The energy-storing human Achilles tendon and equine superficial digital flexor tendon (SDFT) show no adaptation to exercise unlike muscle and bone, and are prone to injury. Injury involves microdamage accumulation until there is sufficient weakening for rupture to occur during normal athletic activity. Anatomically opposing positional tendons, such as the common digital extensor tendon (CDET) in the horse rarely suffer exercise–induced injury. Tenocytes maintain the extra-cellular matrix, but in energy-storing tendons they appear unable to adequately repair microdamage as it occurs. Tenocytes have been classified subjectively into 3 subtypes on the basis of histological nuclear morphology. Long, thin type 1 cells are thought to be less synthetically active than cigar-shaped type 2 cells, but their exact morphology and relative proportions in different tendon sites and ages has not been clearly defined. We hypothesised that tenocytes are separable into morphologically distinct subtypes, reflecting differences in age and functional requirements within and between specific tendons. Samples were taken from tensional and compressed regions of the SDFT and CDET of 5 neonates, 5 foals (1–6 m), 5 young adults (2–6 y) and 5 old horses (18–33 y) Cell nuclei were counted and measured in digital images from histological sections by computerised image analysis. Total tenocyte densities and proportions of the 3 subtypes were calculated for each age group, as were nuclear length:width ratios. Length:width ratio distributions for all horses were evaluated using a normality test followed by a paired t-test. There was a significantly higher total cellularity in the SDFT than the CDET, with a higher proportion of type 1 tenocytes in the CDET. With age, total cellularity decreased in all tendon sites and an increase in the proportion of type 1 tenocytes was observed in tensional regions. Foal and neonatal tendons contained significantly higher proportions of type 2 tenocytes than older tendons. The morphology of the two main subtypes in all age groups was significantly different; type 1 tenocytes had a higher nuclear length:width ratio (mean ± SD = 9.6 ± 2.5) than type 2 (mean ± SD =4.7 ±1.1) (p< 0.001). We were able to objectively separate tenocytes into 3 distinct subtypes based on nuclear length:width ratio measurements. There were significant differences in proportions of subtypes with tendon site and age. The positional tendon had significantly lower cellularity and a higher proportion of type 1 tenocytes; these cells may be less functionally active but sufficient to maintain the matrix in a tendon which is not subjected to high levels of strain. The SDFT continues to grow up to 2 years of age and is subjected to high strains, explaining the need for relatively higher proportions of type 2 cells. There is however an age-related increase in type 1 cells in both tendons which may explain an inability of the adult energy-storing tendon to adapt to exercise and to repair microdamage. Understanding the stimulus for age-related changes in tenocyte subtype proportions in tendons with different functions may help us understand the pathogenesis of exercise-induced tendon injury and to develop more appropriate training regimens.

