The Role of Nitric Oxide in Chondrocyte Models of Osteoarthritis
Dissertation
zur Erlangung des akademischen Grades eines Doktors der Naturwissenschaften (Dr. rer. nat.) des Fachbereichs für Biologie an der Universität Konstanz
vorgelegt von Anna Helena Mais
Tag der mündlichen Prüfung: 31.1.2006 1. Referent: Prof. Dr. Volker Ullrich 2. Referent: Priv. Doz. Dr. Christian Schudt
ACKNOWLEDGEMENTS
This work was performed in the Department of Biochemistry (RDR/B1), ALTANA Pharma AG, Konstanz.
I would like to express my deep and sincere gratitude to my supervisor Prof. Dr. Volker Ullrich for so much of his time, ideas and support.
I would like to thank Dr. Christian Schudt for his guidance, inspiring ideas and especially for always keeping his door open. I would also like to thank the organizers of the graduate program “Biomedical Drug Research” Prof. Dr. Albrecht Wendel and Prof. Dr. Klaus Schäfer for giving me the opportunity to participate in this excellent program. My special thanks go to Dr. Gereon Lauer for his support and motivation during the years. His scientific and technical advice was essential for the completion of this dissertation. I would like to thank Dr. Thomas Klein especially for his enthusiasm. Thomas continually stimulated me with new ideas and encouraged me to develop independent thinking not only in the research area. I would like to thank Marion Hirscher for her expert technical assistance and Nadine Kellner for help with chondrocyte culture. I am grateful to colleagues from RDR/PX and RDR/IT especially to Silke Müller, Dr. Sascha Dammeier, Dr. Hubert Paul, Dr. Gordana Bothe, Lydia Willems, Klaus Hägele and Gisela Schüßler for great support performing 2D-electrophoresis and protein identification as well Affymetrix analysis.
Colleagues and friends at ALTANA, especially Kathrin and Claudi, I thank for support at work but also friendship and help to make Germany my home.
Je dis merci beaucoup à Pierre pour son aide avec la langue anglaise.
My husband Georg for being my best friend, his patience and that he always believed in me. Special thanks to my family:
Mamie i Tacie, Babci i cioci Krysi dziękuję, za to że mnie zawsze kochali i wspierali, szczególnie za Nawojową mojego dzieciństwa. Tą pracę dedykuję pamięci Stanisława Ciołkiewicza, który kochał mnie tak jak nikt inny, co dało mi siłę i wiarę w siebie na całe życie.
Konstanz, 8th of December 2005
Some parts of this thesis have been already published:
Mais A, Klein T, Ullrich V, Schudt C, Lauer G: Prostanoid pattern and iNOS expression during chondrogenic differentiation of human mesenchymal stem cells. JCB, in press (2006 Jan 26 Epub
Brenner SS, Klotz U, Alscher DM, Mais A, Lauer G, Schweer H, Seyberth HW, Fritz P, Bierbach U: Osteoarthritis of the knee-clinical assessments and inflammatory markers. Osteoarthritis Cartilage 12: 469-475 (2004)
ZUSAMMENFASSUNG
Zusammenfassung
Osteoarthrose (OA) ist die häufigste musko-skeletale Erkrankung in der westlichen Welt. Die derzeit übliche Therapie der Erkrankung erlaubt zwar eine effektive Schmerzbehandlung aber es gibt keine pharmakologischen Ansätze, die die Progression des Knorpelabbaus verhindern, oder die Gelenkfunktion verbessern würden. Die Entwicklung der OA ist durch eine fortschreitende Zerstörung der Knorpelmatrix und einen fehlgeschlagenen Reparaturmechanismus charakterisiert. Als Grund wird hierbei ein Ungleichgewicht von anaboler und kataboler Aktivität vermutet. Da der Chondrozyte die einzige Zelle des Gelenkknorpels ist, hat jede Einschränkung seiner Funktion bzw. Vitalität eine Auswirkung auf das Knorpelgewebe. Auch bei der Zerstörung des Knorpelgewebes spielen die Chondrozyten eine wichtige Rolle, da sie inflammatorische Mediatoren und MMPs sezenieren. Zusätzlich produzieren die aktivierten Zellen auch große Mengen von NO. Dieses wasser- und fettlösliche, gasförmige Molekül fungiert nun selber als Mediator einer Vielzahl zellulärer Prozessen. Im Knorpel wird die Synthese von NO durch die iNOS vermittelt, die u.A. durch katabolische Zytokine induziert wird. Dem Interleukin-1 (IL-1) kommt in der OA eine zentrale Rolle zu, da es sowohl die iNOS Expression erhöht, als auch andere degenerative Prozesse induziert. Dabei gehen viele Studien davon aus, dass ein großer Teil der negativen Effekte des IL-1 durch NO vermittelt ist. Aus diesem Grund könnte eine pharmakologische Inhibition der iNOS durchaus als Behandlungsoption bei der OA Therapie in Frage kommen. Das Ziel der vorliegenden Arbeit war es den Einfluss von NO auf die Physiologie des Chondrozyten zu untersuchen, um mögliche positive und negative Effekte einer Inhibition der iNOS auf den humanen Gelenkknorpel zu identifizieren. Um die Rolle von NO aufzuklären wurden verschiedene Zellkulturmodelle verwendet: in situ Kultur von Knorpelstücken, isolierte OA-Chondrozyten in Monolayer oder in 3- dimensionaler Alginatmatrix und nicht zuletzt, als Alternative zur Kultur von gesunden Chondrozyten, ein chondrogenes hMSC basiertes Differenzierungsmodell. ZUSAMMENFASSUNG
Dieses Differenzierungsmodell ist nicht in der Lage gewesen „normalen“ Gelenkknorpel zu generieren. Viel mehr zeigten unsere Analysen, dass ein Knorpel–ähnliches Gewebe entsteht, dass sich in den Expressionsanalysen vom hyalinen Knorpel signifikant unterscheidet. Trotzdem konnten wir zeigen, dass die differenzierten hMSCs in der Lage waren iNOS zu Expremieren und NO freizusetzen. Zusätzlich wurde eine genaue Charakterisierung des Prostanoid-Spektrums während der Differenzierung durchgeführt. Da die Übereinstimmungen zwischen diesen Inflammationsmarkern in den hMSCs während der chondrogenen Differenzierung und Chondrozyten sehr gut war, vermuten wir, dass das Differenzierungsmodell in diesem Zusammenhang als tatsächliche Alternative zu primären Chondrozyten dienen kann. Des weiteren wurde die Regulation der iNOS Expression in Chondrozyten untersucht. Dabei konnten wir zeigen, dass sowohl Zellen aus OA-Patienten als auch aus gesundem Knorpel (Proben aus Schweine und Rindergelenken) mikromolare Mengen von NO nach der Stimulation mit IL-1 freisetzen. Zusätzliche Experimente zeigten, dass die Expression der iNOS in humanen Chondrozyten zentral von dem Transkriptionsfaktor NFκB reguliert wird. Auch die Beteiligung von intrazellulären cAMP an der iNOS-Regulation konnte gezeigt werden. Eine Besonderheit der iNOS-Regulation in Chondrozyten stellt die Nichtbeeinflussung durch Glukokortikoide dar. Wir konnten diese Befunde eindeutig bestätigen und zusätzlich zeigen, dass diese „Dex-Resistenz“ nicht nur bei Chondrozyten von OA- Patienten zu finden ist, sondern auch für andere Spezies und gesunden humanen Knorpel gilt. Darüber hinaus fanden wir diesen Effekt auch in dem hMSC-Modell und weitere Analysen zeigten, dass die „Dex-Resistenz“ nicht mit der NFκB oder cAMP- Signaltransduktion assoziiert ist. Anhand dieser Daten kammen wir zu der Schlussfolgerung, dass Induktion durch IL-1, sowie die fehlende Inhibition der NO- Produktion nach Glukokortikoidgabe, neben den bisher verwendeten Knorpelmatrix- Proteinen, als Marker des Differenzierungsstatus von Chondrozyten verwendet werden kann. Auch konnten wir zeigen, dass humane Chondrozyten nach IL-1 Stimulus COX-2 abhängig, große Mengen PGE2 produzieren. Diese COX-2 Expression ließ sich im Gegensatz zu der iNOS-Expression durch Dex hemmen.
ZUSAMMENFASSUNG
Außerdem war kein deutlicher Effekt von NO auf die Prostanoid-Synthese in unseren Modellen zu beobachten. Das COX-2 Enzym ließ sich nicht durch NO oder seine Derivate inhibieren. Im Gegensatz dazu konnte mit Peroxynitrit ein höherer „Peroxid Tonus“ generiert werden, der zur Aktivierung der COX-Enzyme beitrug. Als weiterer Aspekt wurde der Einfluss von NO auf die Apoptose von Chondrozyten untersucht, da frühere Studien hier einen Zusammenhang hergestellt hatten. Wir konnten jedoch zeigen, dass NO alleine keinen zytotoxischen Effekt auf die Zellen ausübt, dem gegenüber führten hohe Dosen von Peroxynitrit tatsächlich zu einer verstärkten Apoptose der Chondrozyten. Oxidativer Stress scheint im Knorpelgewebe eine Rolle zu spielen. So konnten wir Nitrotyrosin im Knorpelproben nachweisen. Zusätzlich konnten wir eine Reihe von Proteinen identifizieren bei denen Tyrosin-Reste nitriert waren. Interessanterweise waren darunter einige Proteine die mit dem Glukose-Metabolismus assoziiert sind. - Leider war es uns jedoch nicht möglich die O2 -Entstehung in Chondrozyten direkt nachzuweisen. Zusammenfassend kann also gesagt werden, dass sich Chondrozyten leicht durch IL-1 zur NO-Freisetzung stimulieren lassen. Außerdem scheinen diese Zellen ein hohes antioxidatives Potential zu besitzen und sind damit relativ resistent gegen die Folgen von oxidativem bzw. nitrosativem Stress. Da NO alleine offensichtlich keinen negativen Effekt auf die Vitalität der Chondrozyten hat, scheint es, als ob cytotoxische Effekte von - NO nur in Kombinationen mit O2 entstehen können. Wir vermuten dementsprechend, dass nicht nur das Gleichgewicht zwischen anabolen und katabolen Prozessen, sondern auch das antioxidative und oxidative Potential zentrale Bedeutung für die uneingeschränkte Funktion der Chondrozyten und damit für den Erhalt des Knorpelgewebes hat. So konnten wir zwar zeigen, dass IL-1 eine Vielzahl von Genen induziert die mit der Pathophysiologie der OA assoziiert sind, aber entgegen früheren Annahmen, sind diese Veränderungen im Expressionsprofil nicht die Folge der NO-Freisetzung, sondern werden direkt durch IL-1 vermittelt. Aus diesen Gründen zeigt die vorliegende Studie, dass eine Inhibition der iNOS bei OA-Patienten, vermutlich keinen positiven Effekt auf die Physiologie und Funktion von Chondrozyten hat.
SUMMARY AND CONCLUSIONS
Summary and conclusions
Osteoarthritis (OA) is the most common form of musculo-skeletal disorders in the Western world. Current OA therapy relieves pain but there is no pharmacological treatment available that retards the disease progression and improves joint function. The development of OA is characterized by excessive cartilage destruction and defective cartilage repair due to imbalance of the anabolic and catabolic activity of chondrocytes. Chondrocyte viability and function are crucial to articular cartilage as this is the only cell responsible for maintenance of this tissue. However, the chondrocyte plays also an active role in cartilage degradation in OA by releasing inflammatory mediators and matrix metalloproteases. Beside this, activated chondrocytes release high levels of nitric oxide (NO). NO is a gaseous water and lipid soluble molecule that serves as a mediator of a number of cellular processes. In cartilage elevated levels of NO are due to induction of inducible NO synthase (iNOS) in response to stimuli e.g. catabolic cytokines. Interleukin 1 (IL-1) has been shown to play a pivotal role in cartilage damage and is also a very potent stimulator of NO production in articular chondrocytes. It has been implicated that NO mediates many of the destructive effects of IL-1 in cartilage. Therefore pharmacological iNOS inhibition has been proposed as a treatment of OA.
The aim of the present study was to investigate the role of NO in the homeostasis of chondrocytes to evaluate potential beneficial or detrimental effects of iNOS inhibition in human articular cartilage.
To study the role of NO several cartilage-related cell culture models were used as OA cartilage in situ, isolated OA chondrocytes in monolayer and in 3-dimensional alginate matrix and as an alternative to a healthy cartilage a hMSCs chondrogenic differentiation model was established. However, differentiation of hMSCs led to the formation of cartilage-like tissue, which phenotypically differs, from hyaline cartilage in terms of gene expression. Interestingly in the present study we have demonstrated for the first time iNOS expression and NO production in hMSCs during chondrogenic differentiation. We SUMMARY AND CONCLUSIONS
have also provided a detailed characterization of prostanoid production during chondrogenesis. Our investigations suggest that hMSCs undergoing chondrogenic differentiation could be used to investigate the regulation of the production of these inflammatory mediators in a cell system relevant to chondrocytes.
We first studied iNOS expression and its regulation in articular chondrocytes. We could show that OA and normal cartilage (bovine and porcine samples) and chondrocytes released micromolar amounts of NO in response to IL-1 stimulation indicating that NO production is characteristic not only for OA chondrocytes. Further investigation on the regulation of iNOS expression revealed that NFκB is a key transcription factor regulating iNOS expression in human chondrocytes. Intracellular cAMP levels are also involved in iNOS regulation in chondrocytes. It was reported that iNOS expression in human OA chondrocytes is glucocorticoid insensitive. We demonstrated that Dex-resistant iNOS expression is not restricted to human OA chondrocytes but is true for human healthy chondrocytes and chondrocytes from different species. Additionally, glucocorticoid-resistant iNOS expression was even true for hMSCs differentiating to chondrocytes. However, glucocorticoid-insensitivity is not related to NFκB and cAMP signaling pathways. The chondrocyte differentiation status was classically categorized via gene expression of cartilage matrix proteins. We propose corticosteroid – insensitive NO production in response to IL-1β stimulation as additional marker of the chondrocyte differentiation status.
Human chondrocytes after stimulation with IL-1 release high levels of PGE2 due to expression of COX-2. In contrast to iNOS COX-2 expression in chondrocytes is glucocorticoid sensitive. In regard to our study under physiologic conditions the effect of NO on prostanoid synthesis in chondrocytes is not very pronounced. COX-2 was not inhibited by NO or its derivates. On the contrary peroxynitrite can provide a higher “peroxide tone” and activate COX-enzymes in human chondrocytes. In many previous studies NO production was correlated with the level of apoptosis in cartilage. Our results implicate that endogenous or exogenous NO is not cytotoxic to SUMMARY AND CONCLUSIONS
chondrocytes, however peroxynitrite at high concentrations can lead to apoptotic cell death.
Oxidative and nitrosative stresses are present in human OA cartilage as we detected nitrotyrosine in cartilage samples. We identified a number of nitrated proteins; interestingly several of them were related to glucose metabolism. However direct - measurement of ·O2 generation in stimulated chondrocytes was impossible.
