Natural chondroitin sulphates increase aggregation of proteoglycan complexes and decrease ADAMTS-5 expression in interleukin 1 beta-treated chondrocytes.

OBJECTIVE: To assess the effect of natural chondroitin sulphate (CS) on the ability of neosynthesized sulphated proteoglycans (PGs) to aggregate in cultured chondrocytes treated with interleukin (IL)1 beta. METHODS: Primary cultured rabbit articular chondrocytes were treated or not with IL1 beta alone or with concentrations of CS for 20 h. Neosynthesized PGs were labelled by incorporation of [35SO(4)]-sulphate and analysed by chromatography on Sepharose 2B columns. Gelatinolytic activity was measured by zymography, and matrix metalloproteinase (MMP)1 mRNA level in chondrocytes underwent real-time PCR. Expression of ADAMTS (for "a disintegrin and metalloproteinase with thrombospondin motifs") -4 and -5 was analysed by real-time PCR and western blotting. RESULTS: The production of [35SO(4)]-labelled PGs was significantly increased with 10 microg/ml CS in the cellular pool rather than in the incubation medium. The addition of CS to IL1 beta-treated cells inhibited in part the disaggregation of sulphated PGs induced by IL1 beta. This inhibitory effect of CS is associated with a significant decrease in ADAMTS-5 expression at the mRNA and protein levels. No effect of CS was observed on IL1 beta-induced gelatinolytic activity, MMP1 mRNA expression or ADAMTS-4 expression. CONCLUSION: CS increases the production of functional sulphated PGs in the direct environment of chondrocytes in vitro. This beneficial effect of CS in IL1 beta-treated cells is associated with decreased expression of ADAMTS-5.

Posttranscriptional effect of insulin-like growth factor-I on interleukin-1beta-induced type II-secreted phospholipase A2 gene expression in rabbit articular chondrocytes.

Large amounts of type II-secreted phospholipase A2 (type II sPLA2) are secreted into inflammatory synovial fluid and they are believed to induce the synthesis of lipid mediators by articular chondrocytes. Preliminary experiments showed that insulin-like growth factor-I, which counteracts cartilage degradation in arthritis, inhibits interleukin-1beta-induced type II sPLA2 gene expression in rabbit articular chondrocytes (Berenbaum, F., G. Thomas, S. Poiraudeau, G. Bereziat, M.T. Corvol, and J. Masliah. 1994. FEBS Lett. 340: 51-55). The present study showed that IL-1beta induced the sustained synthesis of prostaglandin E2 and a parallel increase in type II sPLA2 gene expression (assessed by enzymatic activity and Northern blot analysis), but no increase in cytosolic PLA2 gene expression (assessed by Northern and Western blot analysis) or cytosolic PLA2 activity in rabbit articular chondrocytes. IGF-I inhibited both IL-1beta-stimulated PGE2 synthesis and type II sPLA2 gene expression, but had no effect on cytosolic PLA2 gene expression. Nuclear run-on experiments revealed that IL-1beta stimulated the transcription rate of type II sPLA2 gene, giving rise to long-lived mRNA in cells treated with actinomycin D. IGF-I did not affect transcription rate, suggesting that it acts as a post-transcriptional step. Sucrose density gradient analysis of the translation step showed no effect of IGF-I on the entry of type II sPLA2 mRNA into the polysomal pool or on its distribution into the various polysomal complexes, suggesting that IGF-I does not act on the translation of the mRNA. Lastly, IGF-I strongly decreased the half-life of IL-1beta-induced type II sPLA2 mRNA (from 92 to 12 h), suggesting that IGF-I destabilizes mRNA. These data demonstrate that IL-1beta stimulates the transcription rate of the type II sPLA2 gene and gives rise to a very stable mRNA. In contrast, IGF-I decreases the half-life of the type II sPLA2 message.

Renin biosynthesis by human tumoral juxtaglomerular cells. Evidences for a renin precursor.

