If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
IL-1β inhibition combined with cholesterol-lowering therapies decreases synovial lining thickness and spontaneous cartilage degeneration in a humanized dyslipidemia mouse model
Address correspondence and reprint requests to: P.L.E.M. van Lent, Radboud University Medical Center, Geert Grooteplein 28, 6525 GA Nijmegen, the Netherlands. Tel: 31-243610512; Fax: 31-243540403.
Both systemic inflammation and dyslipidemia contribute to osteoarthritis (OA) development and have been suggested as a possible link between metabolic disease and OA development. Recently, the CANTOS trial showed a reduction in knee and hip replacements after inhibition of IL-1β in patients with a history of cardiovascular disease and high inflammatory risk. In this light, we investigated whether inhibition of IL-1β combined with cholesterol-lowering therapies can reduce OA development in dyslipidemic APOE∗3Leiden mice under pro-inflammatory dietary conditions.
Materials and methods
Female ApoE3∗Leiden mice were fed a cholesterol-supplemented Western-Type diet (WTD) for 38 weeks. After 14 weeks, cholesterol-lowering and anti-inflammatory treatments were started. Treatments included atorvastatin alone or with an anti-IL1β antibody, and atorvastatin combined with proprotein convertase subtilisin-kexin type 9 (PCSK9) inhibitor alirocumab without or with the anti-IL1β antibody. Knee joints were analyzed for cartilage degradation, synovial inflammation and ectopic bone formation using histology at end point.
Results
Cholesterol-lowering treatment successfully decreased systemic inflammation in dyslipidemic mice, which was not further affected by inhibition of IL-1β. Synovial thickening and cartilage degeneration were significantly decreased in mice that received cholesterol-lowering treatment combined with inhibition of IL-1β (P < 0.01, P < 0.05, respectively) compared to mice fed a WTD alone. Ectopic bone formation was comparable between all groups.
Conclusion
These results indicate that inhibition of IL-1β combined with cholesterol-lowering therapy diminishes synovial thickening and cartilage degeneration in mice and may imply that this combination therapy could be beneficial in patients with metabolic inflammation.
Osteoarthritis (OA) is the leading cause of disability worldwide and is characterized by joint pain and stiffness. Treatment options are currently limited to lifestyle interventions and analgesics, and patients ultimately need whole joint replacement at end-stage disease. OA is a complex and heterogeneous disease that affects all joint structures such as cartilage, synovium, subchondral bone and ligaments
. MetS is characterized by a cluster of metabolic conditions (visceral obesity, diabetes and insulin resistance, dyslipidemia and hypertension) and its prevalence has increased over the last decades. An increased prevalence of MetS in the OA population was observed compared to the non-OA population
Dyslipidemia is frequently described as an increase of circulating lipid levels, such as cholesterol and triglycerides. Several clinical studies showed an association between high cholesterol levels and the development of OA
Osteoarthritis development is induced by increased dietary cholesterol and can be inhibited by atorvastatin in APOE∗3Leiden.CETP mice--a translational model for atherosclerosis.
. Additionally, we and others have shown that dietary cholesterol also exacerbates disease development in both the collagenase-induced OA (CiOA) model and a surgically induced OA model
Synovial macrophages promote TGF-beta signaling and protect against influx of S100A8/S100A9-producing cells after intra-articular injections of oxidized low-density lipoproteins.
Cholesterol accumulation caused by low density lipoprotein receptor deficiency or a cholesterol-rich diet results in ectopic bone formation during experimental osteoarthritis.
. Although cholesterol-lowering therapies successfully reduce the risk of cardiovascular disease, the exact role of high cholesterol and cholesterol-lowering therapies on OA progression remains unclear. Results from clinical studies using statins are variable, showing beneficial or no effects on the incidence or progression of OA
. These inconsistent results imply that additional mechanisms, combined with high cholesterol levels, are likely involved in diet-induced OA development.
Low-grade systemic inflammation, associated with metabolic disease, is thought to contribute to OA development and has been suggested as a connection between MetS and OA
. Increased levels of circulating inflammatory markers and adipokines are observed in OA patients and have been associated with increased joint inflammation
Synovitis: a potential predictive factor of structural progression of medial tibiofemoral knee osteoarthritis–results of a 1 year longitudinal arthroscopic study in 422 patients.
, and OA patients frequently present with various degrees of synovial inflammation. Although inflammation is believed to be involved in the pathogenesis of OA, clinical trials targeting inflammatory factors proved disappointing in ameliorating disease progression. Interleukin-1β (IL-1β) is an important driver of inflammation and is thought to be involved in OA pathology. Despite the fact that IL-1β was thought to play a critical role in the pathogenesis of OA, several pre-clinical and clinical studies were not able to demonstrate any disease-modifying effects after inhibition of IL-1β
Gene deletion of either interleukin-1β, interleukin-1β–converting enzyme, inducible nitric oxide synthase, or stromelysin 1 accelerates the development of knee osteoarthritis in mice after surgical transection of the medial collateral ligament and partial medial meniscectomy.
