If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Correspondence Address correspondence to: Shiue-Cheng Tang, PhD, National Tsing Hua University, Department of Medical Science, 101, Section 2, Kuang Fu Road, Hsinchu 30013, Taiwan. fax: (886) 3-571-5934.
Interstitial cells of Cajal (ICC) closely associate with nerves and smooth muscles to modulate gut motility. In the ICC microenvironment, although the circulating hormones/factors have been shown to influence ICC activities, the association between ICC and microvessels in the gut wall has not been described. We applied three-dimensional (3D) vascular histology with c-kit staining to identify the perivascular ICC and characterize their morphologic and population features in the human colon wall.
Full-thickness colons were obtained from colectomies performed for colorectal cancer. We targeted the colon wall away from the tumor site. Confocal microscopy with optical clearing (use of immersion solution to reduce scattering in optical imaging) was performed to simultaneously reveal the ICC and vascular networks in space. 3D image rendering and projection were digitally conducted to illustrate the ICC–vessel contact patterns.
Perivascular ICC were identified in the submucosal border, myenteric plexus, and circular and longitudinal muscles via high-definition 3D microscopy. Through in-depth image projection, we specified two contact patterns—the intimate cell body-to-vessel contact (type I, 18% of ICC in circular muscle) and the long-distance process-to-vessel contact (type II, 16%)—to classify perivascular ICC. Particularly, type I perivascular ICC were detected with elevated c-kit staining levels and were routinely found in clusters, making them readily distinguishable from other ICC in the network.
We propose a new subclass of ICC that closely associates with microvessels in the human colon. Our finding suggests a functional relationship between these mural ICC and microvessels based on the morphologic proximity.
This work explores the association between interstitial cells of Cajal (ICC) and the microvascular network in the human colon wall using a novel high-definition three-dimensional microscopic approach. The authors propose the existence of a new subclass of ICC, the perivascular ICC.
In the colon as well as other parts of the digestive tract, interstitial cells of Cajal (ICC) integrate with nerves and muscles as part of the machinery to regulate motility.
the signal transduction pathway of the latter—that is, from the hormones to ICC—is poorly understood. In an attempt to establish the connection, lines of evidence have emerged to indicate that ICC may be directly influenced by the intestinal hormones: (1) the expression of the gut hormone cholecystokinin receptors on ICC
Particularly, regarding the serotonin pathway, although the source of serotonin and the function(s) of the 5-HT receptors on ICC are unclear, the receptors are hypothesized to participate in the slow wave initiation and/or alteration based on their known functions.
In addition to modulating the slow-wave activity, circulating hormones and factors have also been shown to maintain the population of ICC in the gut wall. For example, the signaling pathway initiated by the binding of the stem cell factor (SCF) to its receptor tyrosine kinase c-kit is essential for the ICC development and survival.
On the other hand, activation of the 5-HT2B receptor on ICC by serotonin has been shown to increase the ICC proliferation in vitro, while lack of this activation reduced the density of ICC in adult mice.
), study of the ICC microenvironment ideally should include the vasculature to integrate the local ICC microenvironment with the circulation. From a therapeutic perspective, examining the vascular network around ICC provides an opportunity to evaluate the efficiency of therapeutic agents to reach ICC through blood vessels in treating motility disorders such as diabetic gastroparesis
However, reports concerning the interaction between ICC and blood vessels are lacking. This is largely because of the dispersed nature of the ICC and vascular networks, which cannot be easily portrayed by the standard microtome-based two-dimensional histology to characterize their spatial relationship.
To visualize the intestinal tissue networks, we previously developed a penetrative 3D imaging method based on preparation of transparent tissues with optical clearing
In this research, we simultaneously targeted the ICC and vascular networks using paired c-kit and CD31 immunostaining to reveal their association in the human colon wall. Importantly, a new subclass of ICC, the perivascular ICC, is presented in this report with qualitative and quantitative analyses to identify their unique features and population density. The physiologic implications of the mural ICC are discussed.
Materials and Methods
Collection and use of human tissues were approved by the institutional review board of National Taiwan University Hospital, Hsinchu Branch, with written consent from the patients to use the obtained tissues. Colonic tissues were obtained from colectomies performed for nonobstructive carcinoma. Samples of normal muscularis were obtained at least 5 cm apart from the carcinoma. The removed tissues were first perfused with phosphate-buffered saline (PBS) through the mesenteric artery to flush the residual blood and then with ice-cold 4% paraformaldehyde for 30 minutes for fixation. Afterward, the tissues were postfixed in 4% paraformaldehyde solution overnight at 4°C before being transferred to 0.1% paraformaldehyde for preservation. Specimens were later sectioned to 300 μm in thickness by vibratome before being applied for tissue labeling.
