The dynamics of blood feeding

The first study to demonstrate unequivocally the participation of the gut in schistosome nutrition, involved the injection of ³H-leucine-labelled reticulocytes into infected mice [28]. Worms harvested by perfusion four days later were found to have incorporated the radiolabel, the females substantially more than males in keeping with their known greater appetites [28]. The leucine was extensively distributed in the tissues of the schistosomes, while the gut wall displayed radioactivity, suggesting amino acid uptake across the schistosome gastrodermis [28]. Later work with adults [29] and schistosomula [30] confirmed the finding. Note, however, that the autoradiographic technique used in these studies reveals the site(s) where labeled amino acid is incorporated into protein (i.e., biosynthesis), not the site of uptake, so the evidence is circumstantial.

Live S. mansoni worms viewed in the portal vasculature of an anaesthetized mouse were observed to face against the direction of blood flow, and feed in an energy efficient manner by simply opening and closing the oral cavity every 1–2 seconds [31]. Such feeding movements were either continuous for the observed period (60 minutes) or lasted ∼10 minutes to be repeated after a rest period of about the same duration [31]. Recent video recordings of S. mansoni males in vitro at 37°C revealed that erythrocytes were acquired from a dilute suspension by a “rapid grabbing motion” of the oral sucker, approximately four times per second [26]. A bolus of ingested blood accumulated in the lumen of the anterior esophagus. A wave of peristalsis then passed down the esophagus to drive the bolus to the posterior compartment and on into the anterior gut. The actual rate of blood consumption in sentient mice has been determined in a single elegant study involving intravenous administration of ⁵¹Cr-radiolabeled erythrocytes [32]. Male and female adult S. mansoni ingested some 39,000 and 330,000 erythrocytes per hour, equating to daily intakes of 105 nl and 880 nl of whole blood, respectively. Adjusted for packed cell volume, the female ingests ∼484 nl of plasma fluids per day, equivalent to 32.3 gut volumes, or 4.4 body volumes. The values for the male are 58 nl of plasma fluids equivalent to 7.7 gut volumes or 0.2 body volumes of fluid (Table S1). Time course measurements revealed that a steady state, where ingestion of label was balanced by its expulsion, was reached in approximately eight hours [32].

The esophageal gland

Recent work on the functions of the esophageal gland has revealed that the esophagus has a central role in blood processing, not acting simply as a conduit [26]. Erythrocytes are rapidly lysed upon entry to the posterior esophageal compartment so that intact cells are seldom observed there [26], [27]. Remarkably, leucocytes are somehow tethered in the posterior lumen and can be observed, in varying states of degeneration, as a stationary plug (Figure 2C) [26]. This trapping could represent a mechanism to prevent, e.g., neutrophils from making a destructive oxidative burst in the gut lumen. Esophageal gland cell-bodies synthesize large numbers of crystalloid vesicles and release their contents into the lumen [26], [33]–[35]. By a combination of whole mount in situ hybridization and immuno-cytochemistry, a small number of esophageal gland-specific proteins has been identified. These include the Micro Exon Gene (MEG) products 4.1, 4.2 and 14 [26] and Venom Allergen Like (VAL) protein 7 [36]. They likely play roles in the cell tethering process and/or the extensive host cell lysis observed in the esophagus.

Proteins

Digestion of lysed cell material and plasma constituents in the gut is largely extracellular. The pH of the gut has been measured at 5.0 or lower [37] and this may aid both in the unfolding of ingested proteins and in the activation of gut proteases (many of which function more efficiently in an acid environment) [38]. In this context, it has been suggested that lysosomes in the gastrodermal syncytium secrete their contents directly into the lumen to aid in the digestion of incoming blood components [11]. A battery of hydrolytic proteinases, all with an acidic pH optimum, has been identified in the regurgitated gut contents (vomitus) by proteomic analysis [11]. These include the cysteine endopeptidases, cathepsin L, cathepsin B, asparaginyl endopeptidase (legumain), and a cathepsin D-like aspartyl endopeptidase [39]; the localization of these proteases in the gastrodermis has been confirmed by immunocytochemistry [38], [40]–[43]. Exopeptidases, such as dipeptidylpeptidase (cathepsin C) [44] and proline carboxypeptidase homologs have also been identified [11]. A leucine aminopeptidease (LAP), with optimal activity at neutral pH, has been immunolocalized to the gastrodermal cells lining the lumen of adult schistosomes; LAP-immunoreactivity was considerably stronger within the gut of females versus males [45]. It has been proposed that this enzyme cleaves peptides that are taken in from the lumen into the cells. The resulting free amino acids are then distributed to other internal tissues [45].

