Evidence for a Trypanosoma brucei Lipoprotein Scavenger Receptor

African trypanosomes are lipid auxotrophs that live in the bloodstream of their human and animal hosts. Trypanosomes require lipoproteins in addition to other serum components in order to multiply under axenic culture conditions. Delipidation of the lipoproteins abrogates their capacity to support trypanosome growth. Both major classes of serum lipoproteins, LDL and HDL, are primary sources of lipids, delivering cholesterol esters, cholesterol, and phospholipids to trypanosomes. We show evidence for the existence of a trypanosome lipoprotein scavenger receptor, which facilitates the endocytosis of both native and modified lipoproteins, including HDL and LDL. This lipoprotein scavenger receptor also exhibits selective lipid uptake, whereby the uptake of the lipid components of the lipoprotein exceeds that of the protein components. Trypanosome lytic factor (TLF1), an unusual HDL found in human serum that protects from infection by lysing Trypanosoma brucei brucei, is also bound and endocytosed by this lipoprotein scavenger receptor. HDL and LDL compete for the binding and uptake of TLF1 and thereby attenuate the trypanosome lysis mediated by TLF1. We also show that a mammalian scavenger receptor facilitates lipid uptake from TLF1 in a manner similar to the trypanosome scavenger receptor. Based on these results we propose that HDL, LDL, and TLF1 are all bound and taken up by a lipoprotein scavenger receptor, which may constitute the parasite's major pathway mediating the uptake of essential lipids.


Introduction
Exogenous lipids play indispensable roles in trypanosome cell structure and metabolism.
African bloodstream-form trypanosomes are single-celled parasites that appear not to synthesize fatty acids de novo (1)(2)(3), with the exception of myristate (C14 fatty acid). Trypanosomes have an atypical type II fatty acid synthase that utilizes exogenously supplied butyrate to generate myristate which is used exclusively for glycosylphosphatidylinositol anchor biosynthesis (4,5).
Despite having a variety of enzymes that catalyze metabolic lipid-modifying pathways (6)(7)(8)(9), trypanosomes are lipid auxotrophs. They require lipoproteins in addition to other serum components in order to multiply under axenic culture conditions (10,11). Delipidation of the lipoproteins abrogates their capacity to support trypanosome growth. Both major classes of serum lipoproteins, LDL and HDL, are primary sources of lipids, delivering cholesterol esters, cholesterol and phospholipids to trypanosomes (12,13).
Trypanosomes endocytose HDL and LDL through their flagellar pocket (10,13). All endocytosis and exocytosis in trypanosomes occurs at this site. It is a specialized invagination in the cell membrane, which is not lined with the microtubule network that encases the rest of the cell that precludes any vesicular fusion or fission. At physiological concentrations (~1 mg/ml), specific binding and uptake of the protein component of both LDL and HDL has been demonstrated (12)(13)(14). In contrast, at sub-physiological concentrations (1-50 µg/ml) there was no detectable uptake of the apolipoproteins themselves, whereas the lipid components of HDL and LDL were taken up at rates that exceeded fluid phase endocytosis by 1000-fold, suggesting that "specific binding sites were probably involved" (15). A putative LDL receptor protein has been by guest on March 24, 2020 http://www.jbc.org/ Downloaded from Trypanosome Lipoprotein Scavenger Receptor 4 purified but not yet cloned (16,17), while there has been no molecular identification of an HDL receptor in bloodstream-form trypanosomes.
Trypanosome lytic factors are HDL related particles found in human plasma. TLF1 contains lipid, apolipoprotein A-I (apoA-I), paraoxonase and haptoglobin related protein (Hpr) (18), while TLF2 is a lipid-poor molecule that contains apoA-I, Hpr, and IgM (19). Both high and low affinity binding sites for TLF1 on trypanosomes have been reported in experiments using purified preparations of TLF1 (20). The low affinity binding site can be competed by HDL whereas the high affinity binding site is partially competed by reconstituted nonlytic HDL containing Hpr (21), which led to the proposal that Hpr can mediate TLF1 binding to trypanosomes through a haptoglobin-like receptor.
Many lipoprotein receptors have been characterized in eukaryotes, to date only cubilin (22) and members of the CD36 superfamily of scavenger receptors (23)(24)(25) bind native HDL (without requiring ApoE as a component). The CD36 superfamily of scavenger receptors bind and take up both native HDL and LDL as well as other polyanionic ligands, including oxidized and acetylated LDL (26). Some of these scavenger receptors mediate bi-directional lipid flux and exhibit a process called selective lipid uptake. In polarized cells selective lipid uptake is characterized by receptor-mediated uptake of the lipoprotein, distribution of the lipid within the cell, and recycling of the apolipoprotein to the cell surface (27). In non-polarized cells there does not appear to be any uptake of the holo-particle, rather binding to the surface of the cell via lipoprotein scavenger receptors facilitates the transfer of lipid from the lipoprotein into cell membranes and intracellular vesicles (28). After lipid transfer, the lipid-depleted particle is by guest on March 24, 2020 http://www.jbc.org/ Downloaded from Trypanosome Lipoprotein Scavenger Receptor 5 released intact from the cell. One of the members of this family, SR-BI (scavenger receptor class BI), mediates the highest level of selective lipid uptake analyzed to date (29,30).
While studying trypanosome lytic factors, which are by definition lipoproteins, we decided to revisit lipoprotein receptors. We found evidence that T. b. brucei has a lipoprotein scavenger receptor that mediates the selective uptake of lipid over the protein component of both HDL and LDL. The same receptor can also mediate the uptake of oxidized lipoproteins. TLF1 is also bound and endocytosed by this lipoprotein scavenger receptor. We show that HDL and LDL compete for the binding and uptake of TLF1 and therefore attenuate the trypanosome lysis mediated by TLF1.

