ABSTRACT
The mechanical stress state of an organ is a critical, but still poorly understood, driver of organogenesis and regeneration. Here we report a chip-based regulated environment for micro-organs (REM-Chip) that enables systematic investigations of the crosstalk between an organ’s mechanical stress environment and biochemical signaling under a multitude of genetic and chemical perturbations. This method has enabled us to identify essential conditions for generating organ-scale intercellular calcium (Ca2+) waves (ICWs) in Drosophila wing imaginal discs that are also observed in vivo. Spontaneous ICWs require the presence of components in fly extract-based growth serum (FEX). Using the REM-Chip, we demonstrate that the release and not the initial application of mechanical compression is sufficient but not necessary to initiate ICWs. Further, the extent of the Ca2+ response is heterogeneous between discs and correlates with the degree of spontaneous ICWs activity in the pre-stress state. This system and method enable detailed examinations of the interplay between mechanical stress state, biochemical regulatory networks, and physiology in complex, hierarchically organized organ cultures.
SIGNIFICANCE STATEMENT
We present a first-of-class microfluidic chip that can perturb a developing organ both chemically and mechanically. Here we advance the field of organogenesis by presenting precise mechanical perturbation methods for whole-organ explants. This enables researchers to systematically interrogate critical relationships between mechanical stress state and biochemical signaling. Our methods advance available modes of studying Drosophila wing disc development, a powerful model for examining pathways critical for human development. We elucidate the causative links between mechanical perturbations and ICWs. Spontaneous ICWs require fly extract (FEX), a growth serum derived from flies, to be included in the culture media. Further, we find that the extent of ICW response to mechanical perturbation is determined by the spontaneous ICW activity prior to stimulation.
Organs develop in a diverse landscape of signals. However, the connections between exogenous forces, gene expression, and signal transduction in organ culture are still poorly understood. The calcium ion (Ca2+) has been demonstrated as a universal second messenger that regulates and coordinates a diverse range of intracellular processes such as proliferation and morphogenesis (1). Dysregulation of Ca2+ signaling via genetic and epigenetic modifications has been implicated in human diseases including cardiomyopathies (2), cancer metastasis (3), and neurodegenerative disorders (4). Furthermore, Ca2+ signaling has been shown to act as a central signal integrator in stem cell proliferation and homeostasis (5). The ubiquity of Ca2+ signaling makes it difficult to identify causative mechanisms of Ca2+ regulation and function in a given biological context (1). In particular, the roles of Ca2+ in mechanical signal integration has not been systematically investigated in epithelial tissues, despite observations that extreme mechanical perturbations such as wounding (6) and gross mechanical deformations (7) excite intercellular Ca2+ transients. Analogous to Ca2+ signaling, there is compelling evidence that implicates mechanical stress as a regulator of the cell cycle, differentiation, and cell survival (8–11). However, testing hypotheses on the relationships between the Ca2+ signaling, genetic background and mechanical signaling in organs is technically challenging because of the inability to apply precise mechanical perturbations to a specimen with current organ culture practices.
Microsystems have been increasingly used to mechanically perturb individual cells and cell populations. Microsystem designs range from structures with mechanical linkages for microscale manipulation (12) to contained vessels with integrated actuators for in situ perturbation (13). Microfluidic devices are an important class of microsystem that have been used to control chemical perfusion (14) and temperature (15) profiles. More recently, microfluidic devices have been used to mechanically perturb cells and confluent monolayers through deformation of microfluidic channel walls (16). In contrast to monolayers of a homogeneous cell type, organ-like cultures are particularly important for the study of development because they have innate morphogenetic diversity, hierarchical organization, and intact extracellular matrix (ECM). While organ-like cultures better replicate the in vivo environment, no device exists for combined mechanical manipulation and live-imaging of intact cultured organs (17). Here, we present a device designed for detailed mechanical manipulation of organ cultures.
The microsystem and method presented here is tuned to regulate the organ culture environment. Organ culture permits precise environmental perturbations with enhanced image quality compared to the in vivo context. However, organs are more difficult to culture and mechanically perturb than cells as they are generally less robust and less adherent to culture substrates. Our microfluidic device, termed the Regulated Environment for Micro-organs Chip (REM-Chip) is designed to test organs in an ex vivo context that closely mimics the in vivo microenvironment. The key features of the REM-Chip are: a gentle organ loading procedure; integrated fluidic channels to deliver growth media or other chemical constituents; deformable diaphragms to apply a compressive stress to an organ culture; and compatibility with small working distance objectives to enable real-time measurement of fluorescently labeled sensors.
