Synergistic combinations of diet-derived bioactives to maintain joint health and prevent osteoarthritis

Lead Research Organisation: University of East Anglia
Department Name: Biological Sciences


Osteoarthritis is a degenerative joint disease which is a leading cause of disability in the UK. Approximately 8.5 million people in the UK suffer from moderate to severe OA. These patients are predominantly older than 45 years of age, with the disease often becoming a significant burden in patients over 60 years of age . There is a lack of effective drugs available to treat osteoarthritis and the drugs given for pain relief are often not sufficient . Joint replacement is the only treatment offered to patients at later stages of the disease with 66,436 hip and 77,578 knee replacements due to OA performed in the UK in 2011. Given the current trend toward an older population, osteoarthritis, for which age is an important risk factor, will be an increasing problem for society with 17 million OA patients predicted for 2030 . The ability to slow or stop progression of the disease would significantly improve both quality of life and the economic burden of osteoarthritis.

Despite many years of research, there are no drugs to slow or stop the progression of osteoarthritis. In part, this is because drugs for use in any disease which is not life threatening must be very safe. The costs of running clinical trials in osteoarthritis are prohibitive. There is a need to develop new strategies to combat osteoarthritis.

The connection between diet and osteoarthritis has been explored to some extent. High intake of some foodstuffs has been linked to slower progression of the disease in large population-based studies and these show that diet can influence joint health. However, we need to know much more about how compounds found in the diet work, what are the optimal foods to eat, in what amount and in what combination. When we know this, we can then carefully design controlled human studies to provide clear data from which to advise the public on optimal dietary strategies to both prevent and delay the progression of osteoarthritis.

There are many compounds in food with biological activities. To date, we have focused on sulforaphane, found at high amounts in broccoli and related vegetables. Sulforaphane can prevent cartilage destruction in laboratory models and we have confirmed this ability in a mouse model of osteoarthritis. We have identified the main pathway by which sulforaphane acts in human cartilage cells. We have also identified a number of other compounds, derived from the diet, which have activity in the laboratory assays.

The current application is to gain knowledge of the mechanism(s) by which each diet-derived compound acts to protect cartilage. We will then use this to test mixtures of active compounds to identify the best combination which can protect cartilage and slow or prevent osteoarthritis. We will then test this for its impact upon the human osteoarthritic joint.

When we have collated and analysed all these data, we will then design a trial to investigate the best way to deliver this combination of compounds in the diet in order to protect the joint from osteoarthritis.

Technical Summary

Osteoarthritis (OA) is a degenerative disease of the joints characterised by degradation of articular cartilage, thickening of subchondral bone and osteophyte formation. It is a major cause of disability in the UK. There are no effective disease-modifying drugs to treat OA and pain relief is often insufficient. Approx 8.5 million people suffer from OA in the UK, with 71% of these in constant pain. Increasing age and obesity are major risk factors and changing demographics will lead to a large rise in incidence of OA.

Many phytochemicals have been proposed to have positive benefit on joint health and OA, but predominantly these have not been studied in man. The impact of dietary compounds on the joint needs to be studied with increased rigour at scientific and clinical levels to prove efficacy. Our previous research has focused on sulforaphane (SFN), an isothiocyanate from cruciferous vegetables, where we demonstrated efficacy in vitro in cartilage cell and explant assays and in vivo in a mouse model of OA. We are currently undertaking a proof-of-concept trial in man to prove dietary SFN can impact upon the human joint.

We have identified several diet-derived compounds that show efficacy in a surrogate screen of cartilage protection in human chondrocytes. Our hypothesis is that these compounds will display synergy to protect cartilage and prevent or slow the progression of OA. We propose to: (i) investigate mechanism-of-action of 11 diet-derived compounds; (ii) assay synergy between compounds, where this is more likely between compounds which act on different pathways; (iii) test key combinations of compounds in assays of cartilage destruction in vitro; (iv) test one combination in a proof-of-concept human trial to demonstrate action in the human joint.

The identification of a synergistic combination of compounds will allow us to protect intellectual property and design a clinical trial to provide evidence of prevention and slowing OA progression.

Planned Impact

These studies will add to our understanding of cartilage protection and the mechanisms of cartilage destruction. They will inform future research on the impact of diet on joint health and osteoarthritis and in optimizing clinical trial designs to address these key health issues. The research will benefit: the academic community (including the named researchers); the food and drink industry; patients with osteoarthritis and any sector of the population at risk from developing osteoarthritis (e.g. after sports injury); clinicians treating osteoarthritis and providing advice on prevention; government agencies setting dietary reference values; charities involved in patient information and education.