. Recently, a secondary analysis of the Canakinumab Anti-inflammatory Thrombosis Outcome Study (CANTOS) trial showed that systemic inhibition of IL-1β resulted in a decreased incidence of total knee and hip replacements in patients with CVD
Effects of interleukin-1β inhibition on incident hip and knee replacement: exploratory analyses from a randomized, double-blind, placebo-controlled trial.
. Interestingly, this IL-1β effect was found on top of cholesterol-lowering treatment, since over 93% of patients enrolled in the CANTOS study received cholesterol-lowering therapy at the time the anti-IL1-β treatment started. This study indicates that inhibition of IL-1β could be beneficial in a subset of patients with high inflammatory risk or metabolic disease.
In a previous study we have shown that therapeutic cholesterol-lowering treatment in hypercholesterolemic mice successfully attenuated atherosclerosis progression
. In this study, we investigated the effects of IL-1β inhibition on top of high-intensive cholesterol-lowering treatments on spontaneous OA development in dyslipidemic APOE∗3Leiden mice with elevated systemic inflammation.
Materials and methods
Animals
ApoE∗3Leiden (E3L) and ApoE∗3Leiden.CETP (E3L.CETP) transgenic mice fed fat- and cholesterol-containing diets have high translatability for lipoprotein metabolism and are well-established models to study dyslipidemia, atherosclerosis and metabolic disease
Modulation of very low density lipoprotein production and clearance contributes to age-and gender-dependent hyperlipoproteinemia in apolipoprotein E3-Leiden transgenic mice.
Innovative pharmaceutical interventions in cardiovascular disease: focusing on the contribution of non-HDL-C/LDL-C-lowering versus HDL-C-raisingA systematic review and meta-analysis of relevant preclinical studies and clinical trials.
The APOE3-leiden heterozygous glucokinase knockout mouse as novel translational disease model for type 2 diabetes, dyslipidemia, and diabetic atherosclerosis.
. The study was carried out in female E3L and E3L.CETP transgenic mice on a C57BL/6 background (8–12 weeks of age), obtained from the in-house breeding colony (TNO Metabolic Health Research, Leiden, the Netherlands). Female mice were used since they are more susceptible to cholesterol-supplemented diets and develop more pronounced atherosclerosis due to higher systemic cholesterol and triglyceride levels compared to male mice
Modulation of very low density lipoprotein production and clearance contributes to age-and gender-dependent hyperlipoproteinemia in apolipoprotein E3-Leiden transgenic mice.
. All groups had 15 mice per group at the beginning of the study. Two animals of the Western-Type diet (WTD) + atorvastatin group were excluded from the study, one mouse was found dead in cage and one mouse was taken out of the study because of human endpoints. The study was approved by the Governmental Central Committee on animal experiments (AVD5010020172064) and the Institutional Animal Care and Use Committee of TNO (TNO-400) and was in compliance with European Community specifications regarding the use of laboratory animals.
Diet and treatments
8–12 week old E3L and E3L.CETP mice were fed a cholesterol-supplemented WTD for 38 weeks. After 14 weeks of WTD feeding, mice were matched into several treatment groups based on age, body weight, plasma total cholesterol (TC), plasma total triglycerides (TG) and cholesterol exposure (mmol/L∗weeks). Treatments included atorvastatin (10 mg/kg bw/d) (A), atorvastatin combined with proprotein convertase subtilisin-kexin type 9 (PCSK9) inhibitor alirocumab (10 mg/kg bw/wk) (AA) or with a mouse IL-1β blocking monoclonal antibody (10 mg/kg bw/wk) (AI), or triple treatment (AAI). For more details about the diet and treatments, see the online supplementary methods.
Determination of systemic cholesterol and inflammatory factors
Plasma cholesterol levels were monitored throughout the study. Peripheral blood was collected via tail vein incision using EDTA-coated tubes (Sarstedt) after 4 h of food deprivation or by heart puncture at sacrifice. Total cholesterol levels were determined throughout the study using an enzymatic assay (Roche Diagnostics) according to manufacturer's instructions. Plasma levels of Serum amyloid A (SAA), E-selectin and Monocyte chemoattractant protein-1 (MCP-1) were determined in individual plasma samples at end point using ELISA kits from R&D (Minneapolis, MA, USA), and SAA with an ELISA kit from Tridelta Development Limited (Maynooth, County Kildare, Ireland). All assays were performed according to the manufacturer's instructions.
Histology and histological analysis
Murine knee joints were fixed in formalin and decalcified using formic acid. Joints were subsequently embedded in paraffin and cut in 7 μm sections. Sections were stained using Safranin-O/Fast Green and Hematoxylin/Eosin for histological analysis. Cartilage damage in the joint was quantified using a more detailed version of the OARSI scoring system, as described previously
(0 = no damage, 30 = maximal damage). The cartilage damage score represents the assessment of the grade (progression of damage into the cartilage; 0–6) multiplied by the grade (percentage of the damaged cartilage surface; 0–5). Five sections from various depths, representing the entire knee joint, were scored and averaged per joint after blinding. Osteophyte formation and maturation were determined using an arbitrary scoring system as described previously
Resemblance of osteophytes in experimental osteoarthritis to transforming growth factor β–induced osteophytes: limited role of bone morphogenetic protein in early osteoarthritic osteophyte formation.