Renin biosynthesis was studied in a juxtaglomerular cell tumor. The tumoral tissue had a high renin content (180 Goldblatt Units/g of tissue), was heavily stained by immunofluorescence using human renin antiserum, and exhibited numerous characteristic secretory granules by electron microscopy. In one series of experiments, renin biosynthesis was studied in tissue slices, by following the incorporation of radiolabeled amino acids into specific immunoprecipitable renin. Time course studies showed that renin was first synthesized in a high molecular weight form, 55,000 mol wt, i.e., 10,000 mol wt higher than that of active renin, and was then converted into a 44,000-mol wt form. In a second series of experiments renin tumoral cells were cultured. Small, round, birefringent cells obtained after collagenase digestion produced renin in both primary culture and subculture media. After 5 d most of the renin found in the culture medium was inactive, but could be activated by trypsin treatment. The tumoral tissue exhibited a strong renin immunofluorescence and numerous secretory granules were observed by electron microscopy. In contrast, the renin-producing cells isolated from this tumor and grown in culture showed little renin immunofluorescence and no secretory granule could be observed. The renin-producing cells in primary culture and subculture were pulsed with radiolabeled amino acids, and immunoprecipitable radiolabeled renin was found in the culture media, thus demonstrating the actual biosynthesis of the enzyme. This renin was not stored inside cultured cells but was rapidly released into the medium and had a molecular weight of 55,000. No conversion of this inactive high molecular weight renin into the active, 44,000 mol wt form of renin was observed. We postulate the existence of two pathways for the processing, packaging, and secretion of renin in the tumoral cells: in juxtaglomerular cells of tumoral tissue renin is synthesized as a preprorenin and rapidly converted into prorenin (55,000 mol wt), which is in turn packaged in secretory granules where it is processed into active renin (44,000 mol wt) and finally secreted; in the cultured tumoral cells renin is still biosynthesized as a preprorenin molecule and then converted into prorenin, but is neither stored as granules nor processed into active renin. In this case the renin is released in an inactive form.

A material decoy of biological media based on chitosan physical hydrogels: application to cartilage tissue engineering.

The cartilage tissue has a limited self-regenerative capacity. Tissue-engineering represents a promising trend for cartilage repair. The present study was aimed to develop a biomaterial formulation by combining fragments of chitosan hydrogel with isolated rabbit or human chondrocytes. We first reported the properties of the constructs elaborated with rabbit chondrocytes and pure chitosan physical hydrogels with defined molecular weight, acetylation degree and polymer concentration. Morphological data showed that chondrocytes were not penetrating the hydrogels but tightly bound to the surface of the fragments and spontaneously formed aggregates of combined cell/chitosan. A significant amount of neo-formed cartilage-like extracellular matrix (ECM) was first accumulated in-between cells and hydrogel fragments and furthermore was widely distributed within the neo-construct. The optimal biological response was obtained with hydrogel fragments concentrated at 1.5% (w/w) of polymer made from a chitosan with a degree of acetylation between 30 and 40%. Such hydrogels were then mixed with human chondrocytes. The phenotype of the cells was analyzed by using chondrocytic (mRNA expression of mature type II collagen and aggrecan as well as secretion of proteoglycans of high molecular weight) and non chondrocytic (mRNA expression of immature type II collagen and type I collagen) molecular markers. As compared with human chondrocytes cultured without chitosan hydrogel which rapidly dedifferentiated in primary culture, cells mixed with chitosan rapidly loose the expression of type I and immature type II collagen while they expressed mature type II collagen and aggrecan. In these conditions, chondrocytes maintained their phenotype for as long as 45 days, thus forming cartilage-like nodules. Taken together, these data suggest that a chitosan hydrogel does not work as a scaffold, but could be considered as a decoy of cartilage ECM components, thus favoring the binding of chondrocytes to chitosan. Such a biological response could be described by the concept of reverse encapsulation.

Effect of relaxin on the phenotype of collagens synthesized by cultured rabbit chondrocytes.