A summary of routes to impact include:
1. Research coming from this study will be disseminated via meetings, publications and collaborations. This includes the DRINC dissemination meetings where both academic and industry partners will have access to early data.
2. Data coming from these studies will inform the development of novel ingredients for functional foods or beverages containing optimal doses of bioactive compounds which protect cartilage. In order to be commercially viable, any intervention has to have both efficacy and associated intellectual property. By identifying synergistic mixtures of compounds, we will gain maximum efficacy in cartilage protection and have the ability to protect this.
3. Data from this project may also allow us to refine the five-a-day message and provide more specific advice to consumers on optimal intake of specific fruits and vegetables for the prevention of osteoarthritis. We would aim to collaborate with government agencies to translate these findings to understandable public health messages. We will also aim to work in patient education in this area with e.g. Arthritis Research UK, the British Society for Rheumatology and appropriate international charities.
4. It may also be possible, using either the techniques of genetic engineering or with traditional plant breeding, to enhance the levels of bioactive compounds present in a standard strain. This has already been achieved using plant breeding to create a high glucosinolate, so-called 'super broccoli'.
5. There will be training opportunities for the RA in the areas of dissemination, engagement and exploitation of the study, supported by the investigators and local professional development schemes.

As proof of concept, a mixture, predominantly of flavonoids, baicalin from Scutellaria baicalensis and catechin from Acacia catechu are marketed by Primus Pharmaceuticals for osteoarthritis as Limbrel. These were chosen as inhibitors of cyclooxygenase-2 and 5-lipoxygenase and the preparation has been denoted a 'medical food' by the FDA despite the fact that major sources of these flavonoids are not part of the diet (and indeed Limbrel has recently been associated with severe liver toxicity). This product shows the potential of flavonoids to impact upon the market in OA. This product is only licenced for osteoarthritis patients by prescription but we would see diet-derived compounds showing efficacy in prevention, rather than solely as a therapeutic agent. Several diet-derived flavonoids will be tested in the current application. Such compounds would be best delivered through the diet, in a similar way to sterol-fortified products which impact upon levels of cholesterol. E.g. EFSA recently approved cocoa flavanols to help maintain endothelium-dependent vasodilation, contributing to normal blood flow. They accepted evidence that to gain the claimed effect "200 mg of cocoa flavanols should be consumed daily. This amount could be provided by 2.5 g of high-flavanol cocoa powder or 10 g of high-flavanol dark chocolate, both of which can be consumed in the context of a balanced diet", with the general population as target population. Therefore, it is possible to gain sufficient evidence for EFSA approval of health claims.
Description Initially, dose-response curves across 2.5 - 20µM were conducted for 11 compounds (luteolin, isoliquiritigenin, curcumin, genistin, apigenin, EGCG, myricetin, naringen, polydatin, ursolic acid, sulforaphane) +/- interleukin-1, measuring the expression of MMP13, and ADAMTS5 as outcomes in C28/I2 immortalised human chondrocytes. Four compounds showed reproducible dose-dependent inhibition of >1 protease gene, these were luteolin (LUT), isoliquiritigenin (ISO), apigenin (API) SFN) and sulforaphane. Further compounds showed inhibition only at high dose. We later repeated gene expression of MMP13, MMP1, ADAMTS5 and ADAMTS4 in primary human articular chondrocytes at 10µM in cell isolates from three patients in triplicate.

At this early stage in the study, our supply for human cartilage from the Norfolk & Norwich University Hospital largely ceased since elective hip and knee replacements were heavily disrupted. We surmounted this issue by establishing a relationship with the Spire Hospital in Norwich, where NHS patients undergo such surgery in order to address waiting lists. We undertook GCP training and gained honorary contracts, also getting agreement from orthopaedic surgeons and anaesthetists in order to consent patients and collect tissue.

