Spatial relationship between telocytes, interstitial cells of Cajal and the enteric nervous system in the human ileum and colon

Abstract Telocytes (TCs) are recently described interstitial cells, present in almost all human organs. Among many other functions, TCs regulate gastrointestinal motility together with the interstitial cells of Cajal (ICCs). TCs and ICCs have close localization in the human myenteric plexus; however, the exact spatial relationship cannot be clearly examined by previously applied double immunofluorescence/confocal microscopy. Data on TCs and submucosal ganglia and their relationship to intestinal nerves are scarce. The aim of the study was to analyse the spatial relationship among these components in the normal human ileum and colon with double CD34/CD117 and CD34/S100 immunohistochemistry and high‐resolution light microscopy. TCs were found to almost completely encompass both myenteric and submucosal ganglia in ileum and colon. An incomplete monolayer of ICCs was localized between the TCs and the longitudinal muscle cells in ileum, whereas only scattered ICCs were present on both surfaces of the colonic myenteric ganglia. TC‐telopodes were observed within colonic myenteric ganglia. TCs, but no ICCs, were present within and around the interganglionic nerve fascicles, submucosal nerves and mesenterial nerves, but were only observed along small nerves intramuscularly. These anatomic differences probably reflect the various roles of TCs and ICCs in the bowel function.

plexus between the inner and outer circular muscle (ICC-DMP) in ileum, [10][11][12] and the submuscular plexus (ICC-SMP) at the submucosal-circular muscle border in colon. [11][12][13] Immunohistochemical studies have since the 1990s revealed another fibroblast-or ICC-like cell in similar localization as ICCs, but with different phenotype: negative for c-kit (CD117) and positive for CD34 and platelet-derived growth factor receptor α (PDGFRα). 12,14,15 The term telocyte (TC) was coined in 2010 by Popescu and Faussone-Pellegrini for these ICC-like cells, and their long and very thin cytoplasmic projections were called telopodes. 16 TCs have been localized in several other organs as well. 17 Several investigations have suggested that TCs and PDGFRα-positive cells influence the intestinal motility, among many other functions. 12,[18][19][20][21] Electron microscopic studies showed that both ICCs and fibroblast-like cells were present around the myenteric ganglia of the human small and large intestine. 5,9 Double immunofluorescence, combined with confocal microscopy, proved that CD34immunoreactive, fibroblast-like cells were distinct but closely attached to ICCs both in human and mouse stomach, small intestine and colon. 14,22 These CD34-positive cells were also S100negative and accompanied S100-positive glial cells and nerve fibres. 14 The exact spatial relationship between the different cellular components in the myenteric plexus is, however, better analysed in conventional microscope after double immunostainings. The few studies performed regarding ICCs and TC and their relation to the submucosal ganglia have shown varying results. 6,8,11,12 Therefore, the aim of the present study was to examine double immunohistochemistry for ICCs, TCs and glial/Schwann cells in the light microscope, in regard to the connections between these cells and the neural plexi of the ENS and the small nerves in the muscle layers of human ileum and colon.

| MATERIAL AND ME THODS
The study was performed according to the Declaration of Helsinki and approved by the Ethical Review Board at Lund University (2012/527, date of approval 25/10/2012). Since the samples were unidentified samples used as controls, with only sex and age as known variables, the subjects did not have to leave informed consent, according to Swedish ethical rules. The study is not a part of a clinical trial.
Four specimens of normal human ileum and two specimens of normal colon were used for the analyses. The samples were taken from macroscopically normal ileum of right-sided hemicolectomy resections due to carcinoma (3 men of 55, 60 and 84 years of age, respectively, and one woman, 60 years old) and normal sigmoid colon also resected due to carcinoma (2 men, 65 and 71 years old). The specimens were fixed in buffered formalin overnight and embedded in paraffin. Serial sections of 4 μm thickness were deparaffinized and stained by haematoxylin & eosin. For immunohistochemistry, CC1 ph 8.5 buffer (Ventana Medical Systems) was applied for the deparaffinized unstained sections before applying the antibodies.
For detection of the immunostaining, Ultra View DAB Kit (brown colour) and Ultra View Fast Red Kit (red colour) (both Ventana Medical Systems) were used in Ventana's BenchMark ULTRA automated immunostainer system.
With double immunohistochemistry, either CD117 was detected first with Fast Red Kit followed by CD34 with DAB Kit, or CD34 was detected first with DAB Kit followed by S100 with Fast Red Kit.
The double stainings were also performed in reverse order. As internal control served endothelium of blood vessels (CD34 + ) and mast cells (CD117). As negative controls, the primary antibodies were omitted and replaced by serum. Nuclear counterstain was Mayer's haematoxylin.