The fixed specimens were immersed in 2% Triton X-100 solution for 2 hours at 15°C for permeabilization. The primary antibodies used to reveal ICC were a monoclonal rabbit anti-c-kit antibody (cat. no. 1522–1; Epitomics, Burlingame, CA) and a polyclonal rabbit anti-Ano1 antibody (also known as anti-TMEM16A, cat. no. ab53212; Abcam, Cambridge, MA). A monoclonal mouse anti-CD31 antibody (cat. no. MS-353-S1; Thermo Scientific, Fremont, CA) was used to reveal the vasculature.
Before we applied the antibodies, the tissue was rinsed in PBS. This was followed by a blocking step, incubating the tissue with the blocking buffer (2% Triton X-100, 10% normal goat serum, and 0.02% sodium azide in PBS). The primary antibodies were then diluted in the dilution buffer (1:100, 0.25% Triton X-100, 1% normal goat serum, and 0.02% sodium azide in PBS) to replace the blocking buffer and incubated overnight at 15°C. Alexa-Fluor 647-conjugated goat anti-rabbit secondary antibody and Alexa-Fluor 546-conjugated goat anti-mouse secondary antibody (1:200; Invitrogen, Carlsbad, CA) were used to reveal the immunostained structures. Afterward, nuclear staining by SYTO16 (4 μmol/L; Invitrogen) was performed at 15°C for 1 hour. The labeled specimens were then immersed in the optical clearing solution FocusClear (CelExplorer, Hsinchu, Taiwan) before being imaged via confocal microscopy.
A Zeiss LSM510 Meta confocal microscope (Carl Zeiss, Jena, Germany) equipped with 10× Fluar objective lenses (numerical aperture [NA]: 0.5) was used to acquire the gross view of the ICC network (Figure 1, tile-scanning mode: 2 × 2 image stacks; optical section: 10 μm; z-axis increment: 5 μm). The 40× LD C-Apochromat water immersion lenses (NA: 1.1; working distance: 620 μm; optical section: 2 or 3 μm; z-axis increment: 1 or 1.5 μm) were used to acquire the detailed ICC and vascular features. The 25× LD Plan-Apochromat glycerin immersion lenses (NA: 0.8; working distance: 570 μm; optical section: 5 μm; z-axis increment: 2.5 μm) were used to acquire the image stacks for ICC quantitation. The laser-scanning process was operated under the multitrack scanning mode to sequentially acquire signals in multiple channels, including the transmitted light channel. The Alexa Fluor 647-labeled tissue structures were excited at 633 nm, and the fluorescence was collected by the 650- to 710-nm band-pass filter. The Alexa Fluor 546-labeled tissue structures were excited at 543 nm, and the signals were collected by the 560- to 615-nm band-pass filter. The SYTO16 signals were excited at 488 nm, and the fluorescence was collected by the 500- to 550-nm band-pass filter.
Image Processing and Projection
The Avizo 6.2 image reconstruction software (VSG, Burlington, MA; operated under a Dell T7500 workstation), the Zen software (Carl Zeiss), and the LSM510 software (Carl Zeiss) were used for processing, projection, and analysis of the confocal image stacks. Avizo’s noise-filtering algorithms were applied for background noise reduction. Signal tracing, image segmentation, and feature extraction were performed by the Label Field function of Avizo.
Quantitation of ICC was performed using the image stacks derived from the transverse sections of circular muscle. The Landmark function of the Avizo software was applied to digitally mark and count the ICC in the image stack. ICC were defined as the c-kit-positive cells with at least two processes extending from the cell body, which was specified by the c-kit enclosed nuclear signal. Supplementary Video 4 gives an example of identifying the type I and type II perivascular ICC based on the contact patterns between c-kit and CD31 signals. Overall, 59 image stacks were used in quantitation (Table 1). The c-kit staining levels of different ICC subclasses were assessed by the Histogram module of the Zen software with a 50-μm projection depth to estimate the fluorescence signal intensities.
