A number of studies have reported molecular interactions between viruses and lipoproteins (LPs), apolipoproteins or their receptors, with an underlying basis for these links being the similarities in location, size and function of LPs and virus particles (reviewed in ). One of the most studied such connections to date is the influence of the APOE gene on the outcome of infection [2–7]. APOE codes for the protein apolipoprotein E (apoE), which like other apolipoproteins is a constituent of LP particles. We have shown that the influence of APOE on infection may be mediated through direct anti-infective activity of the cationic receptor-binding region of apolipoprotein E (apoE141-149). We reported that tandem repeat peptides derived from this region - in particular apoE(141-149)dp (apoEdp) - and N-terminal truncated apoE-4 have broad antiviral activity in vitro [8, 9]. Recently another group have shown such peptides have direct antiviral activity in vivo, reducing HSV1 titres in an experimental herpes simplex keratitis model .
We have suggested that the broad antiviral action of this heparan sulphate proteoglycan (HSPG) and low density lipoprotein receptor (LDLR) binding region may relate to the ubiquity of the former receptor as an initial attachment site to cells for many viruses, along with the ability of some viruses to enter cells using LDLR family receptors. Thus apoE may compete for binding to these receptors preventing viral attachment . In addition apoEdp and related peptides exert a virucidal activity on viruses and also show broad antibacterial activity, which may more closely relate to the high numbers of cationic and basic hydrophobic residues within apoEdp and related peptides [8, 9]. It is at present unclear whether full-length or N-terminal truncated apoE proteins also show such antibacterial activity.
Other researchers have focused on antiviral activity of peptides derived from another apolipoprotein - apolipoprotein AI (apoAI) - with a peptide referred to as 18A being shown to block fusion of virus particles to cells. This peptide is a consensus domain constructed from the 22 mer amphipathic helical domains within apoAI - and its activity appears to involve its amphipathic -helical structure leading to blockade of specific viral envelope proteins .
The biophysical properties of apoE-derived or apoAI-derived anti-infective peptides are shared by similar regions within other apolipoproteins. In particular many other apolipoproteins contain amphipathic -helical regions. Apolipoprotein AII (apoAII) contains a region - apoAII(18-30) - which has previously been stabilised as an α-helical peptide by addition of a 5 mer motif to the C-terminal to promote α-helical structure . In a separate study apolipoprotein J (apoJ) was predicted to contain five amphipathic -helical regions, which together allow this apolipoprotein to act as a biological detergent .
Similarly several other apolipoprotein regions contain binding domains analogous to the region we previously investigated in apoE. A peptide based on a domain within apolipoprotein B (apoB) - apoB(1000-1016) - has been reported to function as an arterial binding domain . ApoB also contains two LDL receptor binding domains comparable to the receptor binding region of apoE, namely region A (apoB3147-3157) - and region B (apoB3359-3367). As region B is more uniformly conserved across species it has been considered to be more important for receptor binding . Previous studies have shown that heparin-binding activity within apolipoprotein H (apoH or beta-2-glycoprotein) depends on three Lys residues in its fifth domain (in positions 284, 286, and 287 of apoH), and that an octomer peptide apoH281-288 competitively inhibits apoH binding to heparin . Finally a second heparin-binding region with apoE has been linked to residues 211-218 .
Given the ready access of LPs to extracellular pathogens, and the existence of many domains in apolipoproteins similar to the two previously reported to have anti-infective activity, in the present study we tested the potential activity of a variety of apolipoprotein-derived peptides, grouped by similarity to the previously reported apoE or apoAI anti-infective peptides. Specifically we looked at either (i) possible antiviral action of amphipathic α-helical peptides (comparable with the apoA1-derived 18A peptide) or (ii) broader anti-infective activity of tandem repeats of receptor binding domains (similar to the cationic non-amphipathic α-helical apoE peptides).
The first group of peptides included several of the 22 mer amphipathic alpha helical domains from apoA1 (which had never previously been tested) , apoAII 18-30 with the artificial helix promoting region (apoAII(18-30)+) or the equivalent peptide containing an additional five residues from the apoAII sequence (apoAII(18-35)). We also synthesised peptide mimetics of four of the predicted amphipathic helical domains of apoJ . In the second group we examined the activities of apoB(1000-1016), apoB(3359-3367)dp, apoH(281-288)dp and apoE(211-219)dp and apoE(213-221)dp, along with some tandem repeats and non-repeat peptides related to the apoE(141-149) region. In addition we examined the activity of apoE(141-149)dp derived from either the bovine or murine apolipoprotein E sequence. Tandem repeat peptides were synthesised to allow the structure of these short regions to more likely mimic that found in the full length protein, using the strategy we previously utilised for apoE(141-149) derived peptides  and for other heparin-binding domain related AMPs . Finally we devised and tested mutant peptides related to apoB(3359-3368)dp, modified to more closely resemble the apoE-derived peptide mutants we previously reported .