SEARCH

SEARCH BY CITATION

Keywords:

  • liver;
  • CD117/c-kit;
  • portal limiting membrane;
  • telocytes

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. LITERATURE CITED

Portal interstitial cells of Cajal (PICCs), acting as vascular pacemakers, were previously only identified in nonhumans. Moreover, there is no evidence available about the presence of such cells within the liver. The objective of the study is to evaluate whether or not PICCs are identifiable in humans and, if they are, whether or not they are following the scaffold of portal vein (PV) branches within the liver. We obtained extrahepatic PVs and liver samples from six adult human cadavers, negative for liver disease, in accordance with ethical rules. They were stained with hematoxylin-eosin (HE) and Giemsa, and then we performed immunohistochemistry on formalin-fixed paraffin-embedded specimens for CD117/c-kit, a marker of the Cajal's cells. Immune labeling was also performed for S-100 protein, desmin, glial fibrillary acidic protein (GFAP), neurofilaments, α-smooth muscle actin (α-SMA), and CD34. c-kit-Positive PICCs were identified within the extrahepatic PV, in portal spaces, and septa. On adjacent sections, these PICCs were negative for all the other antibodies used. In conclusion, our study confirms the presence of extrahepatic PICCs on humans, which may act as a possible intrinsic pacemaker in the human PV. However, the intrahepatic PICCs, which were evidenced here for the first time, are in need for further experimental studies to evaluate their functional role. A promising further direction of the study is the PICCs role in the idiopathic portal hypertension. Anat Rec, 2011. © 2011 Wiley-Liss, Inc.

Interstitial cells of Cajal (ICCs), first described more than 100 years ago and initially thought to be present only in the gastrointestinal tract, are now considered to be responsible for the generation and/or propagation of slow waves (Povstyan et al., 2003; Huang and Xu, 2010). It is now known that ICCs play important roles in the regulation of gastrointestinal motility.

ICCs outside the gastrointestinal tract are usually known as interstitial Cajal-like cells (ICLCs). More recently, ICLCs with long and moniliform processes were termed telocytes and the respective processes were named telopodes (Popescu and Faussone-Pellegrini, 2010).

ICCs are known to be immunopositive for c-kit, a proto-oncogene that encodes a tyrosine kinase receptor (Mei et al., 2009), making it a useful marker for these cells (Ward and Sanders, 2001; Povstyan et al., 2003; Harhun et al., 2005; Sanders et al., 2006; Popescu and Faussone-Pellegrini, 2010). However, it is not mandatory for all ICCs to be labeled with c-kit antibodies (Huizinga and Faussone-Pellegrini, 2005). ICCs from the small intestine are divided in two subgroups, one located within the myenteric plexus (ICC-MY), with a pacemaker function and another closely associated with the deep muscular plexus (ICC-DMP), playing a role in the mediation of neural inputs (Mei et al., 2009; Huang and Xu, 2010). These two subgroups of ICCs are known as c-kit positive (Chen et al., 2007).

c-kit positive ICCs were for the first time identified in the wall of the portal vein (PV) in rabbits by Povstyan et al. (2003); further studies found that these portal interstitial cells of Cajal (PICCs) may act as pacemakers, controlling rhythmical contractions of the PV (Harhun et al., 2004).

Like ICCs but unlike smooth muscle cells, ICLCs are positively stained with methylene blue (Formey et al., 2011). ICLCs were detected in the myocardial sleeves of the human pulmonary veins (Gherghiceanu et al., 2008; Morel et al., 2008), especially in patients with atrial fibrillation, and were presumed to act as pacemaker cells in these veins (Morel et al., 2008).

No studies were yet performed to check the presence of PICCs in humans, and no presence of the portal hepatic ICCs was identified, even though it is known that the ICLCs are present in rhythmically active structures (Huizinga et al., 2009) such as the heart (Hinescu and Popescu, 2005; Hinescu et al., 2006; Popescu et al., 2006), PV (Harhun et al., 2004), the upper urinary tract and the urinary bladder (Lang et al., 2006), and the gallbladder (Hinescu et al., 2007).

Our objectives for this study were (1) to evaluate the presence of PICCs in humans and (2) to check whether or not they follow the portal venous scaffold within the liver. The study was designed as a qualitative one.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. LITERATURE CITED

PV tissue samples were obtained from six human adult cadavers, negative for any liver disease, (aged 58–72 years, male:female ratio 2:1), from the following topographic locations: retropancreatic, retroduodenal, and intraomental. Liver samples were also taken.

The study was performed according to the national laws regarding human cadaver manipulation, medical legal practice, and the reform of medical system, and Council of Europe, Committee of Ministers, Recommendation No. R (99) 3 on the Harmonization of Medico-Legal Autopsy Rules. The investigation was in agreement with the principles outlined in the Declaration of Helsinki. The Institutional Ethics Committee granted the approval.

