The role of Interleukin-6 in the metabolism of the ApoB-containing lipoproteins in rheumatoid arthritis

Robertson, Jamie Summers (2016) The role of Interleukin-6 in the metabolism of the ApoB-containing lipoproteins in rheumatoid arthritis. PhD thesis, University of Glasgow.

Full text available as:
[thumbnail of 2016Robertsonphd.pdf] PDF
Download (6MB)
Printed Thesis Information:


Patients with rheumatoid arthritis (RA) carry an increased risk of cardiovascular disease and cardiovascular death compared to age- and sex-matched controls. This risk appears to be related to cumulative inflammatory burden, and can be at least partially ameliorated by successful treatment of the disease with conventional or biologic disease-modifying anti-rheumatic drugs (DMARDS). However, RA patients typically exhibit reduced serum levels of cholesterol, which can be increased following DMARD therapy; this is in contrast to the general population, where serum cholesterol is directly proportional to cardiovascular risk. The magnitude and nature of this increase varies between therapeutic agents. Blockade of interleukin-6 (IL-6) signalling with the drug tocilizumab conveys perhaps the most profound lipid changes, leading to average increases in LDL-cholesterol (LDL-c) of around 20% as well as changes in HDL-cholesterol (HDL-c) and triglycerides.
The mechanisms behind this so-called “lipid paradox”, and its impact on cardiovascular outcomes following RA therapy, are not fully understood. Animal studies have shown that hypercatabolism of LDL can lead to reduced circulating LDL-c, possibly due to increased consumption by the reticulo-endothelial system. A deeper understanding of the lipid paradox, and its implications for cardiovascular risk, is vital to allow physicians to provide optimal management of both articular RA and its cardiovascular manifestations. Using IL-6 blockade as a molecular tool, I attempted to unravel the physiological processes underlying the lipid changes observed in RA, and to understand what these changes might mean in term of cardiovascular risk for patients.
1) To investigate lipid changes in patients treated with tocilizumab as assessed by nuclear magnetic resonance (NMR) spectroscopy
2) To determine whether increased LDL-c following IL-6 blockade is due to increased production or reduced catabolism of LDL
3) To explore the lipid-handling behaviour of macrophages in response to IL-6
Methods: The MEASURE study was a placebo-controlled, randomised controlled trial evaluating NMR lipid profiles in patients with severe active RA following treatment with tocilizumab or placebo, with the placebo group switching to open-label tocilizumab after 24 weeks. Results to week 12 have been published previously. Data on all patients to 52 weeks using an updated NMR platform were evaluated, and changes in lipid values were correlated with markers of disease activity, including the acute phase markers C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR), and the composite clinical score DAS28. NMR analysis also yielded the novel marker GlycA, a composite NMR signal reflecting a measure of the acute phase response. GlycA levels were compared to other markers of disease activity, and GlycA’s utility as a marker of future response to therapy and of persistent disease activity in those with normal ESR or CRP was assessed.
The KALIBRA study analysed the kinetics of the apoB-containing lipoproteins before and after IL-6 blockade. Patients with severe active RA (defined as DAS28 ≥5.1) and who were eligible for tocilizumab therapy underwent kinetic modelling of VLDL, IDL and LDL at baseline and again following at least three months’ treatment with tocilizumab. The primary outcome measure was the fractional catabolic rate of LDL, though LDL production rate was also assessed as well as a variety of other lipid parameters. Changes in LDL production and catabolic rates were correlated with serum LDL-cholesterol ester content and with measures of disease activity.
To provide information on the cellular processes underlying kinetic changes, macrophages were generated in vitro from the THP-1 monocyte cell line or from healthy human donor monocytes. These cells were then exposed to IL-6 and assessed for signs of response, including phosphorylation of STAT3 and production of TNF-α. Macrophage lipid loading (and subsequent foam cell formation) following stimulation with 10ng/ml IL-6 for 24 hours was assessed by staining with oil red O and fluorimetry following culture with fluorescently-labelled oxidised LDL. Altered expression of genes involved in lipid metabolism, including surface receptors for both native and oxidised LDL, was assessed at the RNA level using quantitiative polymerase chain reaction (qPCR) and Taqman Low-Density Array (TLDA) plates, with fold change of ≥2 considered significant. Outcomes from these experiments were then validated at the protein level by flow cytometry.
