There is a resurgence of interest in the Drosophila midgut on account of its potential value in understanding the structure, development and function of digestive organs and related epithelia. The recent identification of regenerative or stem cells in the adult gut of Drosophila has opened up new avenues for understanding development and turnover of cells in insect and mammalian gastrointestinal tracts. Conversely, the physiology of the Drosophila gut is less well understood as it is a difficult epithelial preparation to study under controlled conditions. Recent progress in microperfusion of individual segments of the Drosophila midgut, in both larval and adult forms, has enabled ultrastructural and electrophysiological study and preliminary characterization of cellular transport processes in the epithelium. As larvae are more active feeders, the transport rates are higher than in adults. The larval midgut has at least three segments: an anterior neutral zone, a short and narrow acid-secreting middle segment and a long and wider posterior segment (which is the best studied) that secretes base (probably HCO3–) into the lumen. The posterior midgut has a lumen-negative transepithelial potential(35–45 mV) and a high resistance (800–1400 Ω.cm2)that correlates with little or no lateral intercellular volume. The primary transport system driving base secretion into the lumen appears to be a bafilomycin-A1-sensitive, electrogenic H+ V-ATPase located on the basal membrane, which extrudes acid into the haemolymph, as inferred from the extracellular pH gradients detected adjacent to the basal membrane. The adult midgut is also segmented (as inferred from longitudinal gradients of pH dye-indicators in the lumen) into anterior, middle and posterior regions. The anterior segment is probably absorptive. The middle midgut secretes acid (pH<4.0), a process dependent on a carbonic-anhydrase-catalysed H+ pool. Cells of the middle segment are alternately absorptive (apically amplified by ≈9-fold, basally amplified by >90-fold) and secretory (apically amplified by >90-fold and basally by ≈10-fold). Posterior segment cells have an extensively dilated basal extracellular labyrinth, with a volume larger than that of anterior segment cells, indicating more fluid reabsorption in the posterior segment. The luminal pH of anterior and posterior adult midgut is 7–9. These findings in the larval and adult midgut open up the possibility of determining the role of plasma membrane transporters and channels involved in driving not only H+ fluxes but also secondary fluxes of other solutes and water in Drosophila.
The insect gut carries out some of the most vital functions of nutrition and solute and water balance of the organism. It is the first line of defence against ingested pathogens and also the portal of entry of viruses and parasites for which insect species are major vectors. The gut epithelium, with its luminal contents, has a fascinating dynamic structure and nutrient circulation pattern, the importance of which is not completely understood. The diet of insects is of such wide variety in terms of texture, composition,fluidity and mechanical properties, from liquid plant sap to solid bark, from whole blood to decomposing insects, that the digestive system of each insect,in larval or adult form, male or female, appears specialized in overall structure and in biochemical machinery to handle its staple ingested material. The uptake of solutes and water by the gut is also vital for the maintenance of the composition, pH and osmolarity of the haemolymph within a permissible range. However, the integrated role of segments of the gut (midgut and hindgut) and the Malpighian tubules in overall fluid and electrolyte balance is far from clear in most insects. Whereas fluid secretion and excretion of normal nitrogenous metabolic end-products, toxins, pesticides, etc. by Malpighian tubules have been studied extensively(Beyenbach, 2001; Dow et al., 1997; Dow and Davies, 2001; Maddrell and O'Donnell, 1992; O'Donnell and Spring, 2000),the midgut and hindgut of most species have not been as amenable to study,possibly on account of their musculature.
Insect organs such as the midgut are appealing as the epithelium is single-layered and the number and type of cells are limited. This permits in vitro physiological studies with excellent control of key variables: (1) solution composition at the apical or basal surfaces with minimum diffusion barriers, (2) hydrostatic pressure, particularly across the intercellular septa or tight-junctions, and (3) electrical potential(Tripathi and Boulpaep, 1989). Furthermore, the cells should be accessible with conventional and ion-selective microelectrodes, which permit evaluation of the cellular handling of ingested primary electrolytes such as K+,Na+, H+, Cl–,HCO3–, and possible organic substrates co-transported with these ions, and also patch-clamp and fluorescence studies of the cytosol. Combined with other studies such as imaging of transparent gut structures with pH-indicator dyes, localization of membrane transport proteins with fluorescent-labelled antibodies, and scanning of extracellular fluid adjacent to individual cells for ion gradients (the SIET technique), it is possible to get valuable information on the functional organization of the gut, its developmental changes and its ultrastructural design(Boudko et al., 2001a; Onken et al., 2006).