The samples were embedded in paraffin, sectioned at 3 μm, and stained with hematoxylin-eosin (HE) and Giemsa stain, which is a mixture of methylene blue, eosin, and azure B. Histologically, the samples were negative for liver diseases. Immunohistochemistry on formalin-fixed paraffin-embedded tissues was performed, and adjacent sections were labeled with antibodies for CD117/c-kit, CD34, S-100 protein, desmin, glial fibrillary acidic protein (GFAP), α-smooth muscle actin (α-SMA), and neurofilaments (Table 1) using the ABC method. External positive controls were specifically labeled (see Table 1) and sections treated without primary antibodies served as negative controls.

Table 1. Antibodies used for the study of the PICCs
AntibodiesTypesSpeciesClonality, cloneSource, codeExternal positive controls
CD117/c-kitIgGRabbitMonoclonal, Y145BIOCARE PME 296 AASkin (mast cells), testis (germ cells)
CD34IgG1MouseMonoclonal, QBEnd/10BIOCARE PME 084 AATonsil, skin
S100 Protein cocktailIgG1 + IgG2aMouseMonoclonal, 15E2E2 + 4C4.9BIOCARE PME 089 AAMelanoma
DesminIgG1/kappaMouseMonoclonal, D33BIOCARE PME 036 AAUterus
Glial fibrillary acidic protein (GFAP[P])N/ARabbitPolyclonal, N/ABIOCARE PP 040 AAbrain
Smooth muscle actin (SMA)IgGRabbitMonoclonal, C04018BIOCARE PME 305 AAUterus, blood vessels
NeurofilamentIgG1 kappaMouseMonoclonal, NE14BioGenex, AM073-10MBrain

Microscopic slides were analyzed and snapshots were taken and scaled by using a ZEISS working station consisting of an AxioImager M1 microscope with an AxioCam HRc camera and the digital image processing software AxioVision.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. LITERATURE CITED

Histological Evaluation of the Portal Vein

PVs, distinctively identified on slides, consisted of a distinct intima, a narrow tunica media mainly composed of circular smooth muscle fibers and a thick adventitia, the later embedding longitudinal bundles of smooth muscle cells between two sublayers of collagenic extracellular matrix, which we have termed inner adventitia (IA), one between the longitudinal muscle and the media and outer adventitia (OA), the sublayer covering the longitudinal bundles of smooth muscle cells.

Giemsa stain distinctively identified dark blue multipolar cells with moniliform processes (Fig. 1) closely related to muscle bundles of the extrahepatic PVs. These were clearly differentiated from mast cells, which lack processes and are usually present in perivascular locations.

thumbnail image

Figure 1. Giemsa stain. Multipolar cell (arrow), with moniliform processes (arrowheads), closely related to a longitudinal muscle bundle of the extrahepatic PV wall.

Download figure to PowerPoint

PICCs of the Extrahepatic PV

Within PV walls, CD117/c-kit–positive cells (Figs. 2–8) were identified at all topographic levels (retropancreatic, retroduodenal, and intraomental). Within the PV wall, these cells were topographically located either within the IA, or the OA and were, closely but not exclusively, apposed to the smooth muscular bundles, both longitudinal (Figs. 2 and 3) and circular (Fig. 4).

thumbnail image

Figure 2. Oblique cut through the retroduodenal portal vein; positive CD117 immunolabeling (arrows) of cells in the vicinity of longitudinal muscle bundles (LMBs). The portal inner adventitia (IA) and media (M) are identified.

Download figure to PowerPoint

thumbnail image

Figure 3. PICC (arrow) of the extrahepatic PV, closely related to a longitudinal muscle bundle (lmb), and sending off moniliform processes (arrowheads).

Download figure to PowerPoint

thumbnail image

Figure 4. PICC (arrow) of the extrahepatic PV, closely related to a circular muscle bundle (cmb), and sending off a moniliform process (arrowheads).

Download figure to PowerPoint

thumbnail image

Figure 5. Immune labeling on successive sections for c-kit, desmin, and CD34. Within the wall of the extrahepatic portal vein c-kit positive cells (arrows) are desmin- and CD34-negative. However, adjacent c-kit negative cells are either desmin- and CD34-negative (arrowhead) or desmin-positive and CD34-negative (*).

Download figure to PowerPoint

thumbnail image

Figure 6. Immune labeling on successive sections for c-kit and CD34. c-kit-Positive cells with moniliform processes (arrows) are CD34-negative. Adjacent c-kit-negative cells (arrowhead) are CD34-positive.