Resemblance of osteophytes in experimental osteoarthritis to transforming growth factor β–induced osteophytes: limited role of bone morphogenetic protein in early osteoarthritic osteophyte formation.
. Synovial thickness was scored using a scoring range from 0 to 3 in a blinded fashion (0 = no thickening, 1 = low thickening, 2 = mild thickening, 3 = moderate thickening). Three sections per joint were scored and averaged using a fixed position in the joint.
Immunohistochemistry
For immunohistochemical analysis, knee joint sections were deparaffinized and endogenous peroxidase was blocked with H2O2 in methanol. Antigen retrieval was performed in citrate buffer pH 6.0. Sections were stained with polyclonal antibodies against S100A9 (kindly provided by Thomas Vogl, Institute of Immunology, University of Muenster, Germany) and MMP13 (ab219620) or non-relevant rabbit IgG control (R&D systems). Biotinylated anti-rabbit IgG was used as a secondary antibody. Subsequently, sections were stained with avidin-streptavidin-peroxidase (Elite kit, Vector Laboratories) and diaminobenzidine (Sigma–Aldrich) was used for visualization of peroxidase staining. Counterstaining was performed using hematoxylin (Merck).
Statistical analysis
We determined the statistical power of our study based on the main readout parameter, cartilage degradation. The mean differences between groups detectable with a power of 0.8 for each analysis was determined using 15 mice per group, a two sided t-test, a 95% confidence interval and the observed SD, which is 1.7 for our main read-out parameter cartilage degeneration. This resulted in a detectable difference of 1.8. Statistical analysis was performed using GraphPad Prism version 9. Normality was assessed using a Shapiro–Wilk test. A nonparametric Kruskal–Wallis test followed by a Dunn's Multiple Comparison Test was used for comparisons of the control group with different treatment groups. P-values below 0.05 were considered significant. For cholesterol exposure, a parametric One-Way ANOVA followed by a Bonferroni post hoc test to was used. Results are expressed as individual data points with mean ± 95% confidence intervals.
Results
Dyslipidemic E3L mice show an increased inflammatory profile compared to E3L.CETP mice
We first validated the most optimal translational mouse model with diet-induced systemic inflammation to study cholesterol-lowering treatment combined with inhibition of IL-1β in mice. Therefore, we compared systemic cholesterol levels and the inflammatory profile in plasma of E3L and E3L.CETP mice fed a pro-atherogenic hypercholesterolemic diet. After 14 weeks of WTD feeding, plasma cholesterol levels were comparable between E3L and E3L.CETP mice (25.7 mmol/L and 27.2 mmol/L, respectively). E3L mice fed a cholesterol-supplemented WTD showed higher systemic levels of SAA and E-selectin when compared to E3L.CETP mice, while MCP-1 levels were comparable between both strains [SAA: 3-fold difference; P < 0.001, E-selectin: 1.3-fold difference; P < 0.001; Fig. 1(A)–(C)]. Hence, we selected the more pro-inflammatory E3L model for further analysis of the treatments and analyzed spontaneous OA development in their knee joints.
Fig. 1E3L mice show higher levels of inflammatory markers when fed a cholesterol-supplemented WTD compared to E3L.CETP mice. E3L and E3L.CETP mice were fed a cholesterol-supplemented WTD for 14 weeks. Plasma levels of SAA, E-selectin and MCP-1 were determined after sacrifice. (A) SAA levels were significantly higher in E3L mice compared to E3L.CETP mice (P < 0.001). (B) E-selectin were significantly higher in E3L mice compared to E3L.CETP mice (P < 0.001). (C) MCP-1 levels were comparable between both mouse strains. Results are expressed as individual data points with mean ± 95% confidence intervals.
Cholesterol-lowering treatments successfully reduce systemic cholesterol levels and inflammatory markers, which was not further affected by inhibition of IL-1β
We first determined the efficacy of the cholesterol-lowering treatments throughout the study. Mice were fed a cholesterol-supplemented WTD for 38 weeks and treatments were started after 14 weeks [Fig. 2(A)]. Atorvastatin treatment successfully reduced systemic cholesterol levels and cholesterol exposure (A: −34% compared to WTD control; P < 0.0001, AI: −28%; P < 0.0001, AA: −55%; P < 0.0001, AAI: −54%; P < 0.0001), which was further reduced when combined with inhibition of PCSK9 (P < 0.0001 [Fig. 2(B) and (C)]. In addition, we checked whether the anti-IL-1β antibodies lowered IL-1β levels properly. Since IL-1β levels in the circulation are very low and at the limit of detection, we measured its levels in liver homogenates. Treatment with the anti-IL-1β antibodies significantly reduced hepatic IL-1β levels at endpoint compared to the WTD control (AI: −83%; P < 0.0001, AAI: −82%; P < 0.0001) and to mice that received cholesterol-lowering therapy alone (A-AI: −82.2%; P < 0.001), AA-AAI: −67.1%; P < 0.05) [Fig. 2(D)]. We next determined whether the treatment combinations reduced circulating inflammatory markers by quantification of SAA, E-selectin and MCP-1 at end point. Atorvastatin treatment significantly reduced levels of all inflammatory markers compared to the WTD group, and there were no additional effects of more intensive cholesterol lowering by PCSK9 inhibition and by inhibition of IL-1β [SAA: A: −51%; P < 0.0001, AI: −48%; P < 0.0001, AA: −33%, AAI: −39%; E-selectin: A: −30%; P = 0.006, AI: −27%; P = 0.01, AA: −23%; P = 0.05, AAI: −15%; MCP-1: A: −31%; P = 0.03, AI: −39%; P-0.004, AA: −46%; P = 0.0002, AAI: −40%; P = 0.003; Fig. 2(E)–(G)]. We observed a significant correlation with cholesterol exposure and plasma MCP-1 levels, but no significant correlation was observed for SAA or E-selectin [Fig. 2(H)–(J)].