Table 1Ratio of Perivascular ICC in ICC Population in Circular Muscle
Perivascular ICC (%)
ICC not associated with blood vessels (%)
Type I (cell body-to-vessel contact)
Type II (process-to-vessel contact only)
NOTE: Number of ICC counted from each subject: A, n = 438; B, n = 409; C, n = 408; D, n = 428; E, n = 381; and F, n = 391. Density of ICC in circular muscle (ie, sum of the three subclasses of ICC per tissue volume) of each subject: A, 2091 ± 890 cells/mm3 (n = 14 image stacks); B, 2545 ± 763 cells/mm3 (n = 8); C, 2063 ± 621 cells/mm3 (n = 10); D, 2261 ± 416 cells/mm3 (n = 9); E, 2751 ± 1139 cells/mm3 (n = 9); and F, 2667 ± 415 cells/mm3 (n = 8).
The results are expressed as mean ± standard deviation. The data of the intrasubject ICC subgroups were evaluated by unpaired Student t test (Figure 6E). Data of intersubject ICC subgroups were evaluated by regression analysis. P < .05 was considered statistically significant.
We conducted a regression analysis on the ICC of the six subjects to assess whether type I perivascular ICC are indeed brighter than the other two ICC subgroups. In the analysis, we regress the c-kit staining levels on dummy variables for type I, type II, and the six subjects using the ICC not associated with blood vessels in Subject A as the benchmark group (ie, the mean c-kit staining level for this benchmark group is a0).
a1: Regression coefficient for the variable “Type I perivascular ICC”
a2: Regression coefficient for the variable “Type II perivascular ICC”
b1: Regression coefficient for the variable “Subject B”
b5: Regression coefficient for the variable “Subject F”
error: The error term of the regression, with mean 0
c-kit staining level: Expressed as percentage of signal saturation (as shown in Figure 6E)
TypeI: A dummy variable of 1 if the ICC is Type I, and 0 otherwise
TypeII: A dummy variable of 1 if the ICC is Type II, and 0 otherwise
SubjectB: A dummy variable of 1 if the ICC is from Subject B, and 0 otherwise
SubjectF: A dummy variable of 1 if the ICC is from Subject F, and 0 otherwise
Cell type adjustment
a0 = 41.32% (t-stat = 32.90)
a1 = 9.81% (t-stat = 7.30)
a2 = 2.80% (t-stat = 1.98)
b1 = 8.88% (t-stat = 5.71)
b2 = 8.99% (t-stat = 4.93)
b3 = 7.25% (t-stat = 4.13)
b4 = 8.40% (t-stat = 4.62)
b5 = −1.44% (t-stat = −0.80)
Tests of Hypotheses
Hypothesis 1: a1 = 0 (Type I perivascular ICC and ICC not associated with blood vessels have the same c-kit staining level.)
Result: t-stat = 7.30 → Reject Hypothesis 1 at the 1% level.
Hypothesis 2: a2 = 0 (Type II perivascular ICC and ICC not associated with blood vessels have the same c-kit staining level.)
Result: t-stat = 1.98 → Reject Hypothesis 2 at the 5% level.
Hypothesis 3: a1 − a2 = 0 (Type I and type II have the same c-kit staining level.)
Result: F-stat = 16.06 → Reject Hypothesis 3 at the 1% level.
Conclusions of Regression Analysis
The estimate of a1 is 9.8% (t-stat = 7.30), indicating that the type I perivascular ICC are on average 9.8% brighter than the ICC not associated with blood vessels, after accounting for differences across subjects. The difference is statistically significant at the 1% level.
The estimate of a2 is 2.8% (t-stat = 1.98), indicating that the type II perivascular ICC are on average 2.8% brighter than the ICC not associated with blood vessel. The difference is statistically significant at the 5% level.
The result also indicates that the type I perivascular ICC are on average 7.0% brighter than the type II perivascular ICC. The difference (9.8%–2.8%) is statistically significant at the 1% level (result of F test), with a F-stat of 16.06.
The three ICC subgroups (type I perivascular ICC, type II perivascular ICC, and ICC not associated with blood vessels) are significantly different in the c-kit staining levels.
Deep-Tissue Microscopy of Interstitial Cells of Cajal and Vascular Networks
To visualize the ICC and vascular networks in space, we prepared transparent human colon specimens by optical clearing to reveal the c-kit and CD31-labeled structures via deep-tissue microscopy. Figure 1 shows the submucosa, circular muscle, myenteric plexus, and longitudinal muscle and their associated ICC populations—the ICC at the submucosal border (ICC-SM), ICC in the circular muscle (ICC-CM), ICC around the myenteric plexus (ICC-MY), and ICC in the longitudinal muscle (ICC-LM)—in these domains. Importantly, through merged projection of c-kit and CD31 signals (Figure 1D), locations with the apparent overlap of the ICC and vascular networks were revealed in the tissue map.