The importance of these proteases for schistosomes has been demonstrated in some cases by suppressing the expression of their genes. For instance, cathepsin D-suppressed schistosomula appear unable to properly digest hemoglobin and do not survive to maturity when they are used to infect mice [43] and cathepsin B1-suppressed worms exhibit retarded growth compared to controls [46].

It has been proposed that the host specificity of blood-feeding organisms like schistosomes is determined by the molecular compatibility between the parasites' hemoglobin-digesting enzymes and the corresponding host species' hemoglobins [47]. In support of this, protease activities from S. japonicum (a schistosome with a very wide host range) cleave hemoglobin from bovine, sheep, and horse blood more efficiently than does the activity from extracts of S. mansoni (a schistosome with a limited host range, essentially confined to humans and some rodents) [48].

The quantitative data on blood ingestion [32] equate to a daily protein intake by adult females of ∼186 µg and males of only ∼18 µg (Table S1). Thus, a female worm ingests more than four times her dry weight in protein per day, whereas the male ingests only ∼0.2 times (Table S1). As the blood proteins are digested, the heme porphyrin ring within each hemoglobin subunit is released and detoxified by polymerisation into an inert, insoluble brown pigment hemozoin (Hz) [49], which gives the otherwise white schistosomes a dark appearance (Figure 2). The total heme intake per day is ∼0.56 µg for males and ∼4.8 µg for females (Table S1). However, the Hz content of a female is less than one-twelfth her daily intake of heme (at ∼0.375 µg heme per worm) while that of a male is about one-quarter (∼0.125 µg). This implies that there is little pooling of Hz pigment in the worm gut at steady state, the intake of hemoglobin being matched by Hz loss through regurgitation. The discovery of three iron-binding ferritin homologs in the vomitus indicates that some of the ingested iron may be sequestered for uptake into the gastrodermis (but see also the suggestion that a divalent metal transporter is localized in the tegument [50]). The identification of a calcium-binding calumenin homolog in vomitus suggests that the gut also plays a role in the regulation of calcium ions [11].

Lipids

Schistosomes cannot synthesize fatty acids or sterols de novo, so must obtain them in the diet and subsequently modify them to meet their metabolic need. The total lipid content of adult schistosomes is just over one-quarter of dry weight, but due to the differences in body mass and size of blood meal, the male ingests only about 2.5% of his lipid content per day, whilst the female ingests 50% (Table S1). For males at least, this implies significant lipid storage capacity, potentially in the parenchyma cells where lipid droplets are very evident. The female ingests more lipid in total than the male but since lipid is the same proportion of dry weight in both sexes, she must metabolize (or expel) more lipid on a daily basis than the male.

Low density lipoprotein (LDL) has been reported to bind to the tegument of S. mansoni schistosomula [51], [52] and LDL-binding proteins of varying molecular weights have been identified in different schistosome extracts [53], [54]. However, endocytosis of surface labeled lipid has never been observed. In addition, no clear LDL-receptor homolog has been identified in any analysis of the schistosome tegumental proteome [15], [18], [19], providing no evidence that lipid import occurs across the tegument. Regarding the gut, proteomic analysis of vomitus identified a Niemann Pick type C2 protein (NPC2) homolog [11]. NPC2 helps to traffic cholesterol and other sterols and glycolipids in eukaryotic cells; its presence in the schistosome gut suggests a similar function. Several saposin homologs were also found, and one of them, designated Sm-SLP-1, has been localized to the adult gastrodermis by immunocytochemistry [55]. By analogy with saposin function in other systems, these proteins may act to sequester lipids in the parasite gut lumen for uptake by gastrodermal cells [11]. These findings suggest that the gastrodermis plays a prominent role in lipid and fatty acid uptake.