The midgut of several insect species with a wide variety of diet has been studied at a cellular level (Anderson and Harvey, 1966; Cioffi,1979; Clements,1992; Dow, 1984; Dow, 1986; Dow and Peacock, 1989; Lehane and Billingsley, 1996; Moffett and Cummings, 1994; Okech et al., 2008; Onken et al., 2008; Patrick et al., 2006; Smith et al., 2007; Smith et al., 2008; Volkmann and Peters, 1989a; Volkmann and Peters, 1989b; Zhuang et al., 1999). These studies have revealed large longitudinal gradients of pH along the length of the gut. However, the cellular basis of generation of these gradients and their biological significance is not completely understood, with several competing models proposed. It is therefore of interest to know if the midgut of Drosophila, given its potential in genomics, provides an opportunity for testing these hypotheses.
Despite the wealth of data that has made Drosophila an invaluable model system for studying development of structure and function of organ systems, there have been few studies on the structure(Dimitriadis, 1991; Dimitriadis and Kastritsis,1984; Filshie et al.,1971; Gartner,1985) and function (Romero et al., 2000; Sciortino et al.,2001) of the gut. The recent resurgence of interest in the structure and function of Drosophila midgut has coincidentally occurred on two fronts. Microperfusion of the midgut and stereological analysis of ultrastructure under controlled conditions(Shanbhag and Tripathi, 2005)has enabled description of the transport properties of the epithelium and analysis of the cellular basis of generation of axial H+ gradients in different segments. The identification of stem cells in adult midgut and hindgut segments (Lin et al.,2008; Micchelli and Perrimon,2006; Ohlstein and Spradling,2006; Takashima et al.,2008) has shown that Drosophila midgut epithelial development and maintenance is analogous to mammalian intestinal crypt development. In this review, we focus on the structural specialization of the epithelium of various regions of the midgut and the mechanisms of acid–base transport by this epithelium in larval and adult stages.
Structure of midgut regions
Insight into the structural organization and regional specialization of the Drosophila midgut was first obtained by Filshie et al.(Filshie et al., 1971). Gartner (Gartner, 1985)attempted the first stereological analysis of the anterior part of the adult epithelium and observed the presence of regenerative cells and the primary epithelial cells of the anterior midgut only, concluding that these were the only two cell types lining the epithelium. Dimitriadis and Kastritsis(Dimitriadis and Kastritsis,1984; Dimitriadis,1991), however, analysed the entire midgut in detail and identified at least three specialized regions of the midgut and made the first attempts to correlate the structure of the individual segments and cell types with function. The identification of the middle segment as an acid-secreting part has been verified and has invited immediate comparison with mammalian epithelia, particularly the gastric epithelium containing parietal cells(Baumann, 2001; Dubreuil, 2004; Dubreuil et al., 1998; Dubreuil et al., 2001; Yao and Forte, 2003).
For any epithelium, a controlled study requires access to both sides of the epithelium, and for a tubular epithelium requires perfusion of the lumen. This has been achieved for the larval Drosophila midgut(Shanbhag and Tripathi, 2005),allowing stereological analysis of epithelial geometry simultaneously with physiological studies. Fig. 1shows the basic organization of the midgut into at least three separate parts:an anterior near-neutral part, a strongly acidic middle segment and a third zone of increasing alkalinity. The posterior midgut epithelium has been best studied, but in principle all segments are amenable to perfusion studies. The epithelium is tight because of long septa between cells(Fig. 2A). Notable among its properties is the amplification of the apical and basal membranes(Fig. 3A,B,C, Table 1). The value of ultrastructure is further shown in its sensitivity in detecting changes produced by physiological manoeuvres such as Na+ replacement, which can produce massive and irreversible loss of plasma membrane(Fig. 3D–I), particularly when done on the basal side. The apical and basal membranes have portasome-like particles (Fig. 2D,E, inset) indicative of H+ V-ATPase location on both membranes.
