Experimental models of osteoarthritis: a review

Experimental

Models

of Osteoarthritis:

A Review

By Henry Troyer

E

XPERIMENTAL studies on joints have been carried out for many years. Several noteworthy studies were already published in the last quarter of the nineteenth century. The studies prior to 1960 have been reviewed by Gardner.’ Moskowitz’ in 1972 reviewed the types of experimental osteoarthritis models that had been used up to that time. It is my intention to critically review the experimental animal studies designed to contribute toward understanding osteoarthritis, concentrating primarily on those published since 1960. Both the individual studies and the general methods of study will be evaluated. IMMOBILIZATION

STUDIES

Since the work of Ely and Menso? in 1933, it is recognized that if a joint is held in an immobile position for a period of time, degenerative changes will occur in the articular cartilage. Many of the immobilization studies have been done because the cartilage lesions so produced resemble, at least superficially, those of osteoarthritis. In both the experimental immobilization arthropathies in animals and in osteoarthritis of humans there is a loss of glycosaminoglycans from the cartilage matrix. This loss can be readily demonstrated with histochemical methods, but the loss is generally less readily demonstrated with biochemical methods. In experimental immobilization anthropathies, and in osteoarthritis, the cartilage shows fibrillation, flaking, fissuring and overall thinning. The chondrocytes, however, do not react similarly in both cases. In osteoarthritis, the chondrocytes often proliferate in clusters of cells, and From the University of Missouri, Kansas City School of Medicine, Kansas City, Missouri. Henry Troyer, B.A., Ph.D.: Associate Professor of Medicine and Dentistry University of Missouri, Kansas City School of Medicine. Supported in part by a Faculty Research Grant from the Ojice of Research Administration, University of Missouri, Kansas City. Address reprint requests to Henry Troyer. Ph.D.. University of Missouri, Kansas City School of Medicine, 2411 Holmes Avenue, Kansas City, Missouri 64108. 0 I982 by Crune & Stratton, Inc. 0049--0172/82/I 103-0005$02.00~0

362

some of these cells remain active into the late stages of the disease. In immobilization arthropathies of adult animals, the chondrocytes, especially of the contact area, undergo degeneration and death within a few weeks or months. Osteoarthritis often develops over a period of time following joint injury and derangement of the joint mechanics, and more often the disease develops over a period of years without any obvious predisposing condition. Experimental immobilization arthropathies, in contrast, are of a more acute onset caused by drastic and sudden alteration of joint function. Degeneration seen in experimental immobilization arthropathies is probably caused most directly by impaired nutrition of the cartilage. Since cartilage is an avascular tissue, the chondrocytes have to depend on nutrients that diffuse through the matrix of the cartilage. Metabolic wastes must likewise diffuse out through the cartilage matrix. It is widely held, however, that simple diffusion is not a sufficiently rapid process to meet the requirements of the chondrocytes. It is thought, with some experimental and theoretical justification4 that cartilage acts somewhat like a sponge that gives off fluid when squeezed and takes up fluid when relaxed. During joint movement, cartilage is alternately squeezed and relaxed which undoubtedly aids the movement of fluids into and out of the cartilage. During experimental immobilization of a joint, the alternate squeezing and relaxing of cartilage is largely prevented and the nutrient supply and removal of metabolic waste products is impaired. It is, therefore, quite understandable that immobilization would have a profoundly adverse effect on the well-being of chondrocytes.

Immobilization

of Rat Knee Joints

Evans, et al.5 and Thaxter, et a1.6 immobilized the knee joints of adult rats in the flexed position with an internal splint (a metal bar) passing through the soft tissue. The splint was secured to the femur and tibia with steel pins. This procedure did not completely immobilize the knee joint; a small amount of sliding movement could still occur, but the angle of flexion was firmly fixed. Both groups of investigators found flat-

Seminars in Arthritis andRheumatism,

Vol. 11, No. 3 (February), 1982

tened deformation of the cartilage in the area of contact between the femur and tibia, but it did not progress into well developed degenerative lesions. The cartilage of the non-contact area underwent progressive degeneration, but evidently as a result of a soft tissue pannus spreading over the cartilage surface and causing erosion. Thaxter, et aL6 also studied ?S incorporation into the cartilage matrix and found it to be depressed in the contact area but normal throughout the rest of the cartilage. Hall’ performed immobilization experiments on adult rats similar to those of Evans, et al.’ except that immobilization was maintained for 6 mo as opposed to 4 mo in the Evans, et al. experiment. The results were similar to those of Evans, et al. except that in the contact area some of the cartilage eroded away because of the prolonged period of immobilization. Hall’ immobilized the knee joints of young rats (mean weight, 95 grams) in the same manner as described. Hall’s results differed from either Evans, et al.’ or Thaxter, et aL6 in that the cartilage of the contact area, rather than becoming flattened and deformed, hypertrophied within 27 days and remained so through 123 days (the end of the experiment). The cartilage of the non-contact areas showed cartilage degeneration as a result of synovial adhesions on the surface and vascular invasion from the underlying bone. In the young rats, these changes developed more slowly than those in the adult rats of Evans, et al.’ and Thaxter, et a1.6 Immobilization of Rabbit Knee Joints

The largest number of immobilization studies have been done in rabbits. These experiments have involved basically three types of immobilization: immobilization in flexion, immobilization in full extension and immobilization in extension plus an additional force applied across the joint with the aid of a clamp. Progressive degenerative lesions of the cartilage are always seen following immobilization, but they develop most rapidly in immobilized joints with additional extrinsic force across the joint. The progressive development of the lesions always follows the same pattern: loss of basophilia in the cartilage matrix, flaking of the cartilage surface, loss of cellularity, vertical fissuring, and fragmentation of the deeper layers of the cartilage. The first

and most severe signs of degeneration are seen in the area of the cartilage where the femur and tibia make contact. The lesions spread progressively to become larger and more severe. Many investigators repeated the immobilization experiments of others and used their individual expertise with different methods in studying the results. Trueta’ in 1956 was the first to study the effects of immobilization in extension with extrinsic force applied across the joint. He found only “degeneration” in the cartilage. Salter and Field” performed a similar but more extensive series of experiments. They described the gross pathological findings and interpreted the results in relation to osteoarthritis and to clinical immobilization of limbs. Trias” described similar lesions and added histological observations. Crelin and Southwick” immobilized rabbits’ knee joints in flexion, some of which retained a small amount of residual movement. They observed that in joints with residual movement some chondrocytes proliferated but in firmly fixed joints the chondrocytes died. This observation can be explained by the concept that the cartilage acts as a sponge; a small amount of movement would encourage the movement of nutrients and cell waste products. Matthiass and Glupe13 fixed rabbit knee joints in complete flexion (at a 1Y angle) which led to cartilage degeneration in the contact area. By 12 wk, all the cartilage of the contact area was lost and a pannus had formed over the remainder of the cartilage surface. They also noticed a marked atrophy of the subchondral bone. The changes that occur in articular cartilage as a result of immobilization were reconfirmed by Thompson and BassettI but added the observation that the subchondral bone hypertrophied during immobilization. This observation is in contrast to the findings of Matthiass and Glupe13 who found atrophy of the subchondral bone rather than hypertrophy. Thompson and Bassett were probably unaware of the earlier work and did not address themselves to the apparent contradiction. The difference in the findings may be explained in terms of the angle at which the knee joints were immobilized. In the Matthiass and Glupe study, the knees were immobilized in full flexion with little transarticular force. The subchondral bone, therefore, probably underwent disuse atrophy similar to the Thompson

