Primary culture of chondrocytes after collagenase IA or II treatment of articular cartilage from elderly patients undergoing arthroplasty

Results presented in this report support our hypothesis that collagenase IA is more effective than collagenase II to isolate chondrocytes from articular cartilage of synovial joints from elderly patients who had undergone total joint replacement surgery.

Those chondrocytes isolated using collagenase IA digestion were highly viable and could be cultured in vitro for 2 passages without changes in chondrocyte morphology. Using sequential digestion with pronase E and collagenase IA, we could isolate approximately 2.5 million chondrocytes from 1 g of wet cartilage. The average viability of those isolated chondrocytes was around 84%, which was higher than that reported by Jakob et al. after a study using various protocols based on combinations of collagenase II, trypsin, hyaluronidase, and tosyl-lysyl chloromethane [13]. To our knowledge, this is the first report of a study to examine directly the difference in human chondrocyte isolation efficiency between widely used collagenase II and rarely used collagenase IA.

As a pathogenic bacterium, Clostridium histolyticum can cause myonecrosis by producing collagenase, a zinc-containing metalloproteinase, to dissolve various types of collagen. Compared with collagenases produced by vertebrates, collagenase from C. histolyticum has a broader substrate specificity and more potent collagenolytic activity [17]. Collagenase IA or collagenase type I, is composed of several collagenase proteins, some of which are C-truncated forms, all of which are encoded by colG. The product of this gene is mature collagenase IA with a molecular weight around 114 kDa. By contrast, collagenase II is encoded by colH and is composed of a mature protein with a molecular weight of 112 kDa and one C-truncated form [17, 20].

Our data indicate that collagenase IA is more suitable than collagenase II for isolating chondrocytes from human cartilage. Bond and Wart found that collagenase IA contained predominantly β-collagenase, while collagenase II only contained a trace. β-Collagenase is highly active against collagen [18]. To maximize the effect of collagenase IA, cartilage needs to be pre-treated with pronase E, mainly to digest proteoglycans, which can expose sites on collagen for collagenase IA to bind. Pronase E is a mixture of 10 proteases that are produced by Streptomyces griseus K-1. Narahashi and Yanagita showed that pronase E was much more effective in digesting casein than trypsin because the former contained serine proteases and peptidases [21, 22].

Comparing chondrocyte yield and examining cell viability, we identified the sequential digestion of cartilage with pronase E and collagenase IA as the most effective way to isolate chondrocytes from human cartilage. The sequential digestion method involving pronase and collagenase has also been used by Diaz-Romero et al. to isolate chondrocytes from human cartilage of femoral condyles. Our findings were consistent with those of these authors, who reported that the method produced a better chondrocyte yield than collagenase alone [23]. However, Diaz-Romero et al. did not specify the type of pronases they used in their study. Collagenase P was used for the second step of enzymatic dissociation. This type of collagenase acts more like collagenase II than I. Because the authors did not provide the actual chondrocyte yields from their study and use cartilage specimens from donors with a wide age range (32–89 years old), direct comparison of chondrocyte yields between their method and ours are difficult.

In published studies on human chondrocyte isolation, protocol specificities, including the concentrations of proteinases, incubation times, and the order in which different proteinases were applied, vary widely. In a study using autologous chondrocytes for articular cartilage restoration in patients with age ranging from 14 to 48 years, cartilage obtained from a minor load-bearing joint was digested with 0.1% collagenase (w/v) for 16 h. Around 900 chondrocytes were isolated from 1 mg of cartilage [15]. Although the authors did not indicate the type of collagenase used in their study, their chondrocyte yield was less than half of what we observed in Group I in our study despite that they used cartilage specimens from much younger patients and 5 times greater concentration of collagenase.

Yonenaga et al. collected cartilage specimens from 3 elderly patients diagnosed with osteoarthritis and tested a series of collagenase concentration and incubation times to identify optimal conditions for articular chondrocyte isolation. They reported that 0.6% collagenase with 24 h incubation were optimal conditions, yielding around 1000 viable chondrocytes per mg of wet cartilage [24]. However, the authors did not specify which type of collagenase they used in their study or the exact anatomical locations from which they harvested cartilage specimens. Possibly because of the higher collagenase concentration and longer incubation time used, their viable chondrocyte yield was lower than the yield that we observed in Group I. Moreover, our results were derived from 11 cartilage specimens obtained from elderly patients of both sexes, while their study used pooled cartilage specimens obtained from only 3 elderly female patients.

We acknowledge that cartilage in 3 enzymatic treatment groups—is not from the same patient—and this may complicate the interpretation of chondrocyte yield data due to individual variation. The reason why we did not use the same batch of cartilage to compare chondrocyte yield among 3 enzymatic methods was that we could only harvest around 3.8 g of wet cartilage from 1 human femoral head and when we divided this amount of cartilage into 3 enzymatic treatment groups, the total number of chondrocytes isolated in each group was too low to be counted accurately. This may be the result of the uniqueness of cartilage tissue, including its relatively high ECM to cell volume ratio, and that only 10% of tissue volume is occupied by chondrocytes [25].

Cartilage samples from more than 1 donor patient could not be pooled because we could only receive 1 femoral head or tibial plateau specimen on a certain day, which was determined by patient enrolment and surgery schedule. We could not refrigerate previously received specimens because refrigeration would cause time-dependent chondrocyte death in cartilage [26] and we could not predict when the next specimens would be collected. Moreover, according to the report by Stockwell, during aging (30–89 years), cell density in human articular cartilage remained constant (at about 10 cells per 0.22 mm²) [27]. This implies that the number of chondrocytes in 1.0 mg wet weight of cartilage (chondrocyte yield) should be independent of age and sex of patients who provided femoral head or tibial plateau cartilage for our study. We observed that chondrocyte yield in Group I was significantly higher than that in Group II or III. This indicates that collagenase IA is more effective than collagenase II in releasing chondrocytes from articular cartilage.

Here, we developed an effective protocol for chondrocyte isolation from the articular cartilage of elderly patients. Chondrocyte yield and viability were superior to that described for other studies using human cartilage to isolate chondrocytes. Because total hip or knee replacement surgery provides the vast majority of human cartilage specimens in clinics, our chondrocyte isolation protocol may facilitate studies of the pathogenesis of osteoarthritis and cartilage tissue engineering.