After digestion the cells were washed, resuspended and viable cells determined.

The cells were then used immediately in primary high-density cultures or plated out for secondary cultures.

We therefore adopted the methylene blue assay to determine tendon-derived cell numbers. Although this assay has the disadvantage of detecting both live and dead cells it is thoroughly reliable and accurate. Using this assay, it was found that treating primary cells with indomethacin lead to a dose related inhibition of cell proliferation. However, rather unexpectedly the relationship was negative with the greatest inhibition being seen at 10 nM and essentially no effect at 100 μM (Figure (Figure3).3). In secondary subcultures of tendon-derived cells, the cells became relatively refractive to treatment with indomethacin showing no significant effect at any concentration (Figure (Figure4)4) and then by the third subculture a biphasic stimulation of proliferation were seen with a maximum at 1–10 μM (data not shown).

Because of the somewhat unexpected inhibition of proliferation with low concentrations of Indomethacin, this work was repeated using direct counts of cell number and viability (as measured by 7-amino-actinomycin D uptake). Although the concentration relationship was not as linear as with the methylene blue assay, this too showed a significant decrease in cell number at lower concentrations of indomethacin whereas treatment with high concentrations had essentially no effect (Figure (Figure5).5). Also the results with secondary sub culture of tenocytes were similar to methylene blue with no effect of indomethacin on proliferation. We did not repeat this with tertiary subculture of tenocytes.

Tenocyte culture provides a uniform, controlled environment in which to study the in vitro effects of Indomethacin, with limitation of other uncontrolled variables and avoidance of the confounding influences that are present in animal models [18]. The results of our in vitro experiments showed that Indomethacin had a dose related effect on tenocyte proliferation. Indomethacin reduced the number of human tenocyte cells at a dose between 0.1 nM to 100 nM. However at higher doses it did not inhibit tenocyte proliferation. This is in contrast to some studies showing inhibition of tenocytes with therapeutic doses of Indomethacin and hence producing a negative effect on healing tendons [6,8]. Other studies have shown beneficial effect on tenocyte healing by increasing the tensile strength to failure [5,7,17]. This is due to increased maturation of collagen fibrils during healing. However it is still debatable whether this beneficial effect is due to proliferation of tenocytes or increased production of collagen and maturation. Our study indicating that Indomethacin in therapeutic doses did not inhibit tenocyte proliferation has clinical relevance. This would be beneficial in tendon healing and repair as tenocyte is the major cell type in tendons and is responsible for the production and maintenance of extracellular matrix.

The negative dose response effect of Indomethacin on tendon derived cell proliferation remains a conundrum and we have no explanation to this. It should however be noted that serum levels of Indomethacin generally reach a peak of 1–10 μg/ml after administration and due to the poor blood supply in tendons, it is likely that tenocytes are exposed to concentrations even lower than this. These concentrations are well within the range at which the above effects were seen in vitro but did not go any way towards explaining the lack of effect at higher concentrations.

Dissimilar culture conditions as in vivo or in vitro, differing species and even changes in cell culture conditions of the same species have shown different tenocyte behaviour [5-7,17-20]. It can thus be stated that interaction between tenocytes and NSAIDS is influenced by different factors, leading to opposing results. It has been previously argued that sub culture of mesenchymal cells do not reciprocate in vivo results [11]. However this effect has not been studied or documented before, with tenocytes. We considered this as an important area of investigation and looked at role of culture conditions on effect of Indomethacin on tenocytes. We noted that sub culture of cells did not show similar results as primary tenocytes. The dose dependent effect of Indomethacin was only seen on primary tenocytes and sub culture of cells did not show this effect. Previous studies have shown Indomethacin to cause both inhibition and proliferation of tenocytes in in vitro studies [6,10,14,17]. However these studies do not clearly indicate if only primary or subcultures of cells were used. Also it has not been mentioned previously if such effect was seen. We postulate that this effect seen in our study may be due to the selection of a highly proliferative population of tenocytes that is present in the primary digest only at low levels which subsequently overgrow the slower growing cells in later passages. We consider this an important finding, as studies, which have used sub culture of cells, need to be reviewed with some caution.

