Report

Erythrocyte G Protein-Coupled Receptor Signaling in Malarial Infection

See allHide authors and affiliations

Science  19 Sep 2003:
Vol. 301, Issue 5640, pp. 1734-1736
DOI: 10.1126/science.1089324

Abstract

Erythrocytic mechanisms involved in malarial infection are poorly understood. We have found that signaling via the erythrocyte β2-adrenergic receptor and heterotrimeric guanine nucleotide–binding protein (Gαs) regulated the entry of the human malaria parasite Plasmodium falciparum. Agonists that stimulate cyclic adenosine 3′,5′-monophosphate production led to an increase in malarial infection that could be blocked by specific receptor antagonists. Moreover, peptides designed to inhibit Gαs protein function reduced parasitemia in P. falciparum cultures in vitro, and β-antagonists reduced parasitemia of P. berghei infections in an in vivo mouse model. Thus, signaling via the erythrocyte β2-adrenergic receptor and Gαs may regulate malarial infection across parasite species.

Plasmodium falciparum is a protozoan parasite that causes the most virulent form of human malaria. It infects both hepatocytes and mature red blood cells, but the erythrocytic stages of infection are responsible for all of the symptoms and pathologies associated with the disease (1). Parasite invasion is a complex, multistep process where the host erythrocyte membrane undergoes involution and deformation, followed by invagination and swelling (2). However, uninfected erythrocytes are incapable of pinocytocis or endocytosis, and host pathways are not known to be involved in signaling the entry of P. falciparum.

Heterotrimeric guanine nucleotide–binding regulatory proteins (G proteins) constitute a well-characterized class of signal transduction proteins in mammalian cells (3). They regulate important cellular processes ranging from transcription, motility, secretion, and contractility (4). G proteins reside at the cytoplasmic face of the cellular plasma membrane, where they can couple with a variety of transmembrane receptors to transduce extracellular signals initiated by many hormones, neurotransmitters, chemokines, and autocrine and paracrine factors to a wide range of effectors within the cell (5). G proteins are activated by guanine nucleotide exchange factors (GEFs), which promote the dissociation of guanosine diphosphate (GDP) from the inactive G protein and replacement with guanosine triphosphate (GTP). G protein–coupled receptors (GPCRs) act as GEFs for large heterotrimeric G proteins. Four major families of α subunits—Gαs, Gαi/o, Gαq/11, and G12/13—have been described and each specifies a distinct set of downstream signals (4). Although G proteins have been intensively studied in a wide range of cells, their functions in mature red blood cells are poorly understood. These cells are enucleated, have no intracellular structures, and are incapable of de novo protein and lipid biosynthesis (6, 7).

The Gαs present in erythrocytes can be isolated in detergent-resistant membrane (DRM) rafts and is recruited to the malarial vacuole (8). Another heterotrimeric protein, Gαq, is also present in the red blood cell but does not concentrate at the malarial vacuole (Fig. 1). The Gαs-coupled β2-adrenergic receptor (β2-AR), which is detected in DRMs (9), was also recruited to the vacuolar parasite (Fig. 1). To determine whether the recruitment of Gαs to the plasmodial vacuole may have functional consequences for infection, we introduced into cocultures of parasites and erythrocytes peptides derived from the C-terminal region of Gαs that block the interaction of this G protein with its activating receptors. The last 11 amino acids of the G protein are critical for interaction with GPCR (10) and competitively block Gα association with GPCRs and abrogate downstream signaling events. Gαs peptide [QRMHLRQYELL (11)] reduced infection of erythrocytes by P. falciparum by 87% (Table 1). In contrast, the scrambled Gαs (Gαscr), which contained the same amino acids as the Gαs peptide but in a different sequence [ELRLQHYMQLR (11)], inhibited infection by less than 5% (Table 1). Thus, the inhibition of infection effected by the Gαs peptide was dependent on its sequence, and the peptide probably blocked Gαs function (12). Database searches failed to reveal Gαs (or other heterotrimeric G proteins, although Rab GTPases are present) in the P. falciparum genome (13). One report suggests the presence of plasmodial heterotrimeric G proteins (14), but there are no parasite heterotrimeric G proteins that are homologous to mammalian G proteins, and the Gαs peptides used here were expected to selectively disrupt host Gαs protein function.

Fig. 1.

Distribution of endogenous Gαs, Gαq, and β2-AR (green) in newly formed P. falciparum-ring-infected erythrocytes as detected by indirect immunofluorescence assays (8). Parasite (P) nucleus is Hoechst-stained (blue); arrowhead indicates red-cell plasma membrane. Scale bar indicates 3 μm.

Table 1.

Effects of C-terminal peptides of human heterotrimeric Gαs on P. falciparum infection of erythrocytes. In vitro infection assays were incubated with the indicated peptides (Gαs or Gαscr) or mock treated (25). Infection is shown relative to that seen with mock cultures; parasitemia in mock cultures was at ∼20%. Standard error is 10%. Cultures incubated with Gαs peptide were inhibited in new ring formation.