To demonstrate the utility of the device, we used the Drosophila wing imaginal disc as a model organ. The wing disc has a powerful genetic toolkit and conserves many of the mechanisms of human development and disease (18). The wing disc develops in a naturally dynamic mechanical environment as larval motion is coupled to whole body deformation. Specifically, our group and others (7) have observed wing disc deformations during in vivo imaging of developing larva. In our studies, intercellular Ca2+ waves (ICWs) are observable in 34% of the discs (20 min of imaging, n = 37 and Wu et al., in preparation), but do not qualitatively correlate with physiological deformation during larval movement. In recent in vivo studies (7), mechanical compression could not be precisely prescribed, nor the Ca2+ response observed during the application of compressive stress. This device and method have enabled us to report that the release, and not the initiation of mechanical compression, stimulates an ICW that traverses the disc, elucidating the direct causative links between mechanical stress and ICWs. Spontaneous and mechanically-induced ICWs occur only if fly extract (FEX), a growth serum derived from flies, is included in the chemical milieu. Compression-release is sufficient to evoke ICWs under these conditions with the extent of ICW response determined by the pre-stress physiological state of the organ as measured by spontaneous ICW activity occurring in the stress free state. This highlights an important relationship between chemical and mechanical signaling revealed by our method. The REM-Chip and corresponding method is extensible to many other organ culture models and provides a powerful assay for measuring response to mechanical stimulation in organ growth, development, and homeostasis.
RESULTS AND DISCUSSION
REM-Chip system design
The REM-Chip is a two-layer microfluidic device developed for organ cultures that modulates the local chemical and mechanical microenvironment of a developing wing disc (Fig. 1). Both layers are made from the flexible, optically transparent, polymer polydimethylsiloxane (PDMS) and are bonded to a glass coverslip (Fig. 1A). Organ culture media and other chemical constituents flow in the bottom layer, and a pneumatically controlled deformable diaphragm in the top layer applies compression perpendicular to the imaging plane. The REM-Chip is integrated into a system composed of: a syringe pump for driving media flow; an array of pressure transducers to control diaphragm deflection; and a spinning-disc confocal microscope with coordinated stage translation and imaging routines (Fig. 1A, Supplementary Information, SI Fig. S1-3, and SI text 1). Earlier iterations of the REM-Chip design are shown in SI Fig. S4 and described in SI text 2.
The wing disc loading protocol is optimized to reduce wing disc stress. The inlet channel of the REM-Chip is large enough (nominally 600 µm x 100 µm) for a wing disc to be loaded by flowing the disc through the channel. The channel is first flushed with organ culture media, and a wing disc is pipetted into the inlet. Next, the disc is positioned in the culture chamber by withdrawing organ culture media through the outlet channel with a pipette (SI Fig. S5 and SI text 3). Once loaded, culture media is flowed at 2 µL/h via a syringe pump to deliver media and chemicals; this low flowrate does not perturb the position of the wing disc. The area of the culture chamber is larger than an average wing disc to allow for variation in wing disc size and diametrical expansion during mechanical compression. The culture chamber size for wing discs is nominally 800 µm in diameter by 100 µm in height, but can be adjusted to accommodate other model systems.
Validation of the REM-Chip
Compressive stress is applied to the wing disc by the REM-chip’s diaphragm, which is controlled via the pressure line. The deflection of the center of the circular diagram increases monotonically with applied pressure, as measured by confocal microscopy of diaphragms labeled by the fluorescent dye Nile red (Fig. 2A-B). A pressure in excess of 20 kPa results in contact between the diaphragm and the bottom of the chamber, setting the upper limit for compression experiments. The deflection-pressure relationship is in close agreement with a mechanics model of the deflection of a circular membrane under a uniform load (19) and is in agreement with the behavior of similar PDMS structures (20) (SI Fig. S6 and SI text 4.2). Deflection as a function of pressure for an individual diaphragm is repeatable to within ± 2 µm (s.d.), suggesting that variability in the system is negligible (Fig. 2B, SI Fig. S6A).