Fig. 2Cholesterol-lowering treatment attenuates dyslipidemia and reduces systemic cytokine levels. E3L mice were fed a cholesterol-supplemented WTD for 38 weeks, treatments were started after 14 weeks. (A) Schematic overview of the study design and the different treatment groups. (B) Plasma cholesterol levels are successfully reduced after cholesterol-lowering treatments, the combination treatments being most effective. (C) Cholesterol exposure based on systemic cholesterol levels in time was reduced upon cholesterol-lowering treatment (A: P < 0.001, AI: P < 0.001, AL: P < 0.001, AAI: P < 0.001). (D) IL-1β levels were quantified in liver homogenates. Treatment with the anti-IL-1β antibodies significantly reduced hepatic IL-1β levels at endpoint compared to the WTD control (AI: P < 0.001, AAI: P < 0.001) (E–G) SAA, E-selectin and MCP-1 plasma levels were determined at end point. Cholesterol-lowering treatment successfully reduced inflammatory mediators, which was not further affected by inhibition of IL-1β (SAA: A: −51%; P < 0.0001, AI: −48%; P < 0.0001, AA: −33%, AAI: −39%; E-selectin: A: −30%; P = 0.006, AI: −27%; P = 0.01, AA: −23%; P = 0.05, AAI: −15%; MCP-1: A: −31%; P = 0.03, AI: −39%; P-0.004, AA: −46%; P = 0.0002, AAI: −40%; P = 0.003) (H–J) Correlation of cholesterol exposure with systemic cytokines showed a mild correlation with MCP-1 (Spearman R: 0.43). ∗ = P < 0.05, ∗∗ = P < 0.01, ∗∗∗ = P < 0.001. A = atorvastatin, AI = atorvastatin + anti-IL-1β, AA = atorvastatin + alirocumab, AAI = atorvastatin + alirocumab + anti-IL1β. n=13–15 mice per group. Results are expressed as individual data points with mean ± 95% confidence intervals.
Intensive cholesterol-lowering combined with inhibition of IL-1β results in a decrease of synovial thickness in knee joints of dyslipidemic mice
Next we determined whether these treatments were able to ameliorate OA pathology in the knee joints of these mice. As inflammatory processes in the joint contribute to the development of structural OA pathology, we first determined whether the treatments affected synovial thickness. Although we observed only mild synovial thickening, intensive cholesterol-lowering with atorvastatin and alirocumab significantly reduced synovial thickness compared to mice fed a cholesterol-supplemented WTD alone, which was not further decreased by inhibition of IL-1β [AA: −52%; P < 0.05, AAI: −62%; P < 0.01, Fig. 3(A) and (B)]. To investigate the inflammatory state of the synovium in more detail, we performed immunohistochemical analysis for S100A9 as a marker for activated monocytes and macrophages. We observed only minor expression of S100A9 in the synovium, supporting the observation that local activation of myeloid cells was minor in this model [Fig. 3(C)].
Fig. 3Synovial thickness is reduced in mice receiving intensive cholesterol-lowering treatment combined with inhibition of IL-1β. E3L mice were fed a cholesterol-supplemented WTD for 38 weeks, treatments were started after 14 weeks. Synovial thickness was determined at end point using H&E stained sections. (A) Combination treatments reduced synovial thickness compared to mice fed a WTD alone (AA P < 0.05, AAI P < 0.01). (B) Representative pictures of synovial thickness. (C) S100A9 staining in the knee joints showed only minor expression in the synovium. ∗ = P < 0.05, ∗∗ = P < 0.01. A = atorvastatin, AI = atorvastatin + anti-IL-1β, AA = atorvastatin + alirocumab, AAI = atorvastatin + alirocumab + anti-IL1β. n = 13–15 mice per group. Results are expressed as individual data points with mean ± 95% confidence intervals.
Ectopic bone formation and maturation is not affected by cholesterol-lowering treatment and inhibition of IL-1β in dyslipidemic mice
We next determined whether the different treatments affected ectopic bone formation by determining the total amount of osteophytes and their maturation stage at end point. Ectopic bone formation was determined at several sites throughout the joint and their maturation stage was determined ranging from chondrogenesis to the formation of mature bone. The treatment regimens did not show an effect on ectopic bone formation regarding the number of osteophytes, or on their maturation stage compared to the WTD control group at end point [Fig. 4(A) and (B)].
