WO2011025367A1

WO2011025367A1 - Method for culturing chondrocytes and uses for treatment of arthritis and chondral lesions

Info

Publication number: WO2011025367A1
Application number: PCT/NL2010/050505
Authority: WO
Inventors: Jahr Holger
Original assignee: Erasmus University Medical Center
Current assignee: Erasmus University Medical Center
Priority date: 2009-08-31
Filing date: 2010-08-10
Publication date: 2011-03-03
Anticipated expiration: 2012-02-29

Abstract

The relates to proliferation and prevention of dedifferentiation of chondrocytes, that can advantageously be performed, when culturing in vitro, by a method comprising culturing said cells in a medium having a physiological level of tonicity. Preferably, said medium has a tonicity of 350 — 480 mOsm, more preferably a tonicity of about 380 mOsm. In another embodiment, said physiological level of tonicity is also applied during isolation of the cells. Advantageously, the proliferation of the cells in such a method comprises passaging of the cells. In a preferred embodiment of the method of the invention, the chondrocytes are isolated from a subject having osteoarthritis or rheumatoid arthritis. Further preferred is the addition of calcineurin inhibitors and/or COX-2 inhibitors to said medium. The medium can further be used in all instances in which chondrocytic repair in vivo is needed, thus also with ACI and MACI protocols.

Description

Title: Methods for culturing chondrocytes and uses for treatment of arthritis and chondral lesions

FIELD OF THE INVENTION

The present invention relates to the field of tissue engineering, more specifically to culturing of chondrocytes for use in the treatment of chondral lesions, osteoarthritis and rheumatoid arthritis.

BACKGROUND TO THE INVENTION

Cartilage, or more precisely, hyaline articular cartilage (HAC) is a connective tissue covering the ends of bones in joints and is composed of specialized cells, chondrocytes, that produce a large amount of extracellular matrix. This matrix is crucial for the unique biomechanical properties of this tissue and it is composed of a collagen fiber network, providing tensile strength and flexibility, and abundant ground matrix, rich in proteoglycans.

The glycosaminoglycan (GAG) side chains of the proteoglycans are sulphated and responsible for a characteristic high fixed negative charge density (Lesperance, L.M. et al., 1992, J. Orthop. Res. 10:1-13), which binds mobile cations, mainly sodium cations. This binding determines the

physiological tonicity (that is, osmotic pressure) of the extracellular fluid around chondrocytes in vivo, but the tonicity indirectly also largely depends on the quality of the collagen network. Extracellular tonicity in healthy cartilage ranges between 350 and 480 mOsm (Maroudas, A., 1976, Nature 260:808-809; Urban, J./P. et al., 1993, J. Cell. Physiol. 154:262-270). This in vivo tonicity is dynamic and changes due to alterations in matrix hydration upon e.g. joint loading during daily activities. During cartilage degeneration (e.g. in osteoarthritis) the collagen matrix degrades and the GAG concentration diminishes, resulting in a severity- dependent decreased tonicity of between 280 and 350 mOsm (Maroudas, 1976, supra; Bank, R.A. et al., 2000, Arthritis Rheum. 43:2202-2210).

Chondral lesions from, for example, trauma or overuse can cause joint pain, immobility and eventually osteoarthritis (OA). Where OA sometimes has clear causes, rheumatoid arthritis (RA) is an auto-immune disease of which the cause has thus far not yet been unraveled, but which can also lead to the similar phenomena as with OA, i.e. degradation of the hyaline cartilage (in joints). Surprisingly, RA mainly affects joints in the fingers, wrists, knees and elbows, while OA is mostly seen in both the larger and the smaller, load bearing joints of the body, including the hands, feet, back, hip or knee.

The high prevalence of arthritis in diseases and injuries of the joints - 60% of all patients undergoing knee arthroscopy are diagnosed with a chondral lesion (Widuchowski, W. et al., 2007, Knee 14:177-182) - and loss of quality of life makes cartilage damage a major personal and economical burden.

Currently, therapy is directed to alleviation of the symptoms (pain, swelling, etc.) and treatment options for chondral lesions are limited. Autologous chondrocyte implnatation is currently the most developed hyaline repair technique for the knee (Brittberg, M et al., 1994, N. Eng. J. Med. 331:889-895). Characterized chondrocyte implantation, employing phenotypical pre- screening prior to implantation, has recently been proven to improve structural repair (Saris, D.B. et al., 2008, Am. J. Sports Med. 36:235-246).

The major drawback of this treatment is chondrocyte dedifferentiation during in vitro expansion for autologous chondrocyte implantation. However, it is almost inevitable in standard monolayer culture: spherical chondrocytes will gradually convert into fibroblast-like cells (Von der Mark, K. et al., 1977,

Nature 267:531-532; Schnabel, M. et al., 2002, Osteoarthritis Cartilage 10:62- 70). This morphological change is accompanied by a shift in collagen

expression towards less collagen type II (COL2) and more collagen type I (COLl). Consequently, dedifferentiated chondrocytes produce fibrocartilage in vivo, with an extracellular matrix of inferior biomechanical properties due to higher collagen (especially type I) content and less proteoglycans compared with normal hyaline cartilage. Three-dimensional culturing systems (using a scaffold which is to be seeded with cells) can partially prevent

dedifferentiation, but are labor and equipment intensive and essentially impair propagation. Chondrocyte dedifferentiation may also play a role in the pathogenesis of OA and RA, as the ability of aging chondrocytes to produce, maintain and repair the extracellular matrix is compromised (Loeser, R. F., 2009, Osteoarthritis Cartilage 17:971-979) and as COLl is shown to be present in chondrocyte clusters in fibrillated areas of late- stage OA cartilage while it is absent in healthy cartilage (Gouttenoire, J. et al., 2004, Biorheology 41:535- 542).

Very recently, it has become apparent that a hypertonic environment may reduce injury to cartilage (Amin, A.K. et al., 2010, Clin. Orthop. Relat. Res. 468:875-884), but this study used a very high tonicity of 600 mOsm.

Tew et al. (2009, Am. J. Cell Physiol. 297:C898-C906) disclosed that

SOX9 mRNA expression was significantly upregulated in freshly isolated and passaged human articular chondrocytes after exposure to hypertonic solutions of 550 mOsm, this overexpression indicating prevention of dedifferentiation.

It has, however, appeared that at the indicated hypertonicity, the proliferation of chondrocytes is inhibited.

Current commercial treatment protocols, generally indicated by the acronym ACI (autologous chondrocyte implantation) or MACI (matrix-assisted ACI), such as the FDA approved CARTICEL® protocol of Genzyme Corp., still suffer from the low yield of autologous chondrocytes, therefore needing ex vivo expansion, and the lack of preservation of the chondrocyte phenotype, leading to production of fibro-cartilage in vivo. A single surgical intervention protocol in which chondrocytes are co-cultured with chondrocyte- mesenchymal stem cells has been developed proposed by CellCoTec, to enhance the expansion of chondrocytes. Thus, there is still need for an improved method for producing

(autologous) chondrocytes that are suitable for ACI and/or MACI treatment.

SUMMARY OF THE INVENTION

The current inventor now have found that proliferation and prevention of dedifferentiation of chondrocytes, can advantageously be performed, when culturing in vitro, by a method comprising culturing said cells in a medium having a physiological level of tonicity. Preferably, said medium has a tonicity of 350 - 480 mOsm, more preferably a tonicity of about 380 mOsm.

In another embodiment, said physiological level of tonicity is also applied during isolation of the cells. Advantageously, the proliferation of the cells in such a method comprises passaging of the cells. In a preferred embodiment of the method of the invention, the chondrocytes are isolated from a subject having osteoarthritis or rheumatoid arthritis.

In another embodiment, additionally a calcineurin inhibitor is added to the culture medium, and preferably said calcineurin inhibitor is chosen from the group consisting of FK506 (tacrolimus/Prograft®), FK520, cyclosporine, tacrolimus, ascomycin and pimecrolimus.

In a further embodiment, additionally a COX-2 inhibitor is added to the culture medium, preferably chosen from the group consisting of celecoxib, rofecoxib, valdecoxib, lumiracoxib, meloxicam, tramadol, etoricoxib, parecoxib, valdecoxib, licofelone, NS-398 and sulphonanilides such as nimesulide. Also preferable, the presence of a COX-2 inhibitor in the medium causes reduction and not total inhibition of the COX-2 activity of the cells in the medium.

Another embodiment of the present invention is a method for autologous chondrocyte implementation (ACI) or matrix-assisted ACI (MACI) in which chondrocytes are proliferated with a method according to the invention. A further embodiment of the present invention is a method for ameliorating osteoarthritis (OA) or rheumatoid arthritis (RA) by intra- articular washing a joint of a subject suffering from said OA or RA with a medium as defined above, preferably wherein said medium also comprises chondrocytes or (osteo-chondro) progenitors.

