Article Text


Effect of interleukin 17 on proteoglycan degradation in murine knee joints


OBJECTIVE To evaluate the effect of murine interleukin 17 (IL17) on cartilage catabolism and joint inflammation by direct intra-articular injection of the cytokine into murine knee joints.

METHODS Knees of normal C57 Bl mice were injected once or repeatedly with recombinant IL17 or IL1β. Inflammation was estimated by technetium-99m pertechnetate (99Tc) uptake and histological scoring of tissue sections. Proteoglycan depletion was evaluated by histological scoring of safranin O stained sections. Effects on proteoglycan synthesis were studied by 35SO4incorporation.

RESULTS A single intra-articular injection of IL17 (10 ng/knee) produced effects very similar to those of IL1β (10 ng/knee). No inflammation was detected at six or 24 hours by 99Tc uptake. However, safranin O staining showed depletion of proteoglycan at 48 hours. Repeated injections of IL17 induced joint inflammation and cartilage proteoglycan depletion as shown by histological scoring. Unlike IL1β, proteoglycan depletion induced by IL17 seemed to be the result of increased degradation only, as no suppression of35SO4 incorporation was seen.

CONCLUSION These findings confirm, in vivo, the catabolic effects of IL17 on cartilage. IL17 is thus the first T cell cytokine showing a direct catabolic effect on cartilage in addition to stimulatory effects on macrophages and synoviocytes, making it a potentially important cytokine in the pathogenesis of arthritis.

  • interleukin 17
  • proteoglycan degradation
  • murine cartilage
  • knee joints

Statistics from

Interleukin 17 (IL17), is a glycosylated homodimeric protein of 150 amino acids cloned from an activated T cell line,1 showing about 57% of homology to the open reading frame of a lymphotrophic virus, Herpesvirus saimiri.1-3IL17 is secreted only by activated memory CD4+ T cells,1 4 but its receptor is widely distributed, in particular in tissues of mesenchymal origin.2 3 5

T cells have a critical role in rheumatoid pathogenesis. IL17 was found to be highly produced by rheumatoid, but not osteoarthritic, synovium.6 Furthermore, most of the T cells within rheumatoid synovium show a Th1 pattern of cytokine production.7 8 Interestingly, only the Th1/Th0, but not Th2, subsets of CD4+ cell clones isolated from rheumatoid synovium produced IL17.9

In rheumatoid arthritis, T cells, macrophages, and mesenchymal cells all interact through autocrine, paracrine, and cell-contact pathways. IL17 may have an important role in this communication and interaction. It has been shown that IL17 can induce stromal cells to produce proinflammatory cytokines such as IL6, IL8, granulocyte colony stimulating factor, and prostaglandin E2.10IL17 induces IL6 and leukaemia inhibitory factor production by rheumatoid arthritis synoviocytes in a dose dependent manner, similarly to, but slightly less efficiently than, IL1β,11 and stimulates the production and expression of IL1β and tumour necrosis factor α by human macrophages.12

Other in vitro studies have also shown that IL17 could potentially initiate a matrix degradative response directly in cartilage through the up regulation of nitric oxide, stromelysin, or IL1β in chondrocytes.13 14

Collectively, these observations suggest that IL17 may be an important factor in the initiation or maintenance of the T cell-dependent inflammation and cartilage matrix degradation in rheumatoid arthritis. The present in vivo study assessed the effects of IL17 on cartilage catabolism after intra-articular injection of the cytokine.

Materials and methods


C57BL/Ola mice, 8 to 12 weeks old, were used for all experiments. The mice, originally obtained from Dr Carmeliet's laboratory (University of Leuven, Belgium), were bred in the animal facility at the Centre Hospitalier Universitaire Vaudois. The animals were fed a standard commercial diet and tap water freely. All animal procedures were done following standard procedures as required by our institution.


Recombinant murine IL17 was kindly supplied by Dr Satwant Narula (Schering-Plough Research Institute, Kenilworth, NJ). Recombinant murine IL1β was purchased from R&D Systems (Abingdon, UK).

A total volume of 5 μl containing either recombinant IL1β or IL17 in phosphate buffered saline (PBS) was injected through the infrapatellar ligament into the left knee joint space. The right knee was injected with 5 μl of PBS as control. Briefly, mice were sedated with methoxyflurane (Metofane, Arovet, Zollikon, CH), knees shaved with scissors, and the patellar ligament visualised through a small skin incision. Knees were then injected through the infrapatellar ligament using a 30G needle.