To conclude, we have shown that chondrocytes are easily stimulated by IL-1. These cells seem to have a high antioxidative potential and are therefore quite resistant against oxidative and nitrosative stress. NO by itself is not cytotoxic to chondrocytes - and can exert detrimental effects only in combination with ·O2 . We suggest that not only a balance between anabolic and catabolic activity of chondrocytes but also in antioxidative and oxidative potential of the cells are critical for proper function of chondrocytes and maintenance of hyaline cartilage. Although we could show that IL-1 is a very potent inducer of OA-related genes, the observed gene expression changes are not mediated by NO, but are directly due to the action of IL-1. Therefore iNOS inhibition in OA would not be beneficial in regard of chondrocyte homeostasis.
1. Introduction...... 1
1.1. Osteoarthritis ……………………………………………………………………… 1 1.2. Articular cartilage …………………………………………………………………. 4 1.2.1. Chondrocytes ...... 4 1.2.2. Collagen …………………………………………………………………… 6 1.2.3. Aggrecan ………………………………………………………………….. 7 1.3. Human mesenchymal stem cells differentiation model ……………………… 8 1.4. Pathophysiology of OA ...... 10 1.4.1. Changes in cartilage by proinflammatory cytokines ………………….. 10 1.4.1.1. Pivotal role of IL-1β in the pathophysiology of OA …………..…... 10 1.4.2. Apoptosis in the pathophysiology of OA ……………………………….. 11 1.4.3. Prostanoids in the development of OA ………………………………… 12 1.4.3.1. Biosynthesis of prostanoids ………………………………………. 12 1.4.3.2. The role of prostanoids in cartilage ………………………………. 14 1.4.3.3. COX inhibition in OA ……………………………………………….. 15 1.4.4. Nitrosative and oxidative stress ………………………………………… 15
1.4.4.1. Nitric oxide and nitric oxide synthases ………………………….…. 15 1.4.4.1.1. iNOS (NOS-2) ………………………………………………………………….. 17 1.4.4.1.2. NOS inhibitors ...... …. 19 1.4.4.2. Peroxynitrite ...... …. 21 1.4.4.3. Formation of nitrotyrosine ……………………………………….…. 23 1.4.5 Interplay of prostaglandins and nitric oxide ……………………………. 24 1.4.6 Nitric oxide and the development of OA ……………………………….. 25 2. Aims of the study ………………………………………………………………..... 26 3. Materials and Methods ……………………………………………………...... 28 3.1. Chemicals …………………………………………………………………………. 28 3.2. Cartilage processing and cell culture ………………………….. ……………. . 28 3.2.1. Cartilage processing………………………………………………………. 28 3.2.2. Cartilage explants in culture ...... …. 28 3.2.3. Monolayer culture …………………………………………………………. 29 3.2.4. Alginate culture …………………………………………………………... 29
3.3. human Mesenchymal Stem Cells from bone marrow ……………...... 29 3.3.1 Chondrogenic differentiation …………………………………………. ... 30 3.3.2. Osteogenic differentiation ……………………………………………… 30 3.3.2.1. Assessment of osteogenic phenotype …………………………… 30 3.3.2.1.1. von Kossa staining ...... 30 3.3.2.1.2. Alkaline phosphatase staining ...... 31 3.3.3. Adipogenic differentiation ……………………………………………… 31 3.3.3.1. Assessment of adipogenic phenotype ………………………….. 31 3.3.3.1.1. Oil red “O” staining ………………………………………………………… 31 3.4. Cell treatment with growth factors, cytokines and inhibitors ………………. 32 3.4.1. Stimulation of chondrocytes …………………………………………… 33 3.4.2. Stimulation of hMSCs ………………………………………………….. 33 3.5. Biochemical methods …………………………………………………………. 33 3.5.1. Determination of cellular viability ……………………………………… 33
3.5.2. Toxicity: release of lactate dehydrogenase (LDH) ………………….. 33 3.5.3. Griess assay …………………………………………………………….. 33
3.5.4. PGE2 ELISA ……………………………………………………………... 34 3.5.5. GC/MS/MS ………………………………………………………………. 34 3.5.6. Cytochrome c assay ……………………………………………………. 34 3.5.7. Electron Spin Resonance (ESR) ……………………………………… 34 3.5.8. SDS-Page ...... 36 3.5.9. Western blotting and protein detection ……………………………. .. 36 3.5.10. Stripping of Western blots …………………………………………… 37 3.5.11. Two-dimensional gel electrophoresis ……………………………… 37 3.5.11.1. Cell extract ...... 37 3.5.11.2. Two-dimensional gel electrophoresis ...... 38 3.5.11.3. Western blot analysis ...... 39 3.5.11.4. Protein identification (MALDI-TOF)...... 39 3.6. Staining of cartilage and chondrocytes ………………………………….... 40 3.6.1. Preparation of cartilage sections ……………………………………. 40 3.6.2. . Histochemical and fluorescence staining of tissue samples ……. 41 3.6.3. Cytospin preparations ………………………………………………… 42
3.6.4. Evaluation of antibodies ……………………………………………… 42 3.6.4.1. iNOS TK2553 ……………………………………………………………………. 42 3.6.4.2. β-actin ……………………………………………………………………………. 42 3.6.4.3. Nitrotyrosine ……………………………………………………………………. 43 3.6.4.4. PCS ...... 43 3.6.5. TUNEL staining ...... 44 3.6.6. Haematoxylin staining ………………………………………………… 44 3.7. Molecular biology methods …………………………………………………. 45 3.7.1. RNA Extraction and TaqMan PCR ………………………………….. 45 3.7.2. Affymetrix and bioinformatics analysis ………………………………. 47 3.7.2.1. Microarray ……………………………………………………………………….. 47 3.7.2.2. Microarray data analysis ………………………………………………………. 47 4. Results …………………………………………………………………………… 49 4.1. Characterization of hMSC differentiation model ……………………………. 49 4.1.2. Collagen subtypes, aggrecan and SOX-9 expression during the course of chondrogenic differentiation …………………………. 53 4.1.3. Regulation of marker gene expression in pellets was similar to chondrocytes in alginate beads ……………………………………………… 56 4.1.4. Affymetrix gene chip characterization of chondrogenic differentiation 58 4.1.5. Comparison of gene expression of healthy cartilage and chondrogenic pellets …………………………………………………… 60 4.2. iNOS expression and NO production in chondrocytes …………………….. 64 4.2.2. Detection of iNOS in human articular cartilage ………………………. 64 4.2.3. Production of NO in response to inflammatory stimuli: IL-1α, IL-1ß, TNFα, and LPS …………………………………………………………... 66 4.2.3.1. The effect of IL-1ß on the iNOS induction in cartilage explants …………… 68 4.2.3.2. Cellular localization of iNOS in primary chondrocytes ……………………… 69 4.2.4. Time course of NO production, iNOS mRNA and protein expression in human chondrocytes after IL-1ß stimulation ………… 70 4.2.5. The effect of NOS inhibitors on NO synthesis in human chondrocytes 72 4.2.6. The effect of cycloheximide and Byk 17790 on iNOS protein expression ………………………………………………………... 73
4.2.7. Nitric oxide production and iNOS gene expression during the course of chondrogenic differentiation …………………………….. 75 4.3. Regulation of iNOS in chondrocytes …………………………………………… 77 4.3.1. The effect of Dexamethasone …………………………………………… 77 4.3.1.1. iNOS and COX-2 are differentially regulated in human chondrocytes …….. 77 4.3.1.2. Dexamethasone does not inhibit IL-1β induced NO formation in OA chondrocytes or in hMSCs undergoing chondrogenic differentiation . 78 4.3.1.3. Dexamethasone effect on NO production is species-independent ……….. 80 4.3.1.4. Dexamethasone effect on NO production is independent on stimuli …….. 81 4.3.2. Dexamethasone and the regulation mechanism of iNOS expression 82 4.3.2.1. NFκB ………………………………………………………………………………….. 82 4.3.2.2. cAMP …………………………………………………………………………………. 85
4.3.3. cAMP but not NFκB regulates PGE2 production in chondrocytes ….. 88 4.4. Effects of NO on the chondrocyte gene expression …………………………. 90 4.4.1. Effects of NO on chondrogenic differentiation ………………………… 90 4.4.2. Effects of NO on OA chondrocytes ……………………………………. 97 4.5. Effects of IL-1 on the chondrocyte gene expression ………………………… 103 4.5.1. Affymetrix gene chip characterization of IL-1β regulated genes in the hMSCs differentiation model ……………………………………… 103 4.5.2. Affymetrix gene chip characterization of IL-1β regulated genes in human chondrocytes, effect of iNOS inhibition …………………….. 104 4.6. Effects of NO on the eicosanoid production ………………………………….. 109 4.6.1. Eicosanoid production in human chondrocytes and hMSCs ………… 109 4.6.2. COX-2 expression during chondrogenic differentiation ………………. 112 4.6.3. COX-2 mRNA and protein expression in human chondrocytes ……… 113 4.6.4. Time course of eicosanoid production …………………………………. 114
4.6.5. Influence of NO-donors and iNOS inhibitors on PGE2 synthesis ……. 115 4.6.6. Experiments with exogenous Arachidonic Acid (AA)………………….. 119 4.6.6.1. AA concentration dependency of prostanoid formation ……………………. 119 4.6.6.2. Effect of NO-donors on AA stimulated prostanoid formation ……………… 120 4.6.6.3. Effect of NO-donors on AA stimulated Isoprostane formation ……………. 122 4.6.7. Prostacyclin synthase in human cartilage …………………………….. 123
4.6.7.1. Prostacyclin has no effect on the expression of extracellular matrix proteins ……………………………………………………………………………… 124
4.6.7.2. Influence of NO-donors and iNOS inhibitors on PGI2 synthesis …………. 125
4.6.7.3. Effect of NO-donors on AA stimulated PGI2 formation …………………….. 126 4.7. The effect of NO on apoptosis (TUNEL) …………………………………….. 128 4.8. Protein nitrotyrosine in chondrocytes ...... 134 4.8.1. Detection of nitrotyrosine in OA cartilage …………………………….. 134 4.8.2. Nitrotyrosine immunostaining correlates with iNOS in human chondrocytes (staining of IL-1β stimulated cells) ……………….. ….. 135 4.8.3. Detection of protein-nitrotyrosine in one- and two-dimensional gel electrophoresis ………………………………………………………….. 136 4.8.3.1. 2D-gel analyses revealed no COX-2 nitration ……………………………….. 140 4.9. Superoxide ……………………………………………………………………… 142 4.9.1. Measurement of ESR signal in chondrocytes supernatant ………… 142 4.9.2. Measurement of ESR signal in cell suspension …………………….. 143 4.9.3. Measurement of ESR signal generated by SIN-1 …………………… 145 4.9.4. Measurement of superoxide production by human macrophages … 146 4.9.5. Cytochrome c assay (human chondrocytes) ………………………… 148 5. Discussion ………………………………………………………………………… 151 5.1. In vitro studies on cartilage metabolism ………………………………………. 151 5.2. NO production and iNOS expression in chondrocytes ……………………… 157 5.3. iNOS regulation in chondrocytes ………………………………………………. 159 5.4. COX-2 and prostaglandin production in human chondrocytes …………….. 162 5.4.1. Non-enzymatic isoprostane formation in excess of AA ……………… 166 5.4.2. Redox-regulation of prostanoid synthesis …………………………….. 167 5.5. Apoptosis in the development of OA and the role of NO in this process ….. 169 - - 5.6. NO, O2 and ONOO ……………………………………………………………… 171 5.7. Protein tyrosine nitration ………………………………………………………… 176 5.8. IL-1 versus NO mediated effects ………………………………………………. 180 5.9. Clinical implications ……………………………………………………………… 183 6. References………………………………………………………………………… 186 7. Supplement ……………………………………………………………………….. 212
1. INTRODUCTION
1. INTRODUCTION
1.1. Osteoarthritis
Osteoarthritis (OA) is one of the most common forms of musculo-skeletal diseases and the most common form of arthritis that affects millions of patients. According to data in the “Ärzte Zeitung” every second person in Germany aged above 60 suffers from OA. In Europe, a joint is replaced due to OA every 1,5 min, in the USA the situation is even worse (Wieland et al., 2005). Therefore Kofi Annan, the Secretary-General of the United Nations signed the declaration to launch the Bone and Joint Decade 2000-2010 for the treatment and prevention of musculo-skeletal disorders. Clinically OA is characterized by joint paint, stiffness and movement limitation, crepitus, swelling, and finally instability and deformity of the affected joint. The prevalence of OA in all joints is strikingly correlated with age. It is uncommon in adults aged under 40 and becomes extremely prevalent in those aged above 60 (over 70% of people who are 65 or older have OA). Additionally OA is more prevalent in women than in men. OA affects both the small and large joints, either singly or in combination. However, there are joints commonly affected by OA as: knee, hip, interphalangeal joints of the hand, first MTP – metatarsophalangeal joint, cervical and lumbosacral spine. Notably, the ankle, wrist, elbow and shoulder are usually spared. The etiology of OA is unknown. At present the concept that OA represents not a single disease entity but a group of overlapping distinct diseases, which may have different etiologies but with similar biologic, morphologic and clinical outcomes is one of the most accepted. OA is a slowly developing degeneration of articular cartilage, but disease processes finally affect the entire joint including the subchondral bone, ligaments, capsule, synovial membrane, and periarticular muscles (Figure 1).
1 1. INTRODUCTION
normal OA Figure 1. Comparison of normal with OA joint. To see changes in cartilage and bone with secondary inflammatory changes, particularly in the synovium of OA joint. Cartilage fibrillation and destruction is observed. Bone may be most altered in subchondral sites: less stiff and less dense. In OA the development of osteophytes is seen very often (this involves the formation of a cap of new peripheral articular cartilage as well as new bone formation as part of an endochondral process. From: Wieland at el. “Osteoarthritis – an untreatable disease? Nature Reviews 2005
Classically, the diagnosis of OA has relied on the characteristic radiographic changes described by Kellgren and Lawrence (Kellgren and Lawrence, 1957), which include as cardinal feature joint space narrowing, changes in subchondral bone and the formation of osteophytes. The most potent systemic risk factors for developing OA are increasing age, obesity and female gender. The primary factors leading to joint susceptibility to OA are listed in the Table 1.
Table 1. Risk factors for development of OA:
- constitutional susceptibility: - mechanical factors: • heredity • trauma • gender / hormonal status • joint shape • race • usage: occupational, • obesity recreational. - aging
2 1. INTRODUCTION
In contrast to rheumatoid arthritis (RA) osteoarthritis is considered a non-inflammatory arthritis, however episodic inflammation of the synovium (synovitis) is also observed in OA. Short comparison between RA and OA is given in the Table 2.