Perivascular Interstitial Cells of Cajal in the Submucosal Border and Myenteric Plexus Populations
Zooming into the submucosal boundary, Figure 2A and B shows a subgroup of ICC-SM closely following the microvessel in space. The cell body and processes of the perivascular ICC contact the vascular walls, showing their intimacy. The perivascular ICC can be distinguished from the c-kit+ mast cells (Figure 2A) because the latter lack the prolonged processes to connect other ICC in the plexus.
An association between ICC and microvessels was also observed in the myenteric plexus (Figure 2C and D). A zoom-in examination of the myenteric ganglion revealed that a subclass of ICC-MY positioned their cell bodies around the ganglion with the processes extending to and docking on the walls of the periganglionic microvessels (Figure 2E and F, and Supplementary Video 1).
Perivascular Interstitial Cells of Cajal in Longitudinal Muscle
In addition to the ICC-SM and ICC-MY, the perivascular ICC can also be found in the ICC populations in the longitudinal and circular muscles (Figure 3). Using the transparent longitudinal muscle, we identified the embedded microvessels and the perivascular ICC by overlaying and pairing the transmitted light and fluorescence signals (Figure 3A and B). Following the focal depth, we specified a subset of the ICC-LM residing on the microvessel wall (Figure 3C). Zoom-in examination of the mural ICC confirmed the aligned feature of ICC with microvessels and their extended contacts along the vessel wall (Figure 3D and Supplementary Video 2).
Perivascular Interstitial Cells of Cajal in Circular Muscle
Among the different layers/domains of the colon wall, the circular muscle provides the best environment to characterize the perivascular ICC because of the defined domain boundary, homogeneity, and large area for examination (Figures 1A and B and 3E). We therefore zoomed into this region and performed qualitative and quantitative analyses of the perivascular ICC to characterize their detailed morphologic and population features.
Before entering the detailed ICC presentation, we would like to stress two aspects of the CD31 immunostaining to specify the revealed vascular structure. First, the CD31 immunohistochemistry identifies the platelet endothelial cell adhesion molecules (or PECAM-1) and thus reveals the endothelial cells of all microvessels, including the capillary, arteriole, and venule. Second, similar to the transparent longitudinal muscle (Figure 3A), in the transparent circular muscle we used both the transmitted light and CD31 signals to identify the blood vessels (Figure 7A) and to distinguish them from the lymphatic vessels (Figure 7B and C; the latter lack rigid vessel walls and primarily reside around the myenteric plexus against the muscle layers).
In qualitative analysis, Figure 3F shows two contact patterns between the ICC and microvessels: (1) the intimate cell body-to-vessel contact and (2) the long-distance process-to-vessel contact. We classify the ICC with the first feature as the type I perivascular ICC (with or without the second feature) and those with the second feature and only the second feature as the type II perivascular ICC. Based on the classification, we quantified the ratios of type I (18.4% ± 2.6%) and type II (15.7% ± 3.7%) perivascular ICC in the ICC-CM population in six individuals (Table 1) (note that the average ICC-CM density derived from the six individuals is 2396 cells/mm3; ICC are defined as the c-kit-positive cells with at least two processes extending from the cell body).
Added together, the overall ratio of the perivascular ICC in ICC-CM was as much as 34.1% (ie, one of three ICC-CM contacting the blood vessels), reflecting the importance of this ICC subgroup in the ICC-CM population. Supplementary Videos 3 and 4 present the ICC analysis in a stereo fashion.
To rule out that the association of ICC with microvessels is due to their random contacts in space, Figure 8 illustrates our approach to evaluating the potential random contacts of the two tissue networks in space. This test was done by digitally swapping the ICC population (ie, the c-kit signals) between two adjacent image stacks and then examining the ICC-vessel contacts in an arbitrary fashion. In five pairs of these images, only 5.7% of the ICC-CM contacted the microvessels in their arbitrary locations. The ratio was statistically significantly lower than the 34.1% ratio of the perivascular ICC in the ICC-CM (P < .001).
