Preferential association of Hepatitis C virus with apolipoprotein B48-containing lipoproteins

Abstract

Hepatitis C virus (HCV) in cell culture has a density comparable to that of other members of the family Flaviviridae, whereas in vivo infectious particles are found partially in low-density fractions, associated with triacylglycerol (TG)-rich lipoproteins (TRLs). In the blood of infected patients, HCV circulates as heterogeneous particles, among which are lipo-viroparticles (LVPs), globular particles rich in TG and containing viral capsid and RNA. The dual viral and lipoprotein nature of LVPs was addressed further with respect to apolipoprotein composition and post-prandial dynamic lipid changes. The TRLs exchangeable apoE, -CII and -CIII, but not the high-density lipoprotein apoA-II, were present on LVPs, as well as the viral envelope proteins. apoB100 and -B48, the two isoforms of the non-exchangeable apoB, were represented equally on LVPs, despite the fact that apoB48 was barely detectable in the plasma of these fasting patients. This indicates that a significant fraction of plasma HCV was associated with apoB48-containing LVPs. Furthermore, LVPs were enriched dramatically and rapidly in triglycerides after a fat meal. As apoB48 is synthesized exclusively by the intestine, these data highlight the preferential association of HCV with chylomicrons, the intestine-derived TRLs. These data raise the question of the contribution of the intestine to the viral load and suggest that the virus could take advantage of TRL assembly and secretion for its own production and of TRL fate to be delivered to the liver.

Fig. 1

ApoB100 and apoB48 are present in purified LVP. A, Nature of apoB in purified LVP. Plasma from HCV-infected or healthy donors were adjusted to a 1.055g/ml density and centrifugated for 4h at 4°C and 543,000×g. LVP were immunopurified from the low-density fraction as described in experimental procedures. Samples of LVP and of the whole fraction were analyzed by 5% SDS-PAGE under reducing conditions and immunoblotted with 1D1 anti-apoB monoclonal antibody. Lane 1 and 2, the d<1.055g/ml fraction and chylomicrons isolated from a healthy blood donor; respectively; lane 3, mocked-prepared LVP from a healthy blood donor; lane 4 and 5, purified LVP from patient H; lane 6 and 7, the d<1.055g/ml fraction from which LVP were purified from the same patient and a healthy subject, respectively. B, Relative proportions of apoB48 and apoB100 in purified LVP and in the d<1.055g/ml fraction prepared from eight infected patients. ApoB48 and apoB100 spots were quantified by videodensitometry and expressed as % of total apoB. The proportion of apoB48 was significantly higher in purified LVP than in the whole d<1.055g/ml fraction (Student T test p<0.01).

Fig. 2

Presence of apoE, CII and CIII in LVP. LVP were immunopurified from the d<1.055g/ml fraction, as described in experimental procedures, analysed by 12% (apoE) or 15% (apoAII, apoCII and apoCIII) SDS-PAGE under reducing conditions, and immunoblotted with anti-apoE, anti-apoAII or anti-apoCIII monoclonal antibodies (Chemicon International) or anti-apoCII polyclonal antibody (Merck Calbiochem). Lane 1 and 2, mocked-prepared LVP and the d<1.055g/ml fraction from a non infected blood donor, respectively;lane 3 and 4, LVP and the d<1.055g/ml fraction from a chronically infected patient, respectively; lane 5, control plasma from a blood donor. ApoE, CII and III were present in the d<1.055g/ml fraction where apoB-containing lipoproteins reside and in purified LVP. ApoAII, a component of HDL, was neither detected in the d<1.055g/ml fraction nor in LVP.

Fig. 3

Presence of envelope glycoproteins in LVP. A, LVP were immunopurified from the d<1.055g/ml fraction, as described in the experimental procedures, analyzed by 10% SDS-PAGE under reducing conditions, and immunoblotted with the A4 anti-E1 (lower panel) or the H52anti-E2 (upper panel) monoclonal antibodies. Lane 1 and 2, lysates of 293T cells expressing or not E1 and E2 glycoproteins, respectively; lane 3, mocked-prepared LVP from plasma of a healthy blood donor; lane 4, purified LVP from a chronically infected patient; lane 5 and 6, the d<1.055g/ml fraction from which LVP were purified from a HCV patient or a healthy subject, respectively. Western blots are representative of experiments performed with plasma from 3 infected individuals. Note that both viral glycoproteins were detected in purified LVP. B, Detection of protein A-captured LVP in an ELISA with anti-E1 and anti-E2 antibodies. The d<1.055g/ml fractions were prepared as described in material and methods from chronically infected patients and from non-infected blood donors. The d<1.055g/ml fraction from infected (closed square) or non infected (open square) patients were revealed by anti-E1 A4 (a) or anti-E2 H47 (b). As a control, the d<1.055g/ml fraction from infected (open triangle) or non infected (open circle) patients, stained by anti-H measles clone 55, are presented. Results are means of duplicates (×). Note that protein A-captured LVP from infected patient were only recognized by anti-HCV E1 and E2 envelope antibodies and not by anti-measles virus H envelope antibodies. No material from non infected control was recognized by any antibody.

Fig. 4

Evolution of TG/apoB mass ratio in LVP between the pre-prandial and the post-prandial periods. (a), Mean (dash) and individual TG/apoB mass ratios (spots) in both fractions from 7 patients.. Lipids from purified LVP and the whole <1.055g/ml fraction were extracted and separated by TLC. TG spots were scraped off, and the fatty acids were transmethylated and quantified by gas chromatography, and the apoB content of purified LVP was determinated by ELISA, as described under “Experimental Procedures”. Note that the TG/apoB was significantly higher in LVP than in their respective whole fraction (Wilcoxon T test, p≤0.05). (b), TG/ApoB mass ratio in LVP increase between the pre-prandial to post-prandial periods. Results are expressed as the ratio between the TG/apoB mass ratio in LVP in the post-prandial period vs that in the pre-prandial period for each patient. Patients are identified as patient A to G and indicated by arrows in (a). Note that the TG/apoB mass ratio was significantly increased in LVP during the post-prandial period (distribution-free Wilcoxon T test, p≤0.05).

Fig. 5

Effect of lipid intake on LVP characteristics. A, Lipid and protein mass ratios in the whole d<1.055g/ml fractions and purified LVP in the pre-prandial and post-prandial periods. Results are means from seven patients. B, Parallel modifications of fatty acid composition of LVP and of the fraction d<1.055g/ml between the pre-prandial to post-prandial periods. n-6 and saturated fatty acids were quantified by gas chromatography as described in material and methods and expressed as mol% in triacylglycerol (TG) and phospholipids (PL) of purified LVP and of the whole d<1.055g/ml fraction. Results are means from 11 patients. Note that the fatty acid composition of TG and PL vary in the same proportion in LVP and the whole d<1.055g/ml density fraction between the pre-prandial and the post-prandial periods.