Peptide Sequence Inhibition of infection
Gαs QRMHLRQYELL (View inline) 87%
Gαscr ELRLQHYMQLR (View inline) 4%

Erythrocytic infection by P. falciparum is initiated when the extracellular merozoite stage enters the red blood cell to form an intracellular ring. The Gαs peptide displayed a dose-dependent inhibition of new ring formation (table S1). Intracellular development of rings through “trophozoite” and terminal “schizonts” stages (that rupture to release merozoites) remained unaffected, suggesting that the addition of the Gαs peptide blocked a step in erythrocyte entry. To determine how peptides gained access to the erythrocyte, we preincubated fluorescent [fluorescein isothiocyanate (FITC)–labeled] forms of Gαs and Gαscr peptides (fig. S1) with red blood cells or late-stage schizonts and segmenters for four hours and subsequently used them in an infection assay. This pretreatment failed to block infection (fig. S2), suggesting that the peptides, though acetylated, could not enter cells directly. However, when FITC-labeled peptides were added to the invasion assay, fluorescence was found in red blood cells associated with parasites (Fig. 2A). Staining with antibodies to MSP1 (a protein on the surface of invasive merozoites) detected extracellular parasites blocked in entry. Although this assay under-estimates the block in parasite entry (because extracellular merozoites are prone to degradation) in incubations containing FITC-Gαs, 75% of the parasites were detected by the antibodies to MSP1 (Fig. 2A, a). In incubations containing FITC-Gαscr, 90% of the parasites were not detected by the antibodies to MSP1, indicating that they were intracellular (Fig. 2A, b). Thus, peptides gained access to red blood cells at the time of parasite entry (Fig. 2B), probably across the nascent vacuole; the vacuolar membrane has been shown to have altered permeability (15).

Fig. 2.

Model for G peptide translocation into erythrocytesduring parasite infection. (A) FITC-Gαs(a) or FITC-Gαscr (Gαs-scrambled) (b) peptides (green) were added to an in vitro infection assay, probed with antibodies to MSP1 (red) to detect extracellular parasites, and scored (25). Parasite (P) nucleus is Hoechst-stained (blue). Scale bar, 3 μm. (B) A schematic drawing of Gαs inhibition of ring formation. On the basis of data in Fig. 2A, table S2, fig. S1, and fig. S2, we propose that the peptide was taken in with the parasite and translocated across the nascent or newly formed vacuole. Presence of the FITC-Gαspeptide prevented intracellular ring formation, where as FITC-Gαscr allows intracellular ring formation.

Peptides that block interaction of Gαs with its receptors also block malarial infection. Thus, activation of Gαs via its receptors may influence malarial infection. The two major Gαsassociated receptors known to be present on red blood cells are the β-ARs and the adenosine receptors. Agonists of both the β2-ARs and adenosine receptors stimulated infection of P. falciparum (3D7 strain) in vitro about twofold (Fig. 3A). Stimulation was dose-dependent (fig. S3). Competitive antagonists blocked this stimulation. A combination of a β2-AR agonist and adenosine receptor agonist showed additive effects in stimulating infection. Gαs peptides blocked infection by 80 to 90%, suggesting that the receptors were mediating their effects via Gαs. The same level of agonist-stimulated infection, and its inhibition by antagonists, was detected when another strain of P. falciparum (FCB) was used (Fig. 3B), suggesting that the mechanisms of G protein regulation of infection have been conserved across independent strains of P. falciparum.

Fig. 3.

Effects of agonists and antagonists of the β-ARs and adenosine receptors on infection of P. falciparum and cAMP production in erythrocytes. (A) P. falciparum (strain 3D7) infection of erythrocytes in cultures that are mock-treated control (C) cultures or treated with agonists and antagonists of the β2-AR [isoproterenol (I)-agonist and racemic (+/–) propranolol (P)-antagonist] or agonists and antagonists of the adenosine receptor [5′-N-ethylcarbox-amidoadenosine (N) and adenosine (A) are agonists; 8-(p-sulfophenyl)theophilline (S) is an antagonist] (25). With the exception of adenosine (which was used at 1 mM), effective concentrations used for agonists and antagonists were at 105 M (suggesting a low abundance of receptors on erythrocytes). Changes are shown relative to control cultures, which achieved parasitemias of 9 to 11% or greater (25). Standard error is10% in data from triplicate assays. (B) P. falciparum (strain FCB) infection of erythrocytes in cultures under conditions described in (A). (C) cAMP production upon activation of β2-ARs and adenosine receptors in infected erythrocytes was measured with the use of the Direct cAMP Enzyme Immunoassay (Assay Designs, Incorporated, Ann Arbor, MI) kit (25). A two- to fourfold stimulation in cAMP production was seen in triplicate assays. (D) Specificity of various β2-AR antagonists for P. falciparum infection. Isoproterenol (I) and racemic (+/–) propranolol (P) were tested, along with inactive propranolol isomer (+P), neutral antagonist alprenol (Al), and inverse agonist ICI (±)-1-[2,3-(dihydro-7-methyl-1H-inden-4-yl)oxy]-3-[(1-methylethyl)amino]-2-butanol (Ic). All agonists and antagonists were used at 105 M. Infection assays were as described in (A).

