Introduction
The growth of knowledge in the biochemistry of insect digestion had a bright start during the first decades of the last century, but slowed down after the development of synthetic chemical insecticides in the 1940s. Later on, with the environmental problems caused by chemical insecticides, new approaches for insect control were investigated. Midgut studies were particularly stimulated after the realization that the gut is a very large and relatively unprotected interface between the insect and its environment. Hence, an understanding of gut function was thought to be essential when developing methods of control that act through the gut, such as the use of transgenic plants to control phytophagous insects.
Applebaum (1985), in his review on the biochemistry of digestion, described the beginning of the renewed growth of the field. He discussed contemporary research showing that most insect digestive enzymes are similar to their mammalian counterparts, but that insect exotic diets require specific enzymes. In the next decade it became apparent that even enzymes similar to those of mammals have distinct characteristics, because each insect taxon deals with food in a special way (Terra and Ferreira, 1994). Since then, the field of digestive physiology and biochemistry has progressed dramatically at the molecular level (Terra and Ferreira, 2005).
The aim of this topic is to review the recent and spectacular progress in the study of insect digestive biochemistry. To provide a broad coverage while keeping the topic within reasonable size limits, only a brief account with key references is given for work done prior to 2000. Papers after 2000 have been selected from those richer in molecular details, and, when they were too numerous, representative papers were chosen, especially when abundant in references to other papers. Throughout, the focus is on providing a coherent picture of phenomena and highlighting further research areas. Amino acid residues are denoted by the one-letter code, if in peptides, for the sake of brevity. When mentioned in text with a position number, amino acid residues are denoted by the three-letter code to avoid ambiguity. For consistency, traditional abbreviations, like BAPA for benzoyl-arginine p-nitroan-ilide, have been changed, in the example to B-R-pNA, because the one-letter code for arginine is R.
The topic is organized into four parts. The first part (sections 11.2 and 11.3) tries to establish uniform parameters for studying insect digestive enzymes, providing an overview of the biochemistry of insect digestion, and discusses factors affecting digestive enzymes in vivo. The second part (sections 11.4—11.7) reviews digestive enzymes and microvillar proteins, with the emphasis on molecular aspects, whereas the third part (sections 11.8 and 11.9) describes the details of the digestive biochemical process alongside insect evolution. Finally, the fourth part (section 11.10) discusses data on digestive enzyme secretion mechanisms.
Overview of the Digestive Process
Initial Considerations
Digestion is the process by which food molecules are broken down into smaller molecules that are absorbed by cells in the gut tissue. This process is controlled by digestive enzymes, and is dependent on their localization in the insect gut.
Characterization of Digestive Enzymes
Enzyme kinetic parameters are meaningless unless assays are performed in conditions in which enzymes are stable. If researchers adopt uniform parameters and methods, comparisons among similar and different insect species will be more meaningful. A rectilinear plot of product formation (or substrate disappearance) versus time will ensure that enzymes are stable in a given condition. Activities (velocities) calculated from this plot are reliable parameters. According to the International Union of Biochemistry and Molecular Biology, the assay temperature should be 30°C, except when the enzyme is unstable at this temperature or altered for specific purposes. Owing to partial inactivation, the optimum temperature is not a true property of enzymes, and therefore should not be included in the characterization. Optimum enzyme pH should be determined using different buffers to discount the effects of chemical constituents of the buffers and their ionic strength on enzyme activity. The number of molecular forms of a given enzyme should be evaluated by submitting the enzyme preparation to a separation process (gel permeation, ion-exchange chromatography, hydrophobic chromatography, electrophoresis, gradient ultracentrifugation, etc.), followed by assays of the resulting fractions. Substrate specificity of each molecular form of a given enzyme should be evaluated and substrate preference quantified by determining Vm/Km ratios for each substrate, keeping the amount of each enzyme form constant. Substrate preference expressed as the percentage activity towards a given substrate in relation to the activity upon a reference substrate may be misleading because, in this condition, enzyme activities are determined at different substrate saturations. The isoelectric points of many enzymes can be determined after staining with specific substrates following the separation of the native enzymes on isoelectrofocusing gels.