Imaging of a wing disc compressed at increasing applied pressures demonstrates that the natural folds at the apical surface near the pouch flatten and expand when the disc is pressed against the glass coverslip by the diaphragm. This behavior is shown by a cross section through the center of a representative wing disc pouch shown in Fig. 2C.
Wing discs cultured in the REM-Chip using the optimized culture media formula WM1 (21) demonstrate no statistical difference in the number of mitotic cells when compared to discs cultured in a glass-bottom dish in the same media (repeated measures ANOVA, p = 0.95) (Fig. 2D). These results also agree with reported organ viability durations of ~12 hours on glass-bottomed wells (21, 22). Because culture in the REM-Chip does not significantly alter ex vivo disc viability relative to established methods, the REM-Chip can be used to study the effects of chemical and mechanical perturbations over the course of multiple hours.
ICWs require a chemical signal
ICWs have been recently reported in wing discs both in vivo and ex vivo (7). We find that whole fly extract (FEX), a protein-rich serum often added to culture media to support growth and proliferation of Drosophila tissues, is necessary for the formation of ICWs ex vivo (Fig. 3 and Wu et al., in preparation). ICWs were first observed in ex vivo cultured wing discs cultured in WM1 media, which contains FEX (7). Removal of FEX serum from WM1 media results in a complete loss of ICW activity. Conversely, addition of FEX serum to ZB media, a chemically-defined culture medium, results in the formation of ICWs that are otherwise not present (Fig. 3). ZB media was engineered to support long-term proliferation and passaging of the wing-disc derived Cl.8 cell line (23). These results demonstrate that specific chemical stimulation by FEX is specifically required for the generation of both spontaneous and mechanically stimulated ICWs.
Release of compression initiates organ-wide ICWs
Mechanical perturbations have recently been implicated as a potential cause of the large ICWs in imaginal discs (7). However, experimental methods to allow wing disc observation during active compression or control the level and duration of the applied compressive stress were previously unavailable. Here, we have created a device that enables direct quantitative tests of the hypothesis that mechanical perturbation stimulates ICWs. A wing disc under a compressive stress applied by the diaphragm diametrically expands with a positive correlation with applied pressure (Fig. 3A-B). The release of this compressive stress initiates an ICW (Fig. 3C-D). ICWs are observed in 81% of compression events (n = 51/63 compressions) on release of compression in ZB media with 15% FEX. Compression arrests the formation of spontaneous ICWs. ICWs were never observed to formed during applied compressions for up to a 10-minute period (n=63 compressions). We observe 52% of discs (n = 11/21 discs) exhibit spontaneous ICW formation without mechanical compression within a 10-minute period. ICWs already underway were not stopped by application of compression (SI Movie 9).
Release of compressive stress exerted by a diaphragm backpressure of 15 kPa initiates an ICW approximately 20 s after release (Fig. 3C). Compression-release in the absence of FEX is not sufficient to initiate an ICW, indicating that mechanical initiation of ICW activity is dependent on the presence of FEX in the media (Fig. 3D). The outer folded region of the nub>GCaMP6f expressing domain of the wing disc is oriented such that the apical-basal direction of cells is in the plane of axial compression, resulting in lateral stretching in this region of the disc when the diaphragm is deflected. This portion of the wing disc shows persistent, active ICWs, suggesting that stretching promotes ongoing Ca2+ activity.