Table 2. Short comparison between OA and RA: OA RA
strictly local cartilage systemic disease damage -eyes, lung, heart, general characteristic kidneys, blood vessels
slowly developing fast developing
prevalence increases with age younger people (>60 years) (prevalence much lower than OA)
origin bio-mechanical autoimmune disease
primary changes in cartilage synovium
inflammation secondary inflammation chronic inflammation
driven by chondrocytes T-cells, macrophages
cytokines IL-1, IL-6, (TNF-α) TNF-α, IL-1, IL-6
drugs with disease not available on the market modifying efficacy
3 1. INTRODUCTION
1.2. Articular cartilage
Articular cartilage is a type of hyaline cartilage (gr. hyalos = glass). Articular cartilage is a highly specialized tissue, 1-5 mm thick that covers the surface of synovial joints allowing smooth movement between adjacent bones. The structure of cartilage matrix determines biomechanical properties of this tissue: tensile strength and the ability to resist compression. Cartilage matrix comprises: - water (75%) - proteoglycans (5%): mainly aggrecan - collagens (20%): type 2, 6, 9, 11.
1.2.1. Chondrocytes Cartilage is synthesized and maintained by chondrocytes, the only type of cell found in the cartilage. In the adult human, these cells may occupy as little as 2% of the total volume of articular cartilage. Chondrocytes are located within lacunae (chondrons), which usually contain few chondrocytes and pericellular matrix bound within a discrete collagenous capsule (Poole, 1997) (Figure 2a). Under the light microscope chondrocytes appear round or oval in the deeper part of the cartilage, and lens-shaped near the surface (Figure 2b). a) b)
tidemark
surface Figure 2. Haematoxylin staining of human articular cartilage. a) lacunae (chondrons) surrounded by extracellular matrix; intensive staining of pericellular matrix due to higher content of proteoglycans b) cartilage cross-section, description in text (Photos: Anna Mais).
4 1. INTRODUCTION
Nutrition and oxygen are supplied to chondrocytes by the synovial fluid as cartilage is avascular, aneural and alymphatic tissue. Synovial fluid (stands for “egg white” in Greek) is produced by the synovial membrane and is similar in composition to blood plasma, except for the addition of hyaluronic acid, which is responsible for its viscosity. Chondrocytes are mesenchymal cells specialized to produce cartilage-specific matrix, with collagen type 2 and aggrecan, that are responsible for the tensile strength and resistance to mechanical stresses. However, the phenotype of chondrocytes shows multiple modulations. Phenotypic changes are occurring in differentiating chondroprogenitor cells of the fetal growth plate cartilage in vivo. The cells can differentiate from progenitor cells characterized by the expression of the alternative splice variant of type 2 collagen, through mature chondrocytes expressing the typical cartilage proteins: collagen type 2, type 9 and type 11 as well aggrecan to hypertrophic chondrocytes, which are characterized by synthesis of type 10 collagen (Aigner et al., 1993; von der Mark et al., 1992). These cells are found in the lowest zone of the cartilage of the fetal growth plate (Sandell and Aigner, 2001). Chondrocyte hypertrophy has been shown to be an initial event leading to cartilage mineralization (Kirsch et al., 1992) and endochondral ossification (Topping et al., 1994). Premature chondrocyte differentiation to hypertrophic chondrocytes was indicated by von der Mark (von der Mark et al., 1992) in OA cartilage.
In vitro, after isolation from cartilage and expansion in monolayer culture chondrocytes shift toward fibroblast-like morphology and lose their characteristic spherical shape. In this culture conditions not only cell shape is altered, but also production of collagens. Instead of cartilage-specific collagen 2, collagen type 1 and type 3 are synthesized. This process is known as dedifferentiation of chondrocytes. To preserve the chondrocyte specific phenotype 3-dimensional cell culture models were developed, as suspension of these cells in alginate beads.
5 1. INTRODUCTION
1.2.2. Collagen network is the major structural component of articular cartilage. Collagen fibers are bio-mechanically very stable and show an extremely long half-life estimated to be about 117 years (Verzijl et al., 2000). Type 2 collagen (80-90% of articular cartilage collagens) is the most important component of these fibers and a molecule specific for cartilage. However, other collagens such as collagen type 6, found mostly in the territorial matrix closer to the cell, collagen type 11 and collagen type 9 are also present in the ECM but to a much lower extent. These last two collagens may have a role in determining the thickness and assembly of the collagen 2 fibers. Some other molecules, which are bound to collagen fibrils, are fibromodulin, decorin, biglycans and other leucine-rich repeat proteins.
The fiber orientation is different in various parts of articular cartilage (see Figure 3). In the superficial (tangential) zone collagen fibers run parallel to the surface, and this layer has the greatest ability to resist shear stress. In the middle zone (transitional) collagen fibers run in variable directions and this structure is responsible for transition between the shearing forces of the surface layer to compression forces in the deeper layer. The middle zone is the richest of proteoglycans. In the deep (radial) zone of cartilage collagen fibers are perpendicular to the surface and anchored in the tidemark. The structure of this zone is important for the distribution of loads and compression resistance. The calcified zone is located between cartilage layer and subchondral bone. This zone contains the Tidemark Layer (basophilic line which straddles the boundary between calcified and uncalcified cartilage). Collagen 10 is charactreristic for calcified cartilage.
6 1. INTRODUCTION
Figure 3. Cross sections cut through the thickness of articular cartilage, showing the four zones of the cartilage: superficial, intermediate, radiate, and calcified. The foreground shows the organization of collagen fibers. From: Joseph M. Mansour Biomechanics of Cartilage
1.2.3. Aggrecan is the major non-collagenous component of articular cartilage. Aggrecan is a very large molecule consisting of a central protein core (about 2000 amino acids), to which numerous glycosaminoglycan chains of chondroitin sulfate are attached. Many molecules of aggrecan monomers are attached to a long, hyaluronic acid chain forming large aggregates. The glycosaminoglycan chains of aggrecan are negatively charged. Therefore they can bind many water molecules and generate an osmotic swelling pressure making articular cartilage resistant to compression. During loading, when cartilage is compressed, water is squeezed out from ECM. However, when the compressive force is removed, aggrecan again binds water molecules, cartilage swells generating a force equal to the compressive force of the next loading. A net loss of proteoglycan content is one of the hallmarks of OA cartilage degradation (Mankin et al., 1971; Mankin et al., 1981).
All matrix molecules of the articular cartilage are synthesized by chondrocytes and this process is controlled by hormones, cytokines and mechanical stimuli. There are several
7 1. INTRODUCTION endogenous anabolic factors released by chondrocytes that stimulate cartilage generation and remodelling. Among them are transforming growth factor β (TGFβ), bone morphogenic proteins (BMPs) and insulin growth factor 1 (IGF1). These factors are also very important during chondrogenesis. In normal adult cartilage matrix remodelling is a very slow process.
1.3. Human mesenchymal stem cells differentiation model
Human mesenchymal stem cells (hMSCs) are resident in the bone marrow throughout adult life and have the capacity to differentiate along a number of connective tissue lineages, including bone, cartilage and adipose tissue (Jaiswal et al., 1997; Johnstone et al., 1998; Mackay et al., 1998; Pittenger et al., 1999; Murphy et al., 2002). The availability of human chondrocytes is limited and they are difficult to expand in vitro. This is obviously a limiting factor for investigations on the regulation of the above- mentioned mediators. Bone marrow human mesenchymal stem cells (hMSCs) can be differentiated into chondrocytes and are able to synthesize cartilage matrix (You et al., 1998; Neumann et al., 2002). What makes hMSCs such an attractive alternative to primary chondrocytes is that they can be relatively easily expanded and harvested (Solchaga et al., 2004). hMSCs differentiated into chondrocytes could thus provide a solution for drug development projects and tissue replacement therapies targeting OA. Indeed, in recent years there has been increasing interest in hMSCs as an alternative to primary chondrocytes for drug discovery projects and tissue transplant therapies (Redman et al., 2005). However, it is known that currently used protocols are not able to induce homogenous differentiation leading to uniform hyaline cartilage, but rather to chondrocytic phenotype specific for OA cartilage or as recently reported intervertebral disc-like cells (Winter et al., 2003; Steck et al., 2005). Still, most studies were focussed on the expression of ECM-markers such as collagen subtypes. Beside this there was only limited interest in the inflammatory mediator production by hMSCs during and following chondrogenesis, although production of high levels of NO and PGE2 after stimulation of IL-1β is specific for chondrocytes.
8 1. INTRODUCTION
Recently subpopulations of mesenchymal progenitor cells were identified in normal and OA cartilage (Alsalameh et al., 2004; Fickert et al., 2004). These cells are potential sources of cartilage tissue repair, but also lead to the development of osteophytes often observed in OA (Gelse et al., 2003).
chondroblast chondrocyte
hMSCs osteoblast osteocyte
preadipocyte adipocyte
Figure 4. The diagram illustrates the sequence of events involved in the formation of cartilage, bone and adipose tissue from adult hMSCs. hMSCs proliferate (i.e., undergo multiple divisions) to generate increased numbers of MSCs. Many of these expanded hMSCs then undergo a commitment step and enter a particular lineage pathway, leading ultimately to the formation of differentiated, tissue-specific cells (adapted from Osiris Therapeutics, Inc.). hMSCs undergo chondrogenic differentiation when cultured in serum-free conditions and stimulated with TGFβ. Several transcription factors have been shown to be involved in the regulation of chondrogenesis. Sox-9 (SRY-related high mobility group-Box gene 9) is a key regulator of chondrogenic differentiation. This transcription factor regulates expression of cartilage specific collagen 2a1 (de Crombrugghe et al., 2001; de Crombrugghe et al., 2000; Lefebvre et al., 2001; Lefebvre et al., 1997) . Mice with mutated Sox-9 show numerous cartilage-derived skeletal disorders (Wright et al., 1995).
9 1. INTRODUCTION
Another transcription factor Runx2 (Cbfa 1) was shown to play a central role in skeletal development, primarily in osteoblast differentiation and bone formation. However recently it has been demonstrated that Runx2 promotes also early chondrogenic differentiation (Stricker et al., 2002). Interestingly, it has been shown that the chondrogenic and adipogenic activity of OA patient-derived hMSCs is reduced in comparison to hMSCs donors without any symptoms of OA (Murphy et al., 2002). The same authors reported the loss of proliferative capacity of cells from OA patients, which was not age- or site-dependent, but associated with the disease.
1.4. Pathophysiology of OA
1.4.1. Changes in cartilage by proinflammatory cytokines In pathological conditions such as OA, cartilage turnover may accelerate, leading to early regenerative changes, such as synthesis of extracellular matrix which are accompanied by chondrocyte proliferation (hyperplasia) as well as chondrocyte enlargement (hypertrophy). However this early attempt to repair the damaged matrix is followed by degenerative changes, such as insufficient synthesis of ECM, chondrocyte cell death and cartilage matrix erosion. Chondrocytes, but to a lesser extent also synovial cells, can release several catabolic cytokines among others interleukin 1 (IL-1) and tumor necrosis factor α (TNFα), which drive the breakdown of articular cartilage. IL-1β is the major pro-inflammatory cytokine found in synovial fluid (Thomas et al., 2002).
1.4.1.1. Pivotal role of IL-1β in the pathophysiology of OA Evidence has accumulated that IL-1 is by far a more destructive mediator for cartilage than TNFα (van Lent et al., 1995; van de Loo et al., 1995; Probert et al., 1995). IL-1β is synthesized by cells as a precursor, which is intracellulary converted by interleukin-1 converting enzyme (ICE) to produce the active form. IL-1 can stimulate its
10 1. INTRODUCTION own production through autocrine and paracrine mechanisms. IL-1 has been shown to activate matrix metalloproteinases (MMPs), which degradate cartilage matrix. MMP1, MMP8 and MMP13 cleave collagen; MMP3 cleaves proteoglycans (Shingu et al., 1995; Mengshol et al., 2000; Fernandes et al., 2002). It has been demonstrated that IL-1β induces the dedifferentiation process of chondrocytes by decreasing the expression of type 2 and 9 collagens and increasing type 1 and 3 collagen expression (Goldring et al., 1988).
IL-1β stimulates the expression of all enzymes of PGE2 biosynthesis pathway: phospholipase A2, cyclooxygenase 2 (COX-2) and prostaglandin E synthase 1 (PGES) resulting in increased PGE2 production (Massaad et al., 2000; Thomas et al., 2000; Masuko-Hongo et al., 2004). IL-1β is a very potent inducer of the expression of inducible nitric oxide synthase (iNOS) and nitric oxide (NO) production. The pathophysiological importance of IL-1β in OA has been confirmed by use of an ICE inhibitor pralnacasan (Vertex/Sanofi-Aventis) in two murine models of OA, and also by gene transfer of IL-1β receptor antagonist, where both treatments reduced joint damage (Rudolphi et al., 2003; Zhang et al., 2004).
1.4.2. Apoptosis in the pathophysiology of OA Several authors have suggested that apoptosis of chondrocytes is characteristic for OA, and is a reason for cartilage degeneration (Blanco et al., 1998; Hashimoto et al., 1998a; Kim et al., 2000; Goggs et al., 2003; Kirsch et al., 2000; Heraud et al., 2000). There are also opinions, contrary to previous suggestions, that apoptotic cell death is not a widespread phenomenon in aging or OA cartilage (Aigner et al., 2001). Chondrocyte apoptosis in vivo seem to differ from “classical apoptosis”. Unlike classical apoptosis chondrocyte apoptosis involves an initial increase in endoplasmic reticulum (ER) and Golgi apparatus, reflecting an increase in protein synthesis. Increase in ER membranes provides compartments within which cytoplasm and organells are digested what leads to the complete self-destruction of the chondrocyte and remaining empty lacunae. This mechanism may be of importance as there are no phagocytic cells within
11 1. INTRODUCTION the cartilage. It has been proposed to give chondrocyte apoptosis even a special name “chondroptosis” (Roach et al., 2004). Previous studies have linked NO and chondrocyte apoptosis, especially the data from animal models of OA suggested that NO could be a signal for apoptosis. Anterior cruciate ligament transaction in dogs resulted in an increase in NO levels in the affected joint and increase in apoptotic cells (Hashimoto et al., 1998b). Treatment with selective iNOS inhibitor (L-NIL) resulted in the reduced level of chondrocyte apoptosis and reduction of OA progression in this model (Pelletier et al., 2000). The data revealed an association between NO and apoptosis in vivo, however it is still unclear if chondrocytes underwent apoptosis in direct response to NO. Interestingly, in the rabbit model of OA hyaluronan injections decreased the level of apoptosis but not NO (Takahashi et al., 2000).
Whether apoptosis or cell disintegration is primary or secondary to the destruction of cartilage matrix is difficult to answer. There is also the possibility that chondrocyte death and matrix loss form a cycle, with the progression of one having effects on the other.
