The animals used, protocols performed and procedures in this study were in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and approved by the Institutional Animal Use and Care Committee at the University of Nevada, Reno (IACUC; Protocol: 1131-1).

GCaMP6f-floxed mice [Ai95 (RCL-GCaMP6f)-D], Myh11-Cre-eGFP (Xin et al., 2002) and C57BL/6 mice, their wild-type siblings, were purchased from Jackson Laboratories (Bar Harbor, MN, USA). Kit-iCre mice (c-Kit^+/Cre-ERT2) were gifted from Dr. Dieter Saur (Technical University Munich, Munich, Germany), Kit^+/copGFP mice were bred in house.

After the mice were anesthetized by isoflurane (AErrane; Baxter, Deerfield, IL, USA) and killed by cervical dislocation, an abdominal incision was made and the GI organs removed and placed in Krebs-Ringer bicarbonate (KRB) solution. Short segments (2 cm) were taken from the small intestine and proximal colon, and cut open along the mesenteric border. Any intraluminal contents were washed away with KRB. Using fine forceps, the mucosa was carefully peeled away to leave the intact tunica muscularis, which was then used in the following experiments.

Intact muscle sheets from wild-type small intestines were pinned to the Sylgard base of a fixing dish. KRB was removed and replaced with 4% paraformaldehyde (30 min, room temp). Tissues were then unpinned and placed in a test tube containing phosphate-buffered saline with 0.1% Tween 20 (PBS-T). The tissues were washed on a rotator for at least 5 h, changing the PBS-T every hour. After washing, the tissues were placed in a blocking solution (10% normal donkey serum in 0.5% Triton-X), to minimize non-specific antibody binding. After block, tissues were directly transferred to a solution of primary antibody and incubated for 48 h at 4°C. The primary antibody used in this study was a polyclonal raised against c-Kit (mSCFR, AF1356, R and D Systems, MN, USA; 1:100 dilution in 0.5% Triton-X working solution). Next, the tissues were washed in PBS-T for at least 5 h before being incubated in a secondary antibody, Alexa Fluor 488-conjugated anti-goat IgG (1:1000 in PBS-T; Jackson ImmunoResearch, #705-545-147, PA, USA), for 1 h at room temperature. Tissues were again washed in PBS-T for at least several hours before being mounted on glass slides with Vectashield Vibrance Mounting Medium (H-1800, Vector Laboratories, Inc., CA, USA). A Nikon A1R confocal microscope (Nikon Instruments Inc., NY, USA) was used to obtain Z-stack images of ICC and Z-stack projections were then produced using Image J software (National Institutes of Health, MD, USA, http://rsbweb.nih.gov/ij). The distance between ICC subtypes was calculated at multiple regions on several tissue preparations. The total number of regions is referred to as n throughout the manuscript.

A piezo unimorph bending actuator (Python™ PUA3008-5H200) from Bimitech Inc. (San Jose, CA, USA) was used. The piezo measured 46 × 8 × 0.54 mm, and is capable of producing a displacement of up to 0.42 mm. We also had a custom piezo created that measured 19 × 8 × 0.54 mm. Python™ actuators are comprised of a piezo layer that is hermetically sandwiched between two durable epoxy resin layers, which is then further coated with two thin silicone rubber layers. This structure makes the piezo water resistant and very durable. The Python™ piezo actuator was operated by a Thor labs (Newton, NJ, USA) piezo controller (MDT693). The MDT693 is a precision, low-noise, low-drift, high voltage controller for piezo actuators. It has three channels that allow it to control multiple devices or multiple axis piezo stages. However, to control our single Python™ piezo bender, only one channel was required. The front panel of the MDT693 features precision 10-turn potentiometers for the precise manual setting of the outputs over the full operating range (0-150V). The device is utilized to produce 2 Z plane imaging in combination with spinning disk confocal microscopy.