Acute joint inflammation was determined using the technetium-99m pertechnetate (99Tc) uptake method as originally described by Lens et al.15 Briefly, animals were sedated by intraperitoneal injection of pentobarbital (50 μg/g of body weight) and 10 mCi of 99Tc in 0.2 ml saline injected subcutaneously in the neck region. Accumulation of the isotope was determined for both knees five minutes after the injection with an external gammacounter. Severity of inflammation was expressed as a ratio between the experimental knee, injected with recombinant IL1β or IL17, and the opposite control knee injected with PBS alone. Ratios higher than 1.1 are indicative of joint inflammation.


Animals were killed, knees fixed in buffered formalin for seven days, and decalcified in 50% formic acid/15% sodium formate for six additional days. After standard embedding in paraffin, 7.5 μm saggital sections were prepared and stained with a trichrome stain—safranin O, fast green, and Weighert's haematoxylin. Sections were scored in a blinded manner using an ordinal scale for inflammation from 0 to +++, where 0 is normal and +++ is extensive inflammation and infiltration. Sections were scored for loss of safranin O staining as a measure of cartilage proteoglycan depletion using a scale from 0 to +++, where 0 represents no depletion and +++ severe and mostly complete depletion of staining in the superficial layer.


Cartilage proteoglycan synthesis was assayed by sulphate incorporation in explants of intact patellae.16 After intra-articular injection, mice where killed and patellae isolated by dissection. Patellae from mice in the identical experimental groups were pooled, washed in RPMI, and then pulsed for three hours with 25 μCi of 35SO4 (NEN, Le Grand-Saconnex, CH) in RPMI. The patellae were then washed thoroughly and fixed overnight in 4% non-buffered formalin. The next day, the patellae were decalcified for four hours in 5% formic acid and the cartilage dissected using a 1 mm diameter punch biopsy tool. The isolated cartilage was transferred to a scintillation vial containing 0.5 ml of Lumasolve Lumal LSD (Lumac LSC BV, Groningen, NL) as a solubiliser and incubated for eight hours at 60°C. Ten millilitres of scintillation fluid (Lipoluma Lipophilic LSC cocktail, Lumac LSC BV, Groningen, NL) was added after cooling and counts determined in a β scintillation counter. Results were expressed as a percentage of inhibition compared with PBS treated patellae.


Statistical analysis for inflammation and proteoglycan depletion scores was carried out with the Kruskal-Wallis rank test and Dunnett's test for multiple comparisons between groups. Comparisons of99Tc uptake and 35SO4 incorporation were carried out using an unpaired t test.



A single injection of 10 ng recombinant murine IL17 or IL1β for each knee produced similar effects. Histological evaluation at 24 and 48 hours after injection showed no oedema or haemorrhage and little leucocyte infiltration into the synovium (table 1). Neither IL1β nor IL17 produced detectable inflammation at six and 24 hours after the intra-articular injection as assessed by 99Tc uptake (table2).

Table 1

Histological scores of knee sections after a single intra-articular cytokine injection

Table 2

Ratio of technetium-99m pertechnetate uptake six and 24 hours after a single injection of IL1β or IL17

Proteoglycan depletion as assessed by safranin O staining was not evident at 24 hours after the intra-articular injection, but statistically significant depletion of staining was seen at 48 hours after injection, for both IL1β and IL17 (table 1). The absence of increased technetium uptake and the low level of inflammation render it unlikely that the proteoglycan loss is secondary to recruitment of extra-articular cells and most likely represents direct effects of IL17 on chondrocytes.


To evaluate the effects of repeated cytokine injections, knees were injected three times at 48 hour intervals with 10 or 1 ng of recombinant murine IL17 and the effects compared with repeated injections of 10 ng of recombinant murine IL1β. Mice were killed 24 hours after the last injection.

The histology of control knees showed, as expected from the trauma of repeated injections, a slight increase in inflammation compared with the controls after a single injection (table 3). More severe inflammation and loss of safranin O staining was seen in animals with repeated as compared with single cytokine injections. The lower dose of 1 ng IL17 for each knee also induced significant effects; more severe changes were seen after injection of 10 ng for each knee (table 3). Figure 1 shows representative examples of the histological changes.