364

and BassettI second (relief of contact) experiment. (See the section on Relief of Contact.) In the Thompson and Bassett immobilization experiment, the knees were in full extension which causes considerable trans-articular force from muscle pull which would stimulate hypertrophy. Troyer” produced evidence that chondrocytes in the knee joints, immobilized in full extension, underwent hypertrophy after only 1 or 2 days and they did so before there was any sign of cartilage degeneration in the matrix. Eronen, et aLI showed that after 10 days of immobilization, the cartilage showed signs of greatly increased ‘?S incorporation while the total content of glycosaminoglycans is decreased, suggesting that immobilization caused a greatly increased rate of glycosaminoglycan turnover. Videman, et al.” found that “S incorporation increased after only 4 days of immobilization. Even though the chondrocytes of the contact area eventually died, the high rate of 35S incorporation continued in the peripheral cartilage as well as in the tissues surrounding the joint such as in the bone, in the ligaments and even in the cartilage of the hip joints. Sood” showed that whatever the ill-effects of nonrigid plaster cast immobilization, the damage is not usually permanent if the joint is remobilized for a long enough period of time. This is contrary to the results of LangenskiBld, et a1.,19 who found that 7 wk or more of immobilization in rabbits regularly caused permanent contractures of the knee. Roy,*’ Finsterbush and Friedman2’ and Refior and Htibner2* added electron microscopic observations. They showed that chondrocytes may at first proliferate, then accumulate glycogen and lipids in the cytoplasm, and finally die and disintegrate. Ginsberg, et a1.23 proported to show that during immobilization, hydroxyproline (a measure of collagen) was more rapidly lost from the cartilage than hexosamine (a measure of glycosaminoglycans.) Their evidence remains unconvincing, however, since there was no consideration of the fact that as the superficial layer of cartilage flaked off, a spuriously high rate of collagen loss would have been expected. The results also disagree with those of other investigators.

HENRY TROYER

In immobilization studies, the two most significant variables are trans-articular force and incomplete immobilization (residual movement). Some investigators applied force across the joint with clamps or rubber band arrangements. Forces so applied can be measured to a certain extent, but that is not to say that if no compression mechanism is used, no trans-articular force exists. The muscles of the limb also pull and create a force. This force is not only difficult to measure, but it diminishes with time as the muscles atrophy with disuse. It should be remembered that the force of muscle pull is always in addition to any extrinsic force that is applied. At least 1 study has indicated that even a small amount of residual movement in an immobilized joint can make a significant difference in the well-being of articular chondrocytes.‘* These are variables that are difficult to control but investigators should be aware of them. Cartilage degeneration resulting from immobilization has no more than a superficial resemblence to osteoarthritis in humans. Experimental immobilization has greater relevance in terms of the consequences of limb immobilization with plaster casts following orthopaedic injuries, which was first suggested by Salter and Field.” Enneking and Horowitz24 have addressed themselves to this matter. They found that in human joints that had been immobilized for extended periods of time, the articular cartilage may become completely eroded and the joint may become filled with fibrous fatty tissue. Immobilization

of Dog Knee Joints

Palmoski, et al.*’ immobilized dog knee joints in 90” flexion for periods of time from 6 days to 8 wk. The cartilage did not lose its organizational integrity even after 8 wk of immobilization. The cartilage was analyzed with some conventional histological methods but they also assessed the chondrocytes’ ability to synthesize proteoglycans. Cartilage shavings from immobilized and control joints were incubated for up to 20 hr in a medium containing ‘SO,. The shavings were then analyzed for total and newly synthesized proteoglycans. They found a marked loss of ability to incorporate 35S04 into the cartilage matrix. Also, after 3 wk of immobilization the newly synthesized proteoglycans no longer

365

formed hyaluronate aggregates. They suggested that the aggregation defect was due to lack of weight bearing on the joint since the cartilage proteoglycans were able to form aggregates quite normally with exogenous hyaluronate.26 In normal cartilage, a high percentage of the proteoglycans form such aggregates.*’ Several weeks after remobilization, the aggregate problem was reversed and the cartilage appeared normal. EXTERNAL FORCE APPLIED TO MOVING JOINTS

Considering the high incidence of osteoarthritis following orthopaedic injuries to joints, it should not be surprising that any kind of mechanical derangement of the forces across a joint leads to degenerative change. Gritzka, et a1.28tested this hypothesis by devising a twospring rocker-arm method of loading the rabbits’ elbow joints so that additional external force was applied to the moving joints. This resulted in progressive rapid degeneration of the cartilage in the weight-bearing areas and proliferative responses were seen in the peripheral areas of the cartilage. Ogata, et a1.29also applied external force across the moving rabbit knee joint but applied it only to the medial side of the joint, creating a varus condition. This led to degenerative lesions in the medial femoral condyle and the medial tibia1 plateau, but not in the lateral compartment of the knee joint. The lesions were similar to those seen in immobilization experiments. These 2 studies help to confirm the suspicion gained from clinical observations that altered forces across joints evoke chondrocyte response and cartilage degeneration. Here, lack of chondrocyte nutrition cannot be implicated as it can in immobilization arthropathies. Probably an entirely different type of biological response is operative. RELIEF OF CONTACT

In all of the experimental arthropathies, the cartilage begins to undergo degeneration in the contact areas, suggesting that pressure, movement or friction are directly related to cartilage degeneration. However, Harrison, et a1.30found that in only 3% of the osteoarthritis cases found at autopsy, the degenerative lesions were

confined to the weight-bearing areas. Most of the degenerative lesions were confined to nonpressure-bearing areas. It is, therefore, suggested by some authors that osteoarthritis is caused or promoted by too little contact. Several investigators have studied the effects of experimental relief of contact on articular condyles in an attempt to test this hypothesis. Hal13’ resected the lateral femoral condyles of rats and allowed them to use the joints freely up to 1 yr. Degenerative changes took place in the lateral tibia1 plateau (unopposed side) but the medial femoral condyle and the medial tibia1 pleateau remained normal. The degeneration of the lateral tibia1 pleateau started with the superficial layer of cartilage and progressed until the cartilage had eroded down to the subchondral bone. Similar results were obtained by Thompson and BassettI with a similar type of operation on knees of skeletally immature rabbits (2.2-2.8 kg.) Engh and Chrisman3* repeated the experiment in mature (3-4 kg.) and immature (1.5-2 kg.) rabbits. Their findings were essentially the same as those of Thompson and Bassett.14 An appropriate sham control operation is an inherent problem in this type of experiment. The experimental operation involves a considerable amount of trauma, and it is impossible to design an operation with comparable trauma while leaving the condyle intact. Thus it is impossible to evaluate how much of the cartilage degeneration is due to tissue trauma, hemorrhage and tissue exudate, and how much is due to the absence of an opposing condyle. INSTABILITY OF JOINTS BY SURGICAL MANIPULATION