Carbohydrates

It is clear that glycogen is a major storage polysaccharide accounting for 10.5% of the dry weight of males but only 2.8% of females [56], perhaps reflecting the male's greater muscular effort when transporting the female. However, the amounts of glucose ingested in blood per day are small, 0.11 µg for the male and 0.88 µg for the female (Table S1). This has to be set against the well-documented consumption of glucose by worms in culture and the concomitant secretion of similar amounts of lactate. For the male, the figure is 426 µg per day and for the female 115 µg, (representing 4.5 and 2.9 times the respective dry weights) [57]. This amounts to a massive 3872-fold greater consumption than ingestion by the male and 137-fold by the female, emphatically confirming the tegument as the site of glucose uptake, with the gut making a negligible contribution overall. Nevertheless, the identification of a glycan 1,4 beta glucosidase homolog in vomitus implies that complex carbohydrates can be digested in the gut [11]. Furthermore, incubating adult worms in vitro with ³H-glucosamine for 30 minutes revealed rapid accumulation of label in gastrodermal cells, suggesting the gut might also be a prominent site for uptake of amino sugars [29], but again with the caveat about autoradiography as a detection method.

Gastrodermal function

The real problem with the investigation of gastrodermal function is the inaccessibility of the tissue; no one has yet reported isolation of intact gastrodermis from the solid schistosome body. However, in a novel approach, slices of gastrodermal tissue from adult female S. mansoni and S. japonicum were isolated using laser microdissection microscopy to obtain gene expression profiles [58], [59]. As might be anticipated, transcripts encoding a range of endo- and exo-peptidases were abundant in the gastrodermis and expression of a number of saposins together with the NPC2 homolog was evident. A single amino acid transporter homolog was detected but no glucose transporter homologs [58], [59]. Genes associated with membrane trafficking through endocytic compartments and lysosomes were also expressed, reflecting the absorptive nature of the gastrodermis. In this context, it is notable that a recent feeding study with fluorescent-labeled dextran (MW 10 kDa) co-administered with an erythrocyte suspension to 28-day-old worms in vitro [11], showed that by one hour the dextran had reached the furthest extremity of the gut. By 24 hours, it had entered the gastrodermis, where it accumulated in ∼2 µm diameter aggregates. This is the first evidence for macropinocytosis, or receptor-mediated endocytosis, at the luminal surface of the gut epithelium [11]. The genes encoding the molecular machinery of endocytosis (e.g., clathrin heavy and light chains, assembly proteins AP2 and AP180) are present in the schistosome genome [60]; are they expressed in the gastrodermis?

Regurgitation

Following hemoglobin digestion in the schistosome gut, peptides and heme are released [49]. Heme is a complex of iron with protoporphyrin IX and an essential molecule for most organisms [61]. Evidence that heme is used in schistosome metabolism comes from parasites fed reticulocytes in which the heme moiety of hemoglobin was radiolabeled; they later display some radioactivity (∼20%) in their protein fraction [30]. However, due to its potentially pro-oxidant effects, the excess heme that may cause cell damage is converted to hemozoin (Hz) [61], envisioned as a lattice of hydrogen-bonded heme dimers, linked in a head-to-tail manner [62]. Hz formation has been proposed to occur at the surface of lipid droplets in the schistosome gut lumen [49]. However, Hz particles can also be seen within hemoglobin aggregates undergoing digestion (Figure 3A). As Hz is lipophilic, its presence in lipid droplets may simply reflect its natural partitioning there. In this situation, we suggest that as the luminal lipid droplets (some ringed with Hz) (Figure 3A, 3B) continue to be degraded by gut enzymes, eventually all that is left is an insoluble Hz core, adding to the dark, inert Hz accumulations in the schistosome gut (Figure 2).