In an alternative embodiment the invention comprises the use of a medium as defined above for treatment of OA or RA. A further embodiment is the use of a medium as defined above for intra-articular injection in a joint after surgery of said joint. In these uses, preferably said medium further comprises chondrocytes, preferably autologous chondrocytes.

A further embodiment of the present invention is a method for isolating chondrocytes from cartilage tissue comprising use of a medium as defined above 1 as buffer. Accordingly, also part of the invention is the use of a medium as defined above as a buffer medium for isolation of chondrocytes from cartilage tissue.

In a further embodiment the invention comprises the use of a medium as defined above for the culture of mesenchymal stem cells or other chondrocyte progenitor cells.

LEGEND TO THE FIGURES

Figure 1. Hypertonic isolation and expansion increased marker gene expression in osteoarthritis human articular chondrocytes. Relative expression of (a) AGCl, (b) SOX9, (c) COL2 and (d) COL2:COL1 ration in primary (PO) and passaged (Pl to P3) chondrocytes cultured at 380 mOsm compared with 280 mOsm. (e) COL2 protein expression and (f) COLl protein expression in PO and Pl osteoarthritis human articular chondrocytes (HACs). Protein levels normalized to α-tubulin. Data are mean ± standard deviation, n=12.

Differences from cells cultured a 280 mOsm are indicated, *P<0.05, **P<0.01 and ***P<0.001. Figure 2. Hypertonic isolation and expansion increased marker gen expression in non-osteoarthritic human articular chondrocytes. Relative expression of (a) AGCl, (b) SOX9, (c) COL2 and (d) COL2:COL1 ration in primary (PO) and passaged (Pl to P3) non-osteoarthritic human articular chondrocytes (NHACs) cultured at 380 mOsm compared with 280 mOsm. (e) COL2 protein expression and (f) COLl protein expression in PO and Pl non- osteoarthritic human articular chondrocytes. Protein levels normalized to α- tubulin. Data are mean ± standard deviation, n=6. Differences from cells cultured a 280 mOsm are indicated, *P<0.05, and **P<0.01.

Figure 3. Hypertonic conditions activate nuclear factor of activated T-cells 5 in osteoarthritis human articular chondrocytes. Relative expression of (a) nuclear factor of activated T-cells 5 (NFAT5) and its target genes (b) S100A4 and (c) SLC6A12 in primary (PO) and passaged (Pl to P3) chondrocytes cultured at 380 mOsm compared with 280 mOsm. Data are mean ±standard deviation, n = 12. Differences are indicated: *P < 0.05, **P < 0.01 and ***P < 0.001.

Figure 4. Nuclear factor of activated T-cells 5 knockdown inhibits

hypertonicity-induced chondrogenic marker expression, (a) Relative expression of nuclear factor of activated T-cells 5 (NFAT5) and its target genes S100A4 and SLC6A12 in transduced chondrocytes either expressing (NFAT5 shRNA) or not expressing (control) Λ7K4T"5-specific shRNAs, 24 hours after increasing tonicity to 380 mOsm. (b) Effects 0ΪNFAT5 knockdown on chondrogenic markers AGCl, SOX9, COL2 and COLl. Data are mean ± standard deviation, n = 6. Differences from cells transduced with control virus are indicated:

*P < 0.05 and **P < 0.01.

Figure 5. Combined effects of calcineurin inhibition and physiological tonicity (hypertonicity) on collagen gene expression in osteoarthritic human

chondrocytes, (a), gene expression of collagen type I and type II, respectively under standard culture tonicity (280 mOsm, 280) and physiological, hypertonic conditions (380 mOsm, 380) in the presence or absence of FK506 (values in ng/niL). n=12, *p<0.05 compared to 280, #p<0.05 compared to 380.

(b) collagen protein expression as determined by Western Blotting (collagen type I, COLl; collagen type II, COL2; α-tubulin, TUB). Values indicate tonicity in mOsm (280 vs 380) and ng/mL of FK506 (0, 50, 500; FK), respectively. Passage number 1 and 2 are indicated by Pl, P2, respectively.

Figure 6. Combined effects of calcineurin inhibition and physiological tonicity (hypertonicity) on chondrogenic and hypertrophic marker gene expression in osteoarthritic human chondrocytes, (a, b), gene expression; (c), enzyme activity. Chondrogenic markers: aggrecan, Sox9 (); hypertrophic markers: collagen type X (COLlO) alkaline phosphatase (ALPL) and matrix

metalloproteinase-13 (MMP13). Passage number 1 and 2 are indicated by Pl, P2, respectively; n=12, *p<0.05 compared to 280, #p<0.05 compared to 380.

Figure 7. Combined effects of calcineurin inhibition and physiological tonicity (hypertonicity) on ECM marker gene expression in osteoarthritic human chondrocytes. Matrix metalloproteinases (MMP-I, -3; i.e. interstitial collagenase or fibroblast collagenase and stromelysin-1, respectively), their inhibitors (TIMP-I and -2, respectively) and aggrecanases 1 and 2 (ADAMTS-4 and -5, respectively), which are responsible for the degradation of aggrecan in cartilage. n=12,.

Figure 8. Combined effects of calcineurin inhibition and physiological tonicity (hypertonicity) on collagen protein expression of P2 osteoarthritic human chondrocytes. Western Blot of collagen type I and type II (COLl and COL2, respectively) under standard culture tonicity (280 mOsm) and physiological, hypertonic conditions (380 mOsm) in the presence or absence of FK506. TUB, α-tubulin loading control. Representative image of three independent blots, n=3. Note that COL2 is strongly up-regulated by 380mOsm in an FK506- dependent fashion, while COLl expression is down-regulated under these conditions.

Figure 9. Combined effects of calcineurin inhibition and physiological tonicity (hypertonicity) on chondrogenic marker genes (COLl, C0L2, AGCl, Sox- 9), hypertrophic marker (COLlO) and NFAT5 in Pl and P2 normal human chondrocytes in monolayer culture. n=12.,

Figure 10. Relative expression of ECM components in cartilage explants cultured in vitro at 380mOsm without or with addition of calcineurin inhibitor FK506 (+, 50ng/mL). (A), chondrogenic markers COL2, AGCl, and Sox-9 next to unwanted COLl; (B) hypertrophic marker expression (COLlO, ALPL and MMP- 13). n=6. Figure 11. ATDC5 cells specifically up-regulate COL2, but not COLl or

COLlO, in response to culture medium of 380 mOsm. Western Blot analysis, 14 days post exposure to either maintaining cells at 280 mOsm or changing to 380 mOsm culture medium. Representative images of three independent blots are shown. TUB, α- tubulin (loading control).

Figure 12. Marker gene expression by ATDC5 cells after 14 days of monolayer culture in media with different tonicities. Chondrogenic COL2al, unwanted COLlal and hypertrophic COLlOaI, next to Cox-2. n=6.Figure 13.

Proteoglycan content of ATDC5 at different culture tonicities. The figure shows Alcian Blue staining for proteoglycans (PG, e.g. aggrecan etc.), next to relative DNA amounts (Crystal- Violet cell proliferation assay). Note that PG

production per cell is maximal at 380 mOsm. Figure 14. Collagen type II and X production by ATDC5 cells with or without COX-2 inhibition at different media tonicities. Western Blot; NS398, specific COX-2 inhibitor. Proteoglycan content of ATDC5 at different culture tonicities. The figure shows Alcian Blue staining for proteoglycans (PG, e.g. aggrecan etc.), next to relative DNA amounts (Cytox- Violet cell proliferation assay). Note that PG production per cell is maximal at 380 mOsm Figure 15. Collagen type II gene expression by ATDC5 cells with or without COX-2 inhibition at different media tonicities. NS398, specific COX-2 inhibitor. Note, that the combination of 380 mOsm + NS398 increases COL2 about 300-fold. n=6.

DETAILED DESCRIPTION

Osteoarthritis (OA) is characterized by an imbalance between matrix synthesis and degradation: increased synthesis of catabolic enzymes (e.g. MMPs and ADAMTS') by hyaline chondrocytes causes collagen (i.e. mainly collagen type II, COL2) and proteoglycan (i.e. aggrecan, AGCl) depletion in the extracellular matrix (ECM). Both molecules are major structural components of cartilage's ECM: COL2 networks provide biomechanical strength, while sulfated proteoglycans ensure the characteristic high fixed negative charge density (FCD) of this tissue. The latter binds mobile cations (e.g. Na⁺) and builds the tissue's hydrostatic pressure by attracting water, which is confined by an intact collagen network limiting tissue swelling. The FCD determines the physiological extracellular tonicity of healthy hyaline articular chondrocytes in vivo to between 350 and 480 mOsm (milliosmoles per kilogram of water), while OA is associated with a severity- depending drop in tonicity to between 280 and 350 mOsm.

