Vertical inhibition of the PI3K/Akt/mTOR pathway for the treatment of osteoarthritis


  • Dr. Jiezhong Chen,

    Corresponding author
    1. Institute of Health and Biomedical Innovation, Queensland University of Technology, Brisbane, Queensland 4059, Australia
    2. Illawarra Health and Medical Research Institute, University of Wollongong, Northfields Avenue, New South Wales 2522, Australia
    • Institute of Health and Biomedical Innovation, Queensland University of Technology, 60 Musk Avenue, Kelvin Grove, Brisbane, QLD 4059, Australia.
    Search for more papers by this author
  • Ross Crawford,

    1. Institute of Health and Biomedical Innovation, Queensland University of Technology, Brisbane, Queensland 4059, Australia
    Search for more papers by this author
  • Yin Xiao

    Corresponding author
    1. Institute of Health and Biomedical Innovation, Queensland University of Technology, Brisbane, Queensland 4059, Australia
    • Professor, Institute of Health and Biomedical Innovation, Queensland University of Technology, 60 Musk Avenue, Kelvin Grove, Brisbane, QLD 4059, Australia.
    Search for more papers by this author


Osteoarthritis is characterized by degenerative alterations of articular cartilage including both the degradation of extracellular matrix and the death of chondrocytes. The PI3K/Akt pathway has been demonstrated to involve in both processes. Inhibition of its downstream target NF-kB reduces the degradation of extracellular matrix via decreased production of matrix metalloproteinases while inhibition of mTOR increased autophagy to reduce chondrocyte death. However, mTOR feedback inhibits the activity of the PI3K/Akt pathway and inhibition of mTOR could result in increased activity of the PI3K/Akt/NF-kB pathway. We proposed that the use of dual inhibitors of PI3K and mTOR could be a promising approach to more efficiently inhibit the PI3K/Akt pathway than rapamycin or PI3K inhibitor alone and produce better treatment outcome. J. Cell. Biochem. 114: 245–249, 2013. © 2012 Wiley Periodicals, Inc.

Osteoarthritis (OA) is a joint degenerative disease affecting mostly the aged population. OA is characterized by the degenerative alteration of articular cartilage although other joint tissues are involved [Pitsillides and Beier, 2011; Jaiprakash et al., 2012; Troeberg and Nagase, 2012]. This results in pain and dysfunction of the affected joint. Cartilage is an avascular structure consisting of chondrocytes and extracellular matrix (ECM). The degenerative alteration of OA includes both the death of chondrocytes and the degradation of ECM. In OA, ECM is degraded by increased matrix metalloproteinases (MMPs), and adamalysin with thrombospondin motif 5 (ADAMTS-5), which is a key destructive process [Prasadam et al., 2012a; Troeberg and Nagase, 2012]. The main components of ECM are aggrecan and type II collagen [Pitsillides and Beier, 2011]. Many MMPs have been found to be increased in OA. Among them MMP-13, MMP-3, and MT1-MMP are considered the most important. MMP-13 not only degrades aggrecan but also preferentially digests type II collagen rather than types I and III [Billinghurst et al., 1997]. Both MMP-3 and MT1-MMP activate proMMP-13 in OA [Knauper et al., 2002]. MMP-3 can also digest aggrecan and activate many other proMMPs including proMMP-2, -7, -8, and -9. In OA, Reactive oxygen species are increased as well as cytokines due to chronic inflammation and both processes result in increasing apoptosis of chondrocytes [Blanco et al., 1998; Facchini et al., 2011]. The macromolecules produced by apoptotic chondrocytes may further increase ROS and lead to more chondrocyte death [Lopez-Armada et al., 2006]. Cell signaling pathways have been reported in the regulation of chondrocyte hypertrophy, chondrocyte and osteoblast interactions, and degradation of ECM [Prasadam et al., 2010a, b, 2012b]. The intracellular signaling pathway PI3K/Akt has also been demonstrated to be involved in both cellular and ECM alterations. Manipulation of this pathway could be a useful therapy for OA, which is currently difficult to cure.


