methods are described in the Online Data Supplement. The heart remained attached to the dorsal cuticle and was readily identified as a beating structure along the midline from abdominal segments A1 to A4 (Figure 1). Dissected fly hearts remained beating for up to 2 hours at room temperature. Fly heart contractions were stopped by incubation in hemolymph buffer that contained 40 uM cytochalasin-D. Multiple trials to stop fly heart contractions with blebbistatin were unsuccessful. In order to image fly hearts on an inverted microscope, each heart preparation was transferred to a confocal dish containing 100 ul of oxygenated hemolymph buffer. The heart specimen was oriented with the dorsal side up and a glass cover slip was gently applied to avoid movement during imaging.

Figure 1. Measurement of Ca²⁺-dependent fluorescence in adult hearts from w¹¹¹⁸; tinC-GCaMP2.

A. Representative bright field image (top) and fluorescence image (bottom) of the heart prepared from an adult w¹¹¹⁸; tinC-GCaMP2. The heart preparation includes the dorsal cuticle and is oriented with abdominal segments (A1, A2, A3, and A4) as shown. B. Representative pseuodocolored isochrone map from a w¹¹¹⁸; tinC-GCaMP2 heart showing caudal to rostral Ca²⁺ wave propagation. Blue and red pixels indicate the earliest and latest time points during the recording, respectively. The pseudocolored line represents the pixel intensities averaged along the vertical columns. C. Fluorescence intensity traces for three individual pixels representing the earliest (blue), middle (green), and latest (red) times during the Ca²⁺ wave propagation and the superimposed traces are shown. A 100 msec time bar is shown.

A Leica 165FC stereomicroscope equipped with an Andor iXon 860 EMCCD camera was used for experiments employing cadmium, EGTA, diltiazem, TTX, caffeine, thapsigargin, and low [Na⁺]_o buffer. Images were analyzed using Andor Solis software (Andor technologies, Inc.).

For experiments to determine the effects of motion artifact caused by cardiac contractions, Ca²⁺ dependent fluorescence was measured at baseline and after abolishing contraction by treatment with 40 uM cytochalasin-D for 10 minutes at room temperature. Hearts from flies expressing GFP under the direct control of tinC (tinC-GFP) were also prepared in the same manner but without cytochalasin-D treatment.

Fluorescence Ca²⁺ imaging

Fly hearts were imaged using a Nikon TE2000-U inverted fluorescence microscope equipped with an Andor iXon 860 EMCCD camera at a rate of 200 frames per second. The spatiotemporal analysis of Ca²⁺ signals was performed using Matlab software as follows. Ca²⁺ signals were detrended by fitting and subtracting a 2^nd-order polynomial and temporally filtered using 12-sample median and mean filters. Non-myocardial pixels inside the recording field-of-view were excluded from further analysis by imposing a threshold on the peak-to-peak range of each pixel’s fluorescence intensity. Ca²⁺ transient activation times were defined along each upstroke when the pixel intensity reached 50% of its peak. These “50% rise” times were expressed relative to the earliest activation time and used to construct isochrone maps of propagation. In addition, the temporal characteristics of each Ca²⁺ transient were assessed using dF/dt_max (the maximum slope of the transient’s upstroke), the time between 50% rise and peak, the half-time of fluorescence decline from peak, the full duration half maximum (FDHM, or time between 50% rise and 50% decay), and the slope of the transient’s decay calculated using a linear fit of the transient from 20% to 80% decay. Since the parameters of the Ca²⁺ transients at the rostral, middle, and caudal regions of the heart were similar and super imposable, all pixels in the recoding over the entire heart were corrected for the time-offset and used to generate an average Ca²⁺ wave for comparison among groups of recordings. The conduction velocity (CV) of each propagating Ca²⁺ transient was calculated using a manually-selected rostral and caudal subregion by dividing the distance between their centers-of-mass by the difference between their mean activation times.

Quantitative Polymerase Chain Reaction (qPCR)

Total RNA samples from the dissected hearts from groups of 60 female w¹¹¹⁸ or hdp² flies 48–72 hours after eclosion were prepared and used for quantitative (real time) PCR as described in detail in the Online Data Supplement.

Statistical analyses

Statistical calculations employed t-tests or analyses of variances (ANOVAs) with Bonferroni corrections for multiple comparisons using GraphPad Prism statistical software (GraphPad Software, Inc.).