Fig. 4Ectopic bone formation is not affected by cholesterol-lowering treatment and inhibition of IL-1β in dyslipidemic mice. E3L mice were fed a cholesterol-supplemented WTD for 38 weeks, treatments were started after 14 weeks. Ectopic bone formation was determined throughout the joint using SafO/FG stained sections. (A) The total number of osteophytes throughout the joint was comparable between all treatment groups. (B) Maturation of ectopic bone (0 = chondrogenic, 10 = mature bone) was similar between all groups and not affected by cholesterol-lowering treatment or inhibition of IL-1β. A = atorvastatin, AI = atorvastatin + anti-IL-1β, AA = atorvastatin + alirocumab, AAI = atorvastatin + alirocumab + anti-IL1β. n = 13–15 mice per group. Results are expressed as individual data points with mean ± 95% confidence intervals.
Inhibition of IL-1β combined with intensive cholesterol-lowering treatment lowers spontaneous cartilage degeneration in dyslipidemic mice
Finally, we determined whether inhibition of IL-1β on top of cholesterol-lowering treatments could ameliorate cartilage pathology in dyslipidemic mice. Similar to previous observations in a diet-induced model of spontaneous OA
Osteoarthritis development is induced by increased dietary cholesterol and can be inhibited by atorvastatin in APOE∗3Leiden.CETP mice--a translational model for atherosclerosis.
, cartilage damage was mild and most cartilage damage occurred at the lateral site of the joint at end point [Fig. 5(A)–(C)]. A significant correlation between synovial thickness and cartilage damage was observed (Spearman r = 0.44, P = 0.0001, Supplementary Fig. 1). Cholesterol-lowering treatments alone did not ameliorate spontaneous cartilage damage compared to the WTD control group. Interestingly, inhibition of IL-1β on top of atorvastatin and alirocumab treatment resulted in a significant decrease in cartilage damage when compared to mice fed a WTD alone (AAI: −47% P < 0.05) [Fig. 5(A) and (C)]. To investigate enzymatic activity in the articular cartilage in more detail, we performed an MMP13 staining. MMP13 expression was absent in the articular cartilage in all groups [Fig. 5(D)].
Fig. 5Cartilage damage is reduced in dyslipidemic E3L mice after intensive cholesterol-lowering treatment combined with inhibition of IL-1β. E3L mice were fed a cholesterol-supplemented WTD for 38 weeks, treatments were started after 14 weeks. Cartilage damage was determined at end point using an adapted OARSI scoring system (0–30). (A) OA scores were mild and cholesterol-lowering alone did not reduce cartilage damage compared to the WTD control. Triple treatment with atorvastatin, alirocumab and an IL-1β antibody significantly reduced cartilage damage compared to WTD controls (AAI: P < 0.05). Cartilage damage score represents an average of the four individual compartments that were analyzed. (B) Most damage was observed at the lateral side of the joint. (C) Representative pictures of cartilage damage. (D) MMP13 staining was absent in articular cartilage ∗ = P < 0.05. A = atorvastatin, AI = atorvastatin + anti-IL-1β, AA = atorvastatin + alirocumab, AAI = atorvastatin + alirocumab + anti-IL1β. n = 13–15 mice per group. Results are expressed as individual data points with mean ± 95% confidence intervals.
The association of MetS and OA has become increasingly recognized and metabolic inflammation has been suggested as a potential link between both diseases. In this study in mimicking the CANTOS Study, we showed that inhibition of IL-1β combined with intensive cholesterol-lowering treatment resulted in a reduction of both synovial thickness and spontaneous cartilage degeneration in knee joints of diet-induced dyslipidemic E3L mice. Lowering of cholesterol levels resulted in a decrease in systemic levels of inflammatory mediators SAA, E-selectin and MCP-1 levels. These findings may imply that inhibition of IL-1β combined with cholesterol-lowering treatment could be beneficial for a subset of OA patients with high inflammatory risk.
This is the first study to show the effects of inhibition of IL-1β combined with cholesterol-lowering interventions on spontaneous OA development in dyslipidemic mice. IL-1β is produced by inflammasomes and is an important driver of inflammation
, making IL-1β a target of interest in diseases associated with MetS, such as cardiovascular disease and OA. Although IL-1β is thought to play a critical role in the pathogenesis of OA, several clinical studies were not able to demonstrate disease-modifying effects after inhibition of IL-1β
Phase IIa, placebo-controlled, randomised study of lutikizumab, an anti-interleukin-1α and anti-interleukin-1β dual variable domain immunoglobulin, in patients with erosive hand osteoarthritis.
. In contrast, a secondary analysis of the CANTOS trial showed that systemic inhibition of IL-1β resulted in a 40–47% decreased incidence rate of total knee and hip replacements compared to the placebo group
Effects of interleukin-1β inhibition on incident hip and knee replacement: exploratory analyses from a randomized, double-blind, placebo-controlled trial.