Investigators have used various types of surgical procedures to make animal knee joints mechanically unstable. Such instability leads to slow but progressive joint pathology which some believe simulates osteoarthritis in human knee joints years after sports injuries. These are, therefore, more natural types of osteoarthritis models than the previously discussed experimental arthropathies. Severing

of Multiple

Ligaments

Hulth, et al.33and Telhag and Lindberg34 were the first to produce and study this type of arthropathy. They excised the medial collateral

HENRY TROYER

ligament, both cruciate ligaments and the medial meniscus of adult rabbits’ knee joints. They began to see clusters of chondrocytes in the superficial layer of the articular cartilage within 15 days of the operation. Flaking and fissuring were seen in 3 mo. Signs of osteophyte formation were first seen at 3 mo and well-formed osteophytes at 5 mo. Ehrlich, et al.35 felt that a biochemical study of the collagen and glycosaminoglycan concentration of cartilage from the Hulth procedure would increase the validity of this osteoarthritic model. Some time after the operation, the rabbits were sacrificed and cartilage samples were removed for biochemical analysis. Unfortunately, it was not indicated whether the cartilage was taken from the entire condyle or just from the affected areas. Neither did the authors indicate to what depth the cartilage was removed; that is, whether all of the cartilage was removed down to or including the calcified cartilage layer. However, the biochemical data indicated only a 13% decrease in glycosaminoglycans after 6 mo over the normal. Their photomicrographs of the cartilage showed highly significant changes which were similar to what Hulth, et a1.33 and Telhag and Lindberg34 originally showed. Bohr36 studied the effects of the Hulth procedure on the vascularization in the subchondral bone. He found an increase in the blood supply to the subchondral bone which compares with what is sometimes seen in osteoarthritis. Lane, et aL3’ showed that the Hulth procedure causes changes in the material properties of the articular cartilage within 24 hr of the operation. The operation caused a 21% increase in the spectrum of retardation and an 11% decrease in the relaxed shear modulus. There was no change in the unrelaxed shear modulus. In terms of cartilage function, this suggests that in an operated knee, a particular load will cause greater deformation of the cartilage and the deformation will be caused more rapidly. No attempt has been made to relate the altered material properties to any chemical changes in the cartilage, and yet it would seem that the change in material properties must have an underlying chemical basis. The Hulth procedure for creating knee joint instabilities is a drastic modification of the joint.

Some investigators have found that it causes too great an instability and have adapted milder surgical procedures.

Severing of the Anterior Cruciate Ligament Degeneration of the joint will develop, even if only the anterior cruciate ligament is severed. This defect causes instability of the knee joint and clinically produces the anterior drawer sign. This model of osteoarthritis has been used to study osteophyte formation as well as cartilage degeneration. These studies have been done on dogs. The first studies have been concerned with studying osteophyte formation. Marshal13* and Marshall and Olson3’ performed an arthrotomy and cut the anterior cruciate ligament and then followed the development of osteophytes for 1 yr. They noted that as some of the knee joints regained stability, the clinical symptoms of osteoarthritis disappeared concurrently. Gilbertsona cut the anterior cruciate ligament in another study of osteophyte development, using a sharp instrument to make a stab into the knee without opening the joint. It was surprising that within 3 days there was evidence of the beginning of osteophyte formation. The first signs were the formation of a mineral deposit in the cartilage immediately outside the cortical bone. This calcified cartilage soon gave way to new bone with its own marrow spaces. Later, as the osteophyte progressed, the cortical bone became remodeled so that the marrow spaces became continuous with the rest of the marrow spaces of the bone. Pond and Nuki,4’ using the closed method of cutting the ligament, studied the degenerative changes of the articular cartilage with gross and histochemical observations. The chemical composition of such cartilage was studied by McDevitt and Muir4* and the results were compared with that of cartilage from a dog with natural osteoarthritis. Although the number of animals was small (3 experimental dogs and 1 with natural osteoarthritis), the results were similar. They found an increased ratio of chondroitin sulfate to keratan sulfate which is similar to the composition of cartilage from young animals. In another study, McDevitt et a1.43 showed that cartilage erosion and changes in cartilage

367

EXPERIMENTAL OSTEOARTHRlTlS

composition and appearance preceded loss of basophilic staining with Safranin 0. This is at variance with numerous other studies showing loss of basophilia with metachromatic dyes as the first visible manifestation of cartilage degeneration. This model of osteoarthritis has been used by Palmoski, et al.“” to study the suppression of proteoglycan synthesis by salicylates in the diseased cartilage. Although no gross osteoarthritic lesions were evident after 9 wk, the rate of proteoglycan synthesis was increased. Salicylates suppressed proteoglycan synthesis to a degree but that of the osteoarthritic cartilage was suppressed by a much greater degree. Xylose on the other hand stimulated 35S0,, incorporation and overcame the inhibitory effect of salicylates. The effects of salicylates and xylose is most likely on the synthesis of precursors of the glycosaminoglycan chain sugars.45 Meniscectomy

Follow-up studies46s47*48 have demonstrated that meniscectomy in human knee joints leads to a high incidence of osteoarthritis later. Krause, et a1.49studied the mechanical function of the human knee joint menisci from an engineering standpoint. They concluded that the menisci served an “important loadtransmitting and energy absorbing function” and that meniscectomy caused significantly increased stress across the knee joint, Moskowitz, et a1.50 used an osteoarthritis model where a portion of the medial meniscus was resected. Within 12 weeks, this operation resulted in the usual foci of degenerative cartilage, and a high percentage of the animals had various degrees of osteophyte formation on the tibia1 plateau margins. Cartilage from such experiments gradually lost its ability to incorporate 35S04 into the matrix.5’ 3H-thymidine and “C-glycine incorporation were initially increased but eventually returned to normal levels, but this response was a response to arthrotomy rather than to meniscectomy. Lutfi” showed that total meniscectomy in monkeys ultimately leads to severe degeneration of the articular cartilage. Only a limited area of the femoral and tibia1 condyles were affected and the rest of the cartilage remained normal.

