2.1. Background
The plasmodial surface anion channel (PSAC) is the prime example of an unusual ion channel present only in
Plasmodium spp. This channel is on the host erythrocyte membrane and serves an essential role in nutrient acquisition for the intracellular parasite (
Figure 2A). It accounts for the increased permeability of infected erythrocytes to a broad range of organic and inorganic solutes, as first identified some 75 years ago and characterized using tracer accumulation, osmotic fragility and other transport assays in numerous studies before 2000 [
2,
3,
4,
5].
These early studies had three key limitations, all of which arose because the transport methodologies depended on macroscopic flux measurements on populations of cells. First, the precise mechanism of solute uptake was unclear, with proposals including one or more parasite- or host-derived ion channels or transporters, lipid defects resulting from parasite invasion, fluid-phase endocytosis, and membranous ducts that could provide direct access to plasma [
4,
35,
36,
37]. Second, because endocytosis and membranous ducts were possible, macroscopic measurements could not determine the subcellular location of the solute flux across membranes. Finally, although there was substantial interest in identifying the molecular basis [
38], these uncertainties prevented systematic studies aimed at gene identification.
A first advance in addressing these questions came from introduction of patch-clamp methods. The first patch-clamp recordings of infected human erythrocytes identified PSAC as an ion channel mechanism of transport [
39]. Cell-attached patch-clamp revealed individual channel molecules on infected erythrocytes that were conserved in divergent
Plasmodium spp. [
39,
40]. Because this method could limit the measurement of ion flux to a small “patch” of the erythrocyte plasma membrane [
41], it also established the host cell membrane as the location of the transport activity. This study also reported whole-cell patch-clamp of infected cells, allowing a quantitative estimate of 1000-2000 functional channel molecules on a mature infected cell. When combined with noise analysis of single-channel and whole-cell recordings [
42], the study also determined that PSAC was the predominant conductive pathway for ion flux on infected erythrocytes.
Interestingly, subsequent studies from other groups confirmed increased anion channel-mediated currents at the host membrane, but suggested multiple distinct channel types [
43,
44,
45,
46]. These studies also proposed various regulators of channel activity, including activation by oxidative stress [
45], cyclic nucleotides [
44], membrane stretch [
47], and a link to various mammalian ion channels [
46,
48].
The completion of the
Plasmodium falciparum whole-genome sequence failed to identify orthologs of known anion channels [
14], raising questions about the molecular basis of the identified channel(s). Because synthesis, trafficking and insertion of parasite proteins at the host membrane was also considered overly complicated, nearly all workers assumed that the observed channels were host proteins that become activated or modified by the intracellular parasite [
43,
49]. There was already evidence for parasite modification of some erythrocyte membrane proteins [
38,
50,
51], so this model appeared to be the most conservative one. It was, however, difficult to reconcile simple modification of preexisting host membrane proteins with formation of a channel having PSAC’s remarkable properties and with the observation that channel seems to be fine-tuned to permit parasite survival and replication.
2.2. Identification of the rhoph genes as PSAC determinants
These unusual properties prompted us to seek parasite genetic elements in PSAC formation. The first experimental evidence supporting parasite genes was the identification of differences in the channel’s voltage-dependent gating [
52], a term that encompasses the process of opening and closing of the pore to ion flux. More compelling evidence included identification of distinct PSAC mutants with altered solute flux, selectivity, single channel gating, and pharmacology [
53,
54,
55]; these mutants were identified through in vitro selection with blasticidin S and/or leupeptin, antiparasitic toxins that require PSAC-mediated uptake to reach their intracellular targets. These studies suggested that continuous cultivation with these toxins selected for outgrowth of mutants with reduced toxin uptake at the erythrocyte membrane.
Based on these findings, we screened a library of > 50,000 small molecules for PSAC inhibitors that produce differential block of channels associated with geographically divergent parasite clones. These high-throughput screens led to the identification of ISPA-28, a unique inhibitor that blocks channels associated with the Dd2 clone with 800-fold higher affinity than those on from other parasite clones such as HB3 [
56]. An available Dd2 x HB3 genetic cross was then used to track inheritance in 34 progeny clones, revealing that most daughter parasites produced channels matching one or the other parental line and providing conclusive evidence for parasite genetic elements. Linkage analysis implicated a single locus near the 5’ end of parasite chromosome 3. Because none of the genes in this locus resembled known ion channel genes, we then used DNA transfection of the Dd2 line to produce merodiploid parasites expressing both the HB3 and Dd2 alleles of each of the 15 genes in this locus. While transfection with 13 of the genes did not change ISPA-28 affinity, complementation to add the HB3 allele of 2 related genes,
clag3.1 and
clag3.2, yielded an intermediate phenotype, as expected if both parental channel types are expressed on the host membrane [
56]. Allelic exchange and a single nonsynonymous mutation at a highly conserved residue in the leupeptin-resistant PSAC mutant further supported a primary role of the encoded CLAG3 protein. CLAG3 localized to the erythrocyte membrane consistent with the site of PSAC activity. A small variant motif was found to be exposed at the host cell surface and later shown to account for the differential block by ISPA-28 and other clone-specific inhibitors [
57,
58,
59]. Variation at this site strongly suggests selection against the exposed CLAG3 loop by host immune responses [
60,
61]. Immune selection is also supported by the high levels of anti-CLAG3 antibodies in endemic populations and by epigenetic silencing of
clag genes [
62,
63,
64,
65,
66].