Mechanically induced ICWs depend principally on the organ’s physiological status
To investigate important factors governing the extent of ICW activity, we performed a series of compression experiments varying duration and magnitude of compression over multiple biological experiments (n =21, 3 experiments). Each wing disc was loaded into the REM-Chip, held at 0 kPa back pressure for the first 10 min, and then subjected to a 10 kPa, 15 kPa, or 20 kPa diaphragm backpressures. A series of three compression/no-compression cycles were performed on each disc. The order of duration of compression was randomized: 30 s, 300 s, and 600 s. The wing disc response to mechanical compression-release events for this data set could be binned into two differing populations. Discs that showed qualitatively consistent spontaneous ICW activity in the first 10 minutes before compression consistently responded to mechanical compression-release stimulation (n = 34/36 compressions) at a global tissue level for all applied backpressure levels and durations (Fig. 4A-C). Discs that are naturally quiescent, exhibiting low-medium spontaneous ICW activity (small bursts) in the first 10 minutes of culture, responded to compression-release events much less frequently (n = 17/27) with low- to medium-ICW activity regardless of backpressure level or durations (Fig. 4A-C). In many cases, Ca2+ burst area due to compression-release covered the same spatial extent as the previous spontaneous pre-compression ICW events and more localized ICWs frequently initiated near the edge of the pouch with an inward direction towards the geometric center of the tissue. Significant population-level effects were observed in the peak mean pouch intensity after release (p<0.01), ICW burst time (p<0.001), and fraction of wing disc pouch occupied by the ICW (p<0.001). Student’s t-test with Bonferroni correction was used to obtain p-values.
In comparison to the sub-population effect, we saw comparatively minor effects of applied backpressure, pouch deflection or backpressure hold time on the burst area, burst time, normalized mean peak intensity in the pouch, velocity or delay of the ICW response over the ranges tested. A small qualitative increase was observed in the effects of applied backpressure on burst time of the ICW. Additionally, there is a small qualitative increase in the max mean pouch intensity as a function of hold time and applied backpressure. These relationships are shown in SI Fig. S9. Compression-release induced ICW velocity was found to be 0.8 µm/s ± 0.6 µm/s, and the time to initiation was found to be 20 s ± 30 s (s.d.) after the compression-release. Taken together, the magnitude of compression-release induction of ICWs is dependent on the physiological state of the wing disc more than on duration or degree of deformation for the conditions tested (Fig. 4D).
Mechanically stimulated ICWs result from IP3-mediated release from intracellular stores and depend on gap junction activity
We performed a targeted genetic screening approach to define the mechanism of ICW formation occurring due to mechanical stimulation compared to wound induced and spontaneous ICWs. IP3 signaling is highly conserved and present in both Drosophila and vertebrate models as an important mediator of Ca2+ release. The IP3 signaling cascade releases Ca2+ from internal stores in the endoplasmic reticulum (ER) by binding with its receptor, IP3R. Previously, we have delineated that laser-ablation induced intercellular Ca2+ flashes that Ca2+ transients in wing discs through IP3-mediated Ca2+ release and subsequent propagation through gap junctions (GJs) (24). This mechanism is well-established for intercellular transients as a result of wounding (6, 25) and is also responsible for waves observed during Drosophila oogenesis (26). To test whether mechanical induction of ICWs also rely on this conserved IP3-mediated mechanism, RNAi was used to knock down key components of this cascade (Fig. 5A). We found that inhibition of GJs through Inx2RNAi, inhibition of the IP3 receptor (IP3RRNAi), and inhibition of the endoplasmic reticulum (ER) pump SERCA (SERCARNAi) all completely abolish the compression-release induced ICWs (Fig. 5B-C). Taken together, these results suggest that IP3-mediated release from intracellular stores and GJ communication between neighboring cells are necessary for the propagation of mechanically induced ICWs. Knockdown of SERCA shows a characteristic increase in basal Ca2+ level consistent with impairment of the ability to uptake cytoplasmic Ca2+ into the ER (Fig. 5C). This suggests that once the concentration of Ca2+ in the ER stores drop below a critical threshold, cells are unable to respond through this mechanism. This is in agreement with recently published work on wound-induced ICWs in Drosophila wing discs (7, 24). Interestingly, although GJ inhibited discs (Inx2RNAi) do not develop ICWs, increased flashing of individual cells is observed on release of compression (Fig. 5D-E). Cumulatively, these results suggest that compression-release increased Ca2+ spiking in individual cells, which was likely mediated by IP3. GJ activity enables transmission of IP3 and Ca2+ throughout the tissue. Additional details on all RNAi lines and their validation can be found in (SI Fig. S7 and SI text 5)
The REM-Chip enables new chemical and mechanical assays for organ cultures
Mounting evidence suggests that both chemical and mechanical cues impact organ development (27, 28). The REM-Chip is a powerful new tool for providing controlled mechanical and chemical microenvironments compatible with long-term organ cultures. Recent observations have linked mechanical stress to ICWs in wing discs (7); however, these experiments relied on manual methods to apply mechanical perturbations. These methods allow little control over magnitude or extent of the applied perturbation. Our REM-Chip design allows reproducible and controlled compressions of wing discs. Importantly, the ICW response to mechanical perturbation is found to be dependent specifically on the release, rather than the application, of compressive stress – a distinction lacking in previous observations of mechanically induced ICWs in both cultured cells and tissues. This observation is significant for two reasons. First, developing wing discs and other organs are constantly deformed via organismal and morphogenetic movements. Second, this observation posits interesting questions for further investigation. For instance, what are the downstream effects of mechanically induced ICWs on transcription, translation and mechanical properties of tissues? While other approaches have enabled mechanical perturbation of Drosophila systems (7, 12, 29, 30), the high fidelity mechanical control enabled by the REM-Chip expands the scope of available mechanical assays for organ culture. More generally, the REM-Chip system provides an attractive characterization tool to study the mechanotransduction properties of heterologous channels and mechanoreceptors that can easily be expressed using current Drosophila genetic tools. In addition the REM-Chip will enable future efforts that seek to develop and test novel pressure-based synthetic biology input/output modules that integrate both chemical and mechanical inputs.