HDL and LDL compete for HDL uptake by trypanosomes.
We labeled HDL and LDL with Alexa, a fluorophore that conjugates to the free amino groups in the protein components of these lipoproteins. We found that trypanosomes accumulated HDL protein (2.25 pmol, calculated based on a molecular mass of 350,000 Da by size exclusion chromatography, 50% of which is protein) and LDL protein (2 pmol uptake from labeled HDL. Fig. 1 panel D shows that 4 times more HDL than LDL was needed to give a 50% reduction in the uptake of HDL protein. This suggests that the putative lipoprotein receptor has higher affinity for LDL than HDL. Although Fig. 1 (Fig. 3 B) with a distribution similar to endocytosed concanavalin A (conA) (Fig 3 C).
Concanavalin A has been shown to distribute within endocytic vesicles of trypanosomes when endocytosed by live trypanosomes (35,36), in contrast conA labels the VSG coat when used on fixed trypanosomes presumably due to the exposure of carbohydrate epitopes upon fixation.
Coincubation with rhodamine-conA and Alexa-HDL revealed colocalization in some endocytic vesicles (yellow) near the flagellar pocket but not all endocytic vesicles (red) (Fig. 3 D). to T. b. brucei (Fig. 4). We did not investigate the effect of LDL on the binding of TLF1 to T. b.

HDL competes for the binding of TLF1 to trypanosomes
brucei, because LDL takes 6 hours to reach equilibrium binding to trypanosomes whereas HDL Given that we observed competition for binding of TLF1 to trypanosomes by HDL, we evaluated the effect of non-lytic bovine HDL on TLF1-mediated trypanolysis. We observed that non-lytic bovine HDL was able to attenuate trypanosome lysis by purified TLF1 (Fig. 5). Nonlytic human LDL was also effective in attenuating trypanolysis by TLF1. stained readily with anti-mSR-BI (Fig. 6, insert). HDL labeled with the fluorescent lipid DiI exhibited lipid uptake into cells expressing mSR-BI that was 30-fold greater than the uptake by the parental ldlA cells (Fig. 6) previously characterized biochemically. These include an LDL receptor (10) and a HDL receptor (13) both of which may be identical to the scavenger receptor described here (see below), a haptoglobin-like receptor which may also be a TLF receptor (21), and a receptor for transferrin which has been molecularly cloned (38)(39)(40)(41)(42)(43).