The effect of porcine relaxin on rabbit articular and growth plate chondrocytes in primary culture was investigated by measurement of total collagen production and analysis of the phenotypes of newly synthesized collagen chains. A 24-h treatment of monolayer articular and multilayer growth plate chondrocytes with 2 micrograms per ml relaxin had no effect on total DNA and did not significantly modify the amount of [3H]proline-labelled collagen chains secreted by the cells. However, polyacrylamide gel electrophoresis demonstrated relevant modifications in relaxin treated chondrocytes. A significant increase was observed in the proportion of type III collagen and in the intensity of the band corresponding to alpha 2I chains. Two-dimensional peptide mapping of CNBr-cleaved molecules indicated that the band that was identified as alpha 1II on monodimensional gels contained a significant proportion of alpha 1I collagen chains, as demonstrated by the presence of alpha 1I cyanogen bromide-digested peptides. The intensity of this band was increased by relaxin treatment. Furthermore, total RNA analysis by slot blot and Northern blot techniques showed a dose-dependent stimulation of alpha 1I and alpha 1III mRNA levels after incubation with increased relaxin concentrations, but no change in the amount of alpha 1II mRNA. These results suggested that when added to cartilage cells in vitro, relaxin modulated the expression of type I, type II and type III collagen genes by amplifying the dedifferentiation process.

Characterization of insulinlike growth factor I produced by fetal rat pancreatic islets.

Pancreatic islets were prepared from 22-day-old rat fetuses. After 5 days of culture in dishes allowing cell attachment, neoformed islets were kept free floating in RPMI-1640 medium (16.5 mM glucose, 1% fetal calf serum). The islets were then pulsed with [3H]leucine and [35S]methionine for 24 h. The conditioned medium was acidified with acetic acid (final pH 2.7), desalted, concentrated, and gel filtered on Bio-Gel P100 in acid conditions. The radioactive material that comigrated with immunoreactive insulinlike growth factor I (IGF-I) produced by the islets was pooled, concentrated, and further characterized by reverse-phase high-performance liquid chromatography on a C18 Bondapak column with a linear gradient of acetonitrile (20-80%). The radioactive material that eluted as pure IGF-I (40% acetonitrile) was further studied by chromatofocusing on a Pharmacia PBE 94 column. A sharp radioactive peak containing [3H]leucine and [35S]methionine was eluted at pH 8.55. This material was immunoprecipitated with an antiserum to IGF-I. This study demonstrated that fetal islet cells synthesize molecules that are, by several criteria, equivalent to native IGF-I.

Peroxisome proliferator-activated receptor gamma and its ligands in controlling interleukin-1beta target gene expression: a confusing story.

Antiinflammatory effects by peroxisome proliferator-activated receptor gamma (PPARgamma) agonists have been previously reported. However, PPARgamma dependency and the molecular mechanism involved in these effects require more investigation to clearly demonstrate whether PPARgamma is a key modulator of the antiinflammatory process. This would permit the design of more specific agonists or antagonists able to address the gamma subtype without cross reactions with other transcription factors, thus preventing undesirable side effects. However, several hurdles need to be taken into consideration, such as the coexpression of several PPAR isotypes in the same cell type. As PPARgamma and -alpha seem to play equal antiinflammatory roles, determining the subset of specific PPAR subtype target genes appears to be crucial. The work described here is our current understanding of the modulations of interleukin-1 target gene expression by PPARgamma and its ligands.

Evidence for a direct in vitro action of sex steroids on rabbit cartilage cells during skeletal growth: influence of age and sex.