Results

To measure Ca²⁺ handling properties in the adult fly heart, we engineered transgenic flies that harbored a non-His-tagged GCaMP2 construct under the control of tinC in a w¹¹¹⁸ background. We designated this fly line as w¹¹¹⁸; tinC-GCaMP2. GCaMP2 is a circularly permuted enhanced GFP that contains a calmodulin Ca²⁺ binding domain and the M13 helix of myosin light chain kinase and has been previously used as a Ca²⁺ sensing fluorescence indicator.^43–45 Adult hearts isolated from w¹¹¹⁸; tinC-GCaMP2 had robust, cardiac-specific fluorescence (Figure 1 and Movie 1). High-speed imaging (200 fps) of fluorescence signals along the heart tube in abdominal segments A2 and A3 were measured for all subsequent studies. Cytochalasin-D was used during the preparation of heart samples to stop cardiac contraction while maintaining Ca²⁺-dependent fluorescence. Therefore, the measured fluorescence signals reflected changes in cytosolic Ca²⁺ and unlikely the result of motion artifact. Heart preparations that were not treated with cytochalasin D demonstrated significant motion artifact during imaging that accounted for at least 50% of the observed changes in fluorescence intensity (Online Figure I). As an additional control, experiments using tinC-GFP fly hearts demonstrated a significant increase in fluorescence intensity during systole (Online Figure I). Of note, we also conducted experiments that employed the application of Ca²⁺-sensitive dyes but observed poor incorporation of dyes into the fly heart with significant non-specific accumulation in non-cardiac tissue rendering us unable to acquire reproducible Ca²⁺ transients (data not shown).

We processed the raw image data from recordings of the fluorescence signals from the adult heart preparations using Matlab software to measure changes in pixel intensities. Each cardiac Ca²⁺ wave was represented by ~5000 pixels per frame for 500 frames of image data corresponding to the abdominal A2 and A3 segments of the fly heart. This region of the heart tube is a single cell layer thick and contains approximately eight cardiomyocytes on each side of the heart.⁴⁶ Since the level of GCaMP2 that was present in each heart can vary and GCaMP2 is not a ratiometric indicator, we normalized fluorescence intensity of each pixel to the maximum intensity at each pixel over all frames obtained during each recording. The changes in pixel fluorescence intensities in the hearts were analyzed from pseudocolored movies that represented changes at each pixel from 50% rise to 50% decay in normalized intensity (Online Movie II and corresponding isochrone map in figure 1). Additionally, the isochrone pixel intensities for each recoding were averaged along the vertical columns and represented as a line (Figure 1B). The vertical column averaging was performed because this processing served as an additional control for motion artifact due to possible subtle fly heart contraction after cytochalasin-D treatment. Analyses of time-offset adjusted pixel intensities from different areas of the heart were similar and super imposable. We therefore used the average pixel intensity traces for each heart beat to quantify the Ca²⁺ transient in each heart (Figure 1C).

We examined 714 propagating Ca²⁺ transients from 32 w¹¹¹⁸ flies and show that 57.9% of generated Ca²⁺ transients were directed anterograde (from the caudal to the rostral direction) while 36.6% were directed retrograde (from the rostral to the caudal direction). Bidirectional generated Ca²⁺ transients occurred in 5.8% of heart preparations (Figure 2). These findings were consistent with the previously description of rostral and caudal pacemakers in the adult fly heart.^{31, 47, 48}

Figure 2. Determination of Ca²⁺ propagation in adult hearts from w¹¹¹⁸; tinC-GCaMP2.

A. The abdominal A2 and A3 segments from a representative w¹¹¹⁸; tinC-GCaMP2 heart in the rostral to caudal orientation is shown. B. Three different representative temporal isochrones with blue and red pixels indicating the earliest and latest times of activation, respectively, are shown (top). Series of consecutive 5 msec frames from temporal isochrone movies are shown. Caudal to rostral (left column), rostral to caudal (middle column), and bidirectional (right column) Ca²⁺ propagation with the corresponding percentages from 714 measures in 32 w¹¹¹⁸; tinC-GCaMP2 hearts are shown. C. Individual pixel intensity traces along the heart from w¹¹¹⁸; tinC-GCaMP2 showing caudal to rostral Ca²⁺ propagation during a 600 msec recording. D. Individual pixel intensity traces along the heart from w¹¹¹⁸; tinC-GCaMP2 showing rostral to caudal Ca²⁺ propagation during a 600 msec recording. E. Individual pixel intensity traces along the heart from w¹¹¹⁸; tinC-GCaMP2 showing bidirectional Ca²⁺ propagation originating from both the rostral and caudal regions during a 600 msec recording.