The description of Hz expulsion from adult worms as “regurgitation” and naming the released material as “vomitus” suggests a violent process. However, worms in vitro are rarely seen voiding Hz unless they are subjected to a chemical, osmotic, or temperature shock [11], [63], [64]. Indeed, the one direct series of observations on adult worms in the intestinal vasculature of mice [31] makes no mention of Hz regurgitation. A simple explanation is that the approximately 26-fold reduction in overall mass when heme polymerises to Hz upon proteolysis of hemoglobin (Table S1) means there is much less material to be regurgitated than was ingested, so it will occur infrequently. In the recent video analysis of male S. mansoni feeding in vitro [26], regurgitation was recorded but did not involve reverse peristalsis of the esophagus. Instead, intense motor activity in the anterior gut, followed by relaxation of the esophageal wall muscles, resulted in the movement of a thin dark line of vomitus up the esophageal lumen and out through the mouth [26] (Figure 3D). It is unclear whether Hz can be selectively eliminated during regurgitation, while important cathepsins, peptidases, saposins, and a variety of carrier protein homologs are retained [11], [63], [64]. However, the Hz expelled into the bloodstream of infected animals [65], [66] has been proposed to exert an immunomodulatory effect on the host [67].

Besides Hz, other molecules are released by schistosomes and can be detected in the circulation of infected individuals [68], [69]. Most prominent among them are two O-linked glycoproteins called circulating anodic antigen (CAA) and circulating cathodic antigen (CCA) [70], which form a mucin-like coat over the gastrodermis. The major O-linked chains of CAA consist of long, negatively charged repeats of 1–6-linked N-acetylgalactosamine residues substituted with 1–3-linked glucuronic acid [71]. The glycan moiety of CCA consists of linear repeats of the Lewis x (Le^x) trisaccharide that comprise galactose β1–4 linked to N-acetlyglucosamine, the latter substituted with α1–3 linked fucose [72]. The polypeptide backbone of CCA is reputed to be a ∼39 kDa protein [73] but that of CAA is not known. Detection of CAA in serum and CCA in urine has been employed for immunodiagnosis of schistosomiasis [74]–[76]. The release of these glycan structures into the host bloodstream may have immunoregulatory consequences. Since carbohydrates with repeating Lewis x units are also found on circulating neutrophils of the host, the antigenic poly-Lewis x polysaccharide of CCA may induce auto-antibodies against granulocytes, causing the mild to moderate neutropenia observed during schistosome infection [72]. CAA has been shown to interact in vitro with the collagen-like domain of the first complement component Clq [77]. In this way it could interfere with complement cascade activation and/or the binding of Clq to its receptor.

The protonephridial system

The protonephridial system in schistosomes, ramifying throughout the parenchyma of both sexes, is a bifurcated highly branched network of tubules (terminal diameter ∼2 um), each branch ending in a flagellated flame cell (∼10 µm by 3 µm) (Figure 3E, 3F). The beating of the ∼50 flagella inside the barrel, composed of interdigitations between the cap cell and first tube cell, is believed to provide the motive force to suck interstitial fluids through the desmosome-like filter into the tubule lumen [78]. Cilia are found in some wider collecting tubules, presumably to aid fluid passage towards the posterior terminal pore [79]. The minute size of these composite elements and the small volume of the system overall means there is a dearth of information about the physiology of the system in any flatworm. From the localization of three proteins to the adult worm tubules, as part of other studies (Sm bone morphogenic protein, ER-60 cysteine protease and protein disulphide isomerase, [80]–[82]) we can infer the presence of signaling systems, and the synthesis of proteins for export. The fluorescent marker resoruffin, which diffuses into the worm across the body surface, is subsequently excreted via the protonephridial tubules [83], providing direct evidence for the involvement of this system in the excretion of metabolic wastes and xenobiotics. Furthermore, this process is likely facilitated by a multi-drug resistance protein [83], two copies of which are present in the S. mansoni genome [84].