A direct effect of sex steroid hormones on in vitro cartilage cell metabolism was demonstrated. Cells were derived from rabbit fetuses on day 20 of gestation, and from male and female rabbits aged from 2 to 80 days. Testosterone (T), dihydrotestosterone (DHT), or 17 beta-estradiol (E2) (10(-9) -10(-9) M) were added to primary culture of epiphyseal articular chondrocytes. They showed an age-dependent stimulatory effect on [35S]sulfate incorporation into newly synthesized proteoglycans. In cultured rabbit fetal cartilage cells, the maximum active concentration of T and DHT was 10(-9) M with a 40% stimulating effect over control values. E2 was even more active with 80% stimulating effect when added at 10(-8) M. Chondrocytes from animals aged up to 5 days responded poorly and those from animals aged 5-30 days not at all. The response of cells from older animals varied with animal age and sex. T and DHT stimulated chondrocytes from males aged 32-55 days and females aged 40-52 days to about the same extent. E2 stimulated cells from animals of the same ages, but the response of female-derived cells was twice that of male-derived cells. The stimulating effect was dose dependent from 10(-11) to 10(-8) M and maximal at 10(-9) M for T and DHT and at 10(-8) M for E2. Puromycin completely abolished the effect.

Vitamin D and cartilage. I. In vitro metabolism of 25-hydroxycholecalciferol by cartilage.

In the present work, the capacity of cartilage to metabolize 25-hydroxycholecalciferol was investigated. Cartilage preparations from growth plate, articular surface, rib, scapula, and ear were isolated from 3-week-old normal rabbits and chickens. Each tissue was separately incubated with tritiated 25-hydroxycholecalciferol (, x 10(-9) M) for 1-24 h. Incubations of kidney and muscle were performed simultaneously for comparison. Similarly, cultured chondrocytes isolated from rabbit growth plate and articular cartilage were incubated for 1 or 20 h in medium free of fetal calf serum. After methanol-chloroform extraction of tissues, cells, and their respective media, chloroform phases were chromatographed on Sephadex LH-20 columns. The results show that kidney and cartilage are able to convert 25-hydroxycholecalciferol into a derivative which migrates in the 24,25-dihydroxycholecalciferol region. Cartilage tissue previously boiled is unable to metabolize 25-hydroxycholecalciferol. The conversion of 25-hydroxycholecalciferol occurs with all types of cartilage and is also observed in incubations of cultured chondrocytes. In the latter, the polar 25-hydroxycholecalciferol derivative is detected as early as 1 h after addition of 25-hydroxycholecalciferol. Two findings suggest that the polar derivative of 25-hydroxycholecalciferol produced by cartilage is 24,25-dihydroxycholecalciferol: 1) the cartilage derivative and 24,25-dihydroxycholecalciferol (synthetic and biosynthetic) comigrate during Sephadex LH-20 and high liquid pressure chromatography; and 2) both the cartilage derivative and 24,25-dihydroxycholecalciferol are sensitive to periodate treatment.

Isolation and characterization of renin-producing human chorionic cells in culture.

Chorionic tissue is one of the major extrarenal sites of renin production, and as such, cultured chorionic cells are a potential model for in vitro studies of renin biosynthesis and regulation. Human chorionic cells were isolated from four chorions and maintained in tissue culture for a total of eight subcultures. Total renin production was considerable in the primary cultures, but fell gradually with successive passages. The cells could be frozen and thawed without losing their ability to divide or produce renin. Both the primary cultures and the subcultures contained a single type of elongated cell containing abundant rough endoplasmic reticulum and myofibrils, but no renin granules, suggesting that the cells had smooth muscle-like features. Immunocytochemistry indicated that they contained both renin and prorenin. The renin produced by the chorionic cells was not stored within the cells, but was released rapidly into the medium. More than 95% of the renin produced was prorenin, which, after activation, had biochemical and immunological properties similar to those of pure human renin. The cells contained a renin mRNA that had the same size as that for renal renin (1.6 kilobases), confirming the synthesis of renin by these cells. The cells were also examined for the presence of other components of the renin-angiotensin system. Angiotensinogen and angiotensin I were not detected, but angiotensin-converting enzyme was present in extracts of primary and secondary cultured cells. beta hCG and progesterone were also found in the medium of primary culture. However, the production of beta hCG and progesterone fell after the primary culture, and beta hCG and progesterone were indetectable in secondary and tertiary cultures, respectively. These experiments suggest that these two hormones do not influence renin synthesis or vice versa. Thus, these cultures of human chorionic cells synthesized considerable quantities of prorenin and can provide a permanent source of nonrenal prorenin-producing cells.