To characterize myocardial Ca²⁺ handling in w¹¹¹⁸ flies we measured parameters associated with the generated Ca²⁺ transients reflecting changes in the intracellular Ca²⁺ concentration. These parameters included the maximal slope of increasing fluorescence intensity (dF/dt_max), the duration from the 50% rise in intensity to peak intensity, the half-time of fluorescence decline from peak, the full duration at half maximal intensity (FDHM), and the linear slope from 80% to 20% intensity decay (Figure 3A). Additionally, conduction velocities for propagating Ca²⁺ transients along the heart tube were determined. Since the majority of generated Ca²⁺ transients occurred in the anterograde and retrograde directions, the properties of the Ca²⁺-induced fluorescence intensities and conduction velocities in both groups were compared. All parameters including conduction velocities were similar whether propagation was anterograde or retrograde (p=NS) (Figure 3B). Additionally, the effects of motion on the generated Ca²⁺ transient parameters were determined by comparing fly hearts treated with cytochalasin-D vs. vehicle alone. The beating fly heart preparations had significant delays the generated Ca²⁺ transient parameters suggesting that contraction artifact can alter interpretation of fluorescence measurements (Online Figure I).

Figure 3. Comparison of anterograde and retrograde directed Ca²⁺ transients in adult hearts from w¹¹¹⁸; tinC-GCaMP2.

A. Representative trace of averaged pixel intensities from a w¹¹¹⁸; tinC-GCaMP2 heart. Definitions of dF/dt_max (A), the duration of the 50% rise to peak fluorescence intensity (B), the duration of the peak to 50% decay in fluorescence intensity (C), FDMH (D), and the slope of linear decay from 80% to 20% fluorescence intensity (E) are shown. B. Comparison between parameters of pixel intensity traces and conduction velocity for anterograde (i.e.; caudal to rostral) (n=25) and retrograde (i.e.; rostral to caudal) (n=31) Ca²⁺ transients in w¹¹¹⁸; tinC-GCaMP2 hearts. There were no statistically significant differences in the parameters measured between anterograde and retrograde Ca²⁺ wave propagations.

Next, we addressed the ability to manipulate fly myocardial Ca²⁺ handling using pharmacological interventions to examine the relevance of our model to the mammalian heart. To examine the contribution of voltage-dependent Na⁺ channels to the generated Ca²⁺ transients we examined the effects of low [Na⁺]_o buffer, TTX, or cadmium (Cd²⁺). Low [Na⁺]_o buffer (2mM) caused a complete block of Ca²⁺ transients followed by a resumed Ca²⁺ transients suggestive of Ca²⁺ waves (SR overload-induced spontaneous Ca²⁺ release) or Na⁺-independent excitability under experimental conditions (Figure 4A). The resumed propagated Ca²⁺ transients had slower conduction velocities (0.71 ± 0.05 cm/sec for 110mM [Na⁺]_o buffer vs. 0.49 ± 0.04 cm/sec in 2 mM [Na⁺]_o buffer, p<0.05, n=9). The administration of TTX (100 uM) did not block the propagating Ca²⁺ transients (data not shown).

Figure 4. Ca²⁺ transients in adult hearts from w¹¹¹⁸; tinC-GCaMP2 are dependent on extracellular Na⁺, extracellular Ca²⁺ and intracellular Ca²⁺ stores.

