Background Rheumatoid Arthritis (RA) is a debilitating inflammatory disease of the joints afflicting around 1% of Western populations. Some of the best treatments are anti-TNF agents, but these only achieve remission in 60% of patients and cause deleterious side effects. Progranulin (PGRN), a cysteine-rich, multi-domain growth factor was reported to bind TNF-receptors (TNFR) blocking pro-inflammatory signalling2, as well as stimulating chondrogenesis3. PGRN is cleavable and its peptides have pleiotropic effects, some of which may be beneficial and others refractory to ameliorating RA. Granulin A (GRN A) was shown to interact with cartilage ECM protein COMP4. Atsttrin – comprising 3 fused PGRN regions, was shown to ameliorate arthritic disease2. The latency associated peptide (LAP) of TGFβ1 can be fused to short peptides and cytokines via a MMP cleavage site to facilitate targeting to inflamed sites such as RA joints, reducing side effects and enhancing in vivo half-life1. We hypothesised that a peptide based on PGRN could be fused to LAP and used to both block TNF and stimulate cartilage regeneration in RA joints in a targeted way.
Objectives To produce a panel of PGRN derivatives fused to LAP. To determine the chondrogenic and anti-TNF capacities of the fragments in the presence and absence of MMP activation. To evaluate their efficacy in the CIA model of RA.
Methods Micromass cultures of C28/I2 chondrocytes and C3H10T1/2 mesenchymal cells were employed to determine the ability of PGRN fragments to stimulate chondrogenesis. HT-29 and WEHI-164 TNF sensitive cells were used to evaluate the anti-TNF capacity of LAP-PGRN fragments. Co-immunoprecipitation was used to characterise interactions between PGRN fragments and TNFRII. DBA/1 fibroblasts transduced with lentivirus encoding LAP-PGRN fusions were delivered to DBA/1 mice with CIA to assess anti-arthritic effects.
Results PGRN, LAP-PGRN, LAP-GRN A and LAP-Atsttrin were cloned and expressed in mammalian expression systems. PGRN elevated sulphated proteoglycan production in micromass cultures of C28/I2 human chondrocytes, and also stimulated proliferation of C3H10T1/2 cells. Furthermore, PGRN reduced TNF-mediated catabolism of extracellular matrix in established chondrogenic C3H10T1/2 micromass cultures. LAP-PGRN, LAP-GRN A and LAP-Atsttrin potentiated BMP2 mediated chondrogenesis in C3H10T1/2 micromasses. PGRN failed to protect WEHI-164 fibroblasts or HT-29 colorectal carcinoma cells from TNF-mediated cytotoxicity despite interacting with TNF receptor in vitro by co-immunoprecipitation. LAP-PGRN, LAP-GRN A and LAP-Atsttrin all failed to protect WEHI-164 cells from TNF-mediated cytotoxicity, even after MMP1 cleavage and release. CIA disease progression was reduced in DBA/1 mice treated with autologous fibroblasts overexpressing murine Etanercept or PGRN relative to control treatment. LAP-Atsttrin was more effective than LAP-PGRN and LAP-GRN A at reducing arthritic symptoms relative to LAP-Empty controls.
Conclusions These findings suggest that PGRN could be used as a targeted dual-function chondrogenic and anti-inflammatory treatment for RA, but these effects are not elicited directly through the TNF pathway.
Adams et al., Nat Biotechnol. 2003 Nov;21(11):1314–20.
Tang et al., Science. 2011 Apr 22;332(6028):478–84.
Feng et al., FASEB J. 2010 Jun;24(6):1879–92.
Xu et al., J Biol Chem. 2007 Apr 13;282(15):11347–55.
Disclosure of Interest None declared