Further evidence for a CLAG3 role in PSAC formation came from independent genetic mapping studies [
58,
67] and from epigenetic silencing of CLAG3 and CLAG2, a paralog encoded by a gene on parasite chromosome 2, in blasticidin S-resistant PSAC mutant [
65,
68]. These and other CLAG paralogs were known to associate with RhopH2 and RhopH3, two unrelated proteins encoded by single copy genes conserved in
Plasmodium spp. [
69,
70,
71]. We and others therefore examined these paralogs for their possible contributions to PSAC formation. While
clag3-knockouts can be produced and propagated in nutrient-rich parasite culture media [
58],
rhoph2 and
rhoph3 could not be disrupted using CRISPR-Cas9 with several high-scoring sgRNAs that efficiently cleave the genome to allow gene-editing [
72]. Conditional knockdowns of these genes were therefore produced, revealing that both RhopH2 and RhopH3 are trafficked to the host membrane and are essential for PSAC formation [
72,
73,
74]. RhopH3, but not CLAG3 or RhopH2, also contributes to host cell invasion. Biochemical studies suggest that each subunit is essential for PSAC formation, but that CLAG3 is dispensable because its paralogs—CLAG2, CLAG8 and CLAG9 in
P. falciparum with at least two paralogs encoded by each examined
Plasmodium species—can compensate for loss of CLAG3. Consistent with this model, RhopH2 and RhopH3 cannot be disrupted as they are encoded by single copy genes in all
Plasmodium spp.
These studies have provided compelling evidence for parasite genetic elements and implicated the above gene products, which together form the RhopH complex. These findings are remarkable as workers had previously assumed this complex functioned in either cytoadherence or erythrocyte invasion [
75,
76,
77,
78]. They are also surprising because none of the subunits have homology to known channel proteins from higher organisms. Most fundamentally, they also lack the number of predicted transmembrane domains generally thought to be present in channel-forming proteins (
Figure 2B). A tantalizing hint comes from the single amphipathic transmembrane domain detected in CLAG3, where helical wheel analysis reveals that polar residue side chains align at one face of the α-helical transmembrane domain with hydrophobic residues at the face [
79]. This arrangement, confirmed by the subsequent determination of the RhopH complex structure by cryo-EM [
80], is often found in pore-forming proteins, suggesting direct formation of the PSAC pore by these unusual proteins. Immunofluorescence studies as well as live-cell FRET reveals that all three members of the RhopH complex traffic together to the host membrane and that at least two of the subunits, RhopH2 and CLAG3, remain tightly associated after insertion at the host membrane [
81], further supporting a model where these unusual parasite proteins form the ion channel.
At the same time, conclusive evidence for direct PSAC formation by the RhopH proteins is still missing. Biochemical studies reveal that the RhopH complex is manufactured as a soluble complex and that it is trafficked and eventually inserted at the host membrane [
80]. The
de novo structure of the soluble complex has been determined by cryo-EM microscopy using protein purified from an engineered
P. falciparum line cultivated in human erythrocytes (
Figure 2C, [
80]), revealing that the single transmembrane domains of each subunit are buried in the trafficking complex. This finding suggests marked conformational changes associated with membrane insertion. These findings were also confirmed without purification of the RhopH complex using a novel cryo-ID method [
82,
83].
An important unanswered question is the structure of the membrane-embedded RhopH complex. This could directly implicate these proteins in PSAC formation, provide insights into how solutes permeate through the channel, and elucidate the mechanisms that underly the channel’s remarkable selectivity properties, as discussed below.
2.4. Essential role in nutrient uptake and a druggable target
Since its discovery, multiple studies have proposed various roles for the increased permeability of infected cells to diverse solutes. These include 1) nutrient acquisition for the developing intracellular parasite [
1,
3,
4,
5,
86], 2) cation remodeling to raise [Na
+] and lower [K
+] in host cell cytosol [
2,
87,
88], 3) volume regulation of infected cells by allowing efflux of excess amino acid production through hemoglobin digestion [
89], 4) timed osmotic lysis of infected cells to allow daughter parasite egress from infected cells at the end of the intracellular cycle [
88], and 5) a nonessential byproduct of infection and intracellular parasite metabolic activity [
43]. Each of these proposals had some merit and was based in an understanding of parasite biology, but experimental evidence was missing and difficult to obtain prior to identification of the channel genes. Gene identification and experimental advances have now clarified the roles as discussed in this section.