A. Representative trace of the generated Ca²⁺ transients in w¹¹¹⁸; tinC-GCaMP2 in buffer containing 110 mM [Na⁺]_o and after exposure to 2 mM [Na⁺]_o buffer. Two second segments of the time course are shown. Inset table shows fold changes in transient rate and conduction velocity after exposure to 2 mM [Na⁺]_o. (*p<0.05 by paired t-test, n=9). B. Representative trace of generated Ca²⁺ transients in w¹¹¹⁸; tinC-GCaMP2 before and after the administration of 5 mM EGTA. C–E. Representative serial traces of the generated Ca²⁺ transients from w¹¹¹⁸; tinC-GCaMP2 before and after the administration of 10 uM diltiazem. F. Representative traces of the generated Ca²⁺ transients from w¹¹¹⁸; tinC-GCaMP2 before and after the administration of 2 uM thapsigargin. G–I. Summary data for the effects of thapsigargin on generated Ca²⁺ transients. The changes in the slopes of fluorescence increases (G), peak fluorescence (H), and slopes of fluorescence decreases (I) are shown. *p<0.05 by paired t-tests for changes in parameters before and after thapsigargin treatment (n=6).

To examine the contributions of extracellular Ca²⁺ and L-type Ca²⁺ channels, we performed experiments in the presence of EGTA or diltiazem. The chelation of free extracellular Ca²⁺ resulted in the complete inhibition of generated Ca²⁺ transients (Figure 4B). Next, we used diltiazem, an inhibitor of L-type Ca²⁺ channels.⁴⁹ The administration of diltiazem resulted in attenuation and subsequent cessation of generated Ca²⁺ transients (Figure 4C–D). Additionally, the generated Ca²⁺ transients were completely blocked by 200 uM Cd²⁺ (data not shown).

We examined the effects of intracellular Ca²⁺ stores depletion on generated Ca²⁺ transients by using thapsigargin, an inhibitor of SERCA.^{50, 51} Fly hearts that were treated with the thapsigargin had reductions in generated Ca²⁺ transients with decreases in peak fluorescence and the slopes of fluorescence increase and decrease (Figure 4F–I). Lastly, we examined the effects of octopamine, a G-protein coupled receptor (GPCR) agonist, and observed a decrease in the FDHM and a more steep linear slope of intensity decay suggesting that the model responds to agonist stimulation (Online Table I).

Since dilated cardiomyopathies and heart failure have been associated with alterations in myocardial Ca²⁺ handling, we next tested whether a dilated heart in the fly would recapitulate the known abnormalities in generated Ca²⁺ transients in mammalian hearts. Previously, we identified a mutant, hdp² that has a point mutation in Troponin I conserved among multiple species including Drosophila and humans, and has a dilated heart phenotype.¹⁷ hdp²; tinC-GCaMP2 flies were generated by genetic crosses. The cardiac-specific expression of GCaMP2 did not significantly affect cardiac function in w¹¹¹⁸ or hdp² as determined by OCT measurements (Online Figure II).

We measured generated Ca²⁺ transients in hearts from adult, age-matched hdp²; tinC-GCaMP2 and w¹¹¹⁸; tinC-GCaMP2 (Figure 5A). Compared to w¹¹¹⁸, hdp² mutant hearts had similar dF/dt_max but significant differences in the duration of the 50% rise in intensity to peak intensity, the half-time of fluorescence decline from peak, FDHM, and the linear slope of decay from 80% to 20% intensity decay (Figure 5B–F). The conduction velocities were not statistically different between w¹¹¹⁸ and hdp² hearts (0.76 ± 0.04 vs. 0.65 ± 0.04 cm/sec, p=NS for w¹¹¹⁸ vs. hdp², respectively). Additionally, the duration of the fluorescent Ca²⁺ transient was significantly different between w¹¹¹⁸ and hdp² hearts over a range of defined intensities (Online Figure III).

Figure 5. The hdp²; tinC-GCaMP2 mutants have altered myocardial Ca²⁺ handling compared to w¹¹¹⁸; tinC-GCaMP2.

A. Representative average pixel fluorescence intensity traces for w¹¹¹⁸; tinC-GCaMP2 (black) and hdp²; tinC-GCaMP2 (red) hearts. B–F. Measurements of ^dF/dtmax (panel B), the duration of the peak to 50% decay in fluorescence intensity (panel C), the duration of the 50% rise to peak in fluorescence intensity (panel D), FDHM (panel E), and the slope of linear decay from 80% to 20% fluorescence intensity (panel F) in w¹¹¹⁸; tinC-GCaMP2 (n=55 recordings from 19 individual preparations) and hdp²; tinC-GCaMP2 (n=81 recordings from 19 individual preparations) hearts. Individual measurements (black circles) and the mean (open circles) with SEM are shown. *p<0.05 for w¹¹¹⁸; tinC-GCaMP2 vs. hdp²; tinC-GCaMP2 for the indicated parameter.