Furthermore, Figure 4 zooms into individual type I perivascular ICC to depict the detailed ICC morphologic features around the microvessels. Intimate ICC-vessel associations, such as two ICC clamping the microvessel (Figure 4A), processes of ICC embracing the vessel wall (Figure 4B), and the ICC residing at the vessel loop (Figure 4C), were displayed in projection and from various angles (Figure 4D–F and Supplementary Video 5). A gallery of 16 additional projections is presented in Figure 9 to illustrate the ICC-vessel contacts with high definition.
Finally, based on the overlap of the fluorescence signals and the resolving power of confocal microscopy (laser lines at 543 and 633-nm were used in the fluorescence illumination), the span between the perivascular ICC and the vessel wall was estimated to be within 300 nm.
Type I Perivascular Interstitial Cells of Cajal in Clusters and with Elevated c-kit Staining Levels
In addition to the intimate cell body-to-vessel contact, Figure 6 illustrates two additional features of type I perivascular ICC on the microvessels. First, the type I perivascular ICC were routinely found in clusters on the vessel wall (27.0% ± 11.4% of type I perivascular ICC in circular muscle in six subjects). Figure 6A–D shows three examples of the ICC cluster on the microvessel with adjacent cell bodies within a span of 50 μm. Second, the type I perivascular ICC were detected with elevated c-kit staining levels in comparison with those of the type II and ICC not associated with blood vessels in the intrasubject ICC analysis (Figure 6E; P < .05 against the latter, unpaired Student t test) and the intersubject regression analysis (see Regression Analysis) (P < .01 against both groups). Particularly, in five of the six subjects, we observed a further increase in c-kit signals of the clustered type I perivascular ICC, making them readily distinguishable from other ICC in the network.
Perivascular Interstitial Cells of Cajal in the Circular Muscle Network
Zooming out from the contacts of ICC with blood vessels, Figure 5A–D and Supplementary Video 6 show the spatial integration of the perivascular ICC with the entire ICC-CM. Importantly, through in-depth projection, we were able to trace the cell bodies and processes of the perivascular ICC to confirm their coupling with other ICC in space. Figure 5E presents a second example of the cell-cell coupling of ICC from various angles.
In addition to the c-kit-labeled ICC network, Figures 10 and 11 present the Ano1-labeled ICC network to confirm the perivascular ICC. Because the Ano1 expression has been associated with the function of the calcium-dependent chloride channel and the generation of the slow waves,
in addition to the c-kit staining, to specify the perivascular presence of ICC in the circular muscle (Figure 10E). Finally, Figure 11 presents the Ano1-labeled ICC imaged with different tissue depths to examine the acquired fluorescence signals. Because the projection depth of the optically cleared tissue (60 μm in Figure 11C) is larger than the thickness of a standard tissue section (15 μm in Figure 11B), such a difference explains why it appears that more ICC are revealed in our images and videos in comparison with the number of ICC derived from the standard immunohistochemistry.
Overall, the 3D imaging, projection, and quantitation of the perivascular ICC in the human colon identify the integration between the ICC and vascular networks and underline the presence of microvessels in the immediate ICC microenvironment.
The classic view of the ICC microenvironment includes the triad of ICC, nerves, and smooth muscles that contribute to the initiation and regulation of the spontaneous peristaltic movements of gut wall.
In addition to the three components, in this research we applied high-definition 3D microscopy to identify the vascular network as a fourth component to directly associate with the ICC plexuses in the human colon wall. The association was established by the ICC cell body-to-vessel contact and/or the processes docking on the vessel wall to integrate the two network systems. The intimate ICC-to-vessel association together with their elevated c-kit staining levels and clustering on the vessel wall prompted us to propose a new subclass of ICC, the perivascular ICC, to highlight their unique morphologic features.
Unlike the conventional ICC classification based on the locations—the ICC-SM, ICC-CM, ICC-MY, and ICC-LM
—the new ICC-to-vessel classification that we propose emphasizes the contact patterns (Table 1) and the potential vascular influence on the perivascular ICC. The morphologic connection of the ICC plexus to the vascular network supports the concept that ICC rely on both the nervous and circulatory signals to modulate the slow-wave activity in response to physiologic cues. The circulatory pathway thus represents a natural route to target ICC for potential pharmacologic intervention to influence the ICC function.
For the ICC maintenance, however, the situation is complicated by the ICC microenvironment. Our high-definition images reveal that the perivascular ICC are potentially the endothelium-associated anchorage-dependent cells, particularly the type I perivascular ICC, which receive the substratum and likely the c-kit-mediated stimulation from the microvessel (the latter was suggested by the high c-kit staining levels) (Figure 6E). Because chronic diseases such as diabetes and hypertension can cause microvascular injury and/or dysfunction,
whether the disturbances of microcirculation lead to the remodeling of perivascular ICC and a change of the population density requires further investigation to understand the potential microvascular influence on the ICC function and motility disorders.
