Portal de Periódicos da CAPES

Bone: Formation by Autoinduction

2002; Lippincott Williams & Wilkins; Volume: 395; Linguagem: Inglês

10.1097/00003086-200202000-00002

1528-1132

Marshall R. Urist, Leonard F. Peltier,

Hip disorders and treatments

Marshall R. Urist (Fig 1) was born on June 11, 1914 in Chicago, IL, where he spent his early life. After obtaining his bachelor’s degree from the University of Michigan, he returned to Chicago where he received a master’s degree from the University of Chicago. He then went to Baltimore where he earned his medical degree from the Johns Hopkins medical school in 1941. After a surgical internship at the Johns Hopkins Hospital, Urist began an orthopaedic residency in Baltimore. This was interrupted by military service (1943–1946) beginning in the California-Arizona Maneuver Area, extending through service in hospitals in Germany, and ending in the Surgeons General’s Office as an Orthopaedic Consultant.Fig 1.: Marshall R. Urist, MD, PhD.After discharge from the army, Urist finished his orthopaedic training at the Massachusetts General Hospital and the Boston Children’s Hospital. In 1947, Urist returned to the University of Chicago as an intructor in the Department of Physiology. It was here that he began his association with Franklin C. McLean, a pioneer in the study of the physiology of bone, which resulted in their authorshop of a monograph on bone physiology. 1 The direction of Urist’s career was determined by his association with McLean, and he never wavered in his zeal to investigate problems posed by bone physiology. Urist moved to Los Angeles, CA, where he opened a laboratory at the University of California Los Angeles. He was one of the early members of the surgical faculty and had an active clinical practice. As a clinical surgeon, Urist made substantial contributions to the development of total hip replacement using metal acetabular and femoral components. His contribution to the teaching program led to his appointment as a professor of orthopaedic surgery in 1969. In addition to participating in many national and international orthopaedic societies and representing his specialty as a consultant on many important national committees, Urist found the time to work on the editorial board of the journal, Clinical Orthopaedics. In 1966, he became editor of this journal and broadened its scope to include papers reflecting the research interests of the orthopaedic community. This led to a change of names, Clinical Orthopaedics became Clinical Orthopaedics and Related Research. To many, Urist’s editorship of the journal was one of his greatest achievements. His 27 years of leadership established the journal as a major source of clinical/material and research for practitioners and students. While juggling all of these balls, Urist continued to patiently and thoroughly study bone physiology in his laboratory. The work from his laboratory was awarded two Kappa Delta Awards (1950, 1981). The focus of his research for the past 50 years has been on bone induction. Therefore, it is appropriate that this symposium begins with one of his papers as the classic article. BONE: FORMATION BY AUTOINDUCTION Marshall R. Urist Abstract. Wandering histiocytes, foreign body giant cells, and inflammatory connective-tissue cells are stimulated by degradation products of dead matrix to grow in and repopulate the area of an implant of decalcified bone. Histiocytes are more numerous than any other cell form and may transfer collagenolytic activity to the substrate to cause dissolution of the matrix. The process is followed immediately by new-bone formation by autoinduction in which both the inductor cells and the induced cells are derived from ingrowing cells of the host bed. The inductor cell is a descendant of a wandering histiocyte; the induced cell is a fixed histiocyte or perivascular young connective-tissue cell. Differentiation of the osteoprogenitor cell is elicited by local alterations in cell metabolic cycles that are as yet uncharacterized. New evidence in favor of the theory of induction can be gathered from the process of bone formation in the interior of an implant of acellular, devitalized, decalcified, bone matrix. Differing from previous demonstrations of induction systems, which produce scanty deposits and less than 30-percent positive results, decalcified bone yields new bone in an amount proportional to the volume of the implant; the percentage of positive experimental results is as high as over 90 percent. The results of some 70 experiments on approximately 300 animals are summarized below to introduce a hypothesis of postfetal osteogenesis by autoinduction. Long bones, excised from adult rabbits and other laboratory