Abnormalities in mRNA transcript levels in key components of Ca²⁺ cycling have been associated with myocardial dysfunction. Therefore, we performed qPCR analyses of several well recognized components involved in intracellular Ca²⁺ cycling in hearts from w¹¹¹⁸ and hdp² mutants (Figure 6A). Interestingly, hdp² mutants had significant reductions in ryanodine receptor transcripts compared to w¹¹¹⁸. There were no significant differences in SERCA, L-type Ca²⁺ channel, IP3 Receptor, or sodium-Ca²⁺ exchanger transcripts between hdp² and w¹¹¹⁸. Additionally, hdp² hearts had reductions in caffeine-augmented generated Ca²⁺ transients (Figure 7B). These findings are consistent with the known changes in a variety of vertebrate animal models and humans with dilated cardiomyopathy.⁶

Figure 6. Hearts from hdp² mutants have reductions in ryanodine receptor receptor transcript levels compared to w¹¹¹⁸.

A. qPCR measurements of transcript expression levels in hearts from hdp² (open bars) relative to w¹¹¹⁸ (closed bars) for L-type Ca2+ channel (LTCC), ryanodine receptor(RYR) sarco/endoplasmic reticulum Ca2+-ATPase (SERCA), Sodium calcium exchanger (NCX) and inositol-3-phosphate receptor(IP3R). n=5 independent experiments with 60 fly hearts per group per experiment. *p<0.05 for hdp² vs. w¹¹¹⁸ by t-test. B. Representative traces of generated Ca²⁺ transients before and after the administration of 2 mM caffeine in w¹¹¹⁸; tinC-GCaMP2 hearts. C. Summary data for the fold changes in the area under the peaks of fluorescence after caffeine administration to w¹¹¹⁸; tinC-GCaMP2 vs. hdp²; tinC-GCaMP2 hearts.

Discussion

We developed a new method based on the cardiac expression of GCaMP2, a genetically encoded Ca²⁺ indicator, to measure myocardial Ca²⁺ cycling in adult Drosophila. Genetically encoded Ca²⁺ indicators have been used to examine cardiac Ca²⁺ handling in other models including the mouse and zebrafish.^52–56 However, the genetic resources to map and rapidly identify mutants that affect myocardial Ca²⁺ handling in these genetic models can be quite laborious. The method that we describe has a number of significant advantages. First and foremost, Drosophila genetics and genomics offers unique resources for gene discovery that are not available in other model systems. These genetic resources facilitate the mapping and identification of gene mutations and have been used to identify a variety of genes that affect different aspects of Drosophila cardiac contractile function.^{13, 19} Flies that have molecularly-defined genomic deficiencies, P-element insertions, or harbor RNAi for specific gene knockdown can be introduced into the tinC-GCaMP2 background to investigate how individual genes potentially affect myocardial Ca²⁺ handling. Alternative approaches that rely on applying Ca²⁺ indicator dyes can lead to non-specific, extra-cardiac tissue distribution. For example, dye incorporation into the non-tinC expressing striated ventral longitudinal muscle can prevent adequate labeling of the fly heart and lead to diminished specificity during cardiac Ca²⁺ measurements. Our method has the advantage of genetically targeted cardiac-specific Ca²⁺ measurements that circumvent this potential problem.

The circulatory system in flies is simple compared to mammals, however the generated Ca²⁺ transients in adult Drosophila hearts share several characteristics with mammalian myocardial Ca²⁺ handling. First, the Ca²⁺-dependent fluorescence wave in tinC-GCaMP2 hearts moves along the heart tube in an organized manner. However, unlike mammalian hearts, the fly heart has two pacemakers that control anterograde and retrograde contraction.^{31, 47, 48} Second, the generated Ca²⁺ transients in fly hearts are dependent on extracellular Ca²⁺ via L-type Ca²⁺ channels and intracellular calcium stores. Fourth, the generated Ca²⁺ transients appear to require sodium channels since low [Na⁺]_o causes a transient cessation of Ca²⁺ transients. Fifth, Ca²⁺ transient parameters undergo changes after stimulation with octopamine, an agonist of the octopamine GPCR. Sixth, the fly heart appears to have altered Ca²⁺ handling in the context of the hdp² mutant. The hdp² mutant has a well characterized deterioration of the indirect flight muscles, progressive myopathy of other muscle groups, and a dilated cardiomyopathic phenotype.^{17, 57, 58} Compared to w¹¹¹⁸, hearts from mutant hdp² have a prolonged duration of 50% rise to peak intensity, but no significant changes in dF/dt_max. This suggests that while the early rise in cytosolic Ca²⁺ concentration is similar, the late rise in cytosolic Ca²⁺ is delayed. Furthermore, in hdp² hearts, the half-time of fluorescence decline from peak is prolonged and the linear slope from 80% to 20% of fluorescence decay is decreased. These findings suggest that Ca²⁺-reuptake into the SR and/or Ca²⁺ extrusion into the extracellular space are impaired in the hearts from hdp² compared to w¹¹¹⁸.