1.4.3. Prostanoids in the development of OA 1.4.3.1 Biosynthesis of prostanoids Prostanoids are a class of saturated fatty acid deriverates containing prostaglandins and thromboxanes. The prostanoid class is a subclass of eicosanoids (gr. Eicos – twenty) containing 20 carbon atoms in the molecular structure. The first step in prostanoid biosynthesis is the liberation of arachidonic acid from the membrane phospholipids by phospholipase A (PLA2). Arachidonic acid (AA) serves as precursor of all prostanoids. AA undergoes the cyclooxygenase reaction in which two molecules of oxygen are added to form a bicyclic endoperoxide with a hydroperoxy group in position 15 (PGG2). This hydroperoxide is then reduced by a functionally coupled peroxidase reaction to form the 15-hydroxy-9,11-endoperoxide (PGH2). This conversion of AA is catalyzed by cyclooxygenase, named also prostaglandin endoperoxide H2 synthase (COX or PGHS, respectively). COX exists in at least two
12 1. INTRODUCTION isoforms: COX-1, expressed constitutively and COX-2, which expression is induced in various cell types after exposure to bacterial endotoxins or proinflammatory cytokines.
PGH2 is subsequently converted to a variety of prostanoids that include: prostaglandin
E2 (PGE2), prostaglandin D2 (PGD2), prostaglandin F2α (PGF2α) , prostacyclin (PGI2) and thromboxane (TX), as ilustrated in Figure 5.
Figure 5. The biosynthesis of prostanoids from arachidonic acid by cyclooxygenase pathway. From: Narumiya et al., “Prostanoid receptors: Structures, Properties, and Functions”, Physiological Reviews, 1999, Vol. 79, 1193-1226
The proportions of the various enzymes of the prostanoid pathway and therefore produced prostaglandins differ according to the cell type. In 1990 Morrow reported prostaglandin-like molecules in humans that were result of peroxidation of AA by a free radical mechanism independent of COX activity (Morrow et
13 1. INTRODUCTION
al., 1990). As these compounds are isomeric to COX-derived PGF2α, they were termed F2-isoprostanes (isoprostanes). Interest in these molecules arises from the fact that they provide an index of free-radical induced lipid peroxidation. The level of isoprostanes can indicate several disorders, for example higher levels of 8-epi-PGF2α have been detected in the plasma of smokers. However COX-dependent formation of isoprostanes has been also postulated (Klein et al., 1997)
1.4.3.2 The role of prostanoids in cartilage Gene expression, cartilage matrix synthesis and proliferation have been shown to be regulated by prostaglandins in cartilage (Geng et al., 1995). In articular cartilage prostaglandins have been discussed as mediators of inflammation and tissue destruction. However recently prostglandins have been shown to play a role in a cartilage formation by stimulating the differentiation of prechondroblasts to chondrocytes (Jakob et al., 2004). Schwartz et al. have proposed that the effect of
PGE2 depends on its concentration, low levels promote differentiation, whereas high doses promote an anabolic response (Schwartz et al., 1998).
In regard to this PGE2 can exert both catabolic or anabolic effects in chondrocytes depending on the microenvironment and which of the four receptor subtypes are present. Indeed, PGE2 has also been shown to be involved in the development of OA.
PGE2 modulates proteoglycan and collagen synthesis (Abramson, 1999; Goldring et al., 1996), stimulates matrix metalloproteinase-2 expression (Choi et al., 2004) and enhances matrix metalloproteinase-3 production (Amin et al., 2000). In addition, PGE2 inhibits chondrocyte proliferation (Blanco and Lotz, 1995b) and induces chondrocyte apoptosis (Notoya et al., 2000; Amin et al., 1997). Strikingly, PGE2 was also the only parameter, which we recently found to correlate with the WOMAC-index scores of patients with knee OA (Brenner et al., 2004).
The overproduction and role of PGE2 in OA cartilage have been reported several times.
However, COX-2 overexpression in OA not only leads to the production of PGE2 but to a variety of prostanoid endproducts that have not been fully characterized in human cartilage.
14 1. INTRODUCTION
1.4.3.3 COX inhibition in OA COX inhibitors represent the most widely used class of drugs. Although aspirin and other non-steroidal anti-inflammatory drugs (NSAIDs) are used for the treatment of inflammation, pain and fewer for more than 100 years the mechanism of their action, namely COX inhibition, was demonstrated for the first time in 1971 by Sir John Vane. NSAIDs inhibit both COX isoforms. COX-2 inhibition accounts for the therapeutic benefits and inhibition of COX-1 for the side effects of NSAIDs. The most common side effect of NSAIDs is irritation damage of gastrointestinal mucosa leading to the development of gastric erosions and ulcerations. This effect is due to COX-1 inhibition. As COX-2 is primarily responsible for increased PG production in inflamed tissues specific COX-2 inhibitors have been developed to minimize the risk of side effects mainly in gastro-intestinal tract. Both classes of COX inhibitors are very often prescribed for the treatment of OA. Specific COX-2 inhibitors: celecoxib (Celebrex, Pfizer) and rofecoxib (Vioxx, Merck) have been shown to be more effective than placebo and similarly effective as standard doses of nonselective NSAIDs (Hochberg, 2005). Both classes of COX inhibitors provide effective relief from pain as a major subjective symptom of OA (Martel-Pelletier et al., 2003). However, clinical trial investigating rofecoxib revealed that patients at higher cardiovascular risk exhibit a significantly higher incidence of myocardial infraction receiving COX-2 specific inhibitor compared with NSAIDs (Bombardier et al., 2000). Interestingly, recently celecoxib was reported to have chondroprotective effects by enhancing proteoglycan content and therefore matrix integrity in OA cartilage (Mastbergen et al., 2005).
1.4.4. Nitrosative and oxidative stress 1.4.4.1. Nitric oxide and nitric oxide synthases Nitric oxide is a gaseous short lived regulatory molecule exerting a broad range of functions in many physiological and pathophysiological cell and tissue responses (Moncada et al., 1991). Under aqueous, aerobic conditions NO spontaneously oxidizes to its stable products nitrite and nitrate.
15 1. INTRODUCTION
The first described physiological effect of NO was activation of soluble guanylyl cyclase (Ignarro et al., 1999). The rise in cGMP (3`,5`-cyclic guanosine monophosphate) level accounts for many of cellular responses to NO. Relatively low reactivity combined with a high lipophilicity allow NO to diffuse away from the point of origin, and therefore to carry out its function as a messenger molecule beyond cell borders (Wink and Mitchell, 1998). NO has the potential to interact directly or indirectly with metals, thiols and oxides, and therefore affects proteins, nucleic acids, lipids and sugars (Davis et al., 2001). Effects of NO depend on its concentration and the redox state of the cell. NO is produced by nitric oxide synthases (NOS). These group of enzymes catalyse the production of NO and L-citrulline from L-arginine, NADPH and O2.
Figure 6. Nitric oxide biosynthesis. NO synthesis from L-arginine is a reaction which involves two separate mono-oxygenation steps. Nω-hydroxy- L-argninie is an intermediate formed by a reaction requiring O2 and NADPH in a presence of BH4. The second step results in the oxidation of Nω-hydroxy- L-argninie to form citrulline and NO.
From: Knowles, (1994) Nitric oxide synthases in mammals. Biochem J 298: 249-258
There are three NOS isoforms identified: endothelial NOS (eNOS, NOS III), neuronal NOS (nNOS, NOSI) and inducible NOS (iNOS, NOSII). NOSes are enzymes active only
16 1. INTRODUCTION
as dimers. They contain tightly-bound cofactors (6R)-5,6,7,8-tetra-hydrobiopterin (BH4), flavine-adenine dinucleotide (FAD), flavine mononucleotide (FMN) and iron protoporphyrin IX (haem)(Knowles and Moncada, 1994; Marletta et al., 1998; Alderton et al., 2001). The activity of eNOS and nNOS is calcium-dependent, as Ca2+ stabilizes the calmodulin binding to its binding site in these two NOS isoforms, thereby initiating NO synthesis (Bredt and Snyder, 1990; Abu-Soud and Stuehr, 1993). However in iNOS calmodulin is tightly bound and therefore its activity is calcium-independent (Vallance and Leiper, 2002; Ruan et al., 1996; Vallance and Leiper, 2002). eNOS and nNOS are regarded as constitutively expressed, however it has been shown that expression of these enzymes is to some extent also regulated at the transcriptional level by local environmental conditions (Xu et al., 1995; Liu et al., 1996; Kleinert et al., 2000; Forstermann et al., 1998). After activation eNOS and nNOS produce nanomolar concentrations of NO, and are active for relatively short periods of time. Both isoforms are principally considered to participate in the regulation of physiological processes in the cardiovascular and nervous system, were they were first found (Marletta et al., 1998). However, their expression has been found later in other tissues and cell types.
1.4.4.1.1. iNOS (NOS-2) iNOS was first described in activated macrophages (Hibbs et al., 1988). This isoform is not usually expressed in healthy quiescent cells, but is rapidly transcriptionally induced in multiple cell types in response to stimulation with bacterial endotoxins or proinflammatory cytokines (Vallance and Leiper, 2002). Once induced, iNOS produces high amounts of NO (about 100x higher concentrations as eNOS and nNOS) for a prolonged period of time (Nathan, 1992). These high levels of NO are important for a host defence against infectious organisms (Nathan, 1997; Stuehr et al., 1991; Schmidt and Walter, 1994; Nathan, 1997). NO produced by iNOS regulates also the functional activity, growth and death of many immune and inflammatory cell types including macrophages, T lymphocytes, antigen-presenting cells, mast cells, neutrophils and natural killer cells (Coleman, 2001). Expression of iNOS has been found in many cell types like in endothelium, epithelium and also in chondrocytes and synoviocytes after induction.
17 1. INTRODUCTION iNOS expression is regulated at the transcriptional level and at the level of iNOS mRNA stability (Kleinert et al., 2000). Activation of transcription factor NF-kappa B (NFκB) seems to be an essential step for iNOS induction in most cells (Förstermann and Kleinert, 1995). Glucocorticoids have been shown to interfere with iNOS expression in many cell types (Di Rosa et al., 1990). Inhibition of iNOS expression by glucocorticoids has been shown to result from inhibition of NFκB activation (Kleinert et al., 1996; Mukaida et al., 1994). cAMP activated transcription factors like CREB, C/EBP and ATF2 seem also to be involved in the regulation of iNOS expression (Bhat et al., 2002; Kleinert et al., 2003). However, regulation of iNOS expression is very complicated and can also involve other transcription factors like AP1 or STAT1.
Table 3. Short characteristics of nitric oxide synthases.
iNOS eNOS nNOS
expression inducible constitutive constitutive
LPS, proinflammatory stimulated by - - cytokines
calcium dependent - + + chondrocytes, macrophages, present in lymphocytes, epithelial endothelial cells neurons cells, smooth muscle cells etc.
The production of high NO levels in inflammation is considered to be responsible for many detrimental effects leading to tissue destruction. The overproduction of iNOS is implicated in a number of pathologies like septic shock (Titheradge, 1999), ulcerative colitis, Crohn`s disease (Cross and Wilson, 2003), asthma (Barnes and Liew, 1995; Ricciardolo et al., 2004; Moncada, 1999), rheumatoid arthritis (Henrotin et al., 2003), but iNOS may be also involved in the pathogenesis of other disorders associated with a low grade chronic inflammation state, such as atherosclerosis (Cromheeke et al., 1999; Behr-Roussel et al., 2000), diabetes (Shimabukuro et al., 1997; Shimabukuro et al., 1998; Zhou et al., 2000) or osteoarthritis (Amin and Abramson, 1998; Studer et al., 1999).
18 1. INTRODUCTION
1.4.4.1.2. NOS inhibitors As L-arginine is a substrate for NOS several NOS inhibitors are analogues of arginine competing for the active site of the enzyme. L-NMMA (NG-monomethyl-L-arginine) and L-NAME (NG-nitro-L-arginine methyl ester) represent the group of NOS unspecific substrate analogues and their inhibitory action on NOS can be antagonized by high concentrations of L-arginine. L-NAME exhibits slightly higher potency towards constitutive enzymes. AMT (2-amino-5,6-dihydro-6-methyl-4H-1,3-thiazine) is a very potent (about 1000-fold more potent than arginine analogues presented here) but a non-selective NOS inhibitor (Nakane et al., 1995; Boer et al., 2000). In recent years a dogma was established that the constitutive forms of nitric oxide synthases, eNOS and nNOS, are critical to normal physiology and inhibition of these enzymes causes damage, whereas induction of the iNOS is harmful and specific inhibition of this enzyme would be beneficial. Therefore, specific iNOS inhibitors have been developed. The first highly selective iNOS inhibitors were the bis-isothioureas reported by Garvey et al. (Garvey et al., 1994), which were ca. 200-fold more selective for iNOS than for eNOS, but only 5-fold in comparison to nNOS (Alderton et al., 2001). Improvement of iNOS selectivity was achieved with 1400W an amidine-derived iNOS inhibitor. 1400W is not only highly selective as an iNOS inhibitor versus both eNOS and nNOS but also penetrates cells and tissues (Garvey et al., 1997). 1400W does exhibit an acute toxicity at high doses, which is likely to prevent its safe therapeutic use in humans, but it can be used as a pharmacological tool in a variety of animal models (Alderton et al., 2001). BYK191023 is an imidazopyridine derivative developed by ALTANA Pharma. Imidazpyridine compounds represent a novel class of NO-synthase inhibitors with high selectivity for the inducible isoform (Strub et al., 2005). BYK191023 is 200-fold more selective for human iNOS versus nNOS and 1000-fold in comparison to eNOS. BYK191023 did not show any toxicity in various rodent and human cell lines up to high micromolar concentrations (Strub et al., 2005).
19 1. INTRODUCTION
Table 4. The selectivity of NOS inhibitors.
Selectivity (fold) Inhibitor Structure iNOS/eNOS iNOS/nNOS nNOS/eNOS
L-NMMA 0,3 0,8 0,3
L-NAME 0,05 0,05 1
AMT 3 0,8 3
1400W 200 20 10
BYK191023 >1000 >200 -
Selectivity was determined on the basis of IC50 values obtained for purified NOS isoforms in the same experimental conditions (except values for BYK191023). Source: Boer et al. The Inhibitory Potency and Selectivity of Arginine Substrate Site Nitric-Oxide Synthase Inhibitors Is Solely Determined by Their Affinity toward the Different Isoenzymes 2000, Molecular Pharmacology; 58:1026-1034. Data for BYK191023 from Strub et al. The Novel Imidazopyridine BYK191023 is a Highly Selective Inhibitor of the Inducible Nitric Oxide Synthase 2005 Molecular Pharmacology;
There are also NOS inhibitors with distinct mechanisms of action than competing with arginine for the active site of these enzymes. For example BH4 analogs, which act at
BH4 binding site, like 4-amino-H4-biopterin (Werner et al., 2003) or inhibitors of NOS dimerization like imidazole derivatives (Ohtsuka et al., 2002).
Clinical trials with NOS inhibitors were started in septic shock patients, however they were not successful till now and had to be stopped due to increased mortality. This was probably due to use of unspecific NOS inhibitors (L-NMMA) (Petros et al., 1991; Petros
20 1. INTRODUCTION et al., 1994). Still highly selective iNOS inhibitors have not yet been tested in published clinical trials.