The degeneration started within 3 wk with the death of the superficial chondrocytes, clustering of the deeper chondrocytes and loss of basophilia. Erosion of the cartilage had started by 12 wk, and the degeneration progressed until there was a total loss of cartilage by 36 wk. Floman, et a1.53and Shapiro and Glimcher54 studied the effects of experimental meniscectomy in rabbits. They either surgically removed the medial meniscus or simply detached it anteriorly (simulating a bucket handle tear). Their results can not be compared with the results of Moskowitz, et a1.50since the surgical procedure was different. In the articular cartilage, Floman, et a1.,53Shapiro, and Glimchers4 found hypercellularity with cell clusters, loss of basophilia, flaking, fissuring and formation of osteophytes. They waited as long as 16 mo to sacrifice the animals, but it is not clear how soon and in what sequence the various aspects of degeneration (cell clustering, osteophyte formation, etc.) appeared. Biochemically, there was a ‘I-fold increase in collagen synthesis, all of which was collagen type II, the type that is normally found in cartilage. They also found an increase in the molar ratio of galactosamine to glucosamine incorporation in the cartilage matrix, which reRected an increase in the synthesis of chondroitin sulfate over keratan sulfate. This aspect of their results was similar to the findings of McDevitt and Muir,42 although Floman, et a1.53 found that after 5 wk the ratio returned to normal.

MENISCAL REGENERATION It is recognized that the body frequently makes an attempt to regenerate a torn or resected meniscus.55g56 Thus Elmer, et a1.57 endeavored to study the effect that regenerated menisci might have in preventing cartilage degeneration following meniscectomy in rabbits. Four out of 5 rabbits in which the menisci had regenerated showed no degeneration of the cartilage, whereas in all 15 rabbits in which the menisci had not regenerated, some form of cartilage degeneration was evident. Cox and coworkers58*59also determined that experimentally produced meniscal tears in dogs lead to degenerative changes in the cartilage only if the tear

HENRY TROYER

interferes with normal joint function, aL6’ noted similar results in rabbits.

Dann.

et

PATELLECTOMY

DePalma and Flynn6’ showed that experimental patellectomy in dogs results in progressive degeneration of the knee joint cartilage and related those changes to the adverse clinical observations following patellectomy in humans. Garr, et a1.62therefore, developed an osteoarthritis model by removing the patella from the rabbit knee joints. After 3 wk, they found changes ranging from minor degeneration to gross ulceration of the femoral groove cartilage with “complete loss of metachromasia.” By 5-8 wk, they found evidence of remodeling (hypertrophy and spur formation) along the periphery of the femoral grooves, as well as the beginning of pannus formation. In some animals, the cartilage was completely eroded by 27 wk. It would appear that this is an authentic representation of osteoarthritis, but the results were more erratic than those of some of the other models. LIMB DENERVATION

STUDIES

Limb denervation studies have probably been the least consistent of any type of experimental joint arthropathy. Thaxter, et a1.6 found no gross evidence of any joint pathology when rat lower limbs were motor-denervated (cutting the ventral roots), whereas Akeson, et a1.63 found that complete denervation of the dogs lower limbs caused a marked decrease in chondroitin sulfate content and an increase in water content. Finsterbush and Friedman64 found that sensory denervation (cutting the dorsal roots) in rabbits caused atrophy and death of the chondrocytes starting in the middle zone of the articular cartilage, and a thickening of the calcified cartilage and subcondral bone. However, no gross ulceration of the cartilage was seen. The change in the chondrocytes was thought to be due to altered nutrition. Changes were also seen in the synovium, which presumably produced abnormal synovial fluid and in turn was responsible for altered nutrition of the chondrocytes. The changes in the cartilage of the complete denervation studies of Akeson, et a1.63 might have been explained in the same way except that they had earlier found similar results in the cartilage of polio victims (Eichelberger, et a1.65).

FORCED OVER-USE OF JOINTS

Dekel and Weissman used a specially-built machine that would simulate over-use (forced exercise) and overloading (repeated impact in extension) of the rabbit knee joints. Simple overuse of the joint did not cause any histological, biochemical or articular surface (studied with scanning electron microscopy) changes, but over-use plus over-loading led to typical osteoarthritis-like changes in the cartilage. Over-use plus over-loading caused an almost immediate fivefold rise in the E prostaglandins in synovial fluid and a reduction of cyclic AMP in the cartilage and subchondral bone. These biochemical findings are difficult to interpret. Both factors might reasonably be expected to act as mediators of the morphological changes seen in the bone and cartilage, but it does not seem reasonable that one should be elevated and the other depressed. Simon, et a1.67 found that repeated impact loading of guinea pig knee joints also causes cartilage degeneration, including loss of basophilia with Safranin 0, loss of cartilage thickness and marginal liping. They also determined that stiffness of the subchondral bone increases at first and then returns to normal as the cartilage continues to degenerate. ADMINISTRATION

OF PAPAIN

It should be possible to produce a model of the disease by artificially depleting the proteoglycan content, inasmuch as one of the earliest manifestations of osteoarthritis is depletion of proteoglycans. This type of model has been used a number of times using papain, a plant protease. In a classical experiment, Thomas68 showed that intravenous injection of crude papain cause rabbits ears to collapse, and that they recovered their normal shape in a few days. Bryant, et a1.69 showed evidence that the papain injection causes liberation of chondroitin sulfate from connective tissue. A series of experiments were then carried out in which the effects of intra-articular injection of crude papain was studied. This usually involved daily injections of a papain solution up to 6 days. Murray” studied the effects of papain on knee joints of young (1 kg.) rabbits. The most severe effect was on the deep strata of the cartilage. It resulted in a complete separation of the superfi-

cial portion of the cartilage, even though the superficial cartilage retained many viable chondrocytes. By 6 weeks, the cartilage had severely degenerated and largely disappeared. Farkas, et al.” also found that papain caused severe degeneration in the knee joints of young rabbits but that the effects were at least partially reversible. In a later study’* they found that the degeneration in older animals was more severe and not reversible. Histologically, the cartilage of older rabbits appeared to make only feeble attempts to repair itself. This observation apparently contradicts the observation of Thomas6’ who found that the ears of old rabbits were more resistant to collapsing by papain than the ears of younger rabbits. Moreover, Spicer and Bryant73 showed that an intravenous injection of papain readily caused depletion of basophilia in cartilage of young rabbits but not in cartilage of old rabbits. Bentley74 performed the papain experiment on rabbit hip joints and found that the enzyme initiated a sequence of degenerative changes which again closely mimicked osteoarthritis. The sequence culminated after 8 wk in a femoral head that was largely denuded of articular cartilage. Havdrup and Telhag” showed that the degeneration of cartilage resulting from papain injections was accompanied by mitosis of the chondrocytes. It is interesting that trypsin injected similarly caused only slight changes in articular cartilage but induced mitosis among the chondrocytes just as papain does.76 EXTRACTION OF DATA FROM TISSUES

A study cannot be any better than the reliability of the methods of handling and analyzing the joints and its tissues. The following comments are offered to strengthen that aspect of future studies. Sampling of Tissues

Sampling of tissues for physical, chemical, histological or other kinds of analysis is well nigh the single most critical process in evaluating the results of experimental procedures. Biochemists are frequently criticized for disregarding the morphological heterogeneity of complex tissues. In sampling articular cartilage, for instance, how large an area does one remove for analysis?