Chemical and genetic perturbations modulate mechanically stimulated ICWs
The REM-Chip has revealed that the release of compression, is a potent inducer of ICWs. ICWs were observed after mechanical compression only in the presence of FEX, indicating that both chemical and mechanical factors play a role in this response. This leads to two possibilities requiring future investigations: 1) either FEX is a permissive growth medium that enables intrinsic Ca2+ oscillations to occur in the wing disc coordinated through GJ communication; or 2) an unknown instructive component of FEX generates ICWs. Ongoing work (Wu et al. in preparation) favors the latter explanation.
RNAi against key components of the IP3 signaling cascade revealed that ICWs result from IP3-mediated release and travel through GJ-mediated communication. While Inx2RNAi does not allow for the formation of ICWs, increased flashing of individual cells is observed on release of compression, suggesting that cells are still able to sense mechanical perturbations. This provokes the question of whether compressive stress is mainly inhibiting production of IP3 rather than closing GJs. While mechanical stimulation is sufficient to initiate ICWs, it is not the primary driver of ICWs, as they can be observed spontaneously in the absence of any observable mechanical perturbation. Rather a chemical signal contained in the FEX serum is the primary source of spontaneous ICWs in wing discs. The exact nature of such chemical factors and whether they may be developmentally regulated is a subject for future research.
The REM-Chip enables long duration studies for a multitude of model organs
The REM-Chip does not inhibit the long-term viability of the wing disc in comparison to established ex vivo methods. Furthermore, chemical and mechanical perturbations are completely automated, paving a clear path for long-term investigation of mechanical and chemical perturbations on organ growth and homeostasis, as previously studied in cell cultures (31, 32), but not in organ cultures. Other potential long-term studies include determining how mechanical and chemical perturbations affect morphogenetic patterning in tissues. Lastly, although this specific REM-Chip detailed here is tailored for the Drosophila wing disc, simple modifications to channel and chamber dimensions will enable identical assays to be applied to other Drosophila organs in addition to model organs excised from Xenopus, zebrafish, and human organoids, to name a few.
MATERIALS AND METHODS
REM-Chip design and fabrication
REM-Chip layer designs were drafted in AutoCAD and each design was converted to a photolithography mask by printing on a transparency film (Fineline Imaging, Inc.). Microfluidic masters, composed of a silicon wafer and SU-8 photopolymer (SU-8 3050, MicroChem Corp.), were fabricated via standard lithographic micromachining methods (33). The microfluidic channels in the REM-Chip were defined using reverse micromolding of PDMS on the microfluidic masters–also termed soft lithography–and bonded together via standard methods for two-layer microfluidic devices (34, 35). Ports were punched in the PDMS device and the entire device was cleaned using isopropyl alcohol and then bonded to a glass coverslip (34). Complete details can be found in our published protocols(34). Completed REM-Chips were stored in covered petri dishes.