The bare ends of the wires in a BNC cable were soldered to the appropriate terminals on the piezo actuator. The bare wire connections were then waterproofed by sealing with silicone caulk (Henkel, CT, USA). Once the caulk was fully cured (after 24 h), the piezo was attached to the bottom of a Sylgard-coated dish with marine-grade waterproof epoxy (J-B Weld, GA, USA). Furthermore, a thin layer of Sylgard (Dow Chemical Company, MI, USA) 35 mm in diameter was used as a platform around the active end of the piezo, to which we could pin tissues and stretch them across the end of the piezo, while keeping the tissues flat. For experiments, small intestine or colon muscle sheets were pinned so that half of their width was stretched across the piezo and then perfused with KRB solution at 37°C. This was achieved using a perfusion peristaltic pump (minipuls 3, Gilson, WI, USA) with an electric in-line heater and controller (SH-27B and TC-344C, Warner instruments, CT, USA). The piezo BNC cable was attached to the output of a piezo controller and turning the corresponding knob varied the applied voltage. This facilitated precise control of the magnitude of piezo bending. Ca²⁺ imaging was performed using a spinning-disk confocal system (CSU-W1; spinning disk, Yokogawa Electric, Tokyo, Japan) mounted on an upright Nikon Eclipse FN1 microscope equipped with several water immersion Nikon CFI Fluor lenses (10 × 0.3 NA, 20 × 0.5 NA, 40 × 0.8 NA, 60 × 0.8 NA and 100 × 1.1 NA; Nikon Instruments, New York, USA). The system is equipped with two solid-state laser lines of 488 nm and 561 nm. The laser lines are combined with a borealis system (ANDOR Technology, Belfast, UK) to increase laser intensity and uniformity throughout the imaging FOV. The system also has two high-speed electron multiplying charged coupled devices (EMCCD) cameras (Andor iXon-Ultra 897 EMCCD Cameras; ANDOR Technology, Belfast, UK) to allow dual-color imaging simultaneously and maintain sensitive and fast speed acquisition at full frame of 512 × 512 active pixels as previously described (Baker et al., 2021). Briefly, image sequences were collected at 33 to 50 fps using MetaMorph software (MetaMorph Inc, TN, USA).

Movies of two layers (2-Z planes) Ca²⁺ transients in ICC and SMCs were imported, preprocessed and analyzed using Fiji/Image J (National Institutes of Health, MD, USA, http://rsbweb.nih.gov/ij); Briefly, movies of Ca²⁺ transients (stacks of TIFF images) were imported into Fiji and background subtracted and smoothed (Gaussian filter: 1.5 × 1.5 μm, StdDev 1.0). The Ca²⁺ signals were plotted in 2D Ca²⁺ occurrence maps (x axis= time, y axis= cell space) that give information on their spatial and temporal activation. Signal segmentation and analysis were preformed using an automated machine learning plugin (STMapAuto: https://github.com/gdelvalle99/STMapAuto) and 4SM software (https://github.com/SharifAmit/CalciumGAN/tree/main) as described previously (Leigh et al., 2020; Kamran et al., 2022; Moghnieh et al., 2022).

All means ± standard errors were calculated, and graphs generated, using GraphPad Prism 8.0.1 (San Diego, CA, USA) All figures were created using Powerpoint (Microsoft, WA, USA).

Many organs, including those in the gastrointestinal system, are composed of spatially organized layers of cells, which makes it challenging to image and visualize dynamic signals across these layers simultaneously. For example, gastrointestinal tissues have distinct layers, each with specialized cell types that perform different functions, complicating efforts to capture real-time interactions between these layers. Gastrointestinal organs are composed of multiple layers: mucosa for absorption and secretion, submucosa for structural support, interstitial cells and nerve bundles for organization of functions, and circular (CM) and longitudinal (LM) muscle layers for motility. Of interest to us are the interstitial cells of Cajal (ICC) that are arranged in anatomically distinct layers within the myenteric region and circular muscle (Komuro, 2006). In the small intestine, there are two primary types of ICC. One type surrounds the myenteric plexus (ICC-MY), while the other is located within the circular muscle at the deep muscular plexus (ICC-DMP; Figure 1). These distinct subtypes provide different inputs and modes of regulation that facilitate different motor functions in the gut (Sanders et al., 2016, 2024; Drumm et al., 2022; Sanders et al., 2023). In the colon, ICC are more complex as there are four subtypes and each is located at different layers within the tunica muscularis: submucosal ICC (ICC-SM) located at the submucosal border, ICC-IM located within bundles of circular muscle, ICC-MY in the myenteric region located within the myenteric plexus and ICC-SS located at the serosal surface.

We developed a technique to image dynamic fluorescence signals in multiple classes of ICC and smooth muscle cells (SMCs) cells in different layers (two Z depths) simultaneously.

Our approach uses a combination of spinning-disk confocal microscopy (Figure 2A) and a piezo actuator (Figure 2B) to image two layers within the bulk of the tissue (2 Z planes) in the same FOV in real-time. The actuator is driven by a piezo controller, to precisely control the degree of displacement by varying the voltage across the piezo stack (Figure 2B, see materials section for details).