Download figure to PowerPoint

thumbnail image

Figure 7. Immune labeling on successive sections for c-kit (A), S100 protein (B), CD34 (C), and desmin (D). Two c-kit-positive cells with moniliform processes (arrow, arrowhead) are S-100-negative (A, B). A neighbor cell (*), c-kit negative, is CD34- and desmin-positive (A, C, and D).

Download figure to PowerPoint

thumbnail image

Figure 8. Immune labeling on successive sections for c-kit (right panel), and α-smooth muscle actin (SMA, left panel). A c-kit-positive cell (arrow) is SMA-negative.

Download figure to PowerPoint

These cells were considered as PICCs, and were further immunolabeled for CD34, desmin, GFAP, neurofilaments, α-SMA, and S100 protein (Figs. 5–8). As it resulted from immune labeling of the adjacent sections (labeling on successive slides), the c-kit positive PICCS were:

  • CD34 negative (Figs. 5–7);

  • desmin negative (Figs. 5 and 7);

  • GFAP negative;

  • neurofilaments negative;

  • α-SMA negative (Fig. 8);

  • S100 protein negative (Fig. 7).

These PICCs appeared as multipolar, either polygonal or rounded/oval-shaped cells, with numerous processes (Figs. 2–8), often moniliform (Figs. 3, 4, 6, 7).

CD34 immunolabeling found portal IA and OA to contain immunopositive cells scaffolding two well- represented mesenchymal layers within those adventitial sublayers.

Intrahepatic PICCs

We have also identified CD117/c-kit–positive cells within the hepatic tissue, distinctively located at the level of the portal spaces and septa (Figs. 9–14):

  • 1
    at the level of the lobular limiting plate, on its portal side (Fig. 9);
  • 2
    within the portal spaces/septa (Figs. 9 and 11), there were also ICCs concentrated around the bile ducts at that level (Fig. 10) but we found no c-kit positive cells neighboring terminal veins (Fig. 11);
  • 3
    in the periphery of the hepatic lobules (Fig. 9).
thumbnail image

Figure 9. Hepatic portal space with CD117/c-kit positive cells on the portal side of the limiting plate (arrowheads), within the portal space (thin arrows) and in the periphery of the lobules (thick arrowheads).

Download figure to PowerPoint

thumbnail image

Figure 10. Peribiliary c-kit positive cells (arrows) of the portal space. One of these is magnified (*, inset). BD, bile duct; PVb, portal vein branch.

Download figure to PowerPoint

thumbnail image

Figure 11. c-kit positive cells (arrows) are identified within the portal septa but not close to a terminal vein (TV).

Download figure to PowerPoint

thumbnail image

Figure 12. c-kit positive ICLCs on the portal side of the limiting plate. Scalebar corresponds to all panels.

Download figure to PowerPoint

thumbnail image

Figure 13. c-kit positive ICLCs of the portal spaces. Scalebars: 20 μm.

Download figure to PowerPoint

thumbnail image

Figure 14. CD34-positive cells (arrows) on the portal side of the limiting plate; on successive sections, these are different to the c-kit positive cells located on the portal side of the limiting plate.

Download figure to PowerPoint

Morphologically, the intrahepatic PICCs were single or grouped multipolar cells, having processes with even or uneven calibers (moniliform processes) (Figs. 10, 12, 13). The morphological pattern was similar for the PICCs of the limiting plate (Fig. 12) and those of the portal spaces and septa (Fig. 13). The PICCs were equally distributed near arteries, PV branches, nerves, and bile ducts within the portal spaces. We have also identified a uniform circumferential distribution of the PICCS around the bile ducts of the portal spaces (Fig. 10).

CD34 immunolabeling (Fig. 14) identified the positive cells mainly located on the portal side of the limiting plate. However, on successive sections, immune labeling for CD34 and CD117 was not concordant and CD34-positive cells were not identified within the periphery of the hepatic lobules.

Finally, the intrahepatic PICCs had a similar immune labeling pattern as the extrahepatic ones—negative for CD34, S100 protein, desmin, GFAP (Fig. 15), and neurofilaments.

thumbnail image

Figure 15. No positive immune labeling for GFAP is evident at the level of the portal spaces. PVb, portal vein branch; BD, bile duct; a, artery.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. LITERATURE CITED

The wall of the portal vein in rabbit (RPV) is structurally comparable to human PV (Harhun et al., 2005). The RPV is a spontaneously active blood vessel (Bolton et al., 2004) and the distribution of the PICCs in the wall of the RPV resembles the distribution of ICCs in the gastrointestinal (GI) tract (Harhun et al., 2005).

Therefore, for the first time, we bring a proof of the presence of the portal ICCs in humans and the first evidence for the intrahepatic ICCs located within the portal spaces and septa, and in the periphery of the hepatic lobules. These PICCs have the morphological feature that allows their definition as portal telocytes, the moniliform telopodes. Therefore, our results bridge the interspecies gap, further allowing a better extrapolation of the experimental results in humans.