animals, were cut in lengths of 1 to 2 cm and decalcified, unfixed, in 0.6 N HCl. Samples of human cortical bone obtained from accident cases, excised under aseptic conditions at autopsy, were lyophilized, and decalcified similarly in sterile solutions of 0.6 N HCl for a period of 5 days. The acid was removed by prolonged washing in sterile 0.15 M NaCl. The chemical composition of HCl-decalcified bone matrix, in millimoles per liter, per kilogram, was total Ca, 4.4±2.6; total P, 17.5±3.0; Na, 16.3±1.8; hexosamine, 41.4±11.5 (mean and standard deviation); in percentage dry weight, total N was 4.4±0.5. The samples of HCl-decalcified homogenous diaphyseal bone were implanted into (i) a pouch in the belly of either the rectus abdominus, quadriceps, or erector spinae muscles of approximately 250 rabbits, 20 rats, 10 mice, and 5 guinea pigs; (ii) a defect in the ulna in 10 rabbits or a bed of bone on the lumbar vertebrae in 3 dogs; (iii) a defect in a bone in various skeletal system disorders in 21 human beings. In an effort to alter the chemically reactive groups associated with the structure of the matrix, samples of bone were (i) decalcified in a series of equimolar solutions of seven different acids; (ii) decalcified in HCl and treated with blocking reagents for carboxyl, ε-amino, and nonterminal amino or sulfhydryl reactive groups; (iii) altered physically by alcohol fixation, heat shrinkage at 70°C, lyophilization, or denaturation and sterilization in β-propiolactone. The effects were assessed by correlated radiographic and histological methods and by determinations of the percentages of positive results in eight to ten implants of each preparation. The typical implant of homogenous HCldecalcified matrix, 3 weeks after implantation in the anterior abdominal wall, was enveloped in loose, highly vascular, inflammatory, and fibrous connective tissue. The trabecular interstices and old vascular channels were infiltrated with wandering histiocytes or macrophages, large and small lymphocytes, and fibroblasts. The matrix was swollen and amorphous in hematoxylin and eosin and showed metachromatic staining with toluidine blue, but the collagen fiber bundles were clearly distinguishable and surprisingly intact with Wilder’s stain. The osteocyte lacunae, however, were empty and often very much enlarged. The invading cells of the host, chiefly wandering histiocytes, were first closely related to the process of resorption of dead decalcified bone matrix but later became arranged in three distinct population groups. One was found in excavation chambers produced by the pressure of sprouting capillaries, the proteolytic activity of macrophages, the gathering of foreign body giant cells, and the multiplication of young connective-tissue cells. Another was found between ribbons of disintegrating collagenous matrix and consisted of giant cells, inflammatory round cells, dilated blood vessels, and fibrinous exudate. Another appeared in closed ends of smooth-walled, old, vascular channels of the decalcified matrix, or between folds of compressed, softened matrix, and consisted of nests of proliferating cartilage cells. The earliest deposits of new bone appeared at 4 to 6 weeks from osteoprogenitor cells in the interior of well-vascularized excavation chambers. Later, between 8 and 16 weeks, bone formed from proliferating connective-tissue cells associated with vascularization, calcification, and replacement of the nests of cartilage by the typical route of endochondral ossification. Bone formation did not occur from foreign body giant cells, plasma cells, small lymphocytes, large dilated capillaries, or cell populations associated with inflammation. Wherever bone induction occurred there was a pool of stem cells, osteoprogenitor cells, and small capillaries, surrounded by palisades of deeply basophilic plump osteoblasts. By appositional formation of new bone, layers of osseous tissue with intermediate cement lines were deposited on the surface of decalcified dead matrix. The deposits were always enclosed in excavation chambers, either within the external surface or deep inside the old matrix; new bone never extended outside the implant. A periosteum-like outer envelope of fibrous tissue surrounded the dead matrix and confined the processes of bone and cartilage induction strictly to spaces created by resorption of the dead matrix. Living new bone was always easy to distinguish from dead, mineral-free matrix. It calcified rapidly almost as soon as it was laid down, and it was always separated from decalcified matrix by a thin line of cement substance. Areas of recalcification of dead matrix were few, and they rarely coincided with areas of osteogenesis. In one experiment the