Next to the above-mentioned MMP and ADAMTS enzymes other molecules are known that can be used as a marker for the (de) differentiation and

proliferation of chondrocytes. Whereas (the expression of) MMPs and

ADAMTS' are generally considered to be rather catabolic markers, other molecules, such as COL2, AGCl, SOX9, and TIMPs can be used as anabolic markers for testing the 'health' and usefulness of (cultured) chondrocytes. Also, as is shown in the Examples, NFAT5 can be used as a positive marker for chondrocyte health and utility. Further, COLl, which is a marker for the unwanted fibrocartilage formation, can be used. Quantitative PCR assays for COL2, SOX9, AGCl and COLl have been reported earlier (Mandl, E.W. et al., 2004, Matrix Biol. 23:231-243). The level of the nuclear factor of activated T- cells 5 (NFAT5), which is a member of the ReI family of transcription factors, and which is upregulated under hypertonic stress, can be assayed in a PCR assay with the primers HsNFAT5_Fw: 5'-gggtcaaacgacgagattgtg-3' and

HsNFAT 5_Rv 5'-ttgtccgtggtaagctgagaa-3'. Assays and primer/probe sets for testing the expression levels of MMPs, and TIMPs are described by Nutall, R.K. et al. (2003, MoI Cancer Res. 1:333-345) and assays for ADAMTS' are commercially available (e.g. from Anaspec, Fremont, CA or Qiagen, Venlo, The Netherlands).

During standard in vitro culture in 280 mOsm medium, chondrocytes tend to rapidly loose their most important marker, COL2, the expression of which decreases with within the first two passages (expansion rounds) by about 1000-fold. As is shown in the experimental part, the expression of COL2 (and further important markers such as SOX9 and aggrecan) is preserved in medium with 380 and 480 mOsm in a tonicity- dependent manner. As is further shown in the exemplary part, we have found that moderately elevated, physiological tonicity (380 mOsm) did not affect chondrocyte proliferation, while higher tonicities dose-dependently inhibited proliferation and diminished cell viability. Physiological tonicity improved expression of chondrogenic markers and NFAT5 and its target genes, while suppressing dedifferentiation marker collagen type I and improving type II/type I expression ratios >100-fold. Effects of physiological tonicity were similar in osteoarthritic and normal (non-osteoarthritic) chondrocytes, indicating a disease-independent mechanism. NFAT5 RNA interference abolished tonicity- mediated effects and revealed that NFAT5 positively regulates collagen type II expression, while suppressing type I.

It thus seems that a tonicity of about 380 mOsm yields an optimum for the expansion of chondrocytes in vitro. With such a tonicity, the chondrocytes do not or only occasionally dedifferentiate, while they are not inhibited in their expansion (see Table 1). It follows from Table 1 that for 4 doublings of the cell count, a culture will have to be cultured on average for about 11-12 days under the currently used regimen of a tonicity of 280 mOsm. With a tonicity of 380 mOsm, which clearly inhibits dedifferentiation, the culturing time for the same amount of doubling is— on average— only slightly larger, i.e. about 13 days. A higher tonicity (480 mOsm), however, does not only inhibit dedifferentiation, but also has a dramatic impact on proliferation of the cells. For the same amount of doubling now on average a period of about 30 days is needed. It is submitted that the higher tonicity levels known from the prior art (such as Tew et al. (supra) and Amin et al. (supra) disclosing media with 600 and 550 mOsm, respectively, will not be suited for proliferation of chondrocytes, although they describe the beneficial effect of a complete block of

dedifferentiation.

Thus, the slightly hypertonic medium of the invention, having a hypertonicity of 350-480 mOsm, preferably 380-480 mOsm, more preferably 350-450 mOsm, more preferably 360-400 mOsm and most preferably about 380 mOsm, provides an ideal medium for proliferation of chondrocytes, whereby

dedifferentiation of said chondrocytes is inhibited and proliferation at 'normal' speed is maintained. The medium can thus be used as a culturing medium for the ex vivo culturing of autologous chondrocytes (to be used for any form of treatment, like ACI or MACI, of OA, RA and other arthritic conditions), but it can also be used as medium in the isolation of such autologous chondrocytes from the subject which is to be treated. This beneficial effect of the medium of the invention is specifically of use in proliferation of chondrocytes which involves multiple passaging or expansion rounds.

As is further shown in the experimental part, there is no difference between these effects with respect to the source of the chondrocytes: chondrocytes that are isolated from subjects suffering from osteoarthritis, whose cells are deemed to be in a pathological state, react similarly to the same order of tonicity with respect to the marker genes tested. This may imply that physiological tonicity, postulated to be around 380 mOsm for chondrocytes, is sensed by OA cells and normal cells in a similar fashion. However, late stage OA cells, isolated from fibrillated areas, are dedifferentiated, flattened cells that have lost their spherical shape as an integral part of the chondrocyte phenotype because of cytoskeletal changes. It appeared that exposing these cells to the medium of the invention as a redifferentiation stimulus induced a more enduring response as compared with spherical, normal chondrocytes. Thus, autologous chondrocyte implantation employing OA cells may benefit relatively more from a treatment protocol involving expansion according to the present invention.

The beneficial effects of the medium on the proliferation and

dedifferentiation of chondrocytes can yet be improved by addition of one or more calcineurin inhibitors to the medium. It is shown in the present application that addition of the FDA-approved calcineurin inhibitors FK506 or FK520 dose-dependently affected the expression of the tested markers.

Aggrecan expression was improved by 80% by application of the 380 mOsm medium and 110% by application of the medium + 50 ng/μl FK506; SOX9 expression was improved by more than 250% in both occasions and COL2 expression increased 10-fold in the medium of the invention, but >20-fold in the medium to which FK506 was added (see Fig. 5).

It has been established that the action of the calcineurin inhibitors on the expression of chondrogenic markers is asserted through endogenous TGFβ production in the chondrocytes. The proof for this has been given by addition of anti-TGFβ antibodies which appeared to block the effect of FK506 on COL2 mRNA expression completely (data not shown). The relation between calcineurin inhibition and induction of TGFβ signalling has been described earlier in several other cell types, for example in renal cells (Gooch, J.L. et al, 2007, Transplantation 83:439), vascular smooth muscle cells (Giordano, A. et al., 2008, Cardiovasc. Res. 79:519) and lymphatic B cells (Romano, S. et al., 2008, Haematologica 93:1039). For many years, TGFβ has been the most commonly used chondrogenic factor to induce redifferentiation (Mandl, 2002, supra; van Osch, G.J. et al, 2001, Plast. Reconstr. Surg. 107:433).

In a further embodiment of the present invention, also TGFβ, more

particularly TGFBl, is added to the medium and will further potentiate the beneficial effects on the proliferation and the inhibition of dedifferentiation of chondrocytes. Addition of TGFβ may be specially be useful when isolating and culturing cells from subjects suffering from OA or RA.

The observation that calcineurin inhibitors have an additional effect on the effects of physiologic tonicity when culturing chondrocytes also opens further possibilities for treatment. Since at least some calcineurin inhibitors, such as FK506, also have anti- inflammatory effects, they have been suggested for treatment of arthritis (e.g. Magari, K. et al., 2003, Br. J. Pharmacol.

139:927-934; US 2009/0074889). The current protocols for ACI and MACI, however, are generally not felt applicable when the joints to be treated are inflammated. When anti-inflammatory calcineurin inhibitors are used, they would have a double effect: treating the inflammation on one side and enhancing proliferation and inhibition of dedifferentiation of chondrocytes on the other side. Another advantage of the use of calcineurin inhibitors is that many also have an inhibitory effect on the immune system, thereby blocking or decreasing an immune response that can be caused by a scaffold seeded with chondrocytes, or by non-autologous chondrocytes when introduced in a joint for repair of chondric lesions, such as in ACI, MACI or other treatments for OA and RA.