If enzyme characterization is performed as part of a digestive physiology study, emphasis should be given to enzyme compartmentalization, substrate specificity, and substrate preference, in order to discover the sequential action of enzymes during the digestive process.
Knowledge of the effect of pH on enzyme activity is useful in evaluating enzyme action in gut compartments (Figure 1) with different pH values. Finally, the determination of the molecular masses of digestive enzymes, associated with the ability of enzymes to pass through the peri-trophic membrane, allows estimation of the pore sizes of the peritrophic membrane. Molecular masses determined in non-denaturing conditions are preferred, since in these conditions the enzymes should maintain their in vivo aggregation states (not only their quaternary structures if present). The method of choice in this case is gradient ultracentrifugation.
Complete enzymological characterization requires purification to homogeneity, and sequencing. Furthermore,details of the catalytic mechanisms, including involvement of amino acid residues in catalysis and substrate specificity, should be determined. This permits the classification of insect digestive enzymes into catalytic families, and discloses the structural basis of substrate specificities; it will also enable us to establish evolutionary relationships with enzymes from other organisms.
Cloning cDNA sequences encoding digestive enzymes enables the expression of large amounts of recombinant enzymes that may be crystallized or used for the production of antibodies. Antibodies are used in Western blots to identify a specific enzyme in protein mixtures, or to localize the enzyme in tissue sections in a light or electron microscope. Enzyme crystals used for resolving three-dimensional (3D) structures (via X-ray diffraction or nuclear magnetic resonance, NMR) need amounts of purified enzymes that frequently are difficult to isolate from insects by conventional separation procedures.
Figure 1 Diagrammatic representation of insect gut compartments. Glycocalyx: the carbohydrate moiety of intrinsic proteins and glycolipids occurring in the luminal face of microvillar membranes.
However, detailed 3D structures are necessary to understand enzyme mechanisms and the binding of inhibitors to enzyme molecules. Alternatively, cDNA may be amenable to site-directed mutagenesis for structure—function studies. Site-directed mutagenesis tests the role of individual amino acid residues in enzyme function or structure. Such knowledge is a prerequisite in developing new biotechno-logical approaches to control insects via the gut. Finally, interference RNA may be used to suppress the expression of one enzyme, in order to test hypotheses regarding its physiological role.
Classification of Digestive Enzymes
Digestive enzymes are hydrolases. The enzyme classification and numbering system used here is that recommended by the Nomenclature Committee of the International Union of Biochemistry and Molecular Biology (Enzyme Commission).
Peptidases (peptide hydrolases, EC 3.4) are enzymes that act on peptide bonds and include the proteinases (endopeptidases, EC 3.4.21—24) and the exopeptidases.Proteinases are divided into subclasses on the basis of catalytic mechanism, as shown with specific reagents or effect of pH. Specificity is only used to identify individual enzymes within subclasses. Serine proteinases (EC 3.4.21) have a serine and a histidine in the active site. Cysteine proteinases (EC 3.4.22) possess a cysteine in the active site, and are inhibited by mercurial compounds. Aspar-tic proteinases (EC 3.4.23) have a pH optimum below 5, due to the involvement of a carboxyl residue in catalysis. Metalloproteinases (EC 2.3.24) need a metal ion in the catalytic process. Exopeptidases include enzymes that hydrolyze single amino acids from the N-terminus (ami-nopeptidases, EC 3.4.11) or from the C-terminus (car-boxypeptidases, EC 3.4.16-18) of the peptide chain, and those enzymes specific for dipeptides (dipeptide hydrolases, EC 3.4.13) (Figure 2).