decalcified bone was primed in a solution of calcium chloride before implantation to initiate recalcification, but this treatment retarded resorption of matrix and hampered osteogenetic induction. Matrix decalcified with ethylenediamine tetra-acetic acid (EDTA), mixed formic and citric acids, or acetic acid produced osteogenesis in the same way as matrix decalcified with HCl, but EDTA produced a slightly lower percentage of positive results. Lactic acid failed to remove all the mineral, and a diffuse deposit that remained seemed to increase inflammation and prevent osteogenesis. Heating to 70°C, sufficiently to produce shrinkage of the collagen fibers of the bone matrix, impeded but did not altogether prevent osteogenetic induction. Matrix decalcified with nitrous acid or nitric acid, and matrix sterilized with β-propiolactone and decalcified with HCl did not induce bone formation in a single instance. Inhibition also occurred from dinitrophenolation (FDNB) and iodoacetamidization (IAA) of matrix decalcified with HCl. Morphologically, nitrous and nitric acid, which deaminated tissue proteins, produced widespread inflammation and total disintegration of the decalcified matrix; FDNB, which blocks ε-amino and other groups, and IAA, which blocks sulfhydryl and other reactive groups, dehydrated and shrank the decalcified matrix and appeared to prevent osteogenetic induction by retarding the cellular ingrowth and excavation of matrix. Toluidine blue, which blocks carboxylic acid groups and binding of calcium ions, prevented recalcification of matrix but did not inhibit osteogenetic induction. Neither did lyophilization or matrix fixation in alcohol. A series of experiments, designed to determine the fate of HCl-decalcified matrix in a host bed of bone tissue, demonstrated that HCl-decalcified bone is an excellent substitute for a bone graft. The dead, decalcified matrix was invaded by new blood vessels and resorbed rapidly; new bone was deposited in pockets or excavation chambers filled with proliferating osteoprogenitor cells. The process began within a few weeks and was complete within a few months. How much new bone could be attributed to the osteoconduction of cells growing in from the walls of the host bone, and how much could be accounted for by osteogenetic induction, was not apparent from experiments on normal, healthy bones. The most important tests were on decalcified matrix in adult human bones with avascular necrosis, large amounts of missing bone substance, or irregular, small, cortical surfaces like those of the lower lumbar vertebrae. Another important test was the capacity of decalcified bone to produce lumbar spinal fusion in a dog; in this area of the skeleton the mechanical system of the wagging tail acts to impede osteoconduction from the host bed. In abnormal host bone beds, either mechanical or pathological, successful results were obtained chiefly in young individuals with active, rapidly proliferating osteoprogenitor cells. In cases in which there was a defect to fill, the most striking difference between undecalcified and HCl-decalcified bone matrix was in the radiographic picture and relatively rapid rate of replacement Undecalcified bone grafts were more radio-opaque than ingrowing new bone and were replaced slowly over long periods of time, ranging from months to years. Decalcified bone implants were at first completely radiolucent but were replaced by low-density, new-bone tissue very rapidly over a period of weeks to a few months. Mechanically, decalcified matrix was unsatisfactory, inasmuch as it did not provide internal fixation during the preinduction phase and did not arthrodese the vertebrae of the moving lumbar spine of a dog. The area of the implant of dead bone matrix is vacant territory and immediately begins to attract wandering histiocytes. Within 3 weeks every accessible microscopic space is occupied by a new cell. The invading cells resorb matrix, release tropic chemical agents, and stimulate capillaries and perivascular connective tissue to excavate large chambers for further occupancy by their offspring and additional cell populations. When one wandering histiocyte (the inductor cell) and one perivascular connective-tissue cell (the responding or induced cell) divide and interact, cellular differentiation occurs by autoinduction to produce two additional cells, one responding cell and one specialized form, either an osteoprogenitor or a chondroprogenitor cell. The chondroprogenitor cell produces cartilage, which later becomes vascularized to set up a second or delayed induction system for bone formation. Whether the progenitor cell produces bone or cartilage is determined by microenvironmental or local factors. Bassett