Partial purification, characterization, and assay of a slightly acidic insulin-like peptide (ILAs) from human plasma.

An insulin radioreceptor assay (RRA) using human placental microsomal membranes was used to measure insulin-like activity (ILA) extracted from human plasma concentrates (Cohn fraction IV-4) by acid ethanol. The soluble activity (ILAs), chromatographed on Sephadex G-75 in 1 M acetic acid, migrated as a small molecule (fractional elution volume, 0.56) ahead of insulin (fractional elution volume, 0.70), whereas at neutral pH, ILAs migrated as a large molecular weight species. The ILAs peak from acid gel filtration on Sephadex was further purified by chromatography on carboxymethyl cellulose (CMC). The ILAs peak from both Sephadex and CMC diluted parallel to the porcine insulin standard in the insulin RRA and was totally unreactive in an insulin RIA. The CMC-purified material was iodinated and purified by binding to and elution from human placental membranes. The binding of [125I]ILAs to human placental membranes was inhibited only minimally by insulin and proinsulin and not at all by epidermal growth factor, nerve growth factor, glucagon, or lactogenic hormones, including human growth hormone. Multiplication-stimulating activity (MSA) inhibited in a manner parallel to ILAs. A Scatchard plot of the binding data was nonlinear. Sephadex ILAs was subjected to isoelectric focusing. The fractions assayed in both insulin and ILAs RRAs yielded comparable results. Peaks of ILA were observed at pHs 5.3, 6.6, and 8.4. When CMC-ILA was subjected to isoelectric focusing in polyacrylamide, a single peak of activity migrating between pH 6.2-6.8 was seen. [125I]ILAs focused at exactly the same pH. Electrophoresis of CMC-ILAs in acid-urea revealed a sharp peak of activity migrating with one of the five protein bands seen after staining. Again, [125I]ILAs comigrated with unlabeled ILAs. The molecular weight of ILAs, as determined on a calibrated Sephadex G-150 column at neutral pH, was 9,000-10,000 daltons. CMC-ILAs stimulated [14C]glucose incorporation into triglycerides of rat adipose tissue and augmented [3H]thymidine incorporation into human fibroblasts, chicken embryo fibroblasts, and BALB 3T3 cells as well as [35S]sulfate incorporation into macromolecules of rabbit chondrocyte culture medium. In summary, ILAs isolated on the basis of a RRA for insulin is a slightly acidic peptide with some of the biological activities expected of a somatomedin.

In vitro biosynthesis by articular chondrocytes of a specific low molecular size proteoglycan pool.

Proteoglycans synthesized by articular and epiphyseal chondrocytes in culture were compared. Proteoglycans extruded by the two types of cells into the culture medium are of identical Mr. On the other hand, the proteoglycans of cells or pericellular matrix synthesized by the articular chondrocytes are characterized by an heterogeneous fraction of low-Mr which is not present in the material derived from epiphyseal chondrocytes. There are at least two components in this fraction: the first seems to be a precursor of aggregated proteoglycans, the other may represent a component of cell coat. Stimulation of the cell cultures with vitamin D metabolites and somatomedin enhances proteoglycan biosynthesis but no modification is observed in the proteoglycan Mr.

Insulin-like growth factors counteract the effect of interleukin 1 beta on type II phospholipase A2 expression and arachidonic acid release by rabbit articular chondrocytes.