Quantification of TCs and ICCs in the myenteric plexus showed 12
TCs per mm vs 24 ICCs per mm. Few telopodes were observed within the ganglia, probably at the beginning of interganglionic nerve fascicles, in contrast to ICC projections, which were never seen within the ganglia ( Figure 1B

| Colon
The pattern of TCs and the localization of ICCs around the ganglia were similar in the colon to those of the ileum with two exceptions: ICCs were not present within the interganglionic fascicles (data not shown).

| Submucosal ganglia in the ileum and colon
Telocytes were present in a single 'TC-sheath' encompassing the submucosal ganglia in both the ileum and colon, but ICCs never occurred in this localization ( Figure 4A-C). Occasionally, telopodes entered the ganglion ( Figure 4B, 4). Within the nerves, the TCs/telopodes showed a pattern similar to the myenteric fascicles, as well as to the nerves within the mesenterium ( Figure 4C).

| Small nerve fibres in the muscle layers
On transversal section within the muscles, TCs formed a network with long telopodes ( Figure 5A, right lower corner). TCs followed the small nerves in both ileum and colon, but intraneural TCs could not be observed within the muscle layers. Nerves larger than 5 μm thick were surrounded by a 'network' of TCs ( Figure 5A-C), whereas TCs/telopodes were present in a scattered way along the smaller nerves ( Figure 5D). contacts with neurons, as described by some, 23 but contradictory to others' description that fibroblast-like cells never penetrated the ganglia. 9 ICCs could not be found within the myenteric ganglia in the present and some former studies, 8,9 whereas some authors have described ICCs and their processes within ganglia in both ileum and colon. [23][24][25] Telocytes entered the myenteric ganglia at the beginning of the interganglionic nerve fascicles and ran on the surface of the fascicles in both ileum and colon, which has already been demonstrated. 22,23 The novel findings were that TCs/telopodes were observed within the fascicles and that small intramuscular nerve fibres were surrounded or followed by TCs/telopodes on their surface.

| TCs and ICCs between the muscle layers, intramuscularly and in the interlamellar space
The almost complete TC layer around submucosal plexi con- Our findings of TCs and ICCs in the interlamellar septa of the circular muscle layer are well-known data. 5,8,23,24,27,28 The numbers of TCs and ICCs can be related to another report which found comparable findings, without given exact localization within the bowel wall. 14 We have to emphasize that the number of samples in our analysis is too low for any certain conclusion regarding the numerical relation between these two cell types.
Thus, the architecture of the two cell types differs markedly, which may reflect the functional role of these cells. Furthermore, the shape and distribution of TC and their location may vary in different wall layers of ileum, as previously described in a large study cohort. 23  ICCs. 24,26 Although the loss of TC and ICC was parallel in inflammatory bowel diseases, 24,26 findings in diabetes mellitus showed that FLC, proven to be identical to TC, 23 were more resistant to degeneration than ICC. 31,32 Hypothetically, these data might explain the histopathological experience that submucosal ganglia are more preserved than myenteric ganglia in dysmotility disorders (personal experience).
Differences between studies may be explained by small sample sizes of most studies. Differences in affinity and specificity of antibodies from different producers and batches may be another explanation to contradictory results. One hypothesis is that malignant diseases may affect tissues at a distance from the local tumour, 33 although the resection margin looks healthy. Thus, one cannot exclude some loss of ICC in samples from patients obtained during surgery for a malignant tumour.
In conclusion, regarding the localization and number of TCs and ICCs, the difference between human myenteric and submucosal ganglia, nerves and muscle layers on the one hand, and between ileum and colon on the other hand might probably reflect the role of these cells in the function of the bowel, that is the maintenance of homeostasis in the microenvironment and the motility of the bowel, respectively. Further investigations are necessary to prove or disprove this hypothesis.