prepared tissue cultures and reviewed the literature to point out that compaction of cells with low oxygen saturation produces cartilage; compaction with high oxygen saturation produces bone; stretching the tissue, or mechanical tension, produces only fibrous tissue. Whether chemical by-products of cell metabolism and bioelectric effects are secondarily produced are matters for future investigation. Except that the repopulation of the implant with mesenchymal cells (both wandering and fixed histiocytes) is intimately associated with the process of absorption of the collagenous material, the mechanism of removal of decalcified matrix is not known. Foreign body giant cells (multinucleated cells larger than osteoclasts) are found in close contact with the cut edges of the trabeculae, but their numbers are too few to account for all the collagen that is removed. Wandering histiocytes (macrophages) appear in very large numbers between the fibers of disintegrating collagen but do not contain particulate material identifiable as phagocytosed matrix. The evidence at present is that colloidal material of ultrastructural, rather than microscopic, dimensions is phagocytosed. It is reasonable to suppose that the cells that make bone-matrix collagenase carry the enzyme to the substrate and break the fibrous proteins down to subultrastructural or even amino acid size particles. Collagenases are not stored and are not extractable from fresh tissues; they accumulate in the media of certain cells in tissue culture. How the giant cells differ from osteoclasts and engage in resorption is still a mystery. They are found in pits similar to Howship’s lacunae in the surface of decalcified and partially recalcifying surfaces indiscriminately, but they are larger and more numerous around bone decalcified with nitric acid, which neither recalcifies nor induces osteogenesis. Irving’s work on resorption of normal and decalcified dead bone matrix demonstrates that the amounts of cytochrome oxidase, acid phosphatase, succinic dehydrogenase, malic, lactic, and glutamic dehydrogenases, diphosphopyridine nucleotide diaphosphorase, and isocitric dehydrogenases are virtually the same in foreign body giant cells as in osteoclasts. Inasmuch as rachitic osteoid does not elicit foreign body giant cell formation, does recalcify, and is not resorbed, and insofar as decalcified normal bone matrix does attract multinucleated cells, Irving concludes, first, that the inorganic phase of bone plays only a minor role in causing osteoclasis; second, that matrix from a nonrachitic animal contains something that makes it resorbable; and third, that there is a tropism for the ends of the implants where the bone had been cut. Irving also described decalcified bone matrix, prepared in EDTA. 0.5 M at p H 7.0, and implanted subcutaneously for 8 weeks, but did not mention bone induction. The migration of wandering histiocytes into the decalcified bone implant resembles Weiss’s patterns of movement of cells during tissue reconstruction as described by Mascona. Undifferentiated motile cells are guided by tissue-specific factors that control migrations, swarming, translocations, and regroupings. It is assumed, from studies on cell exudates, that cells elaborate an extracellular material that has the dual purpose of binding them together and transmitting inductive stimuli. If the migratory stimulus is tissue-specific, it is conceivable that histiocytes with special predisposition to osteogenesis are attracted into the decalcified implant. There is considerable evidence, reviewed by Saxen and Toivonen, that there are heterogenous inductors which are (i) cross-tissue as well as cross-species specific, (ii) obtainable from adult as well as from embryonic material, (iii) active in heterotopic sites, and (iv) resistant even to chemical fixatives. This process is referred to as heterogenous induction and includes a hypothetical mesodermalizing principle. The sum and substance of current literature on heterogenous induction defies all present understanding of the chemistry of life processes. Bone induction in decalcified matrix raises questions that are common to the whole unsolved problem of cellular differentiation. Does matrix produce a specific diffusible chemical agent that induces the cells of the host to differentiate into osteoblasts? The answer is no. The system is more complex than a simple chemical stimulus and direct cell response; its complexity suggests that we should reconsider Spemann’s theory of induction. Spemann’s concept, derived chiefly from experiments upon embryonic tissue, means the influence of one living cell (an inductor) upon another living one (the induced cell) to cause it to differentiate into a