The calcineurin inhibitor(s) may be chosen from the group of FK506 (also known as tacrolimus or Prograft®), FK520, cyclosporine A, voclosporin, ascomycin and pimecrolimus. Further, (genetically engineered) cell-permeablp peptides derived froro the family of naturaJJy occurring regulators of

calcineurin activity, so-called RCANs (Davies et al. FASEB Journal 2007, 21: 3023-28) or those described by Roehrl et al. PNAS 2004, 101 no. 20: 7554-7559) may be used. Other synthetic peptides, such as the IBP import blocking peptide (IBP) described by Hallhuber et al. (Future Cardiol. 2007, 3(l):91-8) preventing the binding of calcineurin to importin and inhibiting its nuclear shuttling may also be included. Other compounds interacting with the NFAT and calcineurin pathway are members of the family of oleanane triterpenoids such as liquidambaric acid, oleanolic acid, 3a-acetoxy-25-hydroxy-olean-12-en- 28-oic acid or lantanolic acid and more preferably 3a-acetoxy-25-hydroxy- olean-12-en-28-oic acid (IC50: 4.63mM) and lantanolic acid (IC50: 12.62mM) as the latter exhibit the strongest inhibitory activity against NFAT

transcription factors (Dat et al. Biol. Pharm. Bull. 2004, 27(3) 426—428). Also members of the pyrazolopyrimidine family such as NCI3 (Sieber et al., Eur, J. Immunol. 2007. 37: 2617-2626) or related compounds which bind to

calcineurin and cause an allosteric change interfering with NFAT activation or NFAT nuclear import inhibitors (Venkatesh et al., 2004, PNAS 101:8969- 8974). Moreover, instead of or additional to using calcineurin inhibitors, the specificity of the treatment may be increased by releasing cell-permeable NFAT-specific inhibitors like 1 IR-VIVIT (H-Arg-Arg-Arg-Arg-Arg-Arg-Arg- Arg-Arg-Arg-Arg-Gly-Gly-Gly-Met-Ala-Gly-Pro-His-Pro-Val-Ile-Val-Ile-Thr- Gly-Pro-His-Glu-Glu-NH₂; Noguchi et al. Nature Medicine 2004 10, 305 - 309; doi:10.1038/nm994; Luoma et al. J Neuroscience, 2008, 28(12):3159 -31;

Aramburu J et al. Science. 1999 Sep 24;285(5436):2129-33; WO 2003/072606, PCT/JP2003/002106). Peptides which potently inhibit NFAT activation and NFAT- dependent gene expression without affecting the expression of other cytokines that require calcineurin but not NFAT, may decrease potential side- effects of administrating calcineurin blockers. More preferably then, the chosen NFAT- specific inhibitors are able to target NFAT isoforms instead of targeting all different NFATs.

In a further embodiment of the present invention, the method of culturing chondrocytes comprises the use of a medium to which one or more COX-2 inhibitors are added. COX-2 inhibitors are known to be useful in the treatment of OA and RA (see e.g. Ehrlich, G. E., 1999, Inflammopharmacol. 7:265-267) and several of them are actually marketed for these indications.

It was very recently shown that hypertonicity induces COX-2 in renal epithelial cells (Favale, N. O. et al., 2009, Biochem. Biophys. Res. Comm.

381:301). It, however, seems that COX-2 inhibitors have a biphasic effect, where completely blocking COX-2 also would give undesired effects on growth and differentiation of the cells. It thus is proposed that only small doses of COX-2 inhibitors should be given, in order to leave at least some residual COX-2 action. It is thus preferred to consider COX-2 reduction and not total inhibition. It has been established that COX activity is involved in normal bone metabolism and that NSAIDS and COX-2 inhibitors have a negative impact on bone repair (Zhang, X. et al., 2002, J. Clin. Invest. 109:1405-1415). On the other hand, the currently presented results show that the specific COX- 2 inhibitor NS-398 more than doubles COL2 expression under hyperosmotic conditions (Fig. 15). Further, it very recently has been shown that COX-2 inhibition by NS-398 blocked chondrocyte cell death that was induced by sodium nitroprusside, a generator of nitric oxide, through the PGE2 pathway (Notoya, K. et al., 2010, J. Immunol. 165:3402-3410).

Preferably, a COX- 2 reducing compound according to the invention may be chosen from the group consisting of celecoxib, rofecoxib, valdecoxib, lumiracoxib, meloxicam, tramadol, etoricoxib, parecoxib, valdecoxib, licofelone, NS398 and sulphonanilides such as nimesulide. Since complete inhibition is not preferred, the concentration in the medium of the invention should not be excessive; and a person skilled in the art is able to choose the appropriate concentration by checking the COL2 expression and general health of the chondrocytes. If the medium is also to be used therapeutically (see below), then careful dosing is even more warranted, since it has been established in the mean time that COX- 2 inhibitors have negative effects on the cardiovascular system. If the medium is only applied locally (e.g. intra- articularly) the same considerations would be needed as for in vitro applications: a dose that is sufficient to stimulate COL2 expression, but does not have detrimental effects on the chondrocytes (or any other cells that are present in the treated area). If the medium also is able to reach other parts of the body, then cardiovascular effects should be prevented by minimizing the dose further. As illustrated above, the method for proliferation and prevention of dedifferentiation of chondrocytes according to the invention, i.e. with a medium as defined above, can be advantageously be applied in methods for autologous chondrocyte implementation (ACI) or matrix-assisted ACI (MACI) such as the CARTICEL® method of GenZyme Corp.

In another embodiment, the medium as defined above can be used for ameliorating OA or RA, or any other form of arthritis, by intra-articular washing or rinsing. Intra-articular washing of a joint of a subject suffering from arthritis is a commonly used procedure to alleviate arthritic complaints (e.g. Tanaka, N. et al., 2005, Clin. Rheumatol. 25:65-69). It is also possible in such a washing or rinsing protocol to use medium to which (autologous) chondrocytes have been added.

The medium of the invention itself, thus, can be used for treatment of arthritic complaints. Preferably the medium also comprises chondrocytes, which may be autologous or not, and which may be seeded on a scaffold or not. For treatment, the medium preferably is injected into the affected joint(s) of a subject, but it may also be applied during surgery for washing or rinsing the joint.

Also part of the invention is a method for isolating chondrocytes from cartilage tissue using the medium according to the invention. This isolation is advantageously performed during surgery (e.g. Akens, K. and Hurtig, M. B., 2005, BMC Musculoskeletal Disorders 6:23; Stoop, R. et al., 2007, Arthritis Res. Ther. 9:R60). The invention further comprises the use of a medium according to the invention for the isolation of chondrocytes from cartilage tissue of a subject.

The focus, however, lies on applications such as ACI, involving chondrocytes or progenitor cells. It has appeared (see the results in the experimental section on ATDC5 cell cultures) that the chondrocytes may also be derived from progenitor cells, such as mesenchymal stem cells. As is shown in the Examples, culturing these cells in the medium of the invention, especially a medium in which a calcineurin inhibitor is added, causes differentiation of these cells to chondrocytes, and prevents further

development of the cells to bone cells (osteo-progenitor cells). Other cell types that would benefit comprise mammalian bone marrow-derived stromal cells or MSCs when aiming at chondrogenic differentiation (see above) or

pathologically altered synoviocytes or tenocytes when aiming at treating OA or RA or tendonitis/tendinosis, respectively.

Of course, in the present invention said mesenchymal stem cells or other progenitor cells may be autologous or heterologous. It is also possible to derive suitable cells from yet less frequently used sources such as the ear (elastic) cartilage or rib (costal) or .A subject according to the invention can be a mammal, but said subject preferably is a race horse, a pet animal or, most preferably, a human being. Said subject can be a healthy subject, but it can also be a subject having an arthritic condition such as OA or RAor suffering from osteochondritis dissecans, ..

In mammals three types of cartilage exist: elastic cartilage, hyaline cartilage and fibrocartilage. All (diseases related to) these tissues would profit from the claimed invention. A method for treatment may also apply for the following surgical interventions: (stem) cell-based knee or hip or spine

(meniscal) regenerative medical surgeries. The invention may also improve chondrectomy and debridement or brasion and microfracture surgery.

Osteoarthritis, as the second leading cause of disability in the elderly population, can be also dealt with by several non- surgical treatments like e.g. using braces or drug therapies, such as anti- inflammatory substances (eg. diclofenac, ibuprofen, and naproxen— which is not very effective) or COX-2 selective inhibitors (see above) which mainly alleviate the pain. However, in the United States, Genzyme Corporation provides the only FDA approved ACI treatment: Carticel— but this treatment is designated for young, healthy patients with medium to large sized (non-OA) damage to cartilage (thus cartilaginous lesions) and not (yet) applicable to OA patients. The reason is that the patient's own chondrocytes (e.g. 1OK— IOOK cells) are removed arthroscopically from a non load-bearing area from either the intercondylar notch or the superior ridge of the medial or lateral femoral condyles. This is still difficult to accomplish, but the invention provides for an enahncement in obtaining cells that would be suitable for such a treatment.. Originally, cells were then grown in vitro at Genzyme for approximately six weeks until reaching 10-12M cells to be implanted in a second surgery under a periosteal flap to hold the cells in place. Approximately 1,500-3,000 Carticel procedures are currently being performed annually (over 10,000 in total) with costs of $20,000-$35,000 ea. CARTICEL II is a second generation procedure using a "fleece matrix" (i.e. a scaffold) into which the grown harvested chondrocyte cells are planted (i.e. MACI). CARTICEL II procedure is about to undergo clinical trials under the supervision of the FDA in the United States. BioTissue Technologies GmbH (Freiburg, Germany) moved the CARTICEL technology forward using a patient's hyaline biopsy to be grown in a 3D resorbable matrix prior to implantation via open or arthroscopic procedures (a modified, simpler MACI, without using a periosteal patch). Other companies offering similar products are FAB (Fidia Advanced Biopolymers), Geistlich

Biomaterials and Arthro Kinetics or co.don AG (URL http://www.codon.de/). The latter recently launched CHONDROSPHERE, a third generation liquid product 3D matrix, which may be optimally benefit from the claimed invention as cell spheroids (lmm diameter) are implanted through a syringe. Autologous mesenchymal stem cell transplantation protocols are currently successfully used in human pilot studies in human knees (about 400 patients, costs > $7,000 ). All aforementioned methods would benefit from the present invention, either by applicability of the medium to isolate cells, to grow cells, or to use as medium in which the matrix is being introduced into the body, as well as for washing and rinsing the area in which surgery is taking place. In addition, osteochondral autografting (OATS) is a technique that requires that the surgeon transplants sections of bone and cartilage. If this employs a bi- phasic scaffold, such a procedure may benefit from the present invention with respect to cartilage (re-)generation. Basically, all three methods of grafting cartilage defects (periosteal grafting, osteochondral grafting (mosaicplasty), and articular cartilage paste grafting) may benefit as the invention has been shown to work on cartilage explants as well [see Fig. 10]. Finally, spinal disc herniation (prolapsus disci intervertebralis) surgeries and specially, but not excluding fibrocartilagenous, load- distributing annulus fibrosus, the nucleus pulposus regenerative surgeries may benefit from the present invention.