1.4.4.2. Peroxynitrite
● ● - - NO reacts very rapidly with superoxide ( O2 ) to form peroxynitrite (ONOO ). The rate constant of this reaction is near the diffusion-controlled limit
9 -1 -1 ● - (k= 6,7x10 M ● s ) (Huie and Padmaja, 1993). Dismutation of O2 by superoxide 7 -1 -1 dismutases (SOD) was observed with k= 10 M ● s , which is 2-3 orders of magnitude slower than the reaction with NO, so formation of peroxynitrite is favored. Superoxide anion can be generated in cells either enzymatically (NADPH oxidases) or by processes that produce reactive oxygen species (ROS) such as the electron transport chain in mitochondria. Cellular sources of superoxide are given in the Figure 7.
• mitochondrial e- transport
• NADPH oxidases SOD
● - • Xanthine oxidase O2 H2O2
• NO-synthases
(Arg, BH4 deficiency)
ONOO- •OH RNS
●NO • NO-synthases
● - Figure 7. Cellular sources and fates of superoxide ( O2 ). Superoxide reacts with NO much faster than its detoxification reaction with superoxide dismutase (SOD) takes place. The product of superoxide and nitric oxide is peroxynitrate, which can undergo further reactions.
21 1. INTRODUCTION
● ● - - Although neither NO nor O2 is a strong oxidant, ONOO is a potent and versatile oxidant that can attack a wide range of biological targets (Pryor and Squadrito, 1995). Some of biologically important reactions of peroxynitrite are listed below.
Biological reactions of peroxynitrite:
- nitration of tyrosine residues of proteins (Haddad et al., 1994; Reiter et al., 2000) - triggering of lipid peroxidation (Radi et al., 1991b) - inhibition of mitochondrial electron transport (Radi et al., 1994) - oxidation of thiol compounds (Radi et al., 1991a) - DNA-strand breakage (Szabo, 2003; Szabo et al., 1996; Szabo, 1996; Zingarelli et al., 1996) - activation of MMP’s (Migita et al., 2005) - inactivation of TIMP (tissue inhibitor for MMP), and α1-proteinase inhibitor (Brown et al., 2004) - modulation of cyclooxygenase activity (Landino et al., 1996; Deeb et al., 2002; Schildknecht et al., 2005; Mollace et al., 2005) - apoptosis (Lin et al., 1995; Salgo et al., 1995; Virag et al., 2003)
22 1. INTRODUCTION
1.4.4.3. Formation of nitrotyrosine Peroxynitrite is a potent protein-nitrating agent. Especially aromatic tyrosine residues of proteins are susceptible to reaction with ONOO-, leading to the formation of 3- nitrotyrosine (Figure 8).
O O N O OH - OH NH N O NH 2 O 2 O O
OH OH
tyrosine 3-nitrotyrosine
Figure 8. Formation of 3-nitrotyrosine.
Protein nitrotyrosine formation alters the structure and function of proteins. It may prevent tyrosine phosphorylation leading to alterations in signal transduction, which has been considered as pathological event, but on the other hand it might be also a part of cellular signaling mechanisms. Increased levels of 3-nitrotyrosine have been reported in many diseases, such as cancer, Parkinson`s disease, Alzheimers disease, COPD, RA and OA. However, the group of Stuehr postulated that protein nitration is observed under normal conditions in all tissues (Aulak et al., 2004). The role protein nitration plays in cell physiology is unclear. Nitrotyrosine formation can be detected by immunostaining and has been used as a “footprint” of peroxynitrite. However a second mechanism of tyrosine nitration via heme peroxidase dependent reactions using nitrite as substrate to generate the nitrating agent nitrogen dioxide have been proposed. Thus nitrotyrosine staining can be used as an indicator for nitrosative stress in the tissue rather than a specific marker of peroxynitrite.
23 1. INTRODUCTION
1.4.5. Interplay of prostaglandins and nitric oxide After stimulation of iNOS and COX-2 expression NO and prostaglandins are released simultaneously in large amounts. There is also a cross-talk between these mediators, however there is more evidence that NO affects PG biosynthesis than vice versa. The effect of NO on PG biosynthesis is dependent on the cell type generating PG, stimulus, NO concentration and redox state of the cell (Mollace et al., 2005). Both augmentation and inhibition of prostaglandin synthesis by NO have been reported. NO could increase COX-1 activity in microsomal sheep seminal vesicles as well as the activity of murine recombinant enzyme (Salvemini et al., 1993). NO was shown to enhance the release of PGI2 from endothelial cells (Salvemini et al., 1996). NO via peroxynitrite increases also the activity of the COX-2 enzyme and the production of prostaglandins in various cellular systems and this effect seems to be independent of the known effects of NO on soluble guanylyl cyclase (Mollace 2005). However many experiments have shown that NO can inhibit prostaglandin production, mainly by inhibition of COX-2. This was demonstrated in a number of cell types like microglial cells (Minghetti et al., 1996; Guastadisegni et al., 1997), vascular endothelial cells (Doni et al., 2005), rat Kupffer cells (Stadler et al., 1993) and macrophages (Swierkosz et al., 1995). Peroxynitrite has been also shown to modulate COX activity. Landino et al. have shown that peroxynitrite activate purified COX enzymes. ONOO- may activate COX enzymes (peroxide tone) leading to the enhanced production of prostaglandin endoperoxides (Landino et al., 1996). Ullrich and coworkers have shown nitration of prostacyclin synthase by peroxynitrite and thereby decreased production of PGI2 (Zou and Ullrich, 1996; Zou et al., 1997; Zou et al., 1998).
There is contradictory evidence on the regulation of prostaglandin synthesis by NO in cartilage. Inhibition of NO production has been suggested to increase PGE2 release in chondrocytes (Fermor et al., 2002; Mathy-Hartert et al., 2002), but on the other hand also reduction in prostaglandin levels after NO inhibition has been proposed (Manfield et al., 1996; Blanco and Lotz, 1995a; Jouzeau et al., 2002).
24 1. INTRODUCTION
1.4.6. Nitric oxide and the development of OA Very low concentrations of IL-1 are sufficient to stimulate NO production in chondrocytes. This contrasts with most other human cell types where multiple stimuli are required for iNOS induction. Therefore, NO has been made responsible for much of IL-1 effects on chondrocytes. Interestingly, the synthesis of NO in human OA cartilage derives from a glucocorticoid- insensitive induction of iNOS expression (Vuolteenaho et al., 2001). It has been suggested that high concentrations of NO or its derivatives contribute to the development of OA by inhibiting the synthesis and promoting the degradation of cartilage extra-cellular matrix (ECM) (Abramson et al., 2001). As shown by Studer NO could inhibit response of chondrocytes to the anabolic growth factor IGF-I (Studer et al., 2000; Studer, 2004). NO was shown to promote joint inflammation by up-regulation of IL-1-converting enzyme (ICE) and proinflammatory cytokine IL-18 synthesis while decreasing the level of the ICE inhibitor: protease inhibitor-9 (Boileau et al., 2002). Furthermore, high levels of intracellular NO have been demonstrated to induce chondrocyte apoptosis (Oliver et al., 2004; Clancy et al., 2001). On the other hand, NO levels in synovial fluid failed to correlate with the clinical classification of OA knees (Brenner et al., 2004). Previous studies have shown that NO - is not cytotoxic to chondrocytes by itself but only in combination with O2 (Del Carlo M Jr, 2002). There are only two studies dealing with the development of OA in iNOS knockout mouse, unfortunately with contradictory results. Van den Berg reported that iNOS deficiency diminished cartilage lesions and osteophyte formation in OA model induced with intra-articular injection of collagenase into the knee joint (van den Berg et al., 1999). In contrary, Clements et al. in an OA model induced by transaction of the medial collateral ligament and partial medial menisectomy observed accelerated cartilage degeneration in iNOS knock-out animals (Clements et al., 2003).
25 2. AIMS OF THE STUDY
2. AIMS OF THE STUDY
Osteoarthritis is the most common musculo-skeletal disease, affecting millions of patients worldwide. To date only symptomatic treatment of OA using primarily analgesics is available. Therefore there is a strong need to develop drugs with disease- modifying efficacy. NO has been implicated as important factor in the pathogenesis of OA because iNOS overexpression and increased NO synthesis was observed in cartilage of OA patients. In recent years several studies dealing with the role of NO in the development of OA have been performed. Use of iNOS inhibitors has been proposed for the attenuation of cartilage degradation observed in OA, although the data on the effect of NO on cartilage degradation is still contradictory.
The aim of the present study was to investigate the role of NO in the homeostasis of chondrocytes and to further evaluate potential beneficial or detrimental effects of iNOS inhibition in human articular cartilage. All experiments presented in this study were performed in different cartilage and chondrocyte related models as cartilage explants, isolated chondrocytes in monolayer and in 3-dimensional alginate matrix as well as in a hMSCs – chondrogenic differentiation model. In brief the results of this study could help to answer the question if pharmacological iNOS inhibition is a promising approach for the treatment of OA.
Due to the complexity of the issue the present study was focused on the following objectives:
1. To evaluate a hMSC chondrogenic differentiation model as an alternative to healthy human articular cartilage as its availability was limited and we were interested in effects of NO on both OA and healthy cartilage.
26 2. AIMS OF THE STUDY
2. To investigate iNOS expression and NO production in OA cartilage and in hMSC during chondrogenic differentiation.
3. To elucidate some aspects of iNOS regulation in articular chondrocytes with focus on the insensitivity of iNOS expression from human OA cartilage to glucocorticoids.
4. To elaborate the effects of NO on the gene expression of chondrocytes. We were especially interested in genes, which are specific for the chondrocytic phenotype or related with OA.
5. To study the effect of NO on prostaglandin synthesis and apoptosis of chondrocytes as they are both considered essential factors in the cartilage degeneration observed in OA.
6. NO can very rapidly react with superoxide to form peroxynitrite. Since peroxynitrite formation is considered to be involved in pathology of several diseases, among others OA, we were interested in oxidative and nitrosative stress in OA cartilage. This implicated measurements of superoxide production and nitrotyrosine formation in cartilage and chondrocytes.
27 3. MATERIALS AND METHODS
3. MATERIALS and METHODS
3.1. Chemicals
Sigma (Munich, Germany), Merck (Darmstadt, Germany), Roche (Mannheim, Germany) and Cayman Chemicals (Ann Arbor, USA) supplied chemicals.
3.2. Cartilage processing and cell culture
3.2.1. Cartilage processing Cartilage specimens were obtained with institutional approval from OA patients undergoing total knee or hip joint replacement in a local orthopaedic hospital. The cartilage was dissected from the underlying bone and fibrocartilaginous areas were discarded. The cartilage surfaces were rinsed several times with DMEM/F12 (Gibco Life Technologies, Eggenstain, Germany), with gentamicin (Gibco). Scalpels were used to cut cartilage in sections 3 mm apart. These tissue pieces were then digested with pronase (Calbiochem, Bad Soden, Germany) 4 mg/ml over 90 min and collagenase P (Roche Biochemicals, Mannheim, Germany) 1 mg/ml overnight in DMEM/F12 supplemented with 5% FBS (Gibco) and gentamicin (50 µg/ml) to isolate chondrocytes. The released chondrocytes were washed with HBSS (Gibco) and filtered through a 70µ nylon membrane. Chondrocytes were than cultured either in monolayer or in a three-dimensional alginate matrix as indicated.
3.2.2. Cartilage explants in culture For in situ experiments cartilage pieces were prepared as for digestion (rinsed and cut), afterwards incubated in the cell culture medium DMEM/F12 supplemented with 20% FBS, 2 mM L-glutamine, 1 mM L-cysteine (Fluka Biochemicals, Buchs, Switzerland), 25 µg/ml ascorbate (Fluka) and 50 µg/ml gentamicin and incubated at 37°C in a humidified gas mixture containing 5% CO2.
28 3. MATERIALS AND METHODS
3.2.3. Monolayer culture Human chondrocytes after isolation were plated at high density (ca. 200 000 celles/cm2) in cell culture medium as for cartilage pieces. Chondrocytes were incubated at 37°C in a humidified gas mixture containing 5% CO2. Chondrocytes were used at confluency in primary culture or after passage.
3.2.4. Alginate culture Chondrocytes were encapsulated in alginate beads immediately after isolation according to the method of Häuselmann et al. (Häuselmann et al., 1994). Briefly, cells were suspended in 1,2% sodium alginate (Sigma-Aldrich Co., Taufkirchen, Germany) in 150 mM NaCl (4x106 cells/ml alginate solution). The chondrocyte suspension was passed drop-wise through a 22-gauge needle into a 102mM CaCl2 solution under constant stirring. Following 10 min polymerization, beads were washed with 150 mM NaCl and with DMEM/F12. Chondrocytes in alginate beads were cultured in the same medium and culture conditions as cartilage pieces and cells in monolayer. When required, alginate beads were dissolved by exposure to solubilization buffer (55 mM sodium citrate and 150 mM NaCl, pH 6,00) at 37°C for 10 -15 min. The cells were recovered by centrifugation.
3.3. human Mesenchymal Stem Cells (hMSCs) from bone marrow
Frozen hMSCs from healthy donors (age 18-30 y.) were purchased from Cambrex (Cambrex Bio Science, Verviers, Belgium). Cells were plated in 75 or 150 cm2 flasks and cultured in Mesenchymal Stem Cells Growth Medium (MSCGM), a low glucose Dulbecco's modified Eagle's medium supplemented with L-glutamine, penicillin, streptomycin and fetal calf serum (all from Cambrex). Cells were fed every 3-4 days. When the cells reached 80% confluency they were passaged. The cells used for chondrogenic differentiation procedure were derived from the 5th passage.
29 3. MATERIALS AND METHODS
3.3.1. Chondrogenic differentiation For chondrogenic differentiation 25x104 hMSCs were placed in a 15-ml conical polypropylene tube, washed with Incomplete Chondrogenesis Induction Medium (ICIM) consisting of Differentiation Basal Medium Chondrogenic (Cambrex) supplemented with 1 mM sodium pyruvate, 0.17 mM ascorbic acid–sodium salt, proline, glutamine, 100nM dexamethasone , 1% ITS+Premix, and antibiotics: Pen/Strep (supplementation available from Cambrex as Chondrogenic SingleQuots). After washing with ICIM cells were resuspended in 0.5-ml complete chondrogenic medium (CCM) consisting of supplemented ICIM and 0,01 µg/ml TGF-β3 (R&D Systems, Wiesbaden, Germany). Cells were centrifuged at 1000 rpm for 5 minutes at RT. The pellets were maintained in culture with 1 pellet/tube and 0.5 ml CCM/tube. Medium was changed every 2–3 days.
3.3.2. Osteogenic differentiation For osteogeneic differentiation, hMSCs were seeded in 6-well plates: 3x104 cells in 2ml MSCGM per well. After 6 hours when cells were adherent MSCBM was replaced by Osteogenic Induction Medium, containing 100nM dexamentasone, 0,05mM ascorbate and 10 mM ß-glycerophosphate (all from Cambrex). Control wells were cultured in MSCBM to the end of the experiment.