Interleukin 1 beta was found to stimulate arachidonic acid release, and the synthesis and secretion of type II phospholipase A2 by rabbit articular chondrocytes in vitro. Interleukin 1 beta had no effect on the level of cytosolic phospholipase A2 mRNA. Insulin-like growth factors, which help stabilize the cartilage matrix, reduced the effect of interleukin 1 beta on type II phospholipase A2 activity and mRNA level, and decreased the Interleukin 1 beta-stimulated arachidonic acid release to the basal values. This suggests that type II phospholipase A2 plays a key role in arachidonic acid release from rabbit articular chondrocytes and that insulin-like growth factors counteract the effect of interleukin 1 beta. They may therefore be considered as potential antiinflammatory agents.

Characterization of specific insulin-like growth factor (IGF)-I and IGF-II receptors on cultured rabbit articular chondrocyte membranes.

The interaction of insulin-like growth factor (IGF)-I and IGF-II with specific type-I and -II receptor sites on rabbit articular chondrocyte membranes was studied. With labelled IGF-I as tracer, half-maximal displacement of the label was obtained with 1.4 ng IGF-I/ml and 22 ng IGF-II/ml. Using IGF-II as labelled peptide. 16 ng unlabelled IGF-II/ml and 200 ng IGF-I/ml were needed to inhibit the binding by 50%. Covalent cross-linking experiments revealed the presence of typical type-I (Mr 130,000 under reducing conditions) and type-II (Mr 260,000) receptor sites. In addition, with 125I-labelled IGF-II a very intense labelled band appeared at Mr greater than 300,000. This band was not found in mouse liver membranes and human placental membranes.

Inhibition of chondrocyte cathepsin B and L activities by insulin-like growth factor-II (IGF-II) and its Ser29 variant in vitro: possible role of the mannose 6-phosphate/IGF-II receptor.

Lysosomal enzymes and IGF-II both bind to the mannose 6-phosphate (M6P)/IGF-II receptor. This receptor targets newly synthesized lysosomal enzymes to lysosomes. The functional meaning of IGF-II binding to this receptor is not well known. We have postulated that IGF-II, the Ser29 IGF-II variant (vIGF-II) and IGF-I on lysosomal cathepsin B and L activities from post-natal rabbit chondrocytes in vitro. This effect was compared with the ability of each peptide to stimulate chondrocyte-sulfated proteoglycan synthesis. The sulfating dose-response relationship of the IGF peptides corresponded to their relative binding affinities for the type I-IGF receptor (IGF-I > IGF-II > vIGF-II). The intracellular cathepsin B and L activities were inhibited in a time- and dose-dependent manner by IGF-II or vIGF-II. Maximal inhibition of cathepsin B and L activities (40 and 30% below controls, respectively) was found after an 8 h treatment with 100 ng/ml IGF-II or vIGF-II. By contrast, IGF-I up to 1 micrograms/ml or insulin up to 2 micrograms/ml had no inhibitory effect. The relative potency pattern corresponded to the binding profile of each ligand for the M6P/IGF-II receptor. A treatment of chondrocytes with IGF-I or insulin transiently increased the binding of radiolabelled IGF-II at the cell surface to approximately 120% of controls, whereas IGF-II or vIGF-II had no effect. Thus, it is unlikely that the inhibition of lysosomal enzyme activities by IGF-II peptides could result from a redistribution of M6P/IGF-II receptors from intracellular compartments to the plasma membrane. We hypothesize that internalized IGF-II peptides could occupy the intracellular M6P/IGF-II binding sites required for targeting of cathepsins B and L to lysosomes.

Testosterone metabolism in prepubertal rabbit cartilage.