Activation of the PI3K/Akt pathway can increase MMP production by chondrocytes via its multiple downstream target proteins. The PI3K/Akt regulates a cascade of changes through its broad target proteins such as mTOR, NF-κB, GSK-3beta, and p53 [Hay, 2005; Chen and Huang, 2009; Kalaany and Sabatini, 2009; Chen, 2010, 2011, 2012]. NF-κB, a transcriptional factor, has been demonstrated to be a key regulator of MMP production [Lin et al., 2011]. NF-κB regulates MMP-13 production via two signaling pathways E74-like factor 3 (ELF3) and Hypoxia-induced factor 2alpha/unt-related transcription factor 2 (HIF2alpha/Runx2) (Fig. 1) [Goldring et al., 2011]. ELF3 can directly bind to the promoter of MMP-13 to regulate synthesis of MMP-13 [Otero et al., 2012]. SiRNA silence of ELF3 abolished IL-1beta induced MMP13 secretion. Runx-2 is correlated with the expression of MMP-13 and is necessary for IL-1beta induced MMP13 [Mengshol et al., 2001]. Runx-2 binds to the promoter of MMP13 and controls its transcription. Inhibition of the PI3K/Akt pathway and NF-κB has been considered as an option for the treatment of osteoarthritis. Indeed, inhibition of pAkt and NF-κB by curcumin decreased IL-1beta stimulating MMP secretion [Schulze-Tanzil et al., 2004; Shakibaei et al., 2007]. In addition, activator protein-1 (AP-1) is also involved in the regulation of MMP-13 with a binding site in the promoter of MMP-13 [Benderdour et al., 2002]. It has been demonstrated that JUN and its downstream target AP-1 are key mediators of IL-1 and TNF-alpha induced MMP-13 expression [Liacini et al., 2002, 2003]. Inhibition of JUN by SP600125 and AP-1 by norhihydroguaiaretic acid (NDGA) resulted in decreased MMP-13 mRNA stimulated by TNF-alpha [Liacini et al., 2003]. Inflammatory factors, which are increased in the progressive stage of OA can lead to ECM degradation. The most commonly increased cytokines are IL1beta, IL-6, IL-6 like cytokine oncostatin M, IL17, and TNF-alpha. Activated pAkt and NF-κB have been shown to play key roles in cytokine-induced secretion of MMPs [Mengshol et al., 2001; Shakibaei et al., 2007]. IL-6 like cytokine oncostatin M (OSM) has been shown to increase MMP-1 and MMP-13 via the PI3K/Akt pathway [Litherland et al., 2008]. PI3K specific inhibitor Ly294002 inhibited OSM-induced production of MMPs in chondrocytes [Litherland et al., 2008].

Figure 1.

PI3K/Akt pathway in cartilage degeneration. PI3K/Akt can activate mTOR to suppress autophagy. PI3K/Akt can also activate NF-κB, which in turn increases the synthesis of MMPs via HIF2α and ELF3. HIF2α inhibits autophagy while ROS increases autophagy. In addition, JUN and AP-1 also regulate MMPs. mTOR can feedback inhibit the activity of PI3K/Akt pathway. PI3K, phosphoinositide 3-kinase; Akt, protein kinase B; NF-κB, nuclear factor-kB; MMPs, matrix metalloproteinases; ELF3, E74-like factor 3; HIF-2α, hypoxia-induced factor 2α; Runx2, Runt-related transcription factor 2; mTOR, mammalian target of rapamycin; ROS, reactive oxygene species; AP-1, activator protein-1.

Another interesting sample is the role of the PI3K/Akt pathway in leptin-induced MMP secretion in chondrocytes [Koskinen et al., 2011; Hui et al., 2012]. Leptin is highly increased in obesity (30-fold) and thus this study is used to explain the mechanism for obesity-associated osteoarthritis which is demonstrated by epidemiological studies and animal models [Gill et al., 2011; Runhaar et al., 2011; Anandacoomarasamy et al., 2012; Killock, 2012]. Leptin, can increase the secretion of MMPs 1, 3, 9, and 13 by chondrocytes [Hui et al., 2012]. The PI3K/Akt pathway has been demonstrated to be important in the production of these MMPs. Ly294002, a PI3K specific inhibitor can abolish the effect of leptin on the production of MMPs in chondrocytes.