3.3.2.1. Assessment of osteogenic phenotype
3.3.2.1.1. von Kossa staining Cells were fixed with 60% acetone buffered with citrate for 30 sec followed by washing twice in deionized water. The fixed cells were incubated with 2% silver nitrate under direct light from a 60-W lamp for 1h and afterwords with 2,5% sodium thiosulfate for 5 min, rinsed with deionized water and counterstained with 0,33% neutral red for 5 min, finally rinsed in tap water and mounted with Aquatex (Merck) prior to examination under the microscope.
30 3. MATERIALS AND METHODS
3.3.2.1.2. Alkaline phosphatase staining Cells in wells of the 6-well plates were fixed with citrate buffered 60% acetone for 30 sec, then gently rinsed in deionized water for 45 sec and stained with a mixture of Fast Violet B salt and Naphthol AS-MX Phosphate Alkaline Solution for 30 min, according to the manufacturer’s instructions (Sigma, Procedure No. 85). Afterwards the cells were counterstained with Mayer’s hematoxylin (Sigma) for 10 min, rinsed in tap water and mounted with Aquatex prior to examination under the microscope.
3.3.3. Adipogenic differentiation hMSCs were plated at 200,000 cells/well in a 6-well plate and grown to confluence in MSC medium. Subjecting confluent monolayers to 3 rounds of adipogenic treatment induced adipogenic differentiation. Each round consisted of 48–72 hours in adipogenic induction medium: DMEM-high glucose (4,5g/l), 1µM dexamethasone, 0,2mM indomethacin, 0,5 mM 3-isobutyl-1-methyl-xanthine, 0,01 mg/ml insulin and 10% FCS, followed by 48–72 hours in maintenance medium: DMEM-high glucose, 10% FCS and 0,01 mg/ml insulin. Cells were assayed after an additional week in maintenance medium. Control wells were cultured in MSCBM to the end of the experiment.
3.3.3.1. Assessment of adipogenic phenotype 3.3.3.1.1. Oil red “O” staining Cells were fixed with 4% formaldehyde in PBS for 10 min followed by washing twice with PBS. The cells were pretreated with 60% 2-propanol for 2min and then stained with 0,2% oil red “O” in 60% 2-propanol for 10 min. The cells were washed with 2-propanol and then with PBS. Finally, to stain nuclei, the cells were treated with Mayer`s hematoxylin for 10 min and washed in tap water. Coverslips were mounted using AquaTex.
31 3. MATERIALS AND METHODS
3.4. Cell treatment with growth factors, cytokines and inhibitors
For experiments human recombinant cytokines and growth factors were used (all purchased from R&D Systems).
TGF-β3 used for chondrogenic differentiation was dissolved in 10mM HCl / 10% ethanol to prepare 20µg/ml stock. Siliconized tips and tubes were used to prevent adhesion of TGF molecules to plastic. TGF-β3 was stored after reconstitution at -80°C for not longer as one month.
Cytokines and LPS (Sigma) were dissolved in PBS and stored at –20°C.
Cytokine Mix “promoter” consisted of: TNFα 500 ng/ml IFN γ 10 000 U/ml IL-1ß 50 nM
NO donors: Stocks of NO donors were freshly prepared prior to cell stimulation. Spermine-NONOate and DETA NONOate (Cayman) were dissolved in 10mM NaOH pH12,5 and SIN-1 (Sigma) was dissolved in PBS.
Arachidonic acid (AA) was purchased from Cayman. AA was supplied as a solution in ethanol. To change the solvent ethanol was evaporated under a gentle stream of nitrogen and dissolved in DMSO. Stock solution was stored at –20°C.
Inhibitors: Inhibitors were dissolved in appropriate solvent. In most cases stock of inhibitor was prepared in DMSO and diluted to required concentration in assay medium. Final concentration of DMSO did not exceeded 0,1%. To exclude any solvent influence on the experiment control cells were treated with vehicle. The cells were incubated with inhibitors 30 minutes before stimulation with cytokines.
32 3. MATERIALS AND METHODS
3.4.1. Stimulation of chondrocytes Chondrocytes in monolayer were seeded in 6-well or 24 well plates chondrocytes in alginate beads and chondrogenic pellets were placed in 24-well dishes. Cells (calculated 1 bead/100µl medium) were stimulated with human 1nM IL-1β for 24h; when necessary cells were preincubated with substances for 30 min and then stimulated for next 24h with IL-1β.
3.4.2. Stimulation of hMSCs At indicated differentiation days hMSCs were stimulated with 1nM IL-1β for 24h (6 pellets/well, 0,75 ml medium/well: Bio Whittaker BE12-917, with 4,5 g/l Glucose, w/o L- Glutamine and Phenol Red + HMSC Chondrogenic SingleQuots). After 24h of incubation, Griess assay was performed and the rest of supernatant was frozen for prostaglandins assessment. Pellets were harvested and resuspended in RNA-isolation buffer (RLT Buffer, Qiagen), than frozen at -80°C.
3.5. Biochemical methods
3.5.1. Determination of cellular viability Viable cells were counted in a Neubauer chamber by the trypan blue exclusion method.
3.5.2. Toxicity: release of lactate dehydrogenase (LDH) The toxicity was determined by measuring the LDH activity using a CytoTox 96 Non- Radioactive Cytotoxicity Assay (Promega, Mannheim, Germany). 50 µl of cell cuture media were used as samples, the assay was performed according to the manufacture's instructions. As positive control cells were incubated with 1% triton X- 100 and the result was calculated as 100% of LDH release (total).
3.5.3. Griess assay 150µl of cell culture supernatant was mixed with 10 µl 1% sulphanilamide (Sigma) in 0.1N HCl and 10 µl 0.1% N-(1naphthyl)-ethylenediamine (Sigma). Absorbance was - read at 544 nm in reference to 690 nm in a microplate reader. Nitrite (NO2 )
33 3. MATERIALS AND METHODS
concentrations were calculated by using a NaNO2 standard curve (0-35 µM) in cell culture medium.
3.5.4. PGE2 ELISA Chondrocyte culture media were collected and frozen at –20°C.
The amount of PGE2 was determined with a commercially available ELISA kit (R&D Systems), according to the manufacture's instructions. Triplicate assays per sample were performed, and the data were converted to μg/ml.
3.5.5. GC/MS/MS
These measurements were performed in Marburg by Dr. Horst Schweer.
In parallel concentrations of COX-products in cell culture supernatants were determined using specific gas chromatography/triple stage quadrupole mass spectrometry (GC/MS/MS) as described by Schweer et al. (Schweer et al., 1994).
3.5.6. Cytochrome c assay Superoxide production was measured using the cytochrome c reduction assay. 75µM cytochrome c was added to the cell culture media of chondrocytes, after incubation time absorption measurements were made in cell culture media using wavelength 550nm (Spectramax 384plus, Molecular Devices), the absorption maximum of reduced cytochrome c.
3.5.7. Electron Spin Resonance (ESR) For measurement of superoxide production chondrocytes grown in monolayer were washed with PBS, scraped from the plastic surface and resuspended in ESR buffer. Equal cell numbers were used for each experiment. The high cell permeable and non-toxic spin trap: 1-hydroxy-3-methoxycarbonyl-2,2,5,5- tetramethylpyrrolidine hydrochloride (CM-H), was freshly prepared under N2 and added to a final concentration of 1mM.
34 3. MATERIALS AND METHODS
Table 5. The composition of ESR buffer (100ml). Substance: Amount: Final concentration:
NaH2PO4 235mg 8,75mM Na2HPO4 761mg 63mM Glucose 100mg 0,1% NaCl 15mg 2,5mM KCl 37mg 5mM
CaCl2 20mg 1,8mM
Samples without cells but with spin trap and with polyethylene-glycolated superoxide dismutase (PEG-SOD) 50U/ml 100U/ml were measured as internal control. Spectra were recorded at room temperature in a 15µl glass-capillary (Brand) using a Bruker e- spin spectrometer (Bruker, Rheinstetten, Germany) with the following parameters: Field Center field: 3487 G Sweep width: 65G Resolution: 512 points Microwawe Frequency: 9.776 GHz Power: 3.964 mW Receiver : Receiver gain: 2.24e+002 Phase: 0.36 deg Harmonic: 1 Mod. frequency: 86 kHz Mod. amplitude: 2.06G Signal channel: Conversion. 10.240 ms Time constant: 10.240 ms Sweep time: 5.243 s
35 3. MATERIALS AND METHODS
3.5.8. SDS-Page Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed as described in the work of Laemmli (Laemmli, 1970). Composition of gels is given in a table below (Table 6). Samples were boiled for 3 min at 95°C in Laemmli buffer and shortly centrifugated. Finally, samples and protein standards were loaded onto SDS- PAGE gels. As a molecular weight marker See Blue Plus2 (Invitrogen) was used. Gels were run at 50-170 V in the standard electrophoresis cell (Biometra) or Novex Mini-Cell (Invitrogen).
Table 6. Recipes for SDS-PAGE gel mixtures. Stacking gel Separating gel Separating gel
5% 8% 10%
H2O 1,4 ml 3,6 ml 3,1 ml Tris/HCl 1,5M, pH 8,8 0,6 ml 1,9 ml 1,9 ml Acrylamide 30% 0,4 ml 2 ml 2,5 ml SDS 20% 0,15 ml 0,15 ml 0,15 ml APS 10% 13 µl 30 µl 30 µl TEMED 3 µl 5 µl 5 µl
3.5.9. Western blotting and protein detection Proteins subjected to SDS-PAGE were electroblotted onto a Hydrobond-C nitrocellulose membrane (Amersham) using a semi-dry blotting system (Multiphor II) with blotting buffer at 1mA/cm2 for 1h. Nonspecific binding sites were blocked with 5% non-fat dried milk (NFDM) freshly made in TBS/Tween rocked on the orbital shaker for 1h. Afterwards blots were rinsed 3 times in TBS/Tween and incubated with the primary antibody, diluted in the TBS + 1% NFDM, overnight at 4°C on the rotating shaker (antibodies and dilution factors are listed in the table 7). On the next day membranes were washed 3 times with TBS/Tween for 10 min and incubated with horseradish peroxidase-conjugated secondary antibody, diluted in TBS/Tween for 1h at room temperature. For visualization, the membrane after washing with TBS/Tween (3x5min) was treated with SuperSignal West Pico or SuperSignal West Femto (used only for nitrotyrosine detection) chemiluminescent substrate (both from Pierce) according to the
36 3. MATERIALS AND METHODS manufacturer's instructions. Image capture was performed using a LAS-1000 Luminescence Image analyser (Fuji Film). Obtained images were analyzed using the Aida Image Analyzer software, version 3.22 (Raytest).
Table 7. Antibodies that were used for protein immunodetection on Western blots primary antibody dilution factor secondary antibody dilution factor iNOS (from Prof. Pfeilschifter) 1:500 GAR-POX (Dianova) 1:50 000 iNOS (Cayman, 160862) 1:800 GAR-POX 1:50 000 COX-2 (Cayman, 160116) 1:500 GAR-POX 1:30 000 nitrotyrosine 1:250 GAM-POX (Dianova) 1 :30 000 (Cayman, 189 542) PCS 3 ( Dr. Thomas Klein) 1:800 GAR-POX 1:50 000 k376 (Dr. Thomas Klein) 1:250 GAR-POX 1:50 000
3.5.10. Stripping of Western blots Bound antibody was removed from Western Blot membranes by incubation in Western blot stripping buffer for 30 min at 60°C. Afterwards membranes were washed three times in TBS and blocked with 5% BSA/TBS for 30 min. After this procedure Western blots were ready for another immunodetection of proteins.
3.5.11. Two-dimensional gel electrophoresis 3.5.11.1. Cell extract. Primary human chondrocytes after isolation were grown in plastic flasks in DMEM/F12 20%FCS for three days. Confluent cells were incubated with IL-1β 0,5nM or SIN-1 250µM in DMEM/F12 10%FCS w/o phenol red. After 24h Griess assay was performed in cell culture supernatant to test the stimulation. The medium was collected and the cells were washed with PBS, loosened by trypsin/ethylenediaminetetraacetic acid (EDTA) and centrifuged (5min, 1200rpm,RT). Afterwards the cells were washed with TBS and TBS/distillated water (50/50%). Then a lysis buffer [7 M Urea, 2M Thiourea, 4% CHAPS, 1% DTT; 2% IPG-ampholytes] was added to the cell pellets.
37 3. MATERIALS AND METHODS
Further analysis (two-dimensional gel electrophoresis and protein identification) was performed at ALTANA Pharma in the department RDR/PX (Head of the Department Dr. Sascha Dammeier).
3.5.11.2. Two-dimensional gel electrophoresis Two-dimensional gel electrophoresis was performed with the IEF system (Amersham Biosceineces; Multiphor II Sytem). The first dimension used 7-cm nonlinear pH 3–10, immobilizedpH gradient (IPG) strips loading 40 µg protein, or 7-cm nonlinear pH 6,2-7,5 strips loading 100µg protein (for COX-2). IPG strips were rehydrated over night as described in the supplier’s manual and then isoelectric focusing was performed using following programs:
IPG strips pH 3-10: Step V mA W Time (h) Vh 1 0 - 200 2 5 0,01 1 2 200 - 3500 2 5 1,3 2405 3 3500 2 5 1,05 3675
total: 2,36 h 6081 Vh
IPG strips pH 6,2-7,5: Step V mA W Time (h) Vh 1 0 - 200 2 5 0,01 1 2 200 - 3500 2 5 1,3 2405 3 3500 2 5 5,00 17500
total.: 6,3 h 19,9Vh
For the second dimension, the IPG strips were equilibrated 12 to 15 min in IPG strip reducing buffer and then alkylated for 5 min with JAA in IPG strip alkylating buffer. IPG strips were placed on top of the 12% SDS-PAGE gel and covered with melted agarose in Laemmli buffer. The second dimension SDS-PAGE was performed essentially according to Laemmli. After completion of the run, the acrylamide gels were soaked in transfer buffer (20 mM
38 3. MATERIALS AND METHODS
Tris-HCl, 96 mM glycine, and 20% methanol) and then partially transferred onto a polyvinylidene difluoride (PVDF) membrane by using a wet transfer apparatus as suggested by the manufacturer. The gels were then stained with colloidal Coomassie blue (Roti Blue stain; Roth), and Western blot analysis was performed on the PVDF membrane.
3.5.11.3. Western blot analysis. PVDF membranes were blocked for 2h by using a blocking buffer [20 mM Tris, 150 mM NaCl (pH 7.5), 0.1% Tween-20, and 5% BSA]. Membranes then were probed overnight at 4°C with a monoclonal antibody against nitrotyrosine (1:5,000; clone 1A6, Upstate Biotechnology; Lake Placid, NY) in blocking buffer. The membranes then were washed four times in washing buffer [20 mM Tris, 150 mM NaCl (pH 7.5), and 0.1%Tween-20] for 15 min each wash. The membranes were then probed with a goat anti-mouse antibody (horseradish peroxidase conjugate, 1:50,000). After the membranes were washed four times in a washing buffer, the immunopositive spots were visualized by using LumiLight-Plus (Roche) as directed by the manufacturer. In experiments where samples needed to be compared, the membranes were blotted on the same membrane and exposed simultaneously.
