Endothelial tip cells in vitro are less glycolytic and have a more flexible response to metabolic stress than non-tip cells.

Formation of new blood vessels by differentiated endothelial tip cells, stalk cells, and phalanx cells during angiogenesis is an energy-demanding process. How these specialized endothelial cell phenotypes generate their energy, and whether there are differences between these phenotypes, is unknown. This may be key to understand their functions, as (1) metabolic pathways are essentially involved in the regulation of angiogenesis, and (2) a metabolic switch has been associated with angiogenic endothelial cell differentiation. With the use of Seahorse flux analyses, we studied metabolic pathways in tip cell and non-tip cell human umbilical vein endothelial cell populations. Our study shows that both tip cells and non-tip cells use glycolysis as well as mitochondrial respiration for energy production. However, glycolysis is significantly lower in tip cells than in non-tip cells. Additionally, tip cells have a higher capacity to respond to metabolic stress. Finally, in non-tip cells, blocking of mitochondrial respiration inhibits endothelial cell proliferation. In conclusion, our data demonstrate that tip cells are less glycolytic than non-tip cells and that both endothelial cell phenotypes can adapt their metabolism depending on microenvironmental circumstances. Our results suggest that a balanced involvement of metabolic pathways is necessary for both endothelial cell phenotypes for proper functioning during angiogenesis.

Rapid combined light and electron microscopy on large frozen biological samples.

The use of large unfixed frozen tissue samples (10 x 10 x 5 mm(3)) for combined light microscopy (LM) and electron microscopy (EM) is described. First, cryostat sections are applied for various LM histochemical approaches including in situ hybridization, immunohistochemistry and metabolic mapping (enzyme histochemistry). When EM inspection is needed, the tissue blocks that were used for cryostat sectioning and are stored at -80 degrees C, are then fixed at 4 degrees C with glutaraldehyde/paraformaldehyde and prepared for EM according to standard procedures. Ultrastructurally, most morphological aspects of normal and pathological tissue are retained whereas cryostat sectioning at -25 degrees C does not have serious damaging effects on the ultrastructure. This approach allows simple and rapid combined LM and EM of relatively large tissue specimens with acceptable ultrastructure. Its use is demonstrated with the elucidation of transdifferentiated mouse stromal elements in human pancreatic adenocarcinoma explants grown subcutaneously in nude mice. Combined LM and EM analysis revealed that these elements resemble cartilage showing enchondral mineralization and aberrant muscle fibres with characteristics of skeletal muscle cells.

Calvarial osteoclasts express a higher level of tartrate-resistant acid phosphatase than long bone osteoclasts and activation does not depend on cathepsin K or L activity.

Bone resorption by osteoclasts depends on the activity of various proteolytic enzymes, in particular those belonging to the group of cysteine proteinases. Next to these enzymes, tartrate-resistant acid phosphatase (TRAP) is considered to participate in this process. TRAP is synthesized as an inactive proenzyme, and in vitro studies have shown its activation by cysteine proteinases. In the present study, the possible involvement of the latter enzyme class in the in vivo modulation of TRAP was investigated using mice deficient for cathepsin K and/or L and in bones that express a high (long bone) or low (calvaria) level of cysteine proteinase activity. The results demonstrated, in mice lacking cathepsin K but not in those deficient for cathepsin L, significantly higher levels of TRAP activity in long bone. This higher activity was due to a higher number of osteoclasts. Next, we found considerable differences in TRAP activity between calvarial and long bones. Calvarial bones contained a 25-fold higher level of activity than long bones. This difference was seen in all mice, irrespective of genotype. Osteoclasts isolated from the two types of bone revealed that calvarial osteoclasts expressed higher enzyme activity as well as a higher level of mRNA for the enzyme. Analysis of TRAP-deficient mice revealed higher levels of nondigested bone matrix components in and around calvarial osteoclasts than in long bone osteoclasts. Finally, inhibition of cysteine proteinase activity by specific inhibitors resulted in increased TRAP activity. Our data suggest that neither cathepsin K nor L is essential in activating TRAP. The findings also point to functional differences between osteoclasts from different bone sites in terms of participation of TRAP in degradation of bone matrix. We propose that the higher level of TRAP activity in calvarial osteoclasts compared to that in long bone cells may partially compensate for the lower cysteine proteinase activity found in calvarial osteoclasts and TRAP may contribute to the degradation of noncollagenous proteins during the digestion of this type of bone.

Endogenous expression and endocytosis of tartrate-resistant acid phosphatase (TRACP) by osteoblast-like cells.

Tartrate-resistant acid phosphatase (TRACP) is produced by macrophages and other cells of the monohistiocytic lineage. In particular, osteoclasts are characterized for a high expression of this enzyme. Yet, several data suggest that other bone cell types, such as osteocytes and osteoblasts, may also express activity of this enzyme. This is particularly obvious at sites were osteoclasts resorb bone, suggesting that osteoclasts (or their precursors) somehow induce TRACP activity in osteoblasts. In the present study, we investigated this by culturing human osteoblast-like cells with and without conditioned medium (MCM) from human blood monocytes (as a source of osteoclast precursors). High levels of TRACP activity were found in osteoblast-like cells cultured with MCM. Depletion of TRACP from this medium resulted in the absence of its activity in osteoblast-like cells, thus suggesting that the TRACP activity in these cells was the result of endocytosed TRACP that was released by the monocytes in the MCM. Osteoblast-like cells cultured in control (non-conditioned) medium contained very low levels of TRACP-like activity. However, the cells expressed TRACP mRNA and incubation of extracts of these cells with active cathepsin B did induce activity of a TRACP-like enzyme. Inhibition of the activity of cysteine proteinases in general and of cathepsin B in particular, completely blocked TRACP activity of the osteoblast-like cells. This TRACP-like enzyme but not the alleged endocytosed fraction of TRACP was inhibited by fluoride, suggesting that the fractions may be different isoenzymes. Our data seem to indicate that osteoblast-like cells may contain two different fractions of TRACP, one that is released by monocytes and subsequently endocytosed by osteoblast-like cells and a second endogenous fraction that is present in an inactive proform. We hypothesize that the capacity of osteoblast-like cells to endocytose TRACP is important for the removal of this enzyme during or following the bone resorptive activity of the osteoclast.