Giemsa stains previously identified ICLCs in the mammary gland stroma (Radu et al., 2005) and the fallopian tube (Popescu et al., 2005a). According to our findings, this stain is also reliable in identifying the PICCs/portal telocytes, before any immunohistochemical labeling.

Most ICCs belong to a spectrum, ranging from cell types very similar to SMCs (“myoid cells”) to cell types with ultrastructural features, hardly distinguishable from classic descriptions of fibroblasts (fibroblast-like or fibroid cells) (Rumessen and Vanderwinden, 2003). Regarding the PICCs we have identified, both the extrahepatic and the intrahepatic PICCs were αSMA-negative, the myoid phenotype of these PICCs being excluded.

It must be taken into account that c-kit may be also considered a marker of primitive, pluripotent cell types, as it was previously documented that CD34-positive stem cells in bone marrow express c-kit (Hibbert et al., 2004). CD34 monoclonal antibodies are known to label endothelial cells, mesenchymal cells and stroma fibrocytes functioning as matrix-producing cells (Leong et al., 2003; Pusztaszeri et al., 2006). However, as the c-kit cells we have identified were CD34-negative on adjacent sections, we excluded this possibility.

The concept of a diffuse stellate cell system in mammals was suggested (Zhao and Burt, 2007). However, hepatic stellate cells (HSCs) are mesenchymal cells, usually GFAP- and/or desmin-positive (Yokoi et al., 1984; Gard et al., 1985; Buniatian et al., 1996; Zhao and Burt, 2007) and were not proven to be c-kit-positive. Therefore, we have a positive differential diagnosis between our PICCs, desmin-, and GFAP-negative, and the HSCs. Moreover, the presence of long and “moniliform” cytoplasmic processes makes a clear difference between ICCs/ICLCs and resting stellate cells (Popescu et al., 2005b).

Dendritic cells (DCs) are known as the most potent professional antigen-presenting cells (Gulubova et al., 2008). DCs may be labeled nonspecifically with c-kit (Chen et al., 2007) but are specifically labeled with S-100 protein (Gulubova et al., 2008). Regarding the PICCs we evaluated, there were c-kit-positive and S-100-negative, so the diagnostic of DCs was excluded.

In our study, extrahepatic PICCs were identified close to the circular and longitudinal portal muscle bundles. These vicinities could qualify the extrahepatic PICCs as possible generators of spontaneous rhythmical local waves, as it was demonstrated for the RPV (Harhun et al., 2005), but this is only speculative at this time and further studies need to be conducted to determine the exact role(s) of the extrahepatic PICCs in humans.

PICCs appeared mostly, but not exclusively, as constituents of the portal stroma, within the liver, at which level the portal venous branches lack muscular layers. As hepatic PICCs were identified on both sides of the lobular limiting plate, a role at this interface may be presumed, and this led us to strongly suppose that these hepatic PICCs can be involved in the mediation of the neural transmission, before portal pacemaking. Nevertheless, a pacemaking function of the hepatic PICCs cannot be ignored as we identified such cells near arteries, nerves, and biliary tracts. Therefore, the intrahepatic PICCs could equally qualify as arterial ICLCs, perineural, or biliary ICLCs, as it was proven in various studies (Daniel, 2001; Lavoie et al., 2007; Ahmadi et al., 2010; Pucovsky, 2010).

The PICCs we identified are ICLCs, as they are ICCs located outside the gastrointestinal tract. Moreover, as these cells have moniliform processes, they could be labeled as telocytes, the only morphological difference between ICLCs and telocytes being the presence of moniliform telopodes (Faussone-Pellegrini and Bani, 2010; Popescu and Faussone-Pellegrini, 2010).

The switch from an ICC to a SMC phenotype was considered an extremely important phenomenon that might be used for therapeutic purposes when ICC number is decreased in human GI motility disorders (Sanders et al., 2000); such phenotypic switch must be evaluated on larger samples of human PV to evaluate whether or not therapeutic approaches may be considered in PV dysfunctions.

A promising direction of research may be if and how these PICCs are involved in the idiopathic portal hypertension (Okudaira et al., 2002; Kitao et al., 2009), when it is possible that a portal venous blood insufficiency to be responsible for hepatic parenchymal damage (Tsuneyama et al., 2002); also, the PICCs should be taken into consideration when the hemodynamic changes associated with hepatic steatosis are evaluated in correlation with the products released by Kupffer cells and sinusoidal endothelial cells (Farrell et al., 2008).

Electron microscopy needs to be further performed to sustain, beyond the pacemaking potential of the PICCs, the definition of the PV as “portal organ.”

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. LITERATURE CITED