The mechanisms that are responsible for the observed changes in cytosolic Ca²⁺ cycling and contractile dysfunction in hdp² hearts remain to be identified. Hearts from hdp² mutants had reductions in RyR and IP3R transcript levels consistent with the prolonged duration of 50% rise to peak intensity that we observed in the GCaMP2 flies. Additionally, hdp² hearts had reductions in caffeine- augmented increases in the generated Ca²⁺ transients indicating a reduced SR Ca²⁺ load. These results are consistent with some previous studies in mammalian models of heart failure.⁶ However, the transcript levels for SERCA were not significantly different between w¹¹¹⁸ and hdp² although the duration of peak intensity to 50 % decay is prolonged and the linear slope from 80% to 20% of fluorescence decay is decreased in hdp² hearts. A possible explanation for these findings is that alterations of Ca²⁺-reuptake in hdp² hearts may result from changes in the post-transcriptional level of SERCA or alterations in proteins that control SERCA function.

Alternatively, the troponin I mutation in hdp² flies may change the structure and Ca²⁺ binding properties of contractile protein complexes thereby altering the buffering capacity of these complexes for cytosolic Ca²⁺. Recent studies using preparations of indirect flight muscle from Drosophila have demonstrated that the site of the hdp² troponin I mutation decreases the concentration of Ca²⁺ necessary for thin filament activation.^57–59 Furthermore, based on comparisons of the conserved site of the mutation in troponin I in hdp² flies and mammals, studies suggest that the change in Ca²⁺ binding to the troponin I/troponin C complex is an indirect effect of the hdp² mutation.⁵⁸ Thus, changes in the interaction between contractile proteins and Ca²⁺ may also explain some of the changes in Ca²⁺ transients in hdp² mutants.

There are limitations to our approach to measuring myocardial Ca²⁺ handling in adult flies. First, GCaMP2 is a nonratiometric Ca²⁺ indicator, similar to several nonratiometric Ca²⁺ indicator dyes.^{55, 60, 61} Additionally, the amount of GCaMP2 is variable in each fly heart. Therefore, the absolute changes in cytosolic Ca²⁺ concentration cannot be determined and relative changes must be used for comparison among experiments. Second, the fluorescence intensity of GCaMP2 is lower than fluorescence dyes since the intensities of the fluorescence dyes depend on the loading conditions.^{45, 55, 56} Thus, GCaMP2 expression in the fly heart is not sufficient enough to permit imaging through the cuticle of the live insect. Therefore, the tissue surrounding the heart must be partially dissected away prior to imaging. The generation of newer genetic indicators with improved signal-to-noise may circumvent the inability to directly measure Ca²⁺-dependent fluorescence intensity through the cuticle. Third, although GCaMP2 and Ca²⁺ indicator dyes are nonratiometric, GCaMP2 has different affinities and kinetics regarding Ca²⁺ binding and unbinding compared to chemical dyes.^{45, 55, 56} The response of GCaMP2 to Ca²⁺ binding and unbinding is slower than chemical dyes and therefore will tend to prolong the Ca²⁺ duration. Fourth, the fly heart must be electromechanically dissociated prior to imaging to minimize changing optical properties in the tissue that occurs during contraction. In fact, in the beating fly heart approximately 50% of the change in fluorescence intensity is the result of motion. Additionally, the parameters associated with the generated Ca²⁺ transients from fly hearts that were electromechanically dissociated are significantly different from hearts that were contracting. Therefore, measurements of myocardial Ca²⁺ in our fly model require pre-treatment with cytochalasin-D to uncouple true Ca²⁺-dependent fluorescence from motion artifact. Unfortunately, blebbistatin did not stop fly heart contractions and therefore the use of cytochalasin-D introduces another potential limitation since cytochalasin-D is known to affect intracellular Ca²⁺.⁶²

