• interstitial Cajal-like cells;
  • human atrial myocardium;
  • CD 117/c-kit;
  • CD34;
  • vimentin;
  • stromal synapse;
  • intercellular signaling;
  • EGF-receptor;
  • CD13;
  • desmin;
  • electron microscopy


We have previously reported (Hinescu & Popescu, 2005) the existence of interstitial Cajal-like cells (ICLC), by transmission electron microscopy, in human atrial myocardium. In the present study, ICLC were identified with non-conventional light microscopy (NCLM) on semi-thin sections stained with toluidine blue and immunohistochemistry (IHC) for CD117/c-kit, CD34, vimentin and other additional antigens for differential diagnosis. Quantitatively, on semi-thin sections, ICLC represent about 1–1.5% of the atrial myocardial volume (vs.±45% working myocytes, ˜2% endothelial cells, 3–4% for other interstitial cells, and the remaining percentage: extracellular matrix). Roughly, there is one ICLC for 8–10 working atrial myocytes in the intercellular space, beneath the epicardium, with a characteristic (pyriform, spindle or triangular) shape. These ICLC usually have 2–3 definitory processes, emerging from cell body, which usually embrace atrial myocytes (260 nm average distance plasmalemma/sarcolemma) or establish close contact with nerve fibers or capillaries (˜420 nm average distance to endothelial cells). Cell prolongations are characteristic: very thin (mean thickness = 0.150±0.1 μm), very long for a non-nervous cell (several tens of μm) and moniliform (uneven caliber). Stromal synapses between ICLC and other interstitial cells (macrophages) were found (e.g. in a multicontact type synapse, the average synaptic cleft was ˜65 nm). Naturally, the usual cell organelles (mitochondria, smooth and rough endoplasmic reticulum, intermediate filaments) are relatively well developed. Caveolae were also visible on cell prolongations. No thick filaments were detected. IHC showed that ICLC were slightly and inconsistently positive for CD117/c-kit, variously co-expressed CD34 and EGF receptor, but appeared strongly positive for vimentin, along their prolongations. Some ICLC seemed positive for α-smooth muscle actin and tau protein, but were negative for nestin, desmin, CD13 and S-100.

In conclusion, we provide further evidence of the existence of ICLC in human atrial myocardium, supporting the possible ICLC role in pacemaking, secretion (juxta- and/or paracrine), intercellular signaling (neurons and myocytes). For pathology, ICLC might as well be ‘players’ in arrhythmogenesis and atrial remodeling.