3.5.11.4. Protein identification (MALDI-TOF). Proteins were identified by matrix-assisted laser desorption ionization - time of flight (MALDI-TOF) mass spectrometric analysis of in-gel tryptic digests of immunopositive spots. The two-dimensional gel spots were excised and cut into ca. 1mm3 cubes, and Coomassie blue was washed away using following protocol: • distilled water MilliQ, ~50µl, 15min, RT • 50% acetonitrile in 50 mM ammonium bicarbonate 15min, RT, shaking • 50% acetonitrile in 15 mM N-ethylmorpholine till gel pieces are shrunken, then acetonitrile removed • rehydration in 50 mM ammonium bicarbonate ~50µl, 5min • 50% acetonitrile ~50µl, 15min • the excess of fluid removed
39 3. MATERIALS AND METHODS
• 50% acetonitrile till gel cubes are shrunken, then acetonitrile removed • dried in Speed Vac, ~5min.
After being destained and dried the gel pieces were rehydrated in 20-25µl ammonium bicarbonate containing 2,5ng/ml trypsin and incubated overnight at 37°C. The in-gel tryptic digest was extracted with 50% acetonitrile containing 1% trifluoroacetic acid. The extract was dried and after resolving in 0.2 % TFA desalted using Zip-Tipµ-C18 (Millipore). The desalted extract was subjected to MALDI-TOF mass spectrometric analysis by using a Bruker Ultraflex instrument in reflector mode equipped with a nitrogen laser (337 nm). α-Cyano-4-hydroxy-cinnamic acid was use as Matrix. The elution solution (1µl of CH3CN/0.2 % TFA 2/1) from Zip-Tipµ-C18 desalting was directly spotted on a thin layer of the matrix. Each spectrum was accumulated for ca. 300 laser shots. Measured peptide masses were used to search with Mascot the MSDB database for protein identifications.
3.6. Staining of cartilage and chondrocytes
3.6.1. Preparation of cartilage sections The cartilage was dissected from the underlying bone and fibro-cartilaginous areas were discarded. Cartilage samples were fixed with 7,5% paraformaldehyde in PBS pH 7.4 and stored at 4° for min. 1 week. After fixation the specimens were washed with tap water for 2 hours and then decalcified in TBD-2 solution, which is a formic acid solution. The decalcifier solution was replaced every 24h. Using this protocol cartilage sections were fully decalcified in 4-7 days. To remove decalcifier from the tissue, cartilage specimens were washed for 12 h in tap water, which was frequently changed. Decalcified cartilage was dehydrated in a graded series of alcohols (70% 2-propanol: 2h 15 min, 80% 2-propanol: 2h 15min, 90% 2-propanol: 2h 15min, 100% 2-propanol: 2h 15min, 100% 2-propanol: 2h 15min, 100% 2-propanol: 2h 15min, 100% 2-propanol: 2h 15min, xylol: 2h, xylol: 2h, xylol: 2h) and embedded in low-melting-point paraffin (paraffin 60°C: 3h, paraffin 60°C: 3h) using a Shandon Citadel 2000 tissue processor.
40 3. MATERIALS AND METHODS
Tissue sectioning: The paraffin blocks were placed in a microtome and trimmed down to the cartilage surface, then 4 or 5 µm cartilage sections were cut. Tissue sections were floated in a water bath at 46°C and collected on Histo Bond glass slides. Afterwards glass slides with tissue sections were dried for 1-2 h at 60°C.
3.6.2. Histochemical and fluorescence staining of tissue samples Tissue sections were deparaffinized in xylene, passed through decreasing gradations of ethanol (100%, 80%, 50%), and washed in TBS. To enhance protein antigen accessibility, the sections were incubated with digestive enzymes. For each antibody four digestion protocols were tested (all in humid chambers): 1. hyaluronidase (2mg/ml PBS, pH 5,5; 15 min at 37°C) and then pronase (1mg/ml PBS, pH 7,5; 30 min at 37°C), 2. protease (0,02mg/ml TBS, pH 7,6; 60 min at 37°C), 3. chondroitinase ABC (0,05U/ml in 0,1M Tris-acetat buffer, pH 7,6; 30 min at 37°C), 4. trypsin-EDTA (the same as for the cell culture, 60min at 37°C). Enzymes were removed with three 2 min washes with TBS or PBS and 5% BSA in PBS for 30min was placed on the sections to block the non-specific binding. Afterwards the sections were incubated with primary antibodies. Primary antibodies were diluted with TBS containing 1% BSA immediately prior to use and placed on the sections, followed by overnight incubation in humid chamber at 4°C. After washing with TBS to remove residual primary antibody, secondary antibody was used to visualise a presence of examined protein: For immunohistochemical staining: the sections were incubated with biotin conjugated secondary antibodies for 30 min in a humid chambers, followed by washes and incubation with streptavidin-avidin-biotin- alkaline phosphatase complex (StreptABComplex/AP, DAKO) for 30 min at room temp. Finally substrate reagent: Fast Red from Sigma was added. After 15-30min reaction sections were washed in distilled water and mounted with AquaTex.
41 3. MATERIALS AND METHODS
For immunofluorescence staining: the sections were incubated with Cy3 or Cy2 conjugated secondary antibody for 30 min at room temp. in a humid chamber. Finally, nuclei were stained with Hoechst (1:10 000, 5 min) and the sections were mounted with Gel Mount. Sections were observed and photographed using a DMLB fluorescence microscope from Leica, pictures were recorded, and if necessary overlayed, using IM1000 software from Leica.
3.6.3. Cytospin preparations Chondrocytes from digested cartilage, dissolved alginate beads or from trypsinized monolayer culture were dispersed by repeated pipetting to obtain a single-cell suspension in PBS. 100µl of cell suspension were placed in cytofunnel chamber and centrifuged for 3 min, 1500rpm at medium acceleration. Cytoslides were then air-dried for 1h, afterwards fixed in 100% methanol for 3min, washed with PBS and blocked with 5% BSA in PBS for 30min. Further staining was performed to the same protocol as cartilage sections.
3.6.4. Evaluation of antibodies 3.6.4.1. iNOS TK2553 Paraffin embedded cartilage: the best results were obtained with chondroitinase digestion. For all immunostainings: 1:100 dilutions in 1%BSA/TBS, as secondary antibody: - immunofluorescence: Cy3 Goat anti-Rabbit (Dianova 111-165-144) diluted in PBS 1:800.
- immunohistochemistry: Biotin-SP-conjugated Affini Pure Donkey Anti-Rabbit IgG (Dianova 711-065-152) diluted in TBS 1:1000
3.6.4.2. β-actin / Sigma A5441 clone AC-15 For immunofluorescence: 1:500 dilutions in 1%BSA/PBS, as secondary antibody: - Cy2 Goat anti-Mouse (Dianova 115-225-146) diluted in PBS 1:200.
42 3. MATERIALS AND METHODS
3.6.4.3. Nitrotyrosine Anti-Nitrotyrosine (rabbit immunoaffinity purified IgG) / Upstate Cat. # 06-284 - immunofluorescence (cytospin): dilution: 0,5µg/ml in 1%BSA/PBS, as secondary antibody Cy2 Goat anti-Rabbit (Dianova 111-225-144) diluted in PBS 1:800. - immunohistochemistry (tissue): dilution: 5µg/ml in 1%BSA/TBS, as secondary antibody Biotin-SP-conjugated Affini Pure Donkey Anti-Rabbit IgG (Dianova 711- 065-152) diluted in TBS 1:1000. Paraffin embedded cartilage: best results obtained with protease or hyaluronidase/pronase digestion.
Anti-Nitrotyrosine, clone 1A6 / Upstate Cat. # 05-233 - immunofluorescence (cytospin): dilution: 1µg/ml in 1%BSA/PBS, as secondary antibody Cy2 Goat anti-Mouse (Dianova 115-225-146) diluted in PBS 1:200.
Anti-Nitrotyrosine, Monoclonal Antibody / Cayman Cat. # 189542 - immunofluorescence (cytospin): dilution: 0,5µg/ml in 1%BSA/PBS, as secondary antibody Cy2 Goat anti-Mouse (Dianova 115-225-146) diluted in PBS 1:200. - immunohistochemistry (tissue): dilution: 10µg/ml in 1%BSA/TBS, as secondary antibody Biotin-SP-conjugated Affini Pure Donkey Anti-Mouse IgG (Dianova 715- 065-150) diluted in TBS 1:200. Paraffin embedded cartilage: the best results obtained with trypsin or chondroitinase digestion.
The best results for nitrotyrosine staining by immunofluorescence technique were observed with Upstate polyclonal antibody and monoclonal antibody from Cayman. For cartilage embedded in paraffin the best results obtained with polyclonal antibody from Upstate after protease digestion.
3.6.4.4. PCS “PCS2” Ab (kindly provided by Dr. Thomas Klein, ALTANA Pharma) Paraffin embedded cartilage: the best results were obtained with chondroitinase digestion.
43 3. MATERIALS AND METHODS
For all immunostainings: 1:50 dilutions in 1%BSA/TBS, as secondary antibody: - immunofluorescence: Cy2 Goat anti-Rabbit (Dianova 111-225-144) diluted in PBS 1:200.
- immunohistochemistry: Biotin-SP-conjugated Affini Pure Donkey Anti-Rabbit IgG (Dianova 711-065-152) diluted in TBS 1:1000
3.6.5. TUNEL staining TUNEL staining was performed using TACS•XL – BLUE LABEL apoptosis detection kit (Trevigen). Briefly, deparaffinized sections or cells grown on chamber slides and fixed with 3,7% paraformaldehyde for 10 min were incubated with proteinase K solution for
30 minutes at 37°C, with 3% H2O2 solution for 5 minutes at room temperature, and then with terminal deoxynucleotidyl transferase (TdT) enzyme for 30min at 37°C. TdT incorporates brominated nucleotide (BrdU) at the sites of DNA fragmentation. The incorporated BrdU was detected using a higly specific and sensitive biotinylated anti- BrdU antibody. Antibody solution was applied to the sections for 30 minutes at 37°C and developed with TACS-Blue Label, which generates an intense blue staining in cells with DNA fragmentation. Nuclei were counterstained with neutral red. As a negative control, staining was done without TdT application. As a positive control, sections were incubated with TACS-Nuclease for 30 minutes at 37°C to confirm that the permeabilization and labeling reaction had worked. Sections were observed and photographed using Leica DMLB microscope, pictures were recorded using IM1000 software from Leica.
3.6.6. Haematoxylin staining Tissue sections were deparaffinized in xylene, passed through decreasing gradations of ethanol (100%, 80%, 50%), and washed in PBS. Afterwards tissue slides were stained for 10 min in Gill`s haematoxylin III solution (Merck), washed in tap water for 10 min, than shortly in distilled water, dehydrated and finally embedded using Entellan (Merck).
44 3. MATERIALS AND METHODS
3.7. Molecular biology methods
3.7.1. RNA Extraction and TaqMan PCR RNA was prepared using RNeasy Mini Kit (Qiagen, Hilden, Germany). Briefly, cell pellets were harvested and homogenized in RLT Buffer (Qiagen). Total RNA was extracted from the homogenate using manufacturer`s instructions. DNA contaminations were removed with DNA-free kit (Ambion, Dresden, Germany). Finally RNA was eluted with 50µl distilled water (Ambion) and used as a template for cDNA synthesis. To generate cDNA, RNA was reversed-transcribed using RAV2-reverse polymerase (Amersham, Freiburg, Germany) and random hexanucleotide primers (Roche Diagnostics, Mannheim, Germany). Equal portions of the first strand synthesis reaction were used for the following quantitative PCR-analysis. Real Time PCR, using the ABI SDS 7900 (Applied Biosystems, Foster City, CA, USA), was performed in a total volume of 25 µl in a 96 well plate. For all genes the final reaction mix contained: TaqMan Universal PCR Master Mix (Applied Biosystems), forward and reverse primers at final concentrations of 0,9 µM for each primer, the corresponding probe at the final concentration of 0,2 µM. The primers and probes listed in the Table 8. were designed using Primer Express or were ordered as assay on demand as indicated (Applied Biosystems). For all PCRs 10 ng cDNA was added to the reaction mix. A no template control (NTC) that contained all the above reagents except for cDNA was also included to detect the presence of contaminating DNA. All experiments were performed in triplicates or quadruplicates. Amplification and fluorescence detection was conducted with a standard program of 40 cycles. A result was found negative where no amplification occurred, i.e. the threshold cycle (Ct) value was greater than 40 cycles. For standardization of the gene expression levels determined by TaqMan analysis mRNA derived cDNA signal in each sample was calculated relative to 18s ratio as an internal control.
Assays on demand (Applied Biosystems): SOX9 Hs00165814 RUNX2 Hs00231692
45 3. MATERIALS AND METHODS
IL-1β Hs00174097 IL-1RN Hs00277299 (IL-1 receptor antagonist) IL-6 Hs00174131 PTGIR Hs00168765 (IP receptor)
Table 8. Sequences of primer and probes used for TaqMan PCR gene sequences
S GGCTCGTGCAGGACTCACA iNOS A GAGCCTCATGGTGAACACGTT
probe ACCTCAGCAAAGCCCTCAGCAGCAT S ACCCGGACAGGATTCTATGGA COX2 A ACTGTGTTTGGAGTGGGTTTCAG
probe AACTGCTCAACACCGGAATTTTTGACAAGA S CCACGAAGCTAACCTTGAGAGATC Agg 1 A CCTCGCCTTCTTGAAATGTCA
probe TGGCCTAGGAGTGAGCGGCAGC
S GCTTCACCTAGAGCGTCACTGT
Col 1 A1 A TGGTTTTGTATTCAATCACTGTCTTG
probe ATGGCTGCACGAGTCACACCGG S GGCAATAGCAGGTTCACGTACAC Col 2 A1 A GATAACAGTCTTGCCCCACTTACC
probe CCTGAAGGATGGCTGCACGAAACATACC S ACCCAACACCAAGACACAGTTCT (MGB)
A CTTGCTCTCCTCTTACTGCTATACCTTT (MGB) Col 10 A1 probe ATTCCCTACACCATAAAGA (MGB)
S GCTGCCACCACACTCAAGACC
ADAMTS5 A CGTAGTGCTCCTCATGGTCATCT probe TTGCAAGTGGCAGCACCAACACAAC S CGGCTACCACATCCAAGGAA
A 18s GCTGGAATTACCGCGGCT probe TGCTGGCACCAGACTTGCCCTC
46 3. MATERIALS AND METHODS
3.7.2. Affymetrix and bioinformatics analysis
Affymetrix analysis was performed at ALTANA Pharma in the department RDR/PX in the Lab of Dr. Hubert Paul and at ARI/USA in the Lab of Dr. Guido Grentzmann.