Future treatment options for osteoarthritic degenerations of the disk

(spondylosis) may include stem cell therapy and the present invention would also be useful therein. Using injections of cell-suspensions with the medium of the invention may actually work better for all these spinal surgeries than for knee or hip surgeries due to anatomic differences and better concealments. Note that autologous mesenchymal stem cells in animal models were successful in arresting intervertebral disc degeneration. For this type of surgeries, high-end physiological tonicities (i.e. 450mOsmo) may even work better.

The therapeutic compounds that are optionally present in the medium of the invention (such as calcineurin inhibitors and COX-2 inhibitors can be given in a formulation in which they are normally administered to diseased subjects. However, to optimally profit from the therapeutic effects of said compounds, it is possible to administer slow release formulations of said compounds; e.g. employing injectable microspheres such as

TnnoCorc^'s proprietary SynBiosys™ biodegradable multi-block co-polymer platform. This platform forms the basis for the development of long-acting, injectable microparticles for systemic and local delivery of active pharmaceutical ingredients. For the present invention injectable microspheres with a size of 25-100 μm, more preferably 50-100 μm. most preferably 50-75 μm are to be used, because spheres <20um are quickly cleared from the synovia (i.e. joint space fluid) and thus compromising drug release dynamics and very small particles (i.e. 0.5-10 μm) are known to stimulate unwanted macrophage activation (i.e. inflammation). Sustained release of active components is possible from a few weeks up to 6 months according to the manufacturer. SynBiosys^m polymer platform allows encapsulating small and medium-sized pharmaceutical actives with a wide range of release profiles to create tailored drug-loaded microparticles. Said microparticles (or any other comparable slow-release pharmaceutical carrier) is advantageously used to release calcineurin inhibitors (chosen from the above-mentioned list) and Cox-2 specific inhibitors.

The person skilled in the art will know how to prepare such slow release formulations and how to include these in the medium according to the invention. EXAMPLE 1

After informed consent was obtained, human articular cartilage was explanted from macroscopically normal areas of the femoral condyles and tibial plateau of nine patients undergoing total knee replacement surgery for OA (medical ethical approval MEC2004-322). In addition to preparation of cartilage explants and isolation of HACs under standard conditions (DMEM, 280 mOsm) as described by Das and colleagues, medium tonicity was also adjusted to 380 mOsm, 480 mOsm or 580 mOsm by addition of sterile NaCl. Enzymatic digestion, removal of undigested fragments and subsequent chondrocyte culture were all reported earlier (Das, R.H. et al., 2010, J. Orthop. Res. 28:537- 545). The 280 mOsm and 380 mOsm isolations were also performed with cartilage obtained from the femoral condyles and tibial plateau of two non-OA donors (further referred to as normal donors) undergoing above-knee amputation surgery after trauma.

Chondrocyte proliferation and DNA measurements

Primary (PO), passage 1 (Pl), passage 2 (P2) and passage 3 (P3) HACs were monolayer expanded in medium corresponding to their isolation tonicity (280 mOsm, 380 mOsm, 480 mOsm or 580 mOsm), with an initial seeding density of 6,000 cells/cm2. Cells were harvested daily for cell counts and DNA assay between days 2 and 6. Experiments were performed in duplicate from three OA donors (n = 6). At each passage, growth curves were established by cell counts using Trypan Blue (catalogue number T8154; Sigma- Al drich, St. Louis, MO, USA) and DNA quantification. DNA measurements were performed according to Karsten and Wollenberger with slight modifications [41].

Doubling times within each passage were calculated from the trend line of the exponential growth phase using the equation:

j = x ( 0 ) exp( kx ( t ) ) with k = \n 2IT

where k is the growth constant and T is the doubling time. Chondrocyte expansion

Primary HACs were cultured for expansion in monolayers at a seeding density of 7,500 cells/cm2 in medium corresponding to their isolation tonicity (280 mOsm, 380 mOsm, 480 mOsm or 580 mOsm). PO cells to P3 cells were seeded in high-density monolayers (20,000 cells/ cm2) and were cultured for an additional 5 days and 7 days before analysis of mRNA (quantitative RT-PCR) and protein expression (Western blotting), respectively. Experiments were performed in triplicate from four OA donors (n = 12). Additional experiments were performed in triplicate from two healthy donors (n = 6) to investigate whether the hypertonic stress response is specific for pathologically altered cells. To exclude sodium-specific or chloride- specific effects, we performed experiments using iV-methyl-d-glucamine chloride (NMDG-Cl) or sucrose to adjust the medium tonicity to 380 mOsm.

Lentiviral NFAT 5 gene knockdown

We used lentiviral vectors for nontransient shRNA-mediated gene silencing in primary chondrocytes. BamΗI/MunI restriction fragments of the parental pLKO.l-puro vector - each containing the U 6 promotor and one out of five different, sequence-verified antihuman NFAT5 shDNAs (MISSION shRNA library) - were subcloned into corresponding restriction sites of recipient vector pRRL.PPT.PGK.GFPpre. This vector was kindly provided by L Naldini (San Raffaele Telethon Institute for Gene Therapy, Milan, Italy) and was optimized by A Schambach (Department of Experimental Hematology, Hannover Medical School, Hannover, Germany) to express enhanced green fluorescent protein (eGFP) from the phosphoglycerate kinase promoter. Lentiviral particles were produced in HEK293T cells by transient transfection using a calcium

phosphate protocol. Cells transduced with a lentiviral vector lacking the NFATδ-specific shRNA expression cassette served as controls. All cells were grown in monolayers. TRCN0000020019 to 23 for humans and TRCN0000085644, TRCN0000085647 for mouse (both from Sigma-Aldrich, Mission shRNA library; http://www.sigmaaldrich.com/life-science/functional- genomics-and-rnai/shrna/shrna-search-and-order.html) was identified as the best performing anti-NFAT5 shRNA clone by quantitative PCR-based knockdown efficiency determination, and was used in subsequent experiments. Pl OA HACs from two donors were seeded (15,000 cells/cm2) and cultured for 4 days in control medium (280 mOsm). Three hours prior to transduction, cells were deprived of antibiotics, and then were transduced for ± 18 hours, refreshed with control medium with antibiotics and cultured for an additional 4 days before harvesting for fluorescence-activated cell sorting (FACS) analyses. Cells were resuspended in PBS with 10% FCS and antibiotics, and were washed. Cells were collected and stained with Hoechst 33258 (1 mg/ml; Molecular Probes/Invitrogen Corp., Carlsbad, CA, USA) to discriminate between dead cells and live cells. FACS was performed on the FACSAria (Becton Dickinson BV, Breda, The Nederlands), and eGFP-expressing cells were collected (>50%, multiplicity of infection ~1) and reanalyzed for purity (>95%) using Cell Quest Pro Software (Becton Dickinson Biosciences BV, Breda, The Nederlands).

The eGFP-expressing populations were seeded (10,000 cells/cm2) and cultured in control medium up to 80% confluency. Cells were then switched to medium of 380 mOsm or were kept on control medium for 24 hours prior to RNA analysis.