Effects of interleukin-1β inhibition on incident hip and knee replacement: exploratory analyses from a randomized, double-blind, placebo-controlled trial.
. Of note, 93% of patients included in the CANTOS trial received lipid-lowering therapy at the onset of the study and cholesterol levels were generally well managed. This study suggested that inhibition of IL-1β could reduce OA development in OA patients with systemic inflammation associated with metabolic disease. As the CANTOS trial was designed to study the effects on cardiovascular events, data on structural outcomes or radiographic evidence were not collected. In the current study, we observed that inhibition of IL-1β combined with cholesterol-lowering treatment resulted in a reduction of structural joint parameters such as synovial thickness and cartilage damage in mice fed a pro-atherogenic WTD. These results substantiate the findings of the CANTOS trial that inhibition of IL-1β together with cholesterol-lowering therapy could reduce OA progression in a subset of patients with systemic inflammation. As structural joint parameters were not assessed during the CANTOS trial, it would be of interest to investigate if inhibition of IL-1β in OA patients with high inflammatory risk reduces structural pathology such as cartilage damage or joint space narrowing. Moreover, IL-1β is believed to play an important role in joint pain via the upregulation of nociceptive mediators and it has been hypothesized that the beneficial effects observed in the CANTOS trial could be attributed to a reduction in joint pain. In future studies where induced OA models such as CiOA or destabilization of the medial meniscus (DMM) are used, it would be of value to include pain measurements.
Low-grade systemic inflammation is associated with dyslipidemia and is thought to contribute to OA development. Several clinical studies showed an association between high cholesterol levels and the development of OA
. The inconsistent effects of cholesterol-lowering therapies on OA development indicate that additional mechanisms, next to high cholesterol levels, are involved
. Recently, we have shown that high-intensive cholesterol-lowering therapy alone does not ameliorate the development of spontaneous OA pathology in dyslipidemic E3L.CETP mice with low inflammatory involvement
. Therefore, we aimed to determine the effects of cholesterol-lowering combined with inhibition of IL-1β in a more pro-inflammatory environment in the current study. We focused on the analysis of OA pathology in the E3L model as these mice developed higher levels of inflammatory cytokines on a hypercholesterolemic diet compared to the E3L.CETP mice. Systemic levels of SAA, MCP-1 and E-selectin were significantly reduced by the cholesterol-lowering treatments. While they were not reduced any further upon inhibition of IL-1β, we observed a local reduction of synovial lining thickness and cartilage damage in the knee joints of mice that received high-intensive cholesterol-lowering treatment combined with inhibition of IL-1β. We found a mild significant correlation of hepatic IL-1β levels with synovial thickening but not with cartilage pathology or ectopic bone formation (Supplementary Fig. 2), which indicates a positive association of systemic IL-1β levels with the inflammatory state of the synovium. To investigate the contribution of systemic inflammation in more detail, we determined OA development in the less inflammatory E3L.CETP model (Supplementary Fig. 3). In the latter model, the treatments did not reduce cartilage damage or synovial lining thickness, which suggests that the treatment is more beneficial in mice with a more pronounced pro-inflammatory profile. Metabolic inflammation can induce activation of the NLRP3 inflammasome which could explain why inhibition of IL-1β proved more beneficial in the E3L model. Although both strains develop metabolic and cardiovascular disease when fed a atherogenic WTD and show human-like responses to cholesterol-lowering therapies, we only observed effects of the treatments in the E3L mice with increased inflammation. This is similar to observations in the CANTOS trial, where patients with high CRP levels showed the most benefit from the IL-1β inhibition. These results suggest that using translatable models, such as the E3L and E3L.CETP mouse models, can increase the predictive value of pre-clinical studies. As all patients in the CANTOS trial had a history of cardiovascular disease, the presence of metabolic inflammation could explain why beneficial effects of IL-1β were observed in this, but not in previous OA trials in which IL-1β was inhibited
Phase IIa, placebo-controlled, randomised study of lutikizumab, an anti-interleukin-1α and anti-interleukin-1β dual variable domain immunoglobulin, in patients with erosive hand osteoarthritis.
. Since these parameters are usually not used as inclusion criteria of OA patients in clinical trials, these results could point towards a subgroup of OA patients that would respond to anti-inflammatory therapy.
Inhibition of IL-1β could reduce OA development via several mechanisms. Possibly, systemic inhibition of IL-1β could act locally, via a reduction of inflammatory mediators or reduced production of cartilage degrading enzymes like matrix metalloproteinases (MMPs) in the knee joint. Moreover, IL-1β inhibits proteoglycan synthesis by chondrocytes
, thereby hampering repair of the cartilage matrix. In addition, IL-1β stimulates cartilage degeneration via an increased production of several MMPs by chondrocytes, which are released in a latent form within the cartilage matrix and become activated in an inflammatory environment. This may explain why the beneficial effects of IL-1β inhibition were observed in the more inflammatory E3L model. To investigate enzymatic activity in articular cartilage in more detail, we determined expression of MMP13 and in articular cartilage. MMP13 staining in articular cartilage was absent, which can be explained by the mild cartilage damage and the spontaneous model used in this study.