To confirm that Gαs-coupled receptors signal in red blood cells, we demonstrated that agonists for both the β2-ARs and adenosine receptors stimulated cyclic adenosine monophosphate (cAMP) accumulation in red blood cells (Fig. 3C). Thus, both GPCRs were functional in erythrocytes, and inhibition of host Gαs signaling in erythrocytes blocked malarial infection. To investigate further the pharmacology of receptor-mediated inhibition of infection, we tested the effects of inactive stereoisomers, neutral antagonists, and inverse agonists of β2-AR. An inverse agonist is a compound with a negative intrinsic activity, i.e., a compound that produces conformational changes in the receptor that are less favorable to activation of G proteins than the ground state (16). Although racemic preparations for the antagonist propranolol were active in blocking (agonist) isoproterenol-stimulated infection, there was no effect with the (+) inactive stereoisomer of propranol (Fig. 3D). Thus, inhibition of infection was because of propranolol acting at the receptor and not any nonspecific membrane effects. The neutral antagonist alprenolol was slightly less active than racemic (+/–) propranolol, whereas the inverse agonist ICI 118,551 (ICI) was more efficacious and could even reduce baseline levels of infection by about 50%. The greater efficacy of ICI suggests the existence of precoupling of receptors to G proteins. Although theoretically such inverse agonists should be more efficacious, this has yet to be shown in many systems (17, 18).

The in vitro studies led us to investigate whether Gαs receptor antagonists could influence parasite proliferation in a mouse model using P. berghei, a rodent malaria parasite (19). For racemic propranolol, the median inhibitory concentration (IC50) appeared to be a dose of 7.5 mg/kg (administered twice daily; fig. S4A). The median lethal dose (LD50) of propranolol for intravenous injection into rodents is 470 mg/kg per day, and the LD50 for intraperitoneal injection (the route used here) was expected to be even higher, suggesting that these compounds were well tolerated. At 7.5 m/kg, the inactive (+) propranolol isomer had no effect on in vivo infection; the neutral antagonist alprenol showed a reduction of ∼30%, whereas inverse agonist ICI inhibited parasitemia by ∼50% (fig. S4B). This trend of inhibition was consistent with the inhibitory effects of these compounds seen on agonist-stimulated in vitro infections of P. falciparum. Thus, signaling via erythrocyte β2-AR and Gαs activation appears to be conserved across parasite species.

Although we found some degree of precoupling of β2-AR to Gαs in erythrocytes (evidenced by efficacy of neutral antagonists in vitro), there appeared to be stimulation of receptors during infection in vivo. How erythrocyte Gαs-coupled receptors are stimulated during malarial infection in vivo remains to be understood. One possibility is that catecholamine levels are augmented during infection, resulting in stimulation of receptors such as β2-AR. An increase of catecholamines may come from an elevation of the sympathetic response or the production of catecholamine-like molecules by the malaria parasite. This may explain why antagonists like propranolol are effective at reducing parasitemias in vivo.

The use of erythrocyte G-protein raft-associated signaling mechanisms in malarial entry and/or establishment of the vacuole may provide a reason why (i) erythrocyte rafts are required and (ii) their resident proteins are internalized in infection (8, 9, 20). Raft-dependent G-protein signaling has been demonstrated in cells (21). Although the physiological functions of G protein receptors and their associated signaling mechanisms in erythrocytes are not well understood, an emerging idea is that they may contribute to interactions with endothelial cells (22). Because both Gαs and β2-AR were internalized and associated with the vacuolar parasite, their activation in malarial infection may regulate a step of vacuole formation that is conserved across parasite species. This may explain why the same antagonists inhibit infection by human malarias like P. falciparum and rodent malaria parasites like P. berghei. Signaling via G proteins rapidly reorganizes the cellular cytoskeleton in nucleated mammalian cells (23, 24). In P. falciparum–infected erythrocytes, nearly all skeletal components and attached integral proteins associated with the host plasma membrane are excluded from the vacuole. Further parasite entry culminating in intravacuolar residence occurs within minutes. Thus, signaling via GPCR may underlie the rapid and dynamic reorganization of submembranous cytoskeleton required for infection of the nonendocytic, mature erythrocyte by this major human pathogen.

Supporting Online Material

www.sciencemag.org/cgi/content/full/301/5640/1734/DC1

Materials and Methods

Figs. S1 to S4

Table S1

References and Notes

View Abstract

Navigate This Article