Table 3

Histological scores of knee sections after multiple intra-articular injections

Figure 1

Histological changes induced by intra-articular cytokine injection. Knee joints were harvested 24 hours after the last cytokine injection. Standard 7.5 μm paraffin embedded saggital knee sections were stained with safranin O, fast green, and Weighert's haematoxylin (×40). (A and C) Control knees, minimum inflammation and proteoglycan depletion; (B and D) moderate to severe inflammation with infiltration and exudation and proteoglycan depletion in an interleukin 17 injected knee.

Further dose-response studies were performed with two injections at 48 hour intervals using doses of 1, 10, 20, and 50 ng of IL17 for each knee and compared with control IL1β (10 ng/knee). At all cytokine doses tested an inflammatory infiltration of the joint and a more severe degradation of cartilage proteoglycan was seen than after a single injection. Maximum effects of IL17 were seen at 10 ng/knee. The higher IL17 doses (20 and 50 ng/knee) did not induce more severe inflammation or proteoglycan loss (four mice per group, data not shown).


IL1β has a longlasting suppressive effect on proteoglycan synthesis after intra-articular injection.17 As IL17 has similar effects on proteoglycan degradation after single or repeated intra-articular injections, proteoglycan synthesis in IL17 treated explant cultures of intact patellae were analysed and compared with the effects of IL1β. Firstly, patellae were incubated for 24 hours ex vivo with either IL1β or IL17 at 10 ng/ml. IL1β induced the expected decrease in proteoglycan synthesis, but no inhibition was seen with IL17. The same observation was obtained in vivo after a single intra-articular injection of 10 ng of IL17 or IL1β in each knee, where proteoglycan synthesis was decreased after injection of IL1β, but not IL17 (table 4).

Table 4

Comparison of IL1β and IL17 effects on proteoglycan synthesis rate ex vivo and in vivo


IL17 and its receptor have no homology with known cytokines and cytokine receptors, suggesting that they represent new gene families. Additional members of this cytokine/receptor family have not yet been identified but are likely to exist. The characterisation of IL17 functions has shown that it can induce the expression of proinflammatory genes, such as cytokines, metalloproteinases, and inducible nitric oxide synthase, in various cell types, including mononuclear phagocytes, synoviocytes, and chondrocytes.

IL17 activates in chondrocytes the mitogen-activated protein kinases cascade and the transcription factor NF-κB, which are known for their role in the induction of inflammatory and matrix degradative responses in arthritis.14 Our study is the first analysis of the effects of IL17 induced by the in vivo administration of the cytokine. IL17 injections in normal murine knee joints reproduce many of the characteristics of IL1, a prototypic proinflammatory and matrix degradative cytokine.17-19 As reported for IL1, depletion of proteoglycan seems to result from a local action of IL17 on cartilage as joint inflammation and cell recruitment are very mild.17 On the contrary, and unlike IL1β, where prolonged suppression of proteoglycan synthesis appears as a key factor,20 our data indicate that IL17 induces proteoglycan loss primarily through increased degradation without a detectable change in biosynthesis rate.

Similar to other proinflammatory cytokines, IL17 also has an indirect catabolic effect on cartilage. IL17 can induce synoviocytes and macrophages to produce several other cytokines, and recent studies have underlined the role of IL17 in mediating interactions between T cells, mesenchymal cells, and macrophages.10-12 From this point of view, IL17 is certainly an interesting target for therapeutic intervention in rheumatoid arthritis. This cytokine, which is exclusively produced by T cells, is present in high amounts in rheumatoid synovial tissue and its inhibition may result in anti-inflammatory T cell targeted treatment. IL17 has been shown to stimulate osteoclast formation and bone resorption in patients with rheumatoid arthritis,21 22 and as the first T cell cytokine that can directly affect cartilage, bone, and induce other mediators of inflammation by stimulating mesenchymal cells, its inhibition may affect the so-called non-T cell pathways in joint destruction and prevent joint erosions. This hypothesis needs to be confirmed by the use of specific blocking agents, such as soluble IL17 receptors, in animal models of arthritis as no direct injection of a single or even combination of cytokines reproduces the complexity of the arthritic process. Such studies would certainly lead to a better understanding of the role of IL17 in the pathogenesis of rheumatoid arthritis.


We thank Véronique Péclat for her valuable technical help with the histology work. This work was supported in part by the Jean and Linette Warnery Foundation for Research on Rheumatoid Arthritis.


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.