Relapsing-remitting multiple sclerosis patients display an altered lipoprotein profile with dysfunctional HDL

Abstract

Lipoproteins modulate innate and adaptive immune responses. In the chronic inflammatory disease multiple sclerosis (MS), reports on lipoprotein level alterations are inconsistent and it is unclear whether lipoprotein function is affected. Using nuclear magnetic resonance (NMR) spectroscopy, we analysed the lipoprotein profile of relapsing-remitting (RR) MS patients, progressive MS patients and healthy controls (HC). We observed smaller LDL in RRMS patients compared to healthy controls and to progressive MS patients. Furthermore, low-BMI (BMI ≤ 23 kg/m²) RRMS patients show increased levels of small HDL (sHDL), accompanied by larger, triglyceride (TG)-rich VLDL, and a higher lipoprotein insulin resistance (LP-IR) index. These alterations coincide with a reduced serum capacity to accept cholesterol via ATP-binding cassette (ABC) transporter G1, an impaired ability of HDL₃ to suppress inflammatory activity of human monocytes, and modifications of HDL₃'s main protein component ApoA-I. In summary, lipoprotein levels and function are altered in RRMS patients, especially in low-BMI patients, which may contribute to disease progression in these patients.

Figure 1. RRMS patients display an inverse relationship between BMI and sHDL-p.

The regression line describing the relationship between sHDL and BMI, distinguishing between groups (i.e. HC (n = 89, black), RRMS (n = 35, red), and Prog MS (n = 25, green)), for females (a) and males (b) is shown. The dotted vertical line shows the threshold for BMI for which RRMS patients and HC differ significantly in their amount of sHDL-p. - p = particle count.

Figure 2. RRMS patients have reduced serum cholesterol efflux capacity via ABCG1 and reduced monocyte ABCG1 mRNA expression compared to healthy controls.

(a) LCAT mass (μg/ml) was measured in freshly thawed plasma samples of HC (n = 21) and RRMS (n = 22) patients using ELISA. (b) Schematic representation of sHDL measured by NMR, based on 2D gelelectrophoresis, and ultracentrifugal separation. (c) Capacity of serum from controls (n = 50), RRMS (n = 26), and progressive MS (n = 17) patients to accept cholesterol via the ABCA1 (left) or ABCG1 (right) transporters from ³H-cholesterol loaded BHK cells transfected with human ABCA1 or ABCG1 respectively. (d) Basal ABCA1 (left) and ABCG1 (right) mRNA expression in monocytes of HC (n = 43), RRMS (n = 33), and progressive MS (n = 7) patients measured using quantitative PCR (qPCR). Results are expressed as fold change of healthy controls. (e), Pearson correlation coefficient for percentage serum cholesterol efflux capacity via the ABCG1 transporter versus sHDL (μmol/L) for RRMS patients with a low BMI (n = 10). *P < 0.05, ****P < 0.0001.

Figure 3. HDL₃ of low-BMI RRMS patients does not suppress inflammation-induced gene expression of TNFα, CD40, IL1β, and IFNγ.

Human monocytes from healthy donors were isolated from fresh blood samples. Monocytes of HC were pre-incubated with pooled HDL₃ (60 mg/dl) isolated from low-BMI control subjects (n = 8) or low-BMI RRMS patients (n = 9) in low serum culture medium for four days followed by an overnight LPS (100 ng/ml) stimulus. Monocyte gene expression was measured using qPCR. Results are expressed as fold change of control LPS conditions without HDL₃. HC = healthy controls; RRMS = relapsing-remitting multiple sclerosis; LPS = lipopolysaccharide. *P < 0.05, ^†versus control LPS ( = 1) (^†P < 0.05, ^††P < 0.01).

Figure 4. ApoA-I of RRMS patients is modified at its tyrosine and tryptophan residues.

(a–d) Collision-induced dissociation (CID) spectra of Trp-50 nitration (a), and Trp-50 mono-, di-, and tri-oxidation (b,c and d respectively). Spectra were acquired on analysis of in-gel tryptic digests of the ApoA-I band from HDL₃ isolated with sequential flotation ultracentrifugation. Modifications were detected in an LC-MS/MS experiment as described under Methods. CID fragment ion annotation was according to the nomenclature by Roepstorff and colleagues. (e) Summary of ApoA-I modification sites. Residues that were modified are indicated with ∆. The numbering of the amino acids cited refers to the amino acid sequence of the mature protein.