Although inhibition of IL-1β on top of cholesterol-lowering treatment ameliorated synovial thickening and cartilage damage, no effect was found on ectopic bone formation. We observed a correlation with systemic cholesterol exposure and synovial lining thickness and cartilage damage, but not with ectopic bone formation in this study (Supplementary Fig. 4). We additionally found a significant correlation with cholesterol exposure and plasma MCP-1 levels, but not with systemic levels of SAA or E-selectin. This indicates that more intensive cholesterol-lowering treatment affects systemic inflammation and OA pathology. However, as cholesterol-lowering treatments alone were insufficient in ameliorating spontaneous OA development, these results suggest that high cholesterol drives other processes involved in OA development such as inflammation or innate immune training. We and others showed that induced post-traumatic OA models show more severe pathology in hypercholesterolemic mice
Cholesterol accumulation caused by low density lipoprotein receptor deficiency or a cholesterol-rich diet results in ectopic bone formation during experimental osteoarthritis.
. In an inflammatory environment, low-density lipoprotein cholesterol (LDL-C) is transformed into oxidized LDL (oxLDL) due to the presence of inflammatory mediators like reactive oxygen species (ROS). In previous studies, we have shown that mainly oxLDL, and not LDL, is responsible for the OA pathology associated with high cholesterol in mice
Synovial macrophages promote TGF-beta signaling and protect against influx of S100A8/S100A9-producing cells after intra-articular injections of oxidized low-density lipoproteins.
. Furthermore, it was shown that lipoproteins like oxLDL can induce reprogramming of monocytes and their myeloid progenitor cells, which is referred to as trained immunity. Trained immunity increases the inflammatory response of monocytes and macrophages to secondary stimuli, like Toll-like receptor (TLR) ligands. Recently, it was shown by Christ et al. that a WTD can induce trained immunity in pro-atherogenic Ldlr−/− mice, which was dependent on the NLRP3 inflammasome/IL-1β pathway
. In OA, these (ox)LDL-primed monocytes could enter the joint during inflammation and instigate an aggravated local inflammatory response. As we observed a reduction of synovial lining thickness and cartilage damage in mice treated with a combination of atorvastatin, alirocumab and inhibition of IL-1β, it is of interest to explore the efficacy of these combination treatments in induced post-traumatic OA models with more pronounced local inflammation, such as DMM or CiOA.
Some limitations of this study are the absence of a chow control group and a group in which only IL-1β was inhibited. Firstly, as this study focused on the effects of different treatments in dyslipidemic mice, a regular chow control group was not taken along so we were unable to study the effect of the WTD only on development of spontaneous OA. However, previous studies showed that a cholesterol-supplemented WTD contributes to OA development in E3L and E3L.CETP mice
Osteoarthritis development is induced by increased dietary cholesterol and can be inhibited by atorvastatin in APOE∗3Leiden.CETP mice--a translational model for atherosclerosis.
. Secondly, we did not include a group in which only IL-1β was inhibited. We find that cholesterol-lowering therapies combined with inhibition of IL-1β reduces cartilage damage in dyslipidemic mice. It would be of interest to investigate whether anti-IL1β alone is also effective under high cholesterol conditions without the use of cholesterol-lowering therapies. Further studies using these treatments in mice that were fed a regular chow diet should proof whether this protective effect is specific for high cholesterol conditions. Another potential limitation if the use of only females in this study. Although in general male mice develop more severe OA pathology that female mice, we chose to use female mice in the current study as female E3L mice are more responsive to cholesterol containing diets by having higher cholesterol and triglyceride levels compared to male mice. However, for the arthroplasty outcomes in the CANTOS trial, similar effects of IL-1β inhibition were observed after the analyses were stratified by sex
Effects of interleukin-1β inhibition on incident hip and knee replacement: exploratory analyses from a randomized, double-blind, placebo-controlled trial.
Taken together, this study supports the hypothesis that systemic inflammation associated with metabolic dysregulation could contribute to OA development. We have shown that inhibition of IL-1β combined with intensive cholesterol-lowering treatment reduces OA pathology in dyslipidemic mice with systemic inflammation. Our results may indicate that targeting IL-1β in combination with cholesterol-lowering therapies could be beneficial for a subset of OA patients with metabolic dysfunction and high inflammatory risk.
Author contributions
HMGP and EJP have designed the study. EJP and YG have carried out experimental procedures and acquired the data. YG has been the primary person responsible for writing the manuscript. NNLK, MHJB, ABB, PvdK, EJP, HMGP and PLEML were involved in drafting the work or revising it critically for important intellectual content. All authors approved the final version to be published.
Conflicts of interest
The authors declare no competing interests.
Sources of funding
The study was funded by TNO as a personal grant to Hans M.G. Princen as part of the Lorentz price, a lifetime achievement award, and by the TNO research program “Biomedical Health”. This manuscript was supported by funding from the Dutch Arthritis Foundation (Grant number 17-1-404).
Role of the funding source
The funding sources had no role in study design, in collection, analysis or interpretation of data, or in writing the manuscript and decision to submit the manuscript.