3.7.2.1. Microarray. Experimental procedures for GeneChip microarray were performed according to the Affymetrix GeneChip Expression Analysis Technical Manual (Affymetrix, Santa Clara, CA). In brief, 3µg of total RNA was used to synthesize double- stranded DNA (Superscript cDNA synthesis kit, Invitrogen, Carlsbad, CA). The DNA was purified by using phenol/chloroform/isoamylalcohol extraction with Phase Lock Gel (Eppendorf) and concentrated by ethanol precipitation. In vitro transcription was performed to produce biotin-labeled cRNA by using a MEGAscript II Kit (Ambion) according to the manufacturer’s instructions. Biotinylated cRNA was cleaned with an RNeasy Mini Kit (Qiagen, Hilden, Germany), fragmented to 50 to 200 nucleotides, and hybridized 16 h at 45°C to Affymetrix HG-U133 A and B arrays, which contain together approximately 39,000 human transcripts. After being washed, the arrays were stained with streptavidin-phycoerythrin (Dianova, Hamburg, Germany). Staining signal was amplified by biotinylated anti-streptavidin (Vector Laboratories) and by second staining with streptavidin-phycoerythrin and then scanned with a GeneArray Scanner (Affymetrix). An Affymetrix Genechip Operating System 1.2 (GCOS1.2) was used to analyze the scanned image and to obtain and scale quantitative information (according to the Affymetrix Statistical Algorithms Reference Guide). The raw data was imported into the Analyst Software package (GeneData, Basel, Switzerland) for further quality and statistical analysis.
3.7.2.2. Microarray data analysis
Performed by Dr. Gordana Bothe at Altana Pharma / RDR/IT
Genedata Expressionist system was used to analyse the scanned images and to obtaine and scale quantitative information. Changes in intensity of signals were calculated.
47 3. MATERIALS AND METHODS
Student`s t-test and absence/presence analysis were used to analyse and determine the significance of changes (in the medians) of studied groups: e.g. day 0 vs. day 7 or control vs. IL-1.
Transcripts with p values calculated in each experimental setup were selected for further analysis and differentially expressed genes were filtered according to the criteria that are designated for each experimental setup in the results.
48 4. RESULTS
4. RESULTS
4.1. Characterization of hMSC differentiation model
Most of studies dealing with human cartilage physiology and the role of nitric oxide in chondrocyte homeostasis were performed on OA chondrocytes, as the availability of normal human samples from young individuals is limited. It has been reported that human mesenchymal stem cells (hMSCs) from bone marrow can differentiate into several mesenchymal lineages, including bone, adipose tissue and cartilage. In the last years the interest for this cells as an alternative to chondrocytes have been arising. Nevertheless, chondrogenic differentiation is still not sufficiently characterized, as this field of stem cell biology is relatively new and studies are time and resources consuming. Therefore, before we started to examine the role of iNOS in chondrogenic differentiation we performed initial experiments dealing with a better characterization of the used model mainly in terms of gene expression. The first approach was to control the multipotency of hMSCs. Thus hMSCs from four donors were subjected to osteogenic and adipogenic differentiation. Alkaline phospatase and von Kossa staining served as a marker of osteogenic differentiation; oil red O assay was used as an indicator of adipogenesis.
Human mesenchymal stem cells HMSCs from bone marrow had the typical spindle morphology of cultured fibroblasts (Figure 9a). These cells were capable of growth and cell division in monolayer culture. However, it has been reported that the differentiation potential of these cells is decreasing with the time of culture and the number of passages (Zhang et al., 2005). Therefore we used the cells from the 5th passage for all differentiation procedures to obtain comparable results.
49 4. RESULTS
Osteogenic differentiation Under osteogenic induction by dexamethasone, ß-glycerophosphate and ascorbic acid cells from all four donors showed changes in cell morphology, from spindle shaped hMSCs to rhomboid shape as they differentiate and mineralize. At the third week of differentiation osteocytes began to delaminate from the culture surface. At this time point, assays assessing osteogenic differentiation were performed. Von Kossa staining characterized calcium deposition in these cells and alkaline phosphatase assay was performed to demonstrate presence of this enzyme. The alkaline phosphatase activity is restricted to bone marrow osteoblasts, endothelial cells and granulocytes, therefore it is a very useful marker of osteogenic differentiation. We obtained positive results with both assays (Figure 9 b-d) indicating differentiation of the hMSCs into osteocytic lineage. Control, not differentiated hMSCs were von Kossa negative and showed no alkaline phosphatase activity (data not shown).
Adipogenic differentiation At 100 % of confluence of hMSCs, three cycles of induction / maintenance were performed to stimulate adipogenic differentiation. During adipogenic differentiation cells acquired morphological features characteristic for adipocytes as accumulation of lipid droplets, which could be microscopically observed (Figure 9e). The amount and size of lipid vacuoles was increasing with the course of adipogenic differentiation. After two weeks of adipogenic differentiation adipogenesis was assayed by staining triglyceride deposits within the cells by oil red O (Figure 9f). This staining revealed that almost all cells were loaded with lipids. In contrast control cells did not have any lipid vacuoles (data not shown).
50 4. RESULTS a) b)
c) d)
e) f)
Figure 9. Morphologic and histochemic analysis of human hMSCs before and after osteogenic and adipogenic differentiation: a) undifferentiated hMSCs in monolayer b) osteoblasts delaminating from the cell culture surface after 3 weeks of osteogenic differentiation, alkaline phosphatase staining c) osteoblasts, von Kossa staining d) osteoblasts, alkaline phosphatase staining (red) e) adipocytes f) oil red o staining of lipid droplets within adipocytes.
51 4. RESULTS
Chondrogenic differentiation To induce chondrogenic differentiation hMSCs were grown as pelleted micromass cultures in serum-free DMEM supplemented with TGF-ß3. These hMSCs cultures increased in size during the first weeks of chondrogenic differentiation. Increased hMSC pellet size was due to extracellular matrix production, because the number of cells after pellet digestion decreased with the time of culture. The differentiation and maturation process of hMSCs is associated with the expression of specific genes. For chondrogenesis these genes include extracellular matrix (ECM) components such as collagen type 2 which is the major extracellular protein in cartilage and essential for normal cartilage structure and function (Tanaka et al., 2000; Sandell and Aigner, 2001). Collagen type 2 is a marker for activated functional chondrocytes. Hypertrophic chondrocytes are characterized by the expression of type 10 collagen (Sandell and Aigner, 2001; von der Mark et al., 1992). Expression of collagen type 10 is also a hallmark for chondrocytes from OA cartilage (Nah et al., 2001).The expression of collagen type 1 is typical for fibroblastic cells and dedifferentiated chondrocytes (Sandell and Aigner, 2001). Except the specific pattern of collagen expression aggrecan, a large aggregating proteoglycan, is another ECM molecule, which is used as a marker of cartilage and chondrocytic phenotype (Kolettas et al., 1995; Sandell and Aigner, 2001). We first characterized the expression of collagens and aggrecan in hMSCs cultivated under conditions promoting chondrogenesis to determine the extent of differentiation. However, except ECM components we decided to characterize the expression of the master transcription factor in chondrocyte differentiation: SOX-9, which has been previously shown to regulate collagen type 2 expression (Bell et al., 1997; Ng et al., 1997).
52 4. RESULTS
4.1.2. Collagen subtypes, aggrecan and SOX-9 expression during the course of chondrogenic differentiation
Collagen 1 expression: The TGF β3 induced differentiation of hMSCs into chondrocytes as expected led to a decrease of collagen 1 gene transcript levels in the first week of differentiation compared to undifferentiated stem cells. Afterwards, the level of collagen 1 mRNA was stable during the remaining course of chondrogenic differentiation (Figure 10).
Collagen 1 expression
2
1 relative expression relative 0 monolayer 7 14 21 28
Figure 10. Relative gene expression levels of collagen 1 during chondrogenic differentiation. The expression level of undifferentiated hMSCs was set as one and fold-exchange was calculated. PCR was performed in quadruplicate. For standardization of the gene expression levels determined by TaqMan analysis mRNA derived cDNA signal in each sample was calculated relative to 18s ratio as an internal control. Data are presented as mean ±SD.
Collagen 2 expression: Collagen 2 was not expressed in hMSCs. The differentiation of hMSCs into chondrocytes led to an expression of cartilage specific collagen 2. Collagen 2 transcripts were present after 7 and 14 days of differentiation, but on relatively low levels in comparison to later differentiation points. Interestingly, the expression of collagen 2 reached highest levels in the third – fifth week of chondrogenic differentiation (Figure 11).
53 4. RESULTS
Collagen 2 expression
2000,00
1500,00
1000,00
500,00 rel. expression 0,00 monolayer 7 14 21 28 32 46
Figure 11. Relative gene expression levels of collagen 2 during chondrogenic differentiation. The expression level of undifferentiated hMSCs was set as one and fold-exchange was calculated. PCR was performed in triplicate, For standardization of the gene expression levels determined by TaqMan analysis mRNA derived cDNA signal in each sample was calculated relative to 18s ratio as an internal control. Data are presented as mean ±SD.
Collagen 10 expression: An increase in collagen 10 mRNA levels was observed already after 7 days of chondrogenic differentiation. Further increase of collagen 10 gene expression was observed after 14 days and this expression level remained constant for the next two weeks (Figure 12).
Collagen 10 expression
100,00
75,00
50,00
25,00 relative expression relative 0,00 monolayer 7 14 21 28
Figure 12. Relative gene expression levels of collagen 10 during chondrogenic differentiation. The expression level of undifferentiated hMSCs was set as one and fold-exchange was calculated. PCR was performed in quadruplicate, For standardization of the gene expression levels determined by TaqMan analysis mRNA derived cDNA signal in each sample was calculated relative to 18s ratio as an internal control. Data are presented as mean ±SD.
Detection of collagen 10 in chondrogenic aggregates indicates that differentiating cells are becoming hypertrophic.
54 4. RESULTS
Aggrecan expression: Interestingly, high amounts of aggrecan transcripts were found in undifferentiated hMSCs. Due to this phenomenon aggrecan levels were calculated not in comparison to undifferentiated hMSCs but to hMSCs after 6 days of chondrogenic differentiation. The highest expression of aggrecan was observed in the first two weeks of chondrogenic differentiation with increase at 7th and 14th day of differentiation, afterwards expression of the aggrecan gene declined (Figure 13). We were surprised to detect high levels of aggrecan transcripts in un- differentiated hMSCs, as until recently this large proteoglycan has been localized predominantly to skeletal tissue and considered a hallmark of cartilage differentiation. We believe, that in our case aggrecan expression is characteristic for multipotential mesenchymal stem cells and not only chondrogenic progenitors, as these cells were able to differentiate also into adipocytes and osteoblasts as shown before. Interestingly, presence of aggrecan has been already reported aside from cartilage in brain development (Schwartz et al., 1996; Milev et al., 1998). Despite reports demonstrating presence of aggrecan in tissues other than cartilage it remains a biochemical marker of the cartilage phenotype thereafter.
Aggrecan expression
14,00 12,00 10,00 8,00 6,00 4,00 2,00 relative expression relative 0,00 6 7 14 18 21
Figure 13. Relative gene expression levels of aggrecan during chondrogenic differentiation. The expression level of hMSCs after 6 days of differentiation was set as one and fold-exchange was calculated. PCR was performed in triplicate, For standardization of the gene expression levels determined by TaqMan analysis mRNA derived cDNA signal in each sample was calculated relative to 18s ratio as an internal control. Data are presented as mean ±SD.
SOX-9 expression: Expression of this transcription factor, which is known to be necessary for chondrogenic differentiation, was significantly upregulated at the
55 4. RESULTS beginning of the chondrogenic differentiation. Levels of SOX-9 transcripts were high in the first week of differentiation, and then the expression of this gene gradually declined and achieved low level after 3 weeks of differentiation (Figure 14). The observed expression pattern of SOX-9 agreed with the concept that pro- chondrocytic transcription factor should be highly expressed as differentiation begins. However expression of collagen 2 did not fully correlate with SOX-9 expression. Indeed, we have seen induction of collagen 2 after 7 days of chondrogenic differentiation, when the measured SOX-9 mRNA reached the highest level, however the maximum of collagen 2 expression was observed 3 weeks later, when levels of SOX-9 transcripts declined.
Sox 9 expression
12,00 10,00 8,00 6,00 4,00 2,00
relative expression relative 0,00 Monolayer 7 14 21 32 46
Figure 14. Relative gene expression levels of SOX-9 during chondrogenic differentiation. The expression level of undifferentiated hMSCs was set as one and fold-exchange was calculated. PCR was performed in triplicate, For standardization of the gene expression levels determined by TaqMan analysis mRNA derived cDNA signal in each sample was calculated relative to 18s ratio as an internal control. Data are presented as mean ±SD.
4.1.3. Regulation of marker gene expression in pellets vs. chondrocytes in alginate beads As a next point of hMSCs model evaluation comparison with human primary chondrocytes cultivated in alginate beads was performed. We were interested, if differentiating hMSCs in micromass culture react in the same way as chondrocytes in alginate to catabolic stimuli. Therefore influence of IL-1ß or a cytokine mix on the expression of the cartilage specific markers collagen 2 and aggrecan was tested. Chondrogenic pellets or chondrocytes were stimulated over 24h with 1nM IL-1ß and
56 4. RESULTS mixture of TNFα, IFNγ and IL-1ß (Cytokine Mix Promotor). In both cell culture models catabolic cytokines had the same effect: significant downregulation of the expression of collagen 2 and aggrecan (Figure 15).
a) c)
Collagen 2 expression Collagen 2 expression
2,00 2,00
1,00 1,00 relative expression relative relative expression relative 0,00 0,00 control IL-1 1 nM CM Promotor control IL 1b CM Promotor
b) d)
Aggrecan expression Aggrecan expression
2,00 2,00
1,00 1,00 relative expression relative expression
0,00 0,00 control IL-1 1 nM CM Promotor control IL-1 1nM CM Promotor
Figure 15. Comparison of relative gene expression levels of collagen 2 and aggrecan in chondrogenic pellets after 21 days of differentiation a),b) and in primary chondrocytes embedded in alginate beads c),d). The expression level of untreated cells (control) was set as one and fold-exchange was calculated. PCR was performed in triplicates. For standardization of the gene expression levels determined by TaqMan analysis mRNA derived cDNA signal in each sample was calculated relative to 18s ratio as an internal control. Data are presented as mean ±SD.
57 4. RESULTS
4.1.4. Affymetrix gene chip characterization of chondrogenic differentiation – general gene expression analysis
While the preliminary results obtained with TaqMan PCR on chondrogenic gene expression in hMSCs pellet model were promising further detailed analysis of gene expression in this model was performed. HMSC from 4 bone marrow donors were differentiated into chondrocytes. During the course of chondrogenic differentiation at days: 0, 7, 11, 14, 18, 20, 33, 53 pellets were harvested for RNA isolation. For gene expression profiling of hMSCs the Affymetrix GeneChip HG-U133A containing 22283 probe sets was used.
Generally the obtained data sets for each differentiation day studied were compared among each other and genes, which were regulated during the differentiation course, were selected. Two different analyses of obtained data were performed: Comparison of all days - ANOVA, p value cutoff 0.01; median fold change cutoff0.7/1.3x - 298 probe sets (ps) obtained - clustering according to regulation (7clusters)