The MEASURE study showed no change in small LDL particles with tocilizumab therapy. Increases were seen in large LDL, small HDL and some VLDL particles in the tocilizumab group, with change seen by the earliest assessment timepoint of 2 weeks. At 52 weeks follow-up, the same changes were observed in placebo patients who had switched to open-label tocilizumab. Greater increases in large LDL and small HDL tended to associate with falls in CRP, but less so with disease activity as measures by CDAI.
GlycA levels correlated significantly at baseline with CRP (r=0.70, p<0.001) and ESR (r=0.44, p<0.001) but not with CDAI, and fell in a similarly precipitous manner following IL-6 blockade. Neither baseline (area under curve =0.60) or week 2 (AUC=0.53) GlycA levels were effective at predicting response at week 24 as measured by CDAI. In treated patients, with CRP<5mg/l, GlycA did not associate with persistent clinical disease activity.
In KALIBRA, 12 patients were recruited of whom complete data was available for 11. As expected, significant increases in mean serum LDL-c (2.90 v 3.40mmol/L, p=0.014) and HDL-c (1.23 v 1.52 mmol/L, p=0.006) were observed after treatment. IL-6 blockade led to a reduction in median LDL fractional catabolic rate (FCR) from 0.53 to 0.27 pools/day, (p=0.006) with median reduction of 30%, and the change in LDL FCR correlated tightly with that of serum LDL cholesterol ester content (r=-0.74, p=0.011). LDL FCR correlated at baseline with CRP (r=0.74 p=0.012) but not CDAI (r=0.04, p=0.91). The degree of change in CRP with treatment showed a trend to association with change in FCR (r=0.46, p=0.15) and LDL cholesterol ester (r=-0.43, p=0.18). LDL production rate did not increase, and in fact fell (median 763.8 v 442 mg/kg/day, p=0.002). No changes were seen in the activity of lipoprotein lipase (LPL), hepatic lipase (HL), cholesterol-ester transfer protein (CETP) or PCSK9.
Human monocyte-derived macrophages (HMDM) displayed altered TNF-α production and increased STAT3 phosphorylation in response to IL-6; THP-1 macrophages did not, and appeared to lose their IL-6 receptor in the process of differentiation from monocytes. Neither cell type displayed increased uptake of oxidised LDL following culture with IL-6. TLDA analysis showed altered expression of scavenger receptors with an increase in the “macrophage receptor with collagenous structure” (MARCO) and a reduction in LOX-1 in HMDM, though this was not observed at the protein level on flow cytometry.
IL-6 blockade in RA elevates numbers of large LDL and small HDL particles, but not the most pro-atherogenic small LDL particles. These changes are maintained up to 52 weeks follow-up, and seem to be more prominent in subjects with greater reductions in the acute phase response. The KALIBRA study showed that the elevation in LDL is almost entirely due to a reduction in LDL fractional catabolic rate, from a baseline state of hypercatabolism in severe active disease to values approximating the population average after treatment. Greater changes in FCR were associated with greater reductions in acute phase reactants, regardless of RA activity as assessed clinically by CDAI. Lipid changes did not appear to be explicable by changes in activity of lipolytic enzymes, CETP or PCSK9. IL-6 does not appear to exert its lipidaemic effects via augmented macrophage lipid metabolism or increased foam cell formation. These findings are consistent with a normalisation of a pathological, IL-6 driven state of hypercatabolism leading to LDL-c increases following IL-6 blockade, with hepatocytes as the possible main effector cell type. This suggests that LDL-c elevations observed during treatment for RA may not be pro-atherogenic or contribute to increased CVD risk, though confirmation of this hypothesis is required with trials reporting clinical cardiovascular outcomes.

Item Type: Thesis (PhD)
Qualification Level: Doctoral
Keywords: Rheumatoid, cholesterol, kinetics, lipid, inflammation, IL-6.
Subjects: Q Science > QP Physiology
R Medicine > R Medicine (General)
Colleges/Schools: College of Medical Veterinary and Life Sciences > School of Infection & Immunity
Supervisor's Name: McInnesIain, Professor Iain
Date of Award: 2016
Depositing User: Dr Jamie S Robertson
Unique ID: glathesis:2016-7960
Copyright: Copyright of this thesis is held by the author.
Date Deposited: 14 Mar 2017 12:15
Last Modified: 16 Oct 2017 09:41

Actions (login required)

View Item View Item


Downloads per month over past year