The technical advance of preparing the transparent human colon specimens allowed us to use 3D microscopy to simultaneously visualize the ICC and vascular networks in an integrated fashion, which otherwise cannot be easily portrayed by the standard two-dimensional histology. Because of the improved optical condition, we were able to apply transmitted light imaging to identify the domain boundaries and microstructures of the colon wall (Figures 1A and 3A), which served as the landmarks to confirm the signals derived from fluorescence labeling and confocal microscopy. However, this imaging approach has the following limitations. First, the tissue clearing condition is not compatible with calcium imaging to identify the functional connection between ICC and microvessels. Second, the deep-tissue imaging approach is not compatible with super-resolution microscopy to examine the membrane structures, such as the potential gap junctions between the endothelium and ICC. Use of electron microscopy to characterize the ultrastructure of the perivascular ICC will be needed in the future to further define the cell body (type I) and process (type II) contacts of ICC with the endothelium. Detailed information about the ultrastructure and contact patterns in the context of disease (such as diabetic gastroparesis) will help link the unique anatomy of perivascular ICC with the physiology and pathophysiology of intestine.
Despite the limitations, 3D microscopy with optical clearing has two technical advantages. First, the method is compatible with the standard tissue fixation and labeling procedures, including immunohistochemistry, to identify the tissue structures and networks of interest. Second, the imaging approach is compatible with tile scanning in confocal microscopy for large-area network mapping. These two advantages have made 3D microscopy with optical clearing an important tool in neuroanatomy for the visualization and study of the central and peripheral nervous systems,
In characterizing the perivascular ICC, we chose to use the CD31 immunostaining to reveal the microvessels because the marker has been routinely used in the analysis of microvessel density to evaluate the tumor angiogenesis.
This practice led us to use the same approach to identify the colonic microvessels in this research. In reporting the ratios of the perivascular ICC (Table 1), we assume that the CD31 immunohistochemistry identified most, if not all, of the microvessels in the specimen (Figure 7A). However, we do not rule out the possibility that the apparent vessel density would change if a different vascular marker was used, which could in turn influence the estimated ratios of the perivascular ICC.
In quantitation of the c-kit staining levels (Figure 6E), it should be noted that the fluorescence signals derived from the c-kit molecules were amplified in the staining and imaging processes. In acquiring the signals, we assumed that the amplification was proportional and minimally influenced by the sensitivity and saturation of the antigen-antibody binding and microscopic detection. In addition, we assumed that the same antibody-to-c-kit binding efficiency was held across the six subjects, although a slightly different ICC population profile appears in subject F in Figure 6E.
In summary, we applied 3D microscopy to identify a new subclass of ICC, the perivascular ICC, in the human colon wall. Before this study, the microvascular network had not been demonstrated in the ICC microenvironment. We performed both survey imaging and high-definition microscopy of the colon wall to characterize the morphologic and population features of the perivascular ICC. The discovery of perivascular ICC lays the foundation for future research into the interplay between the ICC and endothelium to increase our understanding of the mechanisms that regulate ICC activity in health and the change of the ICC microenvironment in motility disorders.
The authors thank the Brain Research Center in the National Tsing Hua University for technical support in confocal imaging and postrecording image processing, and Dr. M. Schemann at Technical University of Munich for helpful suggestions.
(See Figure 3E and F.) First half: 360° projection of perivascular ICC in circular muscle. Green: c-kit. Red: CD31. Blue: nuclei. Second half: projection and illustration of the type I (green) and type II (white) perivascular ICC in space.
(See Figure 3E and F, and quantitation of ICC.) Recording of type I and type II perivascular ICC in circular muscle. Green: c-kit. Red: CD31. Blue: nuclei. Note: Tracing the ICC cell bodies and processes in the image stack allowed us to identify and quantify the type I (yellow arrows) and type II (magenta arrows) perivascular ICC in space.
Conflicts of interest The authors disclose no conflicts.
Funding This study was funded in part by grants from the joint research program between National Taiwan University Hospital, Hsinchu Branch, and National Tsing Hua University [Grant 102N2734E1] and the National Science Council [Grant NSC 102-2628-B-007-002-MY2] (to S.C.T.).