RNA expression analysis

RNA isolation, purification, quantification and cDNA synthesis are described elsewhere (Uitterlinden, A.J. et al.,2006, Osteoarthritis Cartilage 14:250-257). Expression levels of AGCl, SOX9 and COL2 were studied as chondrogenic markers, while COLl was studied as a dedifferentiation marker. Quantitative PCR assays for COL2, SOX9, AGCl and COLl have been reported earlier (Mandl, E.W., supra). To quantify expression oi NFATδ and its target genes, the following primers were tested for similar amplification efficiency and specificity (according to Das et al., supra), and were used as 20 μl SYBRR Green I reactions: HsNFAT5_Fw, GGGTCAAACGACGAGATTGTG and HsNFAT5_Rv, TTGTCCGTGGTAAGCTGAGAA; HsSl00A4_Fw,

GTCCACCTTCCACAAGTACTCG and HsSl00A4_Rv,

TCATCTGTCCTTTTCCCCAAG; and HsSLC6Al2_Fw,

ACACAGAGCATTGCACGGACT and HsSLC6Al2_Rv, CCAGAACTCGTC TCTCCCAGAA. Data were normalized to an index of three reference genes (GAPDH, UBC, HPRTl) that were pre-evaluated to be stably expressed across samples. Relative expression was calculated according to the 2-ΔCT method.

Western blot analysis

Cells seeded at high densities were washed twice with PBS and were lysed in

RIPA buffer with addition of protease inhibitors. The total protein

concentration was quantified by the bicinchoninic acid assay according to the manufacturer's protocol (#23225; Thermo Fisher ScL, Rockford, IL, USA). Aliquots (10 μg) were subjected to 10% SDS-PAGE prior to electroblotting onto nitrocellulose membranes (Protran BA83; Schleicher & Schuell BV, s- Hertogenbosch, The Netherlands). Blots were blocked in 5% low-fat dry milk in Ix PBS, 0.05% v/v NP-40, were incubated with primary antibodies - anti-type II collagen and anti-type I collagen, both 1:100 (SouthernBiotech, Birmingham, Alabama, USA), or 1:10,000 anti-α-Tubulin (Sigma) - were washed, were incubated with secondary antibodies (both 1:1,000; Dako Cytomation,

Heverlee, Belgium) and were chemiluminescently detected. Signals were quantified using ImageJ 1.42 software.

Statistical analysis was performed using SPSS 13.0 software (SPSS Inc., Chicago, IL, USA). Data were compared between groups by Kruskall-Wallis H test and post-hoc Mann-Whitney U test. Results represent the mean ± standard deviation, and P < 0.05, P < 0.01 and P < 0.001 were considered to indicate levels of statistically significant difference. Results

Hypertonicity influences proliferation and survival of chondrocytes

We first determined the influence of tonicity on proliferation: OA HACs isolated at 580 mOsm hardly attached or proliferated, and 2 days after seeding no viable cells were recovered. At 280 mOsm, 380 mOsm and 480 mOsm, respectively, cells did adhere but increasing tonicity induced marked morphological changes: at 280 mOsm, cells appeared fibroblast-like, stretched out and flattened with long filopodia; while at 380 mOsm, cells were more sphere-shaped and had shorter filopodia. At 480 mOsm, cells showed few filopodia and appeared spherical. The differences in appearance remained throughout the de differentiation period (PO to P3), but were most apparent at earlier passages. Using cell counts and DNA assays, doubling times were calculated from growth curves established from each passage at three different tonicities (280 mOsm, 380 mOsm and 480 mOsm). Throughout

dedifferentiation, OA HACs isolated at 480 mOsm showed severely inhibited proliferation compared with cells at 280 mOsm and 380 mOsm (Table 1).

Table 1: Proliferation of chondrocytes isolated and cultured at 280 mOsm, 380 mOsm and 480 mOsm. The data are presented as relative doubling times in percentage of cells cultured at 280 mOsm, mean ±standard deviation. The absolute doubling time ±standard deviation in hours is displayed in brackets. N=6, mOsm is milliosmoles per kilogram of water. (*P<0.05)

Chondrocyte proliferation (%)

Culture Passage 0 Passage 1 Passage 2 Passage 3 condition

280 mOsm 100 (68±28 hrs) 100 (89±54 hrs) 100 (67±48 hrs) 100 (57±11 hrs)

380 mOsm 113±18 89±25 99±9 154±41

480 mOsm 675±405* 180±24* 168±28* 165±81* In contrast, doubling times of OA HACs at 280 mOsm and 380 mOsm never significantly differed (Table 1). All further experiments were therefore performed at 380 mOsm (as high tonicity condition) and compared with 280 mOsm (control condition).

Isolation and expansion of chondrocytes under hypertonic conditions improves their phenotype

Next, we set out to determine whether expansion culture in physiological tonicity improves the chondrocytic phenotype. Physiological tonicity (380 mOsm) during isolation and subsequent passaging of OA HACs significantly increased mRNA levels of both AGCl (Figure Ia) and SOX9 (Figure Ib) at all passages. In expanded P3 chondrocytes in physiological culture, AGCl levels were still higher than in unpassaged PO chondrocytes cultured under the standard culture conditions (280 mOsm). Physiological tonicity also significantly upregulated COL2 levels from 8.5-fold in PO to 11.6-fold in expanded P3 chondrocytes (Figure Ic) compared with controls. In contrast, COLl expression was significantly suppressed in physiological conditions throughout culture. Consequently, we found a significantly improved

COL2/COL1 ratio during chondrocyte expansion (Figure Id), from sevenfold in PO cells to 100-fold in expanded P3 cells. Physiological tonicity also upregulated COL2 protein expression (Figure Ie): levels significantly increased (between 1.5-fold and 2.2-fold) in PO, Pl and P2 chondrocytes.

In contrast, physiological tonicity significantly decreased COLl protein expression (Figure If), from twofold in PO cells to 13-fold in Pl cells.

Physiological tonicity also significantly increased AGCl (Figure 2a) and SOX9 (Figure 2b) mRNA levels in nonosteoarthritic human articular chondrocytes (NHACs). Furthermore, COL2 mRNA levels were significantly upregulated, from 5.8-fold in PO cells to 270-fold in expanded P3 NHACs (Figure 2c). As in OA HACs, hypertonicity also downregulated COLl expression with increasing passage number in NHACs: the COL2/COL1 ratios increased during expansion (Figure 2d), from 6.8-fold in PO cells to 355-fold in expanded P3 cells. Correspondingly, COL2 protein levels increased under these conditions (4.8-fold in Pl cells and 2.9-fold in P2 cells), while the amount of COLl diminished (by 4.7-fold in Pl cells and fivefold in P2 cells) (Figure 2e, f).

Hypertonicity activates NFAT5 in human articular chondrocytes

Compared with 280 mOsm controls, NFAT5 mRNA levels were significantly increased in 380 mOsm OA HAC cultures (Figure 3a), as was the expression of established NFAT5 target genes S100A4 (in all passages; Figure 3b)

and SLC6A12 (until P2; Figure 4c). Similar effects were found in NHACs (data not shown).

NFAT5 knockdown inhibits hypertonicity -induced chondrogenic marker expression

Upon transduction, sorted eGFP-coexpressing OA HACs were switched to 380 mOsm for 24 hours. In controls not expressing NFAT5- specific shRNAs, an approximately twofold increase in NFAT5 mRNA levels was observed upon hypertonic stimulation (Figure 3a, Pl). In contrast, likewise challenged cells expressing anti-NFAT5 shRNAs showed an approximately 75% reduction in NFAT5 levels. Following NFAT5 knockdown, the NFAT5 targets S100A4 and SLC6A12 were also no longer hypertonically inducible: S100A4 expression decreased twofold and SLC6A12 was virtually undetectable upon NFAT5 RNAi (Figure 4a), confirming a functional NFAT5 knockdown. At 380 mOsm, NFAT5 RNAi also downregulated chondrogenic markers: AGCl by 80%, SOX9 by 32% and COL2 by 84%, as compared with non-RNAi controls (Figure 4b).

Interestingly, expression of COLl increased after NFAT5 RNAi in OA HACs to

-300% of control levels (Figure 4b).

EXAMPLE 2

Human articular cartilage was explanted from macroscopically normal areas of the femoral condyles and tibial plateau of 4 patients undergoing total knee replacement surgery for OA. Human articular chondrocytes (HACs) were isolated under standard culture conditions (280 mOsm) and elevated medium tonicity (380 mθm). Primary HACs were culture expanded in monolayer (7,500 cells/cm²) at both isolation tonicities (280 or 380 mOsm). Passage 1 (Pl) and P2 cells were seeded in high-density monolayers (20,000 cells/cm²) and FK506 (0, 50 or 500 ng/ml) was added for 24 hours prior to six days of in vitro culture, followed by mRNA (QPCR), protein analysis by Western Blotting and activity assays. QPCR and Western blotting procedures are described under Example 1.

ALPL activity

Alkaline phosphatase activity in human chondrocytes was determined using the Oolorimetrie Assay Kit (ab$3369; Abeam pic, 330 Cambridge Science Park, Cambridge, CB4 OFL, UK) according to the manufacturer's instructions.