TLF1 binds to mouse Scavenger Receptor Class B type I and donates lipids
The characterization of this trypanosome scavenger receptor serves to unify a variety of disparate data regarding the utilization of lipoproteins and TLF by the parasite. Vandeweerd et.
al., showed that the uptake at 37 o C of radiolabeled lipid components in either HDL or LDL was inhibited (50-85%) by unlabeled HDL or LDL (15). It was concluded that the uptake process did not discriminate between HDL or LDL. In this study we have confirmed and extended these observations. We also found that accumulation of HDL labeled protein or HDL labeled lipid, was inhibited by increasing concentrations of HDL and LDL (Fig 1C and 1D (45), are also ligands for eukaryotic lipoprotein scavenger receptors (46,47).
Taken together, these results suggest the presence of a lipoprotein scavenger receptor in trypanosomes that can bind multiple ligands.
The trypanosome lipoprotein scavenger receptor shares characteristics with certain subclasses of mammalian scavenger receptors. Members of the CD36 superfamily can bind native HDL and LDL and exhibit selective lipid uptake from both lipoproteins. Binding appears to be mediated by a combination of apolipoprotein and lipid. These characteristics most resemble what we have found for the putative trypanosome lipoprotein scavenger receptor. We find that when we correct for the specific activity of each labeled lipoprotein, the lipid components are selectively accumulated more than the protein component (Fig. 2). It is worth noting that although cholesterol/ cholesterol ester is taken up 3-4 fold more than phospholipid, it only comprises 36% of native HDL lipids relative to 55% for phosphatidyl choline. The selective uptake of lipoprotein cholesterol/cholesterol ester over phospholipid has also been characterized in SR-BI scavenger receptors, which are a subclass of the CD36 superfamily (28,48).
Ligands other than native HDL and LDL have been identified for eukaryotic lipoprotein scavenger receptors, such as oxidized lipoproteins (46). Oxidized LDL is a ligand for the trypanosome lipoprotein receptor, in that native HDL and LDL or oxidized lipoproteins (not shown) were effective competitors for uptake. Native HDL was a consistently better competitor than native LDL when measuring oxidized LDL uptake by trypanosomes. On the other hand, native LDL was a better competitor than native HDL when measuring HDL uptake in trypanosomes (Figs. 1C and 1D). It has been shown for CD36 that oxidized LDL competes more effectively than LDL for HDL binding to CD36 (24).
The similar biochemical properties of the trypanosome lipoprotein scavenger receptor and mammalian CD36 superfamily members compelled us to analyze the interaction of TLF1 with the prototypical class B eukaryotic lipoprotein scavenger receptor, SR-BI, which exhibits the highest degree of selective lipid uptake (29,30). We found that like HDL, TLF1 is able to donate lipids via this eukaryotic lipoprotein scavenger receptor (Fig. 6), indicating that TLF1 can bind to and productively interact with SR-BI. Although there was a specific association of TLF1 with CHO cells expressing SR-BI we did not observe any obvious toxicity at physiological concentrations of TLF1 (~20 µg/ml).
Both apolipoprotein and lipid are taken up by the parasite. The lipid is selectively removed from the lipoprotein and distributed throughout the cell (Fig. 3A). The apolipoprotein localizes to by guest on March 24, 2020 http://www.jbc.org/ Downloaded from endocytic vesicles (Fig. 3D). The distribution of protein is very different from that seen for lipid, suggesting that at some point after interaction with a receptor the lipid is selectively removed and dispersed throughout the cell. The current hypotheses for HDL uptake in eukaryotic cells involve either retro-endocytosis or the formation of a non-polar channel, created by the binding of the apolipoprotein at the cell surface (28), through which lipids are delivered. We have demonstrated that HDL apolipoprotein is found inside the trypanosome (Fig. 3D). Because we detect intracellular HDL and we do not detect degradation (not shown), the majority of the endocytosed apolipoprotein may be recycled back to the cell surface. Other researchers have found that trypanosome endocytosis of HDL (13) and more recently TLF1 (49) do not result in the degradation of the apolipoproteins. In contrast, apolipoprotein B of LDL is rapidly degraded during its transit through the endocytic machinery (50). The reason for this difference in proteolytic processing is not known, it may be due to differential routing of the ligands in the endocytic pathway or differential sensitivity to endosomal and lysosomal proteases.
Physiological concentrations of HDL can compete for at least ~80% of the binding of TLF1 to T. b. brucei (Fig. 4). It has been proposed that the remaining ~20% of TLF1 is taken up by another trypanosome receptor that recognizes haptoglobin (21). Irrespective of whether there are one or more receptors for TLF, if there is competition for binding of TLF, there should be competition for uptake. It has been reported that TLF1 exhibits both high affinity (0.75-3.6 µg/ml) and low affinity (80-175 µg/ml) binding to trypanosomes, and that only the low affinity sites are competed by HDL (20,21). We believe, as has been proposed by others for LDL binding to trypanosomes (10,51), that the low affinity sites for TLF1 may represent single receptors along the flagellum and within the pocket, whereas high affinity sites represent by guest on March 24, 2020 http://www.jbc.org/ Downloaded from dimerized or clustered receptors within the flagellar pocket. Complete inhibition of binding or uptake at the receptor level, requires the competing ligand to be at least 100 fold above its own Kd in order to saturate all of the available receptors (52). Therefore complete inhibition of TLF1 binding and uptake at the receptor level would require 2.7-8 mg/ml of HDL (13,21), and 13-33 mg/ml LDL (44). These concentrations are above the physiological levels found in plasma which are ~1 mg/ml. Therefore, in vivo as in our assay, the lipoproteins would be able to attenuate the killing by TLF1 but they would not inhibit the killing. This is illustrated by the attenuation of TLF1 mediated lysis (Fig. 5) in the presence of HDL (1-1.6 mg/ml) and LDL (0.75-1 mg/ml).
Oxidized lipoproteins were in themselves trypanolytic 2 , and we therefore could not evaluate their capacity to attenuate TLF mediated lysis.