Systems control and disc image acquisition
Media flow from the syringe reservoir to the REM-Chip devices was driven using a Harvard Apparatus Pump 11 Elite programmable syringe pump. Pressure signals to the REM-Chip were externally controlled by a custom-built pressure regulator box consisting of a manual regulator to step down the house air pressure, a bank of four electropneumatic pressure regulators (ITV001-3UML, SMC Corp., Tokyo, Japan) connected to the source pressure manifold, and an analog output module and accompanying hardware (NI 9264, National Instruments, Austin, TX) to electronically control the reference pressure. Each analog output channel was controlled using a custom-written LabVIEW (National Instruments, 2014) script (SI text 1).
Image acquisition was performed on a Nikon Eclipse Ti confocal microscope (Nikon Instruments Inc., Melville, NY) with a Yokogawa spinning disc using an iXonEM+ cooled CCD camera (Andor Technology, South Windsor, CT). Image acquisition was controlled using MetaMorph v7.7.9 software (Molecular Devices, Sunnyvale, CA).
Sample preparation
Wing discs were cultured in the chemically-defined, ZB-based media (23) or WM1 (21) media optimized for organ culture. Each culture chamber was prepared by filling a Becton Dickinson (BD) Plastipak 1 mL syringe with 1 mL of ZB-based or WM1 media, evacuating all air bubbles from the syringe, and pre-filling the microfluidic network with media. Culture chambers were visually inspected under a stereomicroscope to ensure no air bubbles were present in the culture chamber or fluid lines prior to wing disc loading. Imaginal discs from wandering third instar larvae were then dissected in the relevant culture media. Immediately following dissection, discs were loaded into prepared REM-Chip culture chambers by pipetting the disc over the fluid inlet in a small droplet of media. The disc was positioned with a dissecting needle such that the dorsal side was facing the inlet and the disc drawn into the culture chamber by manually withdrawing media from the fluid outlet with a pipette (SI Fig. S5). After all discs were loaded into culture chambers, the REM-Chip was affixed to the microscope stage, and fluid inlet lines were connected to the media reservoir and pre-flooded with culture media. Inlet lines were then plugged into the REM-Chip and media were perfused at a flow rate of 2 µL/h for the duration of all experiments.
Diaphragm and wing disc deformation study
Diaphragm deflection was measured from images of a fluorescently labeled diaphragm over a range of applied pressures. The REM-Chip was first perfused with 1.5 mM Nile red dye (Santa Cruz Biotech) in methanol for at least 1 h to label all PDMS channel walls. Next, the Nile red was then evacuated from the culture chamber with deionized water; Nile red is insoluble in water. The diaphragm was imaged using confocal microscopy at applied pressures of 0, 5, 10, 15, 20, and 25 kPa. Confocal images were post-processed in Fiji (36) to stitch adjacent fields of view and obtain xz plane diaphragm images from the confocal z-stacks. Diaphragm deformation profiles were then measured using a custom MATLAB script that filters and thresholds the xz images, detects the diaphragm edges given a user-selected region of interest, and converts pixels to distance via microscope calibrations (see SI text 4.1).
Wing disc deformation under mechanical compression was studied via spinning disc confocal microscopy. Drosophila lines expressing DE-Cadherin::GFP (37), which concentrates in the subapical region of the wing disc and labels apical cell boundaries, were used to image the unbuckling and diametrical expansion of the pouch. Wing discs were excised, prepared, and loaded as per the standard protocol described above. Back pressure (0, 3, 7, 11, 15 kPa) was statically applied to the REM-Chip diaphragm and a confocal z-stack centered on the pouch was acquired at 60X magnification in 1 µm intervals. Each z-stack of images was then processed using Fiji/ImageJ (36) to generate a cross-sectional image along the AP axis of the wing disc.
REM-Chip viability study
Flies expressing Lac::YFP, a green fluorescent marker for cell boundaries, were staged for 2 hours in vials at 25 °C. Larvae were collected for dissection at 120 hours (5 days) after egg laying. Discs were dissected and loaded into the REM-Chip in WM1 media as described above or dissected and loaded into glass bottom dishes for standard culture comparison as previously described (21). Discs were imaged at 40X magnification every 30 minutes for a minimum of 12 hours. A 250x250 pixel region, representing ¼ of the total frame size, was cropped from each video frame, centered on the wing disc pouch. The cropped regions were saved with descriptive filenames. The filenames were then randomized and a key automatically created using a batch script written by Jason Faulkner (SIFiles.zip) to create a blinded set of images. The numbers of mitotic cells per area were tabulated by counting the number of cells with increased apical area. The data were compared for statistical difference using a repeated measures ANOVA test.