The adult midgut is segmented into anterior, middle and posterior regions(Fig. 4). The basic membrane areas and volumes of the anterior and posterior midgut of the adult epithelium(Table 2) reveal that their transport role is modest and correlate with smaller pH gradients. The anterior segment is probably absorptive and confirms the basic observations of Gartner(Gartner, 1985) pertaining to the anterior midgut only but also reveals that there is a degree of variation of the anterior gut along its length (Fig. 4B,C, Fig. 5, Fig. 6). Posterior segment cells (Fig. 4E, Fig. 7) have an extensively dilated basal extracellular labyrinth, with a volume larger than that of anterior segment cells, indicating more fluid reabsorption in the posterior segment (Tripathi and Boulpaep,1989). Much less is known of the role of this epithelium in adults, which requires further study in both ultrastructure and physiology.
The middle adult acidic segment (Fig. 4D; Figs 8, 9, 10) has attracted a good deal of attention for its striking architecture, with alternate cells `facing'opposite directions. It would be appropriate to refer to them as secretory and absorptive cells on account of the great asymmetry of their apical and basal membranes (Fig. 8D–H). Cells of the middle segment are alternately absorptive (apically amplified≈9-fold, basally >90-fold) and secretory (apically amplified by>90-fold and basally ≈10-fold). The terminology suggested is based on the direction of net transport predicted on areal considerations alone. Table 3 shows that the apical and basal membranes of these two cell types are amplified more than 100-fold in either direction; their back-to-back geometry predicts significant recycling of solutes and water and provides a structural basis for bidirectional transport. The function of the secretory cells, composition of the secreted contents, regulation of secretion, and membrane turnover are still not clear, despite many studies(Lehane and Billingsley, 1996; Yao and Forte, 2003). The apical membranes form a cavity that can be seen spontaneously in either `open'or `closed' configuration (Figs 9 and 10) discharging membrane-bound granular material.
Bidirectional transport in individual regions
The identification of morphologically distinct segments of the midgut implies that controlled study of each segment is necessary before one can integrate the information for the entire midgut. Larvae, being voracious feeders, have a highly active midgut epithelium, as evidenced by the generation of much steeper pH gradients compared with adults(Shanbhag and Tripathi, 2005; Shanbhag and Tripathi, 2008)(Figs 11, 12). Similar gradients have been seen in a wide variety of insect midgut epithelia(Dadd, 1975; Boudko et al., 2001b; Corena et al., 2002; Dow, 1986; Moffett and Cummings, 1994; Onken et al., 2008; Zhuang et al., 1999). Regardless of the direction of secretion of acid or base, the energetics of transport in insect epithelia is believed to be primarily driven by vacuolar or H+ V-ATPases in the Malpighian tubule and also the midgut(Beyenbach, 2001; Dow and Davies, 2001). The fluxes of other electrolytes have been proposed to be driven as secondary transport processes. Furthermore, it is also possible that many of the membrane transporters in insects could be isoforms of mammalian transporters that do not bind inhibitors or agonists as they do in mammals. In the posterior larval midgut of Drosophila, a variety of inhibitors had no effect on transport that was still sensitive to bafilomycin-A1(Shanbhag and Tripathi,2005).
The other general feature of acid and base transport in the midgut epithelia of many insects is that there is a carbonic anhydrase (CA)-catalysed pool of H+ from which the H+ V-ATPases pump(Corena et al., 2002; Corena et al., 2005; Ridgway and Moffett, 1986; Seron et al., 2004). The location of this pool is a subject of active study as many enzymes are glycosylphosphatidylinositol-anchored and it is possible to localize these pools to either intracellular or extracellular or both compartments. Strong CA activity was localized at the apical membranes of goblet cells in the anterior and middle midgut region of Manduca sexta that is associated with the lumen alkalinization, but no CA activity was found in the posterior midgut goblet cells. Lepidopteran midgut is also divided into regions that show a substantial degree of structural and functional differentiation(Ridgway and Moffett,1986).