Should one remove the cartilage only from the weight-bearing area or from the whole condyle? How does one determine exactly what the limits of the weight-bearing area are? How deep should one scrape the cartilage? From electron microscopic studies it is evident that the metabolism of the superficial chondrocytes differs from that of the deeper chondrocytes. Histochemical studies show that alkaline phosphatase exists only in the deep chondrocytes immediately above the calcified cartilage while 5’-nucleotidase is found only in the superficial chondrocytes. Results may be readily distorted unless one carefully defines the sampling procedure and carries it out accurately. Processing of Tissues

Considerable misunderstanding and misinformation is evident in the literature concerning the processing of bone and cartilage tissue for histological and histochemical purposes. Weak and unreliable methods are used frequently while good and reliable methods are ignored. This is unfortunate, for it seems almost a moral obligation to extract as much information as possible from tissues after the time and energy has been invested in manipulative experiments on animals. Pearse,77*78an authority in histochemistry, is rarely cited by workers in experimental osteoarthritis. What follows is a critique on some of the methods that have been used and recommendations for procedures to follow in the future: Fixation. Tissues should be fixed with a formalin fixative at a pH close to neutral. Unbuffered formalin has a pH of 2 to 4. Low pH fixatives are not as effective in preserving cell structure as fixatives of a neutral pH. Rapid penetration of the fixative is also important. A considerable length of time is required for fixatives to penetrate to the center of large blocks of tissue, and significant autolytic changes may occur before fixation takes place. If optimal fixation is desired, the blocks of tissue should be no more than 4 or 5 mm in the smallest dimension. (The tissue blocks may have a larger face, but they should be thin.) Fixatives penetrate cortical bone with considerable difficulty. If long bones are to be processed, they should be sawed into slices with a fine-toothed saw such as a jeweler’s saw or a

370

dental saw. If a long bone is fixed whole, marrow elements, osteoblasts and osteoclasts will be poorly fixed. Decalcification. During the decalcification process, cartilage and other carbohydratecontaining tissues may lose some of their contents. Any low pH solution encourages leaching of carbohydrates. The most commonly used decalcifying agent in experimental osteoarthritis is formic-citrate, which has a pH of about 2.5. Other strong mineral acid decalcifiers may be used which have even lower pH’s. A 10% solution of ethylenediaminetetraacetic acid, phosphate buffered to pH 7.0, should be used as the decalcifying agent. Because it can be used at a neutral pH, it is superior to the mineral acid decalcifying agents in preserving tissue carbohydrates and cellular details. Formic-citrate reduces the staining reaction of cartilage with azure A and alcian blue. Parafin Embedding. Embedding tissues in paraffin remains the standard method of giving tissues support so that sections can be cut from the blocks. The first point to be made is the xylene, which is often used as the clearing agent, hardens bone and cartilage as well as other collagenous tissues so that they are hard to section. Chloroform does not harden collagenous tissues and may be used as a substitute clearing agent. The second point is that hot paraffin (or any form of heat) shrinks collagen. It is, therefore, important to keep the temperature of the molten paraffin only a few degrees above its melting point in order to minimize tissue shrinkage. Glycol Methacrylate Embedding. Glycol methacrylate (GMA) has become popular as a histological embedding medium in the last 15 yr. It is no more difficult to embed tissue in GMA than in paraffin and GMA has certain advantages over paraffin. GMA polymerizes (solidifies) at room temperature without significant shrinkage. Thus, tissue distortion caused by hot paraffin and shrinkage of paraffin is avoided when using GMA. GMA sections can be cut at a thickness of 1 or 2~ which can show greater cellular detail. The GMA sections may be stained with many of the same histological and histochemical stains that are used with paraffin sections, although a hematoxylin and eosin-like

HENRY TROYER

staining method has been developed especially for use with GMA sections.79 The GMA methodology is described in detail by Bennett, et al.*’ and Troyer.*’ Staining. Staining of cartilage tissue is usually done to evaluate either the proteoglycan content of the matrix or the general and cellular morphology. It is desirable to estimate the concentration of glycosaminoglycans in the cartilage sections. This is usually attempted by staining with metachromatic dyes which are a class of basic dyes. Nearly all basic dyes could be used and it is somewhat curious that metachromatic dyes have become so popular for this purpose. Perhaps part of the reason is that metachromatic dyes exaggerate the loss of glycosaminoglycans from the cartilage matrix. Metachromasia (meta-after, beyond; chromacolor) refers to the color change that some basic dyes undergo when they bind to tissue components. The intensity of color is frequently but erroneously referred to as metachromasia, and when a section of diseased cartilage stains with a lower intensity than a normal one, it is assumed (erroneously) to be a “loss of metachromasia.” Loss of affinity for basic dye molecules should be referred to as a “loss of basophilia.” Many investigators have believed that the color intensity of safranin 0 staining is proportional to the concentration of glycosaminoglycans. This belief is based on a study by Rosenberg** who showed that in solution, Safranin 0 reacted stoichiometrically with the ionic groups of chondroitin sulfate and keratan sulfate. Others frequently assume that the same stoichiometric relationship holds true in histological sections. A quantitative histochemical study83 showed that this assumption is ill-founded. Metachromatic dyes depend on more than an electrostatic attraction for their affinity to polyanions. There is probably some physico-chemical influence from neighboring dye molecules which mutually increase each other’s affinity for polyanions. At low concentrations of anionic groups in tissue sections, the groups and attached dye molecules maybe spaced too far apart for the neighboring dye molecules to mutually increase their affinity for their substrate. At any rate, it was shown experimentally that partially depoly-

EXPERIMENTAL OSTEOARTHRITIS

merized cartilage glycosaminoglycans cannot be adequately demonstrated by metachromatic dyes. The metachromatic dyes are not entirely alcohol “fast”; that is, some of the dye is readily washed out of the tissue section during dehydration. The loss of dye is gradual and time dependent and is likely to be a source of error. If the section remains in 70% ethanol solution for a period of time, a significant amount of the dye may be eventually washed out, depending on the type of tissue. Therefore, if a metachromatic dye is desired, the cartilage sections should be thoroughly washed following the staining procedure and then air-dried instead of being dehydrated and cleared. Certain non-metachromatic staining methods are more representative of the concentration of cartilage glycosaminoclycans than the metachromatic staining methods. Several of these methods are based on alcian blue, a phthalocyanin dye. It is generally used at pH 1.0 at which it stains only sulfate groups, or at pH 2.7 at which it stains both sulfate and carboxyl groups.84,77At even higher pH’s, other tissue components (nucleic acids, certain proteins) are also stained. A technique has been developeda whereby alcian blue is used at a considerably higher pH and the staining of undesired components is prevented by adding specific concentrations of MgCl,. The staining of hyaluronic acid is prevented at 0.1 A4 MgCl,, RNA and DNA at 0.2M MgCl*, chondrotin sulfate at 0.65M MgClz and heparin at 0.8M MgCI,. Staining of high molecular weight keratan sulfate could not be prevented at even 1.OM MgCl,. This is known as the critical electrolyte concentration technique. Stockwell and Scott86 used this technique to determine the distribution of chondroitin sulfate and keratan sulfate in articular cartilage. They stained the cartilage sections with alcian blue at pH 5.7, and in the presence of 0.4M MgCl, chondroitin sulfate and keratan sulfate both stain and in the presence of 0.9M MgCl,, only keratan sulfate stained. They confirmed the specificity of this method with hyaluronidase controls.87 Another series of methods is based on the condensation of phenylene amines on the reac-