Briefly, to determine intracellular ALPL, cells from one well of a 6-well plate were washed and homogenized in the Assay Buffer, prior to centrifugation to remove insoluble material. Fifty microliters of sample was transferred into 96- well plates and 30 μl of Assay Buffer were added, prior to adding 50 μl ofpNPP substrate Solution to each sample well (and background controls). After mixing, samples were incubated for 60 min at 25⁰C, protected from light, stopped by adding 20 μl Stop Solution to each well and measured at 405 nm in a standard micro plate reader against a standard curve of pNPP. Sample readings were applied to plotted pNP Standard Curve to get the amount of pNP generated by ALP sample. ALP activity of the test samples were then calculated as relative activity and corrected for the amount of DNA per culture well. For reference see: Bessej', O.A., Lowry O. H. and Brock M.J.:(1946) J.Biol. Chem. 164 321). MMP-13 activity MMP- 13 activity in chondrocytes was determined using the SensoLyte® Plus 520 MMP-13 Assay Kit from AnaSpec (AnaSpec, 34801 Campus Drive, Fremont, CA 94555) according to the manufacturer's guidelines. Briefly, the kit is designed to specifically detect MMP-13 activity in culture medium, from which MMP-13 is captured by immobilized MMP-13 antibody, and its proteolytic activity is measured by 5-FAM/QXL™520 FRET peptide. The fluorescence of 5-FAM (fluorophore) is quenched by QXL™520 (quencher) in the intact FRET peptide. Upon MMP-13 cleavage, the fluorescence of 5-FAM is recovered and can be monitored at Ex/Em=490±20 nm/520± 20 nm. Cell culture media were collected and centrifuged for 15 min at 1000xg, at 4°C and stored at -70⁰C until use. MMP-13 standard (10 μg/mL, Component B) was diluted 1:50 in MMP dilution buffer (Component C) to get 200 ng/mL and six 1:2 serial dilutions were then prepared in MMP dilution buffer, next to a blank control. Hundred μL sample/well, MMP-13 standards, and blank control were transferred to the anti human MMP-13-coated microplate, covered and incubated on a shaker (at room temperature for 2h. Afterwards, plate was washed four times in 0.2mL of Ix Wash Buffer prior to adding 100 μL/well MMP-13 substrate solution and measuring fluorescence signals at Ex/Em=490 nm/520 nm after 2h after reactions were stopped by adding 100 μL/well stop solution (Component H) per well.

Results:

Isolation and subsequent passaging (Pl and P2) of OA-HACs under

physiological tonicity (380 mOsm) significantly increased mRNA levels of important anabolic markers such as AGCl, SOX9 and COL 2 compared to standard culture. Under this condition, addition of FK506 dose-dependently increased this expression even further: SOX9 was induced 6-times, COL2 increased to 50-fold in Pl chondrocytes, COL2 protein 14-fold and AGCl up to 2.6-times. In contrast, unwanted COLl mRNA expression was suppressed to 50% under these conditions. In addition, expression of anti-catabolic genes such as TIMP-I and -2 significantly increased 4-times by 380 mOsm and FK506. In contrast, 380mOsm increased the expression of interstitial collagenase MMP-I and stromelysin-1 (MMP-3), but addition of FK506 to this medium could suppress that effect (Fig. 7). Interestingly, physiological tonicity slightly increased mRNA levels of hypertrophy markers COLlO, alkaline phosphatase (ALPL) and MMP- 13. However, in combination with FK506 all these markers were suppressed to or below control levels (280 mOsm).

Moreover, in combination with FK506, also the enzymatic activity of ALPL and MMP13 was reduced (Fig. 6c). Most importantly, addition of FK506 to the 380mOsmo medium could also suppress ADAMTS5 expression to below control levels.

Conclusions:

Physiological tonicity provides a simple, yet effective, means to improve marker expression during cytokine-free isolation and in vitro expansion of human articular chondrocytes. Compared to FK506 alone, elevated tonicity significantly improved chondrogenic marker expression. Combining FK506 and elevated tonicity significantly improved chondrogenic marker expression even further, while suppressing the few tonicity-induced catabolic and hypertrophic marker genes and their activities where measured As a major component of cartilage's extracellular matrix, the proteoglycan aggrecan imparts compressive resistance to the tissue. Although aggrecan theoretically can be cleaved by several members of ADAMTS family of metalloproteases (i.e. 'aggrecanases'), deletion of active ADAMTS5 (aggrecanase-2) has been shown to be sufficient to prevent cartilage degradation in a murine model of osteoarthritis (Glasson et al. 2005 Nature 434, 644-648;

doi:10.1038/nature03369). Suppressing expression of ADAMTS5, as the primary 'aggrecanase' responsible for aggrecan degradation in osteoarthritis, by using a combination of 380mOsmo + FK506 may thus be a strategy to improve ACI/MACI technique beyond repairing merely chondrol lesions and may extent this applications to pathologically altered joints. EXAMPLE 3

After informed consent MEC2004-322, explants were taken from human osteoarthritic articular cartilage from macroscopically normal areas of the femoral condyles and tibial plateau using a 6 mm diameter dermal biopsy punch, and separated from the underlying bone by dissection with a scalpel. After dissection, the explants were all pooled. For each condition, two explants were randomly taken and cultured for 3 days in a 24-well plate with 330 μl medium per explant. The explants were cultured in control medium (280 mOsm) or hypertonic medium (380 mOsm), with or without addition of 50 ng/mL FK506 for 24 hours, 7 and 14 days. After ex vivo culture, the cartilage explants were snap-frozen in liquid nitrogen and then processed in a Mikro- Dismembrator S (B. Braun Biotech International GmbH, Melsungen.

Germany) for 43sec at 2,8(K) rpms prior to RNA isolation and QPCR analyses as described in van der Windt et al. AR&T 2010 or see above.

Results:

Ex vivo cultured cartilage tissue explants basically showed the same trends in marker gene regulation as isolated chondrocytes under the same conditions (see Fig. 9). Also here, explants which were cultured for 3 days in medium of physiological tonicity (380 mOsm) showed only a modest down-regulation in COLl and mild up-regulation of Sox- 9 and COL2 when compared to control medium (Fig. 10). Interestingly, physiological tonicity slightly increased mRNA levels of hypertrophic markers COLlO and MMP- 13, but not alkaline phosphatase (ALPL). However, in combination with FK506 (50ng/uL) all unwanted markers (COLl; ALPL, COLlO and MMP-13) were down-regulated to or below control levels. In addition, chondrogenic markers (Sox-9, AGCl and COL2) were profoundly up-regulated after addition of FK506 (Fig. 10).

Conclusions:

Physiological tonicity, especially in combination with FK506 provides a simple, yet effective, means to also improve marker gene expression during ex vivo culture of human articular cartilage explants. Combining FK506 and elevated tonicity significantly improved chondrogenic marker expression, while suppressing hypertrophic marker genes. EXAMPLE 4

M urme chondrogerπe cells (ATDCo), established from cu ltures of

teratocarcinoma cells, mature m vitro to become hypertrophic and deposit a mineralized extracellular matrix and represent a robust model for

endochondral bone formation. The cυlls are a good model to study chondrogenic differentiation in comparison to clinically relevant human articular

chondrocytes or adult mesenchymal stem cells from human bone marrow. (Tare et. al. 2005 (European Cells Materials 10 suppl. 2, ISSN 1473-2262).

ATDC5 culture

ATDC5 differentiate into chondrocytes and eventually hypertrophic

chondrocytes in the presence of insulin, transferring (iron transporter) and sodium selenite (antioxidant). Cells were cultured as described by Atsumi et al. 1990, Cell Differentiation Development vol. 30 (2): 109-116). In brief, cells were seeded at 6,400 cells/cm² and propagated in proliferation medium

{DMEM F12+ L-glutamine (GIBCO 32500-043), 5% FCS (PAA A15-101), 1% antibiotic/antimycotic solution (Invitrogenl5240-062), 1% non-essential amino acids (Invitrogen 11140-035) until starting differentiation by adding 1:500 insulin (lOμg/ml, Sigma T2643), 1:3000 transferrin (lOμg/ml, Roche

10652202001) and 1:33333 sodium selenite (30 nM, Sigma S9133). Cells were cultured for up to 4 weeks, with day 0 (t=0, time point of switching to differentiation medium) serving as a control. Cells were cultured in standard medium (±280mOsm) and tonicity was increased by adding sterile NaCl as described for human HACs. GAG staining and measurements ATDC5 cells (e.g. IxIO⁵ cells/well in 12 multi-well plate) were rinsed twice with ice-cold 0.9% (w/v) NaCl and then fixed 10 min in 4 % formaldehyde, prior to washing in demineralized water.

Cells were stained over night in 0.1% Alcian blue (Alcian Blue 8 GX, Sigma) in 0.1 N HCl, and washed 6-times in demineralized water and the cell-associated dye was extracted by incubating 2 h with 0.25mL of guanidine-HCl (6 M). 200 μl of extraction buffer was transfered into 96 well platse and measured next to standards (dilution series of Alcian blue in 6M guanidine-HCl at 620-645 nm in a spectrophotometer.