Various cell-counting methods are used to measure tenocyte proliferation. Initial experiments studying the effects of Indomethacin on tendon derived cell proliferation were carried out using the commonly used Alamar blue assay. However, although on visual examination an effect was evident, no effect was seen after analysis of the cell supernatants. This conundrum was subsequently explained by the finding that tendons appear to contain 2 subpopulations of cells; one subpopulation with apparently normal metabolic activity and a second subpopulation of cells with low levels of mitochondrial enzymes and subsequently a low oxidative metabolism. As the Alamar blue assay is dependent on mitochondrial enzyme activity for the reduction of the dye resazurin to resorufin, in cells without these enzymes, no effect was seen. This was confirmed by the use of MTT which is metabolized by the same enzymes to produce a purple product and it was clearly demonstrated that 2 subpopulations of cells exist; one staining intensely indicating a high metabolic activity and one stained relatively mildly indicating a low metabolic activity. This is entirely consistent current knowledge regarding the metabolic activity of tenocytes and tenoblasts. Tenoblasts are known to contain relatively high numbers of mitochondria and tenocytes very few suggesting comparatively high and low oxidative metabolism respectively [21]. This would in turn that the response seen in primary cells is due to the presence of mature tenocytes in the cultures which would have been lost on subsequent passage. As tenocytes are thought to be non-proliferative it is obviously unlikely that this is due to an inhibition of proliferation. One possibility is a toxic effect on specific to tenocytes but not tenoblasts however, we saw no evidence of excessive cell death in these cultures. Another possibility is that the tenocytes play some form of regulatory role controlling the proliferation of the proliferative tenoblasts as is thought to be the case for osteocytes and osteoblasts [22]. This must however remain speculative until further evidence is available.

One consequence of these data is that investigations using Alamar blue or MTT to assess tenocytes numbers should be interpreted with some caution. We therefore adopted the methylene blue assay to determine tendon-derived cell numbers. Although this assay has the disadvantage of detecting both live and dead cells it is thoroughly reliable and accurate. Using this assay, it was found that treating primary cells with indomethacin led to a dose related inhibition of cell proliferation. In secondary subcultures of tendon-derived cells, the cells became relatively refractive to treatment with indomethacin showing no significant effect at any concentration (Figure (Figure4).4). Because of the somewhat unexpected inhibition of proliferation with low concentrations of Indomethacin, this work was repeated using direct counts of cell number and viability (as measured by 7-amino-actinomycin D uptake). Although the concentration relationship was not as good as with the methylene blue assay, this too showed a significant decrease in cell number at lower concentrations of indomethacin whereas treatment with high concentrations had essentially no effect (Figure (Figure55).

The cause of this rather unexpected dose/effect relationship is at present unclear. Indomethacin is thought to act principally by the modulation of arachidonic acid metabolism via the cyclooxygenase and lipoxygenase pathways thus blocking the production of prostaglandins and HETEs/leukotrienes respectively. The relationship is however not simple as indomethacin has been shown to stimulate [23] and inhibit [24] the production of lipoxygenase metabolites. This differential effect is likely to be both tissue and concentration dependent and is further complicated by the possibility that blocking one pathway is likely to shunt metabolites down the other. No drugs are 100% specific for any one particular mechanism of action and indomethacin is no exception giving rise to a number of effects unrelated to cyclooxygenase [25]. Lastly, endocannabinoids are now known to have a number of peripheral effects on connective tissue [26] and preliminary investigations in this laboratory indicate that this is also the case in tendons and ligaments. In relation to this, indomethacin has been shown to inhibit the enzyme fatty acid amide hydrolase (FAAH), which metabolizes endocannabinoids [27]; inhibition of FAAH would therefore result in the build of levels of endocannabinoids. The situation is therefore highly complex and is unlikely to be unraveled until further insights into the relative roles of prostaglandins, leukotrienes and endocannabinoids in tendons are obtained.