371

tive groups.88~8g*g0 An acidic solution of dimethyl phenylenediamine (a mixture of both p and m isomers) is prepared and used with and without a number of variables, notably ferric chloride and prior periodic acid oxidation of the tissues. The procedure that is most satisfactory for staining cartilage is the high-iron method. It stains only sulfated glycosaminoglycans with apparently good stoichiometry between the dye molecules and the sulfate groups. Hale’s colloidal iron reaction is useful in identifying and characterizing the different types of mucosubstances of glands but does not seem to be useful in analyzing cartilage. The periodic acid-Schiff (PAS) reaction stains cartilage matrix rather intensely as well as any glycogen which might be present in the chondrocytes. The PAS reaction stains any compound which has hydroxyl groups on adjacent carbon atoms, which includes primarily neutral sugars. Cartilage contains 4 chemical sites which theoretically should be PAS positive, including (1) the mono- and disaccharide subunits which are covalently bound to collagen, (2) the glycoproteins involved in proteoglycan aggregation, (3) the 2 units of galactose at the glycosaminoglycan-protein linkage region, and (4) certain of the collagen cross linkages. It is impossible to determine how much of the PAS reaction intensity is due to each of these chemical sites, and it is therefore difficult to interpret the significance of the PAS reaction on cartilage. Quantitation. Most workers are content to judge staining intensities visually, but there are times when it would be of great advantage to be able to express staining intensities in exact numerical terms, such as optical density units, micrograms per milligram of tissue, or simply in some type of comparative units. It is possible and sometimes useful to assay the total reaction product of a whole section of tissue. Such methods have been used to quantitate an azo-dye reaction product of an esterase” and a formazan reaction product of dehydrogenases?293,94 If the reaction product contains a metal, it can be readily quantitated with atomic absorption spectrophotometry.” After the desired reaction is carried out on the histological sections, the sections are removed from the

HENRY TROYER

372

slides, weighed and dissolved in a measured quantity of nitric acid. The acid solution is then analyzed for the metal of interest. The method works particularly well with alcian blue-stained tissue sections (analyzing for copper). The method also has been applied to Hale’s colloidal iron method (analyzing for iron) and Gomori’s method for phosphatases (analyzing for lead), although these methods are not highly useful for studying cartilage and other joint tissues. Microspectrophotometers and microdensitometers are commercially available and are extremely useful for analyzing the color intensity of stained tissue sections, or the nature of their color. These instruments, however, are expensive, and it is often difficult to justify their aquisition. Photographic densitometry provides an alternate for quantifying staining intensities of stained tissue sections with a less expensive set-

up. The process essentially involves registering the intensity of the histochemical reaction on a photographic negative film. The negative can either be printed on paper and the silver or pigment analyzed from the positive image96,97,98 or the silver may be analyzed directly from the negative.99 The latter method can be used with remarkable accuracy. Kodak Panatomic X 35-mm roll film is used which is extremely consistant within rolls of film and from roll to roll. It is usually possible to get all the experimentals and controls on a 20- or 36-frame roll and thereby cancelling any variation that might be due to developing the film. The unexposed and fully exposed film ends can serve as the appropriate blanks. The usefulness of this method has been demonstrated in a study of alkaline phosphatase kineticslw and in evaluating the staining intensities of cartilage of arthritic rats.“’

REFERENCES 1. Gardner DL. The experimental production of arthritis. A review. Ann Rheum Dis 1960;19:297-317. 2. Moskowitz RW. Experimental models of degenerative joint disease. Semin Arthritis Rheum 1972;2:95-116. 3. Ely LW, Mensor MC. Studies on the immobilization of normal joints. Surg Gynecol Obstet 1933;57:212-215. 4. Maroudas A, Bullough P, Swanson SAV, Freeman MAR. The permeability of articular cartilage. J Bone Joint Surg [Br] 1968;50:166-177. 5. Evans EB, Eggers GWN, Butler JK, Blumel J. Experimental immobilization and remobilization of rat knee joints. J Bone Joint Surg [Am] 1960;42:737-758. 6. Thaxter TH, Mann RA, Anderson CE. Degeneration of immobilized knee joints in rats. Histological and autoradiographic study. J Bone Joint Surg [Am] 1965;47:567-585. 7. Hall MC. Articular changes in the knee of the adult rat after prolonged immobilization in extension. Clin Orthop 1964;34:184-195. 8. Hall MC. Cartilage changes after experimental immobilization of the knee joint of the young rat. J Bone Joint Surg [Am] 1963;45:3644. 9. Trueta J. Osteoarthritis: An approach to surgical treat-

14. Thompson RC, Bassett CAL. Histological observations on experimentally induced degeneration of articular cartilage. J Bone Joint Surg [Am] 1970,52:435443.

ment. Lancet 1956;1:585-589. IO. Salter RB, Field P. The effects of continuous compression on living articular cartilage. J Bone Joint Surg [Am] 1960;42:3149. 11. Trias A. Effect of persistent pressure on the articular cartilage. An experimental study. J Bone Joint Surg [Br] 1961;43:376-386. 12. Crelin ES, Southwick WO. Changes induced by sustained pressure in the knee joint articular cartilage of adult rabbits. Anat Ret 1964;149:113-134. 13. Matthiass HH, Glupe J. lmmobilisation und Druckbelastung in ihrer Wirkung auf die Gelenke. Arch Orthop Umfall-Chir 1966;60:38&396.

20. Roy S. Ultrastructure of articular cartilage in experimental immobilization. Ann Rheum Dis 1970;29:634-642. 2 I. Finsterbush A, Friedman B. Early changes in immobilized rabbits knee joints: A light and electron microscopic study. Clin Orthop 1973;92:305-3 19. 22. Refior HJ, HiJbner G. Zur Morphologie des hyalinen Gelenkknorpels unter Immobilisation und Remobilisation. Arch Orthop Trauma Surg 1978;9 1:305-3 14. 23. Ginsberg JM, Eyring EJ, Curtiss PH. Continuous compression or rabbit articular cartilage producing loss of hydroxyproline before loss of hexosamine. J Bone Joint Surg [Am] 1969;51:467474. 24. Enneking WF, Horowitz M. The intra-articular

15. Troyer H. The effect of short term immobilization on the rabbit knee joint cartilage. A histochemical study. Clin Orthop I975;107:249-257. 16. Eronen I, Videman T, Friman C, Michelsson JE. Glycosaminoglycan metabolism in experimental osteoarthrosis caused by immobilization. Acta Orthop Stand 1978;49:329-334. 17. Videman T, Michelsson JE, Rauhamaki R, Langenskiold A. Changes in “S-sulphate uptake in different tissues in the knee and hip regions of rabbits during immobilization, remobilization and the development of osteoarthritis. Acta Orthop Stand 1976;47:290-298, 18. Sood SC. A study of the effects of experimental immobilisation on rabbit articular cartilage. J Anat 1971;108:4977507. 19. Langenskiijld A, Michelsson JE, Videman T. Osteoarthritis of the knee in the rabbit produced by immobilization. Acta Orthop Stand 1979;50:1-14.