Injury to the core region of energy-storing tendons is a frequent occurrence in both human and equine athletes, the incidence of which increases with age. Such energy-storing tendons include the human Achilles tendon (AT) and the equine superficial digital flexor tendon (SDFT). By definition, energy-storing tendons experience high strains during high-speed athletic activity. In contrast, anatomically opposing tendons (“positional” tendons), such as the common digital extensor tendon (CDET) in the horse and extensor digitorum longus tendon in man act only to transmit muscular force and rarely suffer exercise–induced injury. Functional adaptation of muscle and bone in response to exercise is well – documented, but there has been no convincing evidence to suggest that the energy-storing tendons in adults have the ability to adapt to exercise. We hypothesised that adaptive increases in tenocyte cellularity would occur in the energy-storing and positional tendons of young horses subjected to three specific exercise regimens. Samples were taken from midmeta-carpal regions of the SDFT (periphery and core) and CDET of young Thoroughbred horses from the following groups. Group 1: 6 horses exercised on a high-speed treadmill for 18 months from 21.3 months of age (SD 1.1) with 6 age-matched controls that underwent walking exercise only (long-term); Group 2: 6 horses exercised on a high-speed treadmill for 18 weeks from 19.4 months of age (SD 0.6) with 6 age-matched controls that underwent walking exercise only (short-term) and Group 3: 6 horses trained on pasture in New Zealand for 18 months beginning at 7–10 days of age, with 6 age-matched controls kept at pasture with no additional enforced exercise (Global Equine Research Alliance). Tenocyte nuclei were counted and measured in digital images from histological sections stained with haematoxylin and eosin, by computerised image analysis. Tenocyte densities (per mm2) for exercised and control groups for each study were evaluated using paired t-tests. Tenocyte density was significantly higher in the CDET of exercised horses in Group 3 (mean ± SD =260.4 ± 23.4) compared with the non – exercised controls (mean ± SD =226.9 ± 23.8) (p < 0.01). There was no such difference in the SDFT (core or periphery). There was also no significant exercise-related difference in tenocyte density in either the SDFT (core or periphery) or CDET for Groups 1 or 2. No previous data is available on the effect of exercise on tenocyte populations in equine tendons. The lack of other adaptive changes in previous studies of mature equine tendons had raised the question as to whether immature tendons would be more able to adapt to mechanical stimuli. In this study we were able to show that beginning training of horses shortly after birth (Group 3) stimulated an adaptive response by tenocytes in the positional CDET but not the SDFT. The inability of energy-storing tendons to show functional adaptation to exercise in immature or mature animals may explain the high incidence of strain-induced injury. Understanding the pathway by which exercise-related increases in tenocyte densities occur in immature positional but not energy-storing tendons may increase our understanding of the pathogenesis of strain-induced tendon injury.

Introduction: Tendons consist of longitudinally running parallel bundles of collagen with rows of tendon cells between them. The tendon cells are linked to one another via gap junctions1 and cytoskeletally associated adherens junctions. Recent in vitro studies indicate that the two types of gap junction present, made of connexin 32 and connexin 43, regulate tendon cell responses to mechanical load2. In view of the importance of cell-cell interactions in tendon cell behaviour, we describe the behaviour of tendon cells in a novel 3-dimensional culture system designed to allow cells to establish cell-cell contacts and deposit matrix in the absence of scaffolds, which would favour cell-substrate interactions, and without disturbance by medium changes.

Methods: Tendon fibroblasts were isolated from chicken tendons by protease and collagenase digestion3, grown to passage 3 in HEPES buffered DMEM/5% foetal calf serum/1% antibiotic/1% L-glutamine at 370C, in 5% CO2 in air. Cells were suspended at 3x107 cells/ ml and 1 ml placed in a Spectrapor DispoDialyzerTM tube (MW cutoff 300,000). This was then placed into a 60 ml centrifuge tube with 40 ml DMEM containing ascorbate (1mg/ml) on a roller. Medium in the large tube was changed every 2 days. At 24 hours, 7, 14 and 21 days the cell aggregates were fixed in 90% methanol (4°C), frozen on dry ice and cryosections cut at 10–15 μm. Sections were labelled by indirect immunofluorescence with monoclonal type I, II, and III collagen, actin, vimentin and decorin, connexins 32 and 43, vinculin, Pan cadherin and N-cadherin.

Results: Cells in suspension culture formed elongated aggregates up to 3cm long. Immunolabels showed that at 7 days type I and III collagens were present, predominantly in the periphery. At 14 days the collagens were uniformly distributed throughout the aggregates and showed parallel longitudinal organisation. It is also clear from propidium iodide label that the cell nuclei have distinct areas of alignment. The aggregates labelled positively for actin stress fibres, N-cadherin, decorin and connexin 32. Type II collagen and connexin 43 showed no conclusive label.

Discussion: The suspension cultures clearly show that tendon cells can form large scale, organised structures in this culture system, and are capable of assembling an organised extracellular matrix in the absence of a scaffold to support them. The structures formed were similar to tendons in their cell and matrix organization. The amount of collagen deposited by the cells increases over time. Therefore, tendon cells could be used in a tissue engineering context to form well organised tissue in the absence of scaffolds in suitable culture systems. The presence of connexin 32 gap junctions and absence of connexin 43 shows the cell aggregates favouring matrix synthesis pathways – in cell cultures connexin 32 junctions promote collagen sythesis whereas connexin 43 inhibits it2.