more specialized form. In oversimplified but realistic terms, the idea of induction comes from the observation that the cells of the “eye” cup derived from the neural plate induce cells of skin-forming ectoderm to bring about the formation of the lens. Similarly, a group of proliferating cells inside an excavation chamber in decalcified matrix induces a group of young connective-tissue cells associated with capillary sprouts to differentiate first into osteoprogenitor cells, then into osteoblasts. A group of proliferating cells in the closed end of an old vascular channel induces a group of their own offspring to differentiate into chondroprogenitor cells (without any preliminary sprouting of capillaries or absorption of dead matrix). Table 2 (not shown) illustrates the sequence of appearance of various cell types in decalcified matrix induction systems for cartilage and bone. Experiments with specific diffusible chemical agents from spinal cord and notochord which act locally to induce chondrogenesis, systems which induce osteogenesis across Millipore filters, and extracts of bone and cartilage that alter tissues to produce bone formation by induction, are not incontrovertible evidence of a diffusible inductor. The reaction of embryonic tissue can always be challenged with questions about the difficulty of assessing the stage of differentiation of cells in the preinduction phase. Experiments with extracts of tissue are often rejected as examples of the unspecific effects of chemical injury. All of these objections are avoided, however, by the hypothesis of autoinduction and the simple assumption that nonspecific substances or degradation products of dead tissue stimulate or attract the wandering histiocytes to migrate into the interior of the implant. Shipley and Macklin describe ingestion of ultramicrons of trypan blue by histiocytes preceding osteogenesis and encourage one to speculate that these cells induce other primitive connective-tissue cells to differentiate into osteoblasts. Inasmuch as decalcified bone produces new bone even after heat shrinkage, fixation in formalin and in alcohol, lyophilization, and toluidine blue staining, and other treatments, only nonspecific acid or neutral salt-insoluble substances, which resist the most harsh treatment, and only nonspecific histiocyte-attracting chemical substances could possibly emanate from dead matrix. Furthermore, intramuscular implants of HCl-treated muscle, tendon, cartilage, and many other tissues unrelated to bone matrix may similarly produce new bone by autoinduction. Thus, RNA and degradation products of various intracellular proteinaceous materials, previously claimed to act as inductors, would enter only into the preinduction process. In the bone-induction process involving, as noted in Table 2, the interaction of living cells under normal conditions in the living animal, the inductor would not need to diffuse beyond subultramicroscopic distances. Cell interaction could take place immediately before mitosis, and the inductor could flow along membranes from one cell to another. The evidence for osteogenesis by autoinduction is entirely morphological but, nevertheless, substantial. Bone formation occurs in extraskeletal implants of decalcified bone matrix in the interior of excavation cavities, and the new osteoblasts are derived, not from elements of the donor tissue, but from proliferating pleuripotent, ingrowing cells of the host. Comparable implants of undecalcified dead bone also produce osteogenesis, but only very rarely, only in scanty amounts, and then only after a latent period of several months. Ray and his associates and Sherrard and Collins describe rapid replacement of decalcified bone in bone defects, but as new bone arises from surrounding bone tissue, the orthotopic system does not offer convincing evidence of induction. In an extraskeletal or heterotopic implant of decalcified dead matrix, cell-induction sequences produce an entirely new ossicle with a marrow cavity as the end-product, and not merely histiotypic osteogenesis. Because differentiation of new-bone marrow occurs concomitantly with the process of new-bone formation by induction, the possibility of implanting decalcified bone matrix in subcutaneous soft parts in individuals with severe anemia, to produce a new site of active hematopoiesis, is important and interesting to contemplate. Uncertainty exists, however, about whether the period of survival of an ossicle in soft parts would be determined by the physiological demands for bone marrow, or mechanical stimuli for bone tissue. In a normal animal, extraskeletal ossicles decrease in size and may be entirely resorbed after a few years in the anterior abdominal wall.