Using thin-layer chromatography, celite column chromatography and recrystallization methods, articular (AR) and growth plate (GP) cartilage tissues and cells from prepubertal rabbits were shown to convert testosterone (T) into at least three main metabolites: dihydrotestosterone (DHT), delta 4-androstenedione and androstanediols. In tissue incubation experiments the amount of each newly formed metabolite per mg of tissue was always greater in AR than in GP cartilage. After a 24 h incubation with AR or GP cartilage tissues, T was mainly converted to DHT and delta 4-androstenedione in approximately equal amounts. The amount of androstanediol metabolites formed was much lower. In a time-course experiment, the conversion of T to DHT and delta 4-androstenedione was shown to increase in a linear fashion, while the conversion to androstanediols was more variable. Using cultured AR cartilage cell incubations, similar results were obtained. In addition, DHT was shown to be the sole metabolite which accumulated in the cellular pool during the first 3 h incubation, as well as during the 24 h incubation when maximum cellular uptake of radioactivity was observed. At this time, the intracellular amount of unmetabolized [3H]T (88 pmoles/100 micrograms DNA) was similar to the amount of [3H]DHT (70 pmoles/100 micrograms DNA) accumulated in the chondrocytes. For both delta 4-androstenedione and androstanediols, 99% of radioactivity was extracted from the incubation medium.

Bone and cartilage responsiveness to sex steroid hormones.

Gonadal steroids influence the skeletal growth and metabolism both during the pubertal growth spurt and in adulthood with aging. It is now generally agreed that sex steroid effect on skeletal tissues is due to indirect and direct actions. In this presentation, in vitro effects of sex steroids on cartilage cells are reported by comparison with those observed on bone cells.

Age-dependent responsiveness of rabbit and human cartilage cells to sex steroids in vitro.

Rabbit epiphyseal cartilage tissue has been shown to convert testosterone (T) to dihydrotestosterone (DHT). In this report, the metabolic conversion of T into DHT is shown to be age-dependent, being most active in cartilage from animal at the age of gonadal maturation. Human cartilage from newborn and prepubertal children is also shown to convert T into DHT and--to a lesser extent--to estradiol. Low concentrations of DHT and 17 beta-estradiol (E2) (10(-11)-10(-9) M) were also shown to stimulate in vitro cartilage cells from boys and girls respectively. As previously shown for cultured rabbit chondrocytes, the stimulating effects of both hormones on human chondrocytes was age-dependent. Cartilage cells derived from children up to one year old did not respond, while cells from boys and girls in the early phase of puberty responded best. These data indicate that human cartilage tissue in vivo, contains both 5 alpha-reductase and aromatase activities during post-natal skeletal growth. Androgens may act on cartilage after their metabolic conversion to estrogens. The mechanism of age-dependency of both cartilage androgen enzymatic activities and chondrocyte responsiveness to sex steroids in vitro remains to be explained.

Phenotypic characteristics of rabbit intervertebral disc cells. Comparison with cartilage cells from the same animals.

STUDY DESIGN: Intervertebral disc cells were extracted from the surrounding matrix, and their metabolic activities and phenotypes were studied. OBJECTIVES: To compare the metabolic activities and phenotypes of cell populations extracted from the intervertebral discs of young rabbits with those of articular and growth plate chondrocytes from the same animals. SUMMARY OF BACKGROUND DATA: The phenotype of intervertebral disc cells has been poorly studied and still is debated. METHODS: The intervertebral discs as well as articular and vertebral growth plate cartilage of rabbits were digested enzymatically. The morphology of freshly isolated cells was examined. Their contents of collagen II and X mRNAs were determined by Northern blot analysis, and their sulfation activity by 35S-sulfate incorporation as chondrocytic markers. Cells were cultured at high density or low density and grown in primary culture. The stability of their phenotype was monitored by evaluating the collagen I and II mRNA ratio. The proteoglycans newly synthesized by the cells also were quantified, and their elution profile analyzed on Sepharose 2B columns. RESULTS: The anulus fibrosus cells were morphologically undistinguishable from articular chondrocytes. The nucleus pulposus contained mainly large vacuolated cells and a few smaller cells. All freshly extracted cells expressed different levels of collagen II mRNA. Anulus fibrosus and nucleus pulposus cells contained, respectively, 22% and 8% of collagen II mRNA compared with that found in articular or growth plate chondrocytes from the same animal. Only growth plate chondrocytes expressed collagen X. When anulus fibrosus cells were incubated for 48 hours at high density, they had collagen II mRNA contents similar to those of articular and growth plate chondrocytes, but synthesized five to six times fewer sulfated proteoglycans. When seeded at low density, anulus fibrosus cells divided more slowly than articular chondrocytes and incorporated four times fewer 35S-sulfate into proteoglycans. Their collagen II mRNA content was 2.75-fold lower than that of chondrocytes, and the procollagen alpha 1II/alpha 1I mRNA ratio was 3.1 for anulus fibrosus cells and 7 for chondrocytes. No collagen X mRNA was detected. When incubated for 48 hours at high density, the nucleus pulposus giant cells had four times less collagen II mRNA content than cartilage cells but synthesized the same amounts of sulfated proteoglycans. They did not divide during 21 days in culture and still contained collagen II mRNA but no collagen X mRNA. CONCLUSIONS: Findings showed that intervertebral disc cells all express cartilage-specific matrix proteins with quantitative differences, depending on their anatomic situation. It is suggested that anulus fibrosus cells are chondrocytic cells at a different stage of differentiation than articular and growth plate chondrocytes. The phenotype of nucleus pulposus cells still is unclear. They could be chondrocytic or notochordal. A definitive answer to this important question requires differentiating markers of notochordal cells.