Acknowledgments
The authors would like to thank Joline Attema, Dorothé van Beugen, Wim van Duyvenvoorde, Nanda Keijzer, Christa de Ruiter, Jessica Snabel, Anita van Straalen-van Nieuwkoop, Alexander Thomas, Nikki van Trigt, Nicole Worms, Birgitte Walgreen and Monique Helsen for their excellent technical assistance.
Appendix A. Supplementary data
The following are the Supplementary data to this article:
Supplementary Fig. 1Synovial thickness correlated to cartilage damage in knee joints of E3L mice fed a cholesterol-supplemented WTD. E3L mice were fed a cholesterol-supplemented WTD for 38 weeks, treatments were started after 14 weeks. Synovial lining thickness correlated to the observed pathology in the knee joints. (P = 0.0001, Spearman r = 0.440).
Supplementary Fig. 2IL-1β correlated with synovial thickening, but not with cartilage damage or ectopic bone formation. E3L mice were fed a cholesterol-supplemented WTD for 38 weeks, treatments were started after 14 weeks. IL-1b produced by the liver showed a mild correlation with synovial thickening. (P = 0.036, Spearman r = 0.25), but not with cartilage damage or ectopic bone formation.
Supplementary Fig. 3Cholesterol-lowering treatment combined with inhibition of IL-1β does not reduce OA pathology in E3L.CETP mice fed a cholesterol-supplemented WTD. E3L.CETP mice were fed a cholesterol-supplemented WTD for 38 weeks, treatments were started after 14 weeks. (A) Synovial thickening was determined using an arbitrary score from 0 to 3. (B) Cartilage damage was determined at end point using an adapted OARSI scoring system (0–30). Cartilage damage was not affected by the treatment groups compared to mice fed a WTD alone. (C) Ectopic bone formation was determined throughout the joint using SafO/FG stained sections. The total number of osteophytes throughout the joint was comparable between all treatment groups. A = atorvastatin, AA = atorvastatin + alirocumab, AAI = atorvastatin + alirocumab + anti-IL1β. Results are expressed as individual data points with mean ± 95% confidence intervals.
Supplementary Fig. 4Cholesterol exposure correlated to synovial thickness and cartilage damage in dyslipidemic E3L mice. Cholesterol exposure was correlated to synovial lining thickness, cartilage degeneration and the total number of osteophytes at end point. (A + B) Cholesterol exposure correlates to synovial lining thickness (P = 0.0004, Spearman r = 0.4) and cartilage damage (P = 0.01, Spearman r = 0.28) observed after 38 weeks of WTD feeding. (C) No correlation was observed between systemic cholesterol exposure and the number of osteophytes.
Osteoarthritis development is induced by increased dietary cholesterol and can be inhibited by atorvastatin in APOE∗3Leiden.CETP mice--a translational model for atherosclerosis.
Synovial macrophages promote TGF-beta signaling and protect against influx of S100A8/S100A9-producing cells after intra-articular injections of oxidized low-density lipoproteins.
Cholesterol accumulation caused by low density lipoprotein receptor deficiency or a cholesterol-rich diet results in ectopic bone formation during experimental osteoarthritis.
Synovitis: a potential predictive factor of structural progression of medial tibiofemoral knee osteoarthritis–results of a 1 year longitudinal arthroscopic study in 422 patients.
Gene deletion of either interleukin-1β, interleukin-1β–converting enzyme, inducible nitric oxide synthase, or stromelysin 1 accelerates the development of knee osteoarthritis in mice after surgical transection of the medial collateral ligament and partial medial meniscectomy.
Effects of interleukin-1β inhibition on incident hip and knee replacement: exploratory analyses from a randomized, double-blind, placebo-controlled trial.
Modulation of very low density lipoprotein production and clearance contributes to age-and gender-dependent hyperlipoproteinemia in apolipoprotein E3-Leiden transgenic mice.
Innovative pharmaceutical interventions in cardiovascular disease: focusing on the contribution of non-HDL-C/LDL-C-lowering versus HDL-C-raisingA systematic review and meta-analysis of relevant preclinical studies and clinical trials.
The APOE3-leiden heterozygous glucokinase knockout mouse as novel translational disease model for type 2 diabetes, dyslipidemia, and diabetic atherosclerosis.
Resemblance of osteophytes in experimental osteoarthritis to transforming growth factor β–induced osteophytes: limited role of bone morphogenetic protein in early osteoarthritic osteophyte formation.
Phase IIa, placebo-controlled, randomised study of lutikizumab, an anti-interleukin-1α and anti-interleukin-1β dual variable domain immunoglobulin, in patients with erosive hand osteoarthritis.
00932-3&id=gr1.jpg){kind=link}
00932-3&id=gr2.jpg){kind=link}
00932-3&id=gr3.jpg){kind=link}
00932-3&id=gr4.jpg){kind=link}
00932-3&id=gr5.jpg){kind=link}
00932-3&id=figs1.jpg){kind=link}
00932-3&id=figs2.jpg){kind=link}
00932-3&id=figs3.jpg){kind=link}
00932-3&id=figs4.jpg){kind=link}