The piezo actuator was placed in a Sylgard-coated dish with GI muscles secured as flat sheets above the surface of the piezo actuator (Figure 2B; see materials section for details).

By increasing the voltage across the piezo actuator using the piezo-electric controller, it was caused to move and lift part of the tissue into a different Z plane (Figure 2C). By imaging the area of tissue just above the end of the piezo we could visualize cells in different Z planes, in the same FOV (Figure 2C).

To validate our approach, we first tested the principles of our method on non-dynamic fluorescent signals. Using our method, we achieved two Z plane imaging of eGFP-positive ICC and eGFP-positive SMCs in small intestine muscles distributed in two distinct tissue layers across the depth of the tissue (Figure 3). We simultaneously visualized ICC-DMP and ICC-MY that are located in different layers of muscle within the same FOV (i.e., in different Z planes; Figure 3A, left panels) and imaged SMCs in circular and longitudinal muscle layers (also located in different Z planes; Figure 3A, right panels). Higher magnification images of ICC-DMP and ICC-MY of ICC in the small intestine show the district morphology of each cell type correlating to each tissue layer (Figure 3B). To estimate how much the tissue was displaced by the piezo actuator, the distance between ICC-MY and ICC-DMP was calculated. To achieve this, we used confocal microscopy to visualize each subpopulation of ICC, noted their location in the Z axis, and then calculated the distance between them. In the small intestine, we found that the average distance between ICC-DMP and ICC-MY was 11.9 ± 0.9 μm (Figure 3C, n = 10). In the colon, distances between the four types of ICC in the colon were summarized in Figure 3D and Table 1.

To further validate our technique, we measured dynamic fluorescent signals, such as Ca²⁺, across two distinct tissue layers simultaneously. High-resolution, real-time Ca²⁺ imaging was conducted in multiple classes of interstitial cells of Cajal (ICC) and multiple classes of smooth muscle cells (SMCs) located in different Z planes and organized into two tissue layers. This was achieved using mice genetically engineered to express optogenetic sensors in both ICC (GCaMP6f, a green Ca²⁺ sensor) and smooth muscle cells (SMCs, RCaMP1.07, a red Ca²⁺ sensor; Figure 4A). This dual-sensor system allowed for precise monitoring of Ca²⁺ dynamics in different cellular populations across spatially separated tissue layers. Using a dual color spinning-disk confocal system (Figure 4B; see material section) we were able to resolve dynamics of intercellular Ca²⁺ in multiple cell populations. For example, ICC subtypes (ICC-MY, ICC-IM) express the green GCaMP signal and SMCs subtypes (circular and longitudinal muscle) express the red RCaMP signal. We were able to visualize all these four cell types that are located within two different layers (Z planes) across the depth of the sample.

Colonic muscles can generate propulsive contractions to move contents in an oral to anal direction. This motility behavior is known as colonic migrating motor complex (CMMC; Koh et al., 2022; Smith et al., 2014). Information on how different cells function and coordinate at different layers within the muscle during CMMC is not known due to limitations in the current imaging approaches. Our method overcomes these limitations by allowing the visualization of Ca²⁺ activity in four types of cells, organized in two layers (Z planes), simultaneously. Imaging of ICC-MY in the myenteric region, ICC-IM within the longitudinal region and imaging of circular and longitudinal SMCs is shown in Figure 5 and Supplementary Video S1. Under basal conditions, ICC-MY fired stochastic, localized Ca²⁺ transients, that occasionally organized into a Ca²⁺ flash across the network (Figures 5A, B). ICC-IM fired stochastic Ca²⁺ transients as discrete events (Figures 5A, B). SMCs in the circular and the longitudinal muscles showed little basal Ca²⁺ activity, but when Ca²⁺ flashes occurred in ICC-MY, Ca²⁺ signals were resolved in SMCs. When CMMC were initiated, Ca²⁺ transients were coordinated in all four cell types as seen in Figures 5A, B and Supplementary Video S1. Ca²⁺ events were tabulated as shown in Figures 5C–E and Table 2. During CMMCs both ICC-MY and ICC-IM respond and fire Ca²⁺ transients that can initiate a Ca²⁺ events in both circular and longitudinal smooth muscles.

In summary, the novel approach described allows real-time, high-resolution imaging across multiple tissue planes simultaneously, via the integration of confocal microscopy and piezo actuators. This method facilitates examination of cell-to-cell coordination in varying tissue depths. The study demonstrates, for the first time, how multiple cell types within the muscle wall of the colon behave and coordinate their functions to produce colonic motor complexes and motility patterns.