Lastly, our studies suggest that the propagating Ca²⁺ transient is mediated by electrical action potential propagation. The generated Ca²⁺ transients in fly hearts are transiently blocked by low extracellular Na⁺ buffer suggesting that sodium-dependent actions potentials are necessary for the observed Ca²⁺ transients. The generated Ca²⁺ transients are also sensitive to Cd²⁺ consistent with a dependence on voltage-sensitive Na⁺ channels similar to that observed in mammalian hearts.^63–65 However, the dose of Cd²⁺ required to block fly heart Ca²⁺ transients is in a range that can also inhibit L-type Ca²⁺ channels. The effects of TTX were modest and suggest that the fly heart may not be sensitive to TTX. Therefore, more definitive studies employing voltage-sensitive dyes will be necessary to understand the contribution of the electrical action potential to the Ca²⁺ transient in the fly heart.

Overall, our results demonstrate the potential utility of Drosophila as a model of myocardial Ca²⁺ handling. Future applications including an examination of fly mutants using this methodology have the potential to identify and characterize new genes that regulate intracardiac Ca²⁺ signaling and may lead to new insights into human cardiovascular diseases.

NOVELTY AND SIGNIFICANCE.

What is known?

• The fruit fly, Drosophila melanogaster, is recognized as a model of mammalian cardiovascular disease and can recapitulate aspects of mammalian cardiomyopathies.

• Abnormalities in myocardial calcium handling are observed in mammalian models of heart failure; however, the myocardial calcium handling in Drosophila has not been well-characterized.

What new information does this article contribute?

• A new transgenic fruit fly harboring a cardiac-specific, reversible calcium-dependent fluorescent protein reporter, GCaMP2, was engineered and myocardial calcium transients were investigated using high speed fluorescence imaging.

• The propagating calcium transients in adult fly hearts originate from two different pacemakers and are dependent on extracellular calcium via L-type-calcium channels and intracellular calcium stores.

• The study results suggest that components that underlie myocardial calcium transients are evolutionarily conserved between Drosophila and mammals and the fruit fly may provide new insights into genes that potentially regulate calcium handing in mammalian hearts.

The study describes a new method to measure calcium transients in the adult fruit fly heart using flies that express a cardiac-specific, reversible calcium-dependent fluorescent protein reporter and a high-speed imaging system. By applying this technique, the propagating calcium transients were measured in adult fly hearts from controls and mutants that have dilated cardiomyopathic phenotypes. Drosophila myocardial calcium transients propagate in two directions and are similar to mammalian myocardial calcium transients in that they require extracellular sodium, extracellular calcium via L-type calcium channels, and intracellular calcium stores. These findings further support the fruit fly as a model of cardiovascular disease. Furthermore, this method suggests that the fly can serve as a genetically tractable system for identifying genes that affect myocardial calcium handling in mammals.

Non-standard abbreviations and acronyms

CV

conduction velocity

dF/dT_max

slope of maximum fluorescence intensity increase

EDD

end-diastolic dimension

EMCCD

electron-multiplying charge-coupled device

ESD

end-systolic dimension

FDHM

full duration half maximum

FS

fractional shortening

GCaMP2

circularly permutated enhanced green fluorescent protein fused with M13 helix of myosin light chain at the N-terminus and calmodulin at the C-terminus

GPCR

G-protein coupled receptor

hdp²

held-up² mutant fly

IP3

inositol-3-phosphate

LTCC

L-type calcium channel

NCX

sodium-calcium exchanger

OCT

optical coherence tomography

RyR

Ryanodine receptor

SERCA

sarcoendoplasmic reticulum calcium ATPase

SR

sarcoplasmic reticulum

tinC

DNA regulatory region of the Drosophila tin gene

tinC-GCaMP2

transgenic fly harboring GCaMP2 under the control of the tinC DNA element

TTX

tetrodotoxin

W¹¹¹⁸

laboratory stock fly