A PSAC role in nutrient acquisition would be consistent with the channel’s high permeability to sugars, purines, key vitamins, and the essential amino acid isoleucine, all of which are required for parasite development and not available in adequate quantities within uninfected erythrocytes [
1]. Although parasite killing by nonspecific PSAC inhibitors supported this and other proposed essential roles [
90,
91], uncertainties about mechanism of killing limited interpretation. Indeed, selection of a resistant mutant with unaltered PSAC activity and inhibitor affinity confirmed that phlorizdin, a commonly used inhibitor, kills parasites through action on unrelated targets [
92]. To address this longstanding uncertainty, we developed a modified medium, termed
PSAC
growth
inhibition
medium (PGIM), with lower, more physiological concentrations of three key nutrients acquired via PSAC [
67]. In contrast, the standard medium used in most labs, RPMI 1640 supplemented with a lipid source, has most nutrients present at concentrations > 10-fold higher than levels in plasma from healthy donors. We found that ISPA-28 had low potency against parasite growth in standard medium, but that it killed Dd2 parasites at nearly 800-fold lower concentrations than HB3 parasites in studies using PGIM, paralleling its clone-specific action against PSAC in these lines. Linkage analysis using the Dd2 x HB3 progeny clones and this difference in growth inhibitory activity mapped the
clag3 locus, establishing that PSAC block accounts for killing by this uniquely specific inhibitor. Other PSAC inhibitors that do not exhibit differential activity against lab clones are also more effective against in vitro parasite growth in PGIM than in standard RPMI 1640-based media, whereas antimalarials acting on unrelated targets have indistinguishable
IC50 values in these media [
67]. Importantly, because these studies required use of a nutrient-optimized medium, they provided the first experimental evidence for an essential role in nutrient uptake.
The cation remodeling role for PSAC is based on the observation that infected cells gradually incur a concomitant increase in [Na
+] and decrease in [K
+] with intracellular parasite development because of the nonzero PSAC permeability to these cations [
2,
85,
88,
93]. The leak of these cations at the erythrocyte membrane dissipates the outward and inward gradients for these respective ions, as maintained by the host cell Na
+/K
+ ATPase pump [
94]. This host cytosol cation remodeling was hypothesized to make the erythrocyte more hospitable for parasite growth, possibly by providing an inward Na
+ gradient for coupled solute uptake at the intracellular parasite plasma membrane [
87]. To explore this role, we designed and used a separate modified medium, 4suc:6KCl, that replaces the Na
+ salts in the RPMI 1640-based medium with K
+ salts and sucrose to preserve infected cell osmotic stability. Growth studies revealed that this medium supports unabated parasite growth without a need for adaptation, a remarkable finding in light of the marked changes in composition. Because it abolishes Na
+ and K
+ gradients across the erythrocyte membrane, parasite cultivation in 4suc:6KCl prevented PSAC-mediated cation leak and cation remodeling, as confirmed with infected erythrocyte ion content measurements [
95]. This study revealed unexpectedly low Na
+, K
+, and Cl
- requirements for parasite development; it also provided compelling evidence against Na
+-coupled phosphate uptake at the parasite plasma membrane and excluded an essential role of K
+ signaling in merozoite activation [
96].
Alternate roles for PSAC have also been examined. Volume regulation of infected cells by allowing PSAC-mediated efflux of excess amino acids generated by hemoglobin digestion remains possible [
89]. One prediction of this hypothesis is that potent PSAC inhibitors should lead to osmotic lysis of infected cells because of blocked amino acid efflux; as this has not been observed [
67,
97], this hypothesis should be considered with some caution. Another hypothesis, timed osmotic lysis of infected cells, proposes that gradual Na
+ and K
+ leak through PSAC leads to osmotic swelling and lysis ~ 44 hours after invasion; as this coincides with the duration of
P. falciparum intracellular development, osmotic lysis may facilitate parasite egress at the end of the erythrocytic cycle [
88]. This role is excluded by the normal developmental cycle in studies using 4suc:6KCl medium, where Na
+ and K
+ leak are abolished [
95]; it is also inconsistent with studies implicating protein kinases in coordinated parasite egress [
98,
99]. Finally, proposals that PSAC is a nonessential byproduct of intracellular parasite development are excluded by
rhoph2 and
rhoph3 knockdown and by an advanced drug discovery and development project [
72,
73,
74,
97], both of which establish this target’s essentiality for bloodstream malaria parasites.