ICW mechanism study
A tester line was created by recombining nub-GAL4 with UAS-GCaMP6f on the second chromosome. This line was crossed with UAS-Gene-of-interestRNAi lines to perturb essential components of the IP3 signaling cascade (full details provided in SI text 5). Discs from wandering third instar larvae from each cross were dissected and loaded into the REM-Chip as previously described. These discs were imaged at 10X magnification on a Nikon Eclipse Ti confocal microscope for a total of 30 minutes, imaging every 10 seconds. While being imaged, the discs were subjected to the following schedule of applied pressures: 300 s at 0 kPa, 30 s at 15 kPa, 300 s at 0 kPa, 30 s at 15 kPa, 300 s at 0 kPa, 30 s at 15 kPa, 300 s at 0 kPa. Pressure fluctuations had a rise time of 1 sec. Videos were qualitatively scored for the presence or absence of a ICW on release of compression for each RNAi cross tested. The control group consisted of uncrossed discs from the tester line, subjected to the same schedule of applied pressures and imaging regiment.
REM-Chip compression study
Wandering third instar wing discs from the tester line were dissected and loaded into the REM-Chip as previously described. Each experiment consisted of a set of three, unique, 3rd instar wandering wing discs tested in parallel. Each wing disc was subjected to either 10 kPa, 15 kPa or 20 kPa diaphragm back pressure for a randomized set of three compression durations consisting of 30 s, 300 s, and 600 s each. Each compression period was preceded and followed by a 600 s quiescent period (0 kPa) to observe the Ca2+ dynamics. Discs were imaged as described above, except the total imaging period was extended to 60 minutes to accommodate the increased backpressure duration. Videos were randomly renamed as described above for blind analysis and the time to initiation, velocity, burst time, and burst area fraction and peak mean pouch intensity of each mechanically induced ICW were quantified using a combination of Fiji and MATLAB, complete details of the analysis are provided in (SI text 6 and SI Fig. S8).
Fly lines and reagents
Ca2+ signaling was visualized using the GAL4-UAS system (38). The Ca2+ signaling sensor UAS-GCaMP6f (39) (Bloomington stock #42747) was recombined with the nubbin-GAL4 driver (Bloomington stock #25754) to provide a readout of Ca2+ activity in the wing disc pouch (nub-GAL4>UAS-GCaMP6f), which we term the GCaMP6 tester line. For genetic perturbation experiments, the GCaMP6f tester line was crossed to TRiP lines for the targeted gene (all genotypes listed in SI text 5). All flies were raised at 25 °C, 70% relative humidity on cornmeal agar (40). All reagents are listed in (SI text 5).
Image analysis routines
Supplemental movies to accompany figures are described in (SI text 7). All scripts used in the analysis of image data are included in (SIFiles.zip) and described in (SI text 8).
AUTHOR CONTRIBUTIONS
D.J.H. and J.J.Z. conceived the project. C.E.N. performed the organ culture studies. N.M.C. performed diaphragm deflection studies. T.J.S. participated in REM-Chip design including the pressure and vacuum regulator system and contributed to Fig. 2. D.J.H. initially designed REM-Chip and integrated system. C.E.N., N.M.C., D.J.H. and J.J.Z designed the experiments, analyzed the results and wrote the paper.
COMPETING FINANCIAL INTERESTS:
The authors declare no competing financial interests.
ACKNOWLEDGEMENTS
The work in this manuscript was supported in part by NSF Award CBET-1403887 and the Notre Dame Advanced Diagnostics & Therapeutics Berry Fellowship. The authors gratefully acknowledge the Notre Dame Integrated Imaging Facility and the Notre Dame Nanofabrication Facility for the use of their imaging and fabrication facilities, respectively, Melinda Lake and Maxwell Kennard for their assistance with REM-Chip fabrication and pressure and vacuum regulator system design, respectively. The authors thank the Developmental Studies Hybridoma bank for antibodies. The authors thank S. Shvartsman, J. Boerckel, S. Restrepo and members of the Zartman lab for feedback on the manuscript.