Sensitivity of anulus fibrosus cells to interleukin 1 beta. Comparison with articular chondrocytes.

STUDY DESIGN: Anulus fibrosus cells from rabbits were grown in primary culture 1) to study their ability to produce prostaglandin E2 and Type II phospholipase A2, and to express stromelysin-1 messenger ribonucleic acid; and 2) to study the effect of interleukin 1 beta on this production and on proteoglycan aggregation. OBJECTIVES: To investigate the potency of anulus fibrosus cells to respond to interleukin 1 beta by producing degradative and inflammatory agents as compared with the potency of articular chondrocytes in the same animal. SUMMARY OF BACKGROUND DATA: Interleukin 1 beta has been implicated in the degradation of intervertebral discs. The way anulus fibrosus cells differ from articular chondrocytes in their responses to interleukin 1 beta remains to be established. METHODS: Anulus fibrosus cells and articular chondrocytes were obtained from young rabbits, grown in primary culture, and incubated with interleukin 1 beta. The newly synthesized proteoglycan was measured by labeling with [35S]-sulfate. Proteoglycan aggregation was analyzed by the elution profile on Sepharose 2B columns. The contents of collagen Type II and stromelysin-1 messenger ribonucleic acid were assessed by Northern blot analysis. The Type II phospholipase A2 activity was measured using a fluorometric substrate. Prostaglandin E2 production was evaluated by radioimmunoassay. RESULTS: Anulus fibrosus cells had 2.5-fold less Type II collagen messenger ribonucleic acid than articular chondrocytes, and interleukin 1 beta had no significant effect on this. Anulus fibrosus cells synthesized and secreted four-fold less proteoglycan than articular chondrocytes. Interleukin 1 beta reduced the anulus fibrosus content of total [35S]-sulfated proteoglycan by 35% (P < 0.01), and that of articular cells by 41% and decreased proteoglycan aggregation. Interleukin 1 beta induced the production of stromelysin-1 messenger ribonucleic acid in both cell types. The stromelysin-1 messenger ribonucleic acid content of anulus fibrosus cells was one half that of articular cells. Interleukin 1 beta increased the production of prostaglandin E2 and caused a dose-dependent secretion of Type II phospholipase A2 activity in both cell types. Its effect was 2.5-fold lower in anulus fibrosus cells than in articular chondrocytes. CONCLUSION: Anulus fibrosus cells can be stimulated by interleukin 1 beta to produce factors implicated in local degradative and inflammatory processes. This production is associated with decreased proteoglycan aggregation. Anulus fibrosus cells respond slightly less well to interleukin 1 beta in vitro than do articular cells.