The localization of CA in the Drosophila midgut has not yet been achieved, but there are several interesting approaches that can be tried. Corena et al. have localized CA in the mosquito midgut(Corena et al., 2004), and a similar approach with membrane permeant and impermeant inhibitor would be a valuable approach. The removal of the peritrophic membrane by microperfusion improves access of the luminal perfusate to the apical membrane. It is possible that the in vivo conditions of the peritrophic space may be altered in terms of enzymes like CA that are located there(Smith et al., 2007), but there is good reason to assume that the intracellular enzyme-catalysed H+ pool is intact in perfused midguts, as shown below. As the ectoperitrophic space is a likely candidate for such a location, it would be important to verify this localization with antibodies to CA9 in both perfused and unperfused preparations where the peritrophic membrane is intact(Fig. 2C). The dissipation of the acid gradient by inhibition of CA with acetazolamide(Shanbhag and Tripathi, 2008)is easily detected on pH paper (Fig. 11D,E). However, the effects of inhibition of CA involved in base secretion in the posterior midgut are not easily detected by this simple method. One then has to rely on a more sensitive method (e.g. ion-selective microelectrodes) to detect acid or base fluxes. Such an approach is shown in Fig. 13.
In Fig. 13A, which shows measurement of intracellular pH, along with other membrane parameters, in the perfused larval midgut, acetazolamide alkalinizes the cell when applied to the bath; these effects are also seen from the lumen. This is clear evidence that extrusion of acid and base is rate limited in the H+ pool. There has been some uncertainty about the localization of the ATPase, particularly in view of data on the Malpighian tubule and the mosquito midgut(Beyenbach et al., 2000; O'Donnell et al., 1996; Wieczorek et al., 1999; Wieczorek et al., 2003). One line of evidence for a basal location of the ATPase in the Drosophilaposterior midgut (Shanbhag and Tripathi,2005), where the lumen is strongly alkaline, is the presence of portasome-like structures in the BEL (Fig. 2E, inset). The intracellular pH being more alkaline than the BEL or unstirred layer of the bath (Fig. 13A) provides even stronger evidence for the predominantly basal location of the H+ V-ATPase. Thirdly, sensitivity to bafilomycin is also greater in this preparation from the bath side, with hardly any effect detected from the lumen. Thus, one can measure acid extrusion rates from the basal side (Fig. 13B), and see its reversible inhibition by acetazolamide, even without the peritrophic membrane (from the lumen or bath) and bafilomycin (from the bath only).
Perfusion of each midgut region is important to know the overall driving forces for ions in each segment. Representative traces for three adult segments are shown in Fig. 14along with basic membrane potential data under free-flow conditions in control Ringer and taking into account the fact that the ends of the segments are damaged but electrically isolating the lumen from the bath. The anterior region has a low transepithelial potential; the middle and posterior segments have a transepithelial potential whose polarity favours net H+flux, which can occur passively for the observed gradients. Therefore, primary and secondary active transport processes have to be independently tested in these segments along with a complete characterization of all electrochemical driving forces and passive properties of individual cell membranes and the paracellular pathway. Whereas the larval gut has a predominant H+V-ATPase as primary transporter, the situation in the adult is far from clear.
Microperfusion and electrophysiological approaches have given us new tools to investigate membrane transport processes in an important genomic organism with many characterized mutations, e.g. labial(Dubreuil et al., 1998), where individual cuprophilic cells may be absent. For a complete understanding of the role of pH gradients in the Drosophila gut, experiments need to be designed to supplement the approach shown here with newer techniques. For instance, it is still not clear how base is secreted across the apical membranes in the posterior midgut in either larva or adult. It would be of interest to examine whether transporters like NDAE1(Romero et al., 2000; Sciortino et al., 2001) are involved in the exit of base in the posterior midgut. Likewise, the transport of other electrolytes and osmolytes needs clarification, as it is very likely to be linked to the transport of acid and base in this epithelium. The approach shown here can contribute to this end.
We thank our colleagues T. V. Abraham and J. N. Parmar for their unfailing support and Professor L. C. Padhy for valuable discussions. Supported by Interdisciplinary Programme11-R&D-TFR-5.02-1106.