X-voilet staining and quantification

As colorimetric proliferation assay, we basically used the protocol by

Bonnekoh et al. (1985, published in Archives Dermatological Research vol. 281(7):487-490; DOI : 1 G.1007/R FCGSl 0085; with slight modifications. Briefly, 0.5 ml 0.1 % (w/v) aqueous solution of crystal violet solution was added to cells seeded as for GAG staining. After 30 min at room temperature, cells were washed with demineralised water and air- dried. The dye was then extracted with 1 ml of 10 % (v/v) acetic acid/water and incubated for 15 min at room temperature with gentle agitation, transferred into in 96 well plates (200 μl each) and measured at 590 nm in a spectrophotometer against a blank (10% acetic acid/water).

RNA expression analysis

Quantitative PCR assays for ATDC5 cells were performed as 20 μl SYBRR Green I reactions employing the following assays: MmNFATcI

(NM_198429)Fw, CGGAACGCTAGGAGCTTATG; Rv,

CTCACTGACCAGCCCTTCTC; MmNFATc2 (NM_001037177) Fw,

CCTCCTGGCTCTGTCTTCAG, RV, AAAGTGACATTGGGCTGGAC;

MmNFATc2 (NM_010899) FW, GGAGAACTGGAGAAACTGCG; Rv,

GAGGAAAACACAGCAAAGGC; MmNFATc3 (NM_010901) Fw, CCTCCACAAGGCTTTGAGAC; RV, TCCTTTGGTAGACAGCAGGC;

MmNFATc4 (NM_023699) Fw, TGGAAGGGACTAAGGGTGTG; Rv,

TCAGGACATCATGTTCAGCC; while NFAT5 was detected using the human assay decsribed above. For detection of murine COLl, COL2, COLlO and Cox- 2, we used the following assays MmB2MG Fw

TCGCTCGGTGACCCTAGTCTTT, RV, ATGTTCGGCTTCCCATTCTCC;

MmCoIlAl Fw GCCAAGAAGACATCCCTGAAG, Rv,

TGTGGCAGATACAGATCAAGC (iVIarkiewicz «t a I. 2007, . J, Dermatol. SoL 47:217-226) and MmCol2 (NM_031163) Fw, GATGACATTATCTGTGAAG; Rv ATCTCTGATATCTCCAGG; MmCoIlO (NM_0099255) Fw

CTTTGTGTGCCTTTCAATCG 3, Rv GTGAGGTACAGCCTACCAGTT; and MmCOX-2 (NM..01H98) Fw CAGGAAGTCTTTGGTCTGGTGCC, Rv

GCTGGTTTGGAATAGTTGCTCATCA. All primers are listed according to the convention in δ' - 3^' direction. Data were normalized and relative expression calculated as described for human cells, but 62-microglobin (mB2MG) expression was used as an interna] control.

Western blot analysis

Protein analyses were essentially identical to those performed for human cells, using the same protocol, buffers, antibodies and dilutions as above.

Results represent the mean ±standard deviation.

Results:

In the course of a 4-week differentiation experiment (day 0-28), ATDC5 cells express all NFAT family members with increasing NFAT5 expression towards the end of culture. Culturing ATDCδs in standard medium of 280mOsm increased COLl, COL2, COLlO and Cox-2 expression at day 14 of culture (Fig. 13) compared to day 0 (t=0). However, culturing ATDCδs at 380mOsm decreased expression of COLl, but further increased that of COL2, COLlO and Cox-2. Interestingly, using 480mOsm culture medium resulted in a decreased expression of all these markers to or below control levels (day 14, 280mOsm). Most strikingly, 380mOsm increased the proteoglycan (GAG) production by ATDCδs (Fig. 14), while 480mOsm suppressed it to below control levels.

ATDCδs are proliferating optimally and equally well at both 280mOsmo and 380mOsmo, respectively. In contrast, 480mOsmo decreased the proliferation rate by 50% (Fig. 14). Importantly, in culture medium of 380mOsmo, ATDCδs produced more GAGs per cell (ratio GAG staining/proliferation assay) than in any other condition. Clearly, 480mOsm proofed to be sub-optimal in this respect. With respect to collagen production (Fig. 12), the same trend was seen as for mRNA synthesis (Fig. 13): when compared to standard culture

(280mOsm) at 380mOsm, COLl decreased while COL2 increased, with COLlO surprisingly not being clearly increased on protein level in contrast to mRNA levels. At 480mOsm, COLl protein slightly increased compared to 380mOsm, while COL2 decreased; COLlO expression was hardly altered between conditions.

Conclusions:

Physiological tonicity (380mOsm) not only improved chondrogenic marker gene expression, while suppressing hypertrophic markers, in human adult chondrocytes but also in murine chondrogenic ATDC5 cells. Again, culturing these cells at 380mOsm proofed to be better than using 480mOsm.

EXAMPLE 5

We have shown that increased tonicity increased Cox-2 expression,

ATDCo cells were incubated with a combination of physiological tonicity and a Cox-2 inhibitor (NS39S: N194, Sigma).

Results:

Addition of 2OuM NS398 increased COL2 protein synthesis by ATDC5 cells compared to standard cultures (280mOsm without NS398; Fig. 15). Culturing cells at solely 380mOsm increased COL2 to levels even higher than those reached at 280mOsm with addition of NS398. The combination of 380mOsm and NS398 increased COL2 levels even further (Fig. 15 and 16). Interestingly, while COLlO is increased at 280mOsm + NS398 and 380mOsm alone when compared to the 280mOsm control, the combination of 380mOsm and NS398 decreased COLlO synthesis (Fig. 15) to almost control levels.

While, at day 14 of culture, using solely 380mOsm increased COL2 mRNA more than 3-fold compared to 280mOsm with or without NS398 (Fig. 15), adding NS398 to 380mOsm up-regulated COL2 about 7-times compared to standard cultures.

Conclusions:

Physiological tonicity provides an effective means to also improve marker expression in ATDC5 cells. Adding NS398 to 380mOsm improves the most important chondrogenic marker COL2 on mRNA and protein level, while it may also increase COLl mRNA levels. Therefore, addition of FK506 might help to prevent induction of hypetrophic markers (as shown for HACs), while (modest) inhibition of NS398 might improve expression of chondrogenic markers.

Claims

1. A method for proliferation and prevention of dedifferentiation of chondrocytes, when cultured in vitro, comprising culturing said cells in a medium having a physiological level of tonicity.

2. A method according to claim 1, wherein said medium has a tonicity of 350 - 480 mOsm.

3. A method according to claim 1 or 2, wherein said medium has a tonicity of about 380 mOsm.

4. A method according to any of claims 1-3, wherein said physiological level of tonicity is also applied during isolation of the cells.

5. A method according to any of claims 1-4, wherein proliferation comprises passaging of the cells.

6. A method according to any of claims 1-6, wherein the chondrocytes are isolated from a subject having osteoarthritis or rheumatoid arthritis.

7. A method according to any of claims 1-7, wherein additionally a calcineurin inhibitor is added to the culture medium.

8. A method according to claim 8, wherein said calcineurin inhibitor is chosen from the group consisting of FK506 (tacrolimus/Prograft®), FK520, cyclosporine, tacrolimus, ascomycin and pimecrolimus.

9. A method according to any of claims 1-8, wherein additionally a COX-2 inhibitor is added to the culture medium.

10. A method according to claim 9, wherein said COX-2 inhibitor is chosen from the group consisting of celecoxib, rofecoxib, valdecoxib, lumiracoxib, meloxicam, tramadol, etoricoxib, parecoxib, valdecoxib, licofelone, NS- 398 and sulphonanilides such as nimesulide.

11. A method according to claim 9 or 10, wherein the COX-2 activity of the cells in the medium is reduced.

12. A method for autologous chondrocyte implementation (ACI) or matrix- assisted ACI (MACI) in which chondrocytes are proliferated with a method according to any of claims 1-11.

13. A method for ameliorating osteoarthritis (OA) or rheumatoid arthritis (RA) by intra-articular washing a joint of a subject suffering from said OA or RA with a medium as defined in any of claims 1-11.

14. A method according to claim 13, wherein said medium also comprises chondrocytes or (osteo-chondro) progenitors.

15. Use of a medium as defined in any of claims 1-11 for treatment of OA or RA.

16. Use of a medium as defined in any of claims 1-11 for intra-articular injection in a joint after surgery of said joint.

17. Use according to claim 15 or 16, wherein said medium further comprises chondrocytes, preferably autologous chondrocytes.

18. A method for isolating chondrocytes from cartilage tissue comprising use of a medium as defined in any of claims 1-11 as buffer.

19. Use of a medium as defined in any of claims 1-11 as a buffer medium for isolation of chondrocytes from cartilage tissue.

20. Use of a medium as defined in any of claims 1-11 for the culture of mesenchymal stem cells or other chondrocyte progenitor cells.