The downstream target protein of PI3K/Akt pathway-mTOR has also involved in the cell death of chondrocytes in OA. In OA, increased cell death is accompanied by autophagy, a cell survival mechanism. Under stress such as nutrient deprivation, the cells can degrade unnecessary cellular components to save energy for cell survival [Levine and Klionsky, 2004]. Therefore, autophagy can reduce macromolecules produced by dead cells from chondrocytes [Sasaki et al., 2012]. These molecules can produce ROS, resulting in more cell death. The decreased autophagy with age may be an explanation for age-related OA [Carames et al., 2010]. In the early stage of OA, there is increased apoptosis of chondrocytes in the superficial and middle zones of the cartilage as indicated by increased active caspase 3 and TUNEL signal [Almonte-Becerril et al., 2010]. Correspondingly autophagy is also increased as indicated by LC3II molecules. Inhibition of mTOR by rapamycin, an mTOR inhibitor has been shown to increase autophagy in several cell types [Sabers et al., 1995; Shigemitsu et al., 1999]. Recently, rapamycin has also been demonstrated to reduce autophagy of chondrocytes in an osteoarthritis animal model [Sasaki et al., 2012; Carames et al., 2012]. Rapamycin is more effective than PI3K specific inhibitor Ly294002 because mTOR is also activated by MAPK pathway, AMPK and serum amino acids. Carames et al. [2012] tested rapamycin in osteoarthritis in C57Bl/6J mice established by transaction of the medial meniscotibial ligament and the medial collateral ligament in the right knee. The effect of rapamycin on the pathway is detected by decreased phosphorylation of ribosomal protein S6. Increased autophagy is indicated by LC3II. Rapamycin was shown to decrease the severity of osteoarthritis-like changes such as cartilage degeneration, loss of surface lamina and fibrillations. The study also showed that ADAMTS-5, an aggrecan degradation proteinase, was decreased and Rapamycin preserved the cellularity and decreased synovial inflammation with decreased IL-1beta. However, the study did not examine MMPs.


We think that a problem in using rapamycin is that it can increase the activity of the PI3K/Akt pathway. It has been demonstrated that mTOR negatively regulates PI3K/Akt activity [Harrington et al., 2005]. Silencing of mTOR with siRNA or inhibition by rapamycin of mTOR resulted in activation of the PI3K/Akt pathway [McDonald et al., 2008]. mTOR can increase p70S6K which in turn phosphorylates IRS1 and inhibits IRS1, a major upstream regulator of the PI3K/Akt pathway [Hay, 2005]. The feedback inhibition of mTOR has been considered as the reason for the limited efficacy of mTOR inhibition in cancer treatment [Cloughesy et al., 2008]. However, it is not well studied about the feedback inhibition of mTOR in the pathway when mTOR inhibitor is used for the treatment of OA. When mTOR is inhibited, the activity of PI3K/Akt pathway increases. Therefore, mTOR inhibition by rapamycin may result in the increase of the activity of the PI3K/Akt pathway. In the event of mTOR inhibition, Akt can activate NF-κB to affect MMP secretion. Therefore, rapamycin may increase autophagy to protect cartilage but it may increase MMP secretion. Evidence shows that rapamycin inhibits mTOR but increases NF-κB. In rat glomerular MC cells, rapamycin further amplified IL-beta-induced NF-κB activation [Osman et al., 2011]. This led to increased MMP-9 promoter activity although rapamycin inhibited MMP-9 mRNA stability via other mechanisms. No study has investigated the effect of rapamycin on the production of MMPs when it is used for increasing autophagy. Another adverse effect of the rapamycin is that NF-κB activated HIF2 is the suppressor of autophagy. It is not examined if this effect decreases the effect of rapamycin. Bohensky et al. [2009] showed that knockout of HIF2 by siRNA resulted in a robust increase of autophagy with decreased mTOR activity. Controversy, NF-κB can activate anti-oxidant enzymes such as superoxide dismutases and glutathione peroxidise to reduce ROS [Mathy-Hartert et al., 2008]. Inhibition of NF-κB can lead to increased ROS, which enhances autophagy [Scherz-Shouval et al., 2007].

We propose that the use of dual inhibitors of PI3K and mTOR could be a promising approach to more efficiently inhibit the PI3K/Akt pathway than rapamycin or PI3K inhibitor alone. The inhibitor of the PI3K may not be sufficient to decrease mTOR because mTOR is also regulated by the MAPK pathway and serum levels of amino acids independent of the PI3K/Akt pathway. However, inhibition of mTOR (by rapamycin) only may lead to increased activity of the PI3K/Akt/NF-κB/MMP pathway. Therefore, inhibitors with dual effect on both mTOR and PI3K/Akt are superior to PI3K or mTOR inhibitor alone. At present several dual inhibitors of PI3K and mTOR have been developed by various companies including XL765, SF1126, PI103, Bez235, BGT226 and GDC0980. Several of them have been used in clinical trials for cancer therapy and demonstrated to be well tolerated, indicating a potential application in the treatment of OA.


In conclusion, the PI3K/Akt/mTOR pathway is important in regulating the production of MMPs by chondrocytes. Therefore, selective inhibition of the pathway may prevent cartilage degeneration by reducing the production of MMPs. Here, we propose that dual inhibition of PI3K and mTOR may be a promising approach, as mTOR inhibitor alone (rapamycin) may increase PI3K/Akt activity due to the inhibitory role of mTOR on PI3K/Akt.