EXPERIMENTAL OSTEOARTHRITIS

373

effects of immobilization of the human knee. J Bone Joint Surg [Am] 1972;54:973-985. 25. Palmoski M, Perricone E, Brandt KD. Development and reversal of a proteoglycan aggregate defect in normal canine knee cartilage after immobilization. Arthritis Rheum 1979;22:508-5 17. 26. Perricone E, Brandt K, Palmoski M. Reduced joint usage leads to reduced proteoglycan aggregation in human articular cartilage. Arthritis Rheum 1980;23:730-731. 27. Hascall VC, Sajdera SW. Protein polysaccharide complex from bovine nasal cartilage: The function of glycoprotein in the formation of aggregates. J Biol Chem 1969;244:2384-2396. 28. Gritzka TL, Fry LR, Cheesman RL, LaVigne A. Deterioration of articular cartilage caused by continuous compression in a moving rabbit joint. A light and electron microscopic study. J Bone Joint Surg [Am] 1973;55:16981720. 29. Ogata K, Whiteside LA, Lesker PA, Simons DJ. The effect of varus stress on the moving rabbit knee joint. Clin Orthop 1977;129:313-318. 30. Harrison MHM, Schajowicz F, Trueta T. Osteoarthritis of the hip: A study of the nature and evolution of the disease. J Bone Joint Surg (Br) 1953;35:5988626. 3 1. Hall MC. Cartilage changes after experimental relief of contact in the knee joint of the mature rat. Clin Orthop 1969;64:6476. 32. Engh GA, Chrisman OD. Experimental arthritis in rabbit knees. A study of relief of pressure on one tibia1 plateau in immature and mature rabbits. Clin Orthop 1977;125:221-226. 33. Hulth A, Lindberg L, Telhag H. Experimental osteoarthritis in rabbits. Preliminary report. Acta Orthop Stand 1970;4 I :522-530. 34. Telhag H, Lindberg L. A method for inducing osteoarthritic changes 1972;86:214-223.

in

rabbits’

knees.

Clin

Orthop

35. Ehrlich MG, Mankin HJ, Jones H, Grossman A, Crispen C, Ancona D. Biochemical confirmation of an experimental osteoarthritis model. J Bone Joint Surg [Am] 1975;57:392-396. 36. Bohr H. Experimental osteoarthritis in the rabbit knee joint. Acta Orthop Stand 1976;47:558-565. 37. Lane JM, Chisena E, Black J. Experimental knee instability: Early mechanical property changes in articular cartilage in a rabbit model. Clin Orthop 1979;140:262-265. 38. Marshall JL. Periarticular osteophytes: Initiation and formation in the knee of the dog. Clin Orthop 1969;62:3747. 39. Marshall JL, Olsson SE. Instability of the knee: A long-term experimantal study in dogs. J Bone Joint Surg [Am] 1971;53:1561-1570. 40. Gilbertson EMM. Development of periarticular osteophytes in experimentally induced osteoarthritis in the dog. Ann Rheum Dis 1975;34:12-25. 41. Pond MJ, Nuki G. Experimentally induced thritis in the dog. Ann Rheum Dis 1973;32:387-388.

osteoar-

42. McDevitt CA, Muir H. Biochemical changes in the cartilage of the knee in experimental and natural ostcoarthritis in the dog. J Bone Joint Surg [Br] 1976;58:94lOl. 43. McDevitt C, Gilbertson E, Muir H. An experimental

model of osteoarthritis: Early morphological and biochemical changes. J Bone Joint Surg [Br] 1977;59:24-35. 44. Palmoski MJ, Colyer RA, Brandt KD. Marked suppression by salicylate of the augmented proteoglycan synthesis in osteoarthritic cartilage. Arthritis Rheum 1980;23:83-90. 45. Palmoski M, Brandt K. Studies of the basis for salicylate-induced inhibition of glycosaminoglycan synthesis in joint cartilage. Arthritis Rheum 1980;23:729. 46. Fairbanks TJ. Knee joint changes after meniscectomy. J Bone Joint Surg [Br] 1948;30:664-670. 47. Tapper EM, Hoover NW. Late results after meniscectomy. J Bone Joint Surg [Am] 1969;51:517-526. 48. Jackson JP. Degenerative changes in the knee after meniscectomy. Brit Med J 1968;2:525-527. 49. Krause WR, Pope MH, Johnson RJ, Wilder DG. Mechanical changes in the knee after meniscectomy. J Bone Joint Surg [Am] 1976;58:599-604. 50. Moskowitz RW, Davis W, Sammarco J, et al. Expcrimentally induced degenerative joint lesions following partial meniscectomy in the rabbit. Arthritis Rheum 1973;16:397405. 51. Moskowitz RW, Goldberg VM, Malemud CJ. Metabolic responses of cartilage in experimentally induced osteoarthritis. Arthritis Rheum 1980,23:724. 52. Lutfi AM. Morphological changes in the articular cartilage after meniscectomy. An experimental study in the monkey. J Bone Joint Surg [Br] 1975;57:525-528. 53. Floman Y, Eyre DR. Glimcher MJ. Induction of osteoarthrosis in the rabbit knee joint: Biochemical studies on the articular cartilage. Clin Orthop 1980;147:278-286. 54. Shapiro F, Glimcher MJ. Induction of osteoarthrosis in the rabbit knee joint: Histologic changes following meniscectomy and meniscal lesions. Clin Orthop 1980;147:287295. 55. Doyle JR, Eisenberg JH, Orth MW. Regeneration of knee menisci: A preliminary report. J Trauma 1966;6:5055. 56. Evans DK. Repeated regeneration of a meniscus in the knee. J Bone Joint Surg [Br] 1963;45:748-749. 57. Elmer RM, Moskowitz RW, Frankel VH. Meniscal regeneration and postmeniscectomy degenerative joint disease. Clin Orthop 1977; 124:30&3 IO. 58. Cox JS, Nye CE, Schaefer WW, Woodstein IJ. The degenerative effects of partial and total resection of the medial meniscus in dogs knees. Clin Orthop 1975;109:178183. 59. Cox JS, Cordell LD. The degenerative effects of medial meniscus tears in dogs knees. Clin Orthop 1977;125:23&242. 60. Dann P, Haike H, Rosenbauer K. Experimentelle Untersuchungen zur Frage der totalen oder partiellen Meniscusresektion. Arch Orthop Umfall-Chir 1969;65:209219. 61. DePalma AF, Flynn JJ. Joint changes following experimental partial and total patellectomy. J Bone Joint Surg [Am] 1958;40:395413. 62. Garr EL, Moskowitz RW, Davis W. Degenerative changes following experimental patellectomy in the rabbit. Clin Orthop 1973;92:296-304. 63. Akeson WH, Eichelberger L, Roma M. Biochemical studies of articular cartilage. II. Values following the dener-