We initially undertook synergy experiments in the same C28/I2 cells, using the 4 compounds, in all combinations and at 6 doses. We determined IC50 for all comounds and used 3 doses either side of this for synergy experiments. When we had a supply of primary human articular chondrocytes from cartilage, we then repeated all these experiments, using at least 3 isolates of primary cells. Using the Talalay-Chou method to calculate synergy, we identified strong synergism between SFN and ISO, and weaker synergy between API and ISO. This synergy in inhibition of IL-1-induced MMP13 expression is also reflected in expression of MMP1, ADAMTS4 and ADAMTS5 . There is also synergy in induction of HO1, an Nrf2-responsive gene, and in inhibition of IL-1-induced A20, an NF?B-responsive gene.

The signalling pathways modulated by these dietary factors were assayed in three isolates of HACs. SFN, API and LUT all inhibit IL-1-induced NF?B signalling in a dose-dependent manner. Only API significantly inhibits TGFß-induced Smad2/3 signalling, whilst both SFN and ISO show a pattern to induction of BMP6-induced Smad1/5/8 signalling (though this does not reach statistical significance). Similarly, SFN may show a weak inhibition of Wnt3a-induced canonical Wnt signalling.

Again in three independent isolates of HACs, ISO and LUT show dose-dependent direct anti-oxidant activity, though this is not shown by SFN and API.

A kinase array has been performed, though the kinases identified in this array have not been successfully replicated across dose- or time-responses using Western blot.

Assays have been performed in the bovine nasal cartilage assay for all individual compounds across a dose-range and in four independent tissues. Proteoglycan degradation, measured by glycosaminoglycan (GAG) release, shows a dose-dependent inhibition of IL-1 -induced GAG release for SFN and LUT, with some inhibition at the highest dose for API and ISO. Collagen degradation is measured using a hydroxyproline assay and this requires hydrolysis of cartilage using 6M HCl. This is then dried using an acid resistant centrifugal evapourator (SpeedVac). Unfortunately, this machine broke and it took many months to purchase a replacement

We have not reached the final objective of feeding the synergistic mixture of compounds to patients ahead of their knee replacement surgery and measuring the levels of compounds or their metabolites in both plasma and synovial fluid in order to show that they penetrate the joint. This is a disappointment to us, but has been the consequence of the problems we have had during the project. However, we have gained key data in identifying additional compounds that have potential to be chondroprotective, preventing or slowing the progression of osteoarthritis. These data are novel and have now been published (Davidson et al Sci Rep. 2018 8(1):17173). We have also submitted a grant application to allow us to pursue these findings in animal models of osteoarthritis and in man.

During the early part of this project, we were also able to complete the analyses of the human broccoli feeding study which was performed during our first DRINC grant but which remained to be analysed at this time. This has also been published (Davidson et al. 2017 Sci Rep. 7:3398) and this has enabled us to gain funding (Arthritis Research UK) to perform a clinical trial to determine the outcome of broccoli consumption on pain and physical function in osteoarthritis which has recently begun.
Exploitation Route The research is not complete yet, but in the long term, we aim to identify a synergistic mix of dietary compounds which will prevent the onset or slow the progression of osteoarthritis. We have submitted funding applications to take this forwards.
Sectors Agriculture, Food and Drink,Healthcare

Description We have been involved in dissemination and engagement activities around this award.
First Year Of Impact 2014
Sector Education,Other
Impact Types Societal

Description Clinical
Amount £146,956 (GBP)
Organisation Versus Arthritis 
Start 04/2018 
End 06/2020
Description Plant derived bioactive compounds 
Organisation Royal Botanic Gardens, Kew
Country United Kingdom 
Sector Charity/Non Profit 
PI Contribution Our collaborator has supplied plant extracts and purified compounds for the Clark lab to test in their models of chondroprotection. She has also helped write grant applications which are currently submitted.
Collaborator Contribution see above
Impact Grant applications under submission
Start Year 2017
Description Proteomics 
Organisation University of Liverpool
Country United Kingdom 
Sector Academic/University 
PI Contribution Supply of samples from clinical trial
Collaborator Contribution Proteomic analysis of synovial fluid
Impact Davidson et al 2017 Sci Rep
Start Year 2014
Description Clinical Trials Day 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Local
Primary Audience Professional Practitioners
Results and Impact Poster presentation to meeting
Year(s) Of Engagement Activity 2015
Description Orthopaedic Walk 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Local
Primary Audience Public/other audiences
Results and Impact Poster presentation at Norwich Orthopaedic Family Fun Day
Year(s) Of Engagement Activity 2015
Description School visit (Wymondham College) 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Local
Primary Audience Schools
Results and Impact Talk and discussion with sixth form students
Year(s) Of Engagement Activity 2014