374

vation of an extremity. J Bone Joint Surg [Am] 1958;40:153-162. 64. Finsterbush A, Friedman B. The effect of sensory denervation of rabbits knee joints. A light and electron microscopic study. J Bone Joint Surg [Am] 1975;57:949956. 65. Eichelberger L, Miles JS. The histological characterization of articular cartilage of poliomyelitis patients. J Lab Clin Med 1952;40:284-296. 66. Dekel S, Weissman SL. Joint changes after overuse and peak overloading of rabbit knees in viva. Acta Orthop Stand 1978; 49:519-528. 67. Simon SR, Radin EL, Paul IL, Rose RM. The response of joints to impact loading. II. In viva behavior of subchondral bone. J Biomech 1972;5:267-272. 68. Thomas L. Reversible collapse of rabbit ears after intravenous papain, and prevention of recovery by cortisone. J Expl Med 1956;194:245-252. 69. Bryant JH, Leder IG, Stetten D. The release of chondroitin sulfate from rabbit cartilage following the intravenous injection of crude papain. Arch Biochem Biophys 1958;76: 122-I 30. 70. Murry DG. Experimentally induced arthritis using intraarticular papain. Arthritis Rheum 1964;7:21 I-219. 7 1. Farkas T, Bihari-Varga M, Biro T. Thermoanalytical and histological study of intra-articular papain-induced degradation and repair of rabbit cartilage. II. Mature animals. Ann Rheum Dis 1976;35:23-26. 72. Farkas T, Bihari M, Biro T. Thermoanalytical and histological study of intra-articular papain-induced degradation and repair of rabbit cartilage. I. immature animals. Ann Rheum Dis 1974;33:385-390. 73. Spicer SS, Bryant JH. Cartilage changes in papaintreated rabbits. Am J Pathol 1957;33:1237-1245. 74. Bentley G. Papain-induced degenerative arthritis of the hip in rabbits. J Bone Joint Surg [Br] 1971;53:324-337. 75. Havdrup T, Telhag H. Papain-induced changes in the knee joints of adult rabbits. Acta Orthop Stand 1977;48:143-149. 76. Havdrup T. Trypsin induced mitosis in the articular cartilage of rabbits. Acta Orthop Stand 1979;.50: 15-I 9. 77. Pearse AGE. Histochemistry: theoretical and applied. Third edition. Vol 1. Boston: Little, Brown and Co., 1968. 78. Pearse AGE. Histochemistry: Theoretical and Applied. Third edition. Vol II. Baltimore: The Williams and Wilkins Company, 1972. 79. Troyer H, Babich E. A hematoxylin and eosin-like stain for glycol methacrylate-embedded tissue sections. Stain Technol, 1981 56:39-43. 80. Bennett HS, Wyrick AD, Lee SW, McNeil JH. Science and art in preparing tissues embedded in plastic for light microscopy, with special reference to glycol methacrylate, glass knives and special stains. Stain Technol 1976;51:71-97. 81. Troyer H. Principles and techniques of histochemistry. Boston: Little, Brown and Co., 1980. 82. Rosenberg L. Chemical basis for the histological use of safranin 0 in the study of articular cartilage. J Bone Joint Surg [Am] 197 1;53:69-82. 83. Troyer H. A microspectrophotometric study of

HENRY TROYER

metachromasia. J Histochem Cytochem 1974;22:11181121. 84. Spicer SS, Horn RG, Leppi TJ. Histochemistry of connective tissue mucopolysaccharides. In: Wagner BM, Smith DE, eds. The connective tissue. Baltimore: The Williams and Wilkins Company, 1967. 85. Scott JE, Dorling J. Differential staining of acid glycosaminoglycans (mucopolysaccharides) by alcian blue in salt solutions. Histochem 1965;5:221-233. 86. Stockwell RA, Scott JE. Observations on the acid glycosaminoglycan component (mucopolysaccharide) content of the matrix of ageing cartilage. Ann Rheum Dis 1965;24:341-350. 87. Scott JE, Stockwell RA. On the use and abuse of the critical electrolyte concentration approach to the localization of tissue polyanions. J Histochem. Cytochem 1967;15:1 I I113. 88. Spicer SS. The use of various cationic reagents in histochemical differentiation of mucopolysaccharides. Am J Clin Pathol 1961;36:393-407. 89. Spicer SS. Diamine methods for differentiating mucosubstances histochemically. J Histochem Cytochem 1965;13:21 l-234. 90. Spicer SS, Jarrel MH. Histochemical reaction of an aromatic diamine with acid groups and periodate engendered aldehydes in mucopolysaccharides. J Histochem Cytochem 1961;9:368-379. 91. Hopsu VK, McMillan PJ. Quantitative characterization of a histochemical enzyme system. J Histochem Cytothem 1964; 12:3 15-324. 92. Altman FP. The quantitative elution of nitro-blue formazan from tissue sections. Histochemistry 1969;17:3 19326. 93. Butcher RG, Altman FP. Studies on the reduction of tetrazolium salts. II. The measurement of the half reduced and fully reduced formazans of neotetrazolium chloride in tissue sections. Histochemistry 1973;37:351-363. 94. McCabe M, Maple AJ, Jones GRN. Variations in dehydrogenase and neotetrazolium reductase (diaphorase) activities in various buffers studied quantitatively in frozen sections. J Histochem Cytochem 1965;13:541~546. 95. Rosenquist TH, Rosenquist JW. A procedure for the use of atomic absorption spectrophotometry in quantitative histochemistry. J Histochem Cytochem 1974;22:104-109. 96. Niemi M. Cytophotometry by silver analysis of photographs. Acta Anat 1958;Suppl 34: l-92. 97. Ornstein L. The distributional error in microspectrophotometry. Lab Invest 1952;1:250-265. 98. den Tonkelaar EM, van Duijn P. Photographic colorimetry as a quantitative cytochemical method. I. Principles and practice of the method. Histochemistry 1964;4:1-9. 99. Troyer H, Rosenquist TH. Atomic absorption spectrophotometry applied to photographic densitometry. J Histothem Cytochem 1975;23:941-944, 100. Troyer H, Fisher OT, Rosenquist TH. Bone alkaline phosphatase kinetics studied by a new method. Histochemistry 1976;50:25 l-259. 101. Nusbickel FR, Troyer H. Histochemical investigation of adjuvant-induced arthritis. Arthritis Rheum 1976;19:1339-1346.