Glycosidases (EC 3.2) are classified according to their substrate specificities. They include the enzymes that cleave internal bonds in polysaccharides and are usually named from their substrates – for example, amylase, cel-lulase, pectinase, and chitinase. They also include the enzymes that hydrolyze oligosaccharides and disaccha-rides. Oligosaccharidases and disaccharidases are usually named based on the monosaccharide that gives its reducing group to the glycosidic bond, and on the configuration (a or P) of this bond (Figure 2).
Lipids are a large and heterogeneous group of substances that are relatively insoluble in water but readily soluble in apolar solvents. Some contain fatty acids (fats, phospholipids, glycolipids, and waxes), while others lack them (terpenes, steroids, and carotenoids). Ester bonds are hydrolyzed in lipids containing fatty acids before they are absorbed. The enzymes that hydrolyze ester bonds comprise: (1) carboxylic ester hydrolases (EC 3.1.1), such as lipases, esterases, and phospholipases A and B; (2) phosphoric monoester hydrolases (EC 3.1.3), which are the phosphatases; and (3) phosphoric diester hydrolases (EC 3.1.4), including phospholipases C and D (Figure 2).
Phases of Digestion and Their Compartmentalization in the Insect Gut
Most food molecules to be digested are polymers, such as proteins and starch, and are digested sequentially in three phases. Primary digestion is the dispersion and reduction in molecular size of the polymers, and results in oligomers. During intermediate digestion, these undergo a further reduction in molecular size to dimers; in final digestion, they become monomers. Digestion usually occurs under the action of digestive enzymes from the midgut, with little or no participation of salivary enzymes.
Any description of the spatial organization of digestion in an insect must relate the midgut compartments (cell, ecto-, and endoperitrophic spaces) to each phase of digestion, and hence to the corresponding enzymes. To accomplish this, enzyme determinations must be performed in each midgut luminal compartment and in the corresponding tissue. Techniques of sampling enzymes from midgut luminal compartments and enzymes trapped in cell glycocalyx have been reviewed elsewhere (Terra and Ferreira, 1994). Microvillar enzymes are discussed in detail in section 11.7.
Frequently, initial digestion occurs inside the peri-trophic membrane (see sections 11.8.1 and 11.8.2), intermediate digestion in the ectoperitrophic space, and final digestion at the surface of midgut cells by integral micro-villar enzymes or by enzymes trapped into the glycocalyx (Figure 1). Exceptions to this rule, and the procedures for studying the organization of the digestive process, will be detailed below.
Role of Microorganisms in Digestion
Most insects harbor a substantial microbiota, including bacteria, yeast, and protozoa. Microorganisms might be symbiotic or fortuitous contaminants from the external environment. They are found in the lumen, adhering to the peritrophic membrane, attached to the midgut surface, or within cells. Intracellular bacteria are usually found in special cells, mycetocytes, which may be organized in groups, mycetomes. Microorganisms produce and secrete their own hydrolases, and cell death will result in the release of enzymes into the intestinal milieu. Any consideration of the spectrum of hydrolase activity in the midgut must include the possibility that some of the activity may derive from microorganisms.
Figure 2 Digestion of important nutrient classes. Arrows point to bonds cleaved by enzymes. (A) Protein digestion; R, different amino acid moieties; (B) starch digestion; (C) p-linked glucoside; (D) lipid digestion; PL, phospholipase; R, fatty acyl moieties.
Despite the fact that digestive enzymes of some insects are thought to be derived from the microbiota, there are relatively few studies that show an unambiguous contribution of microbial hydrolases. Best examples are found among wood- and humus-feeding insects like termites, tipulid fly larvae, and scarabid beetle larvae. Although these insects may have their own cellulases (see section 11.4.3), only fungi and certain filamentous bacteria developed a strategy for the chemical breakdown of lignin. Lignin is a phenolic polymer that forms an amorphous resin in which the polysac-charides of the secondary plant cell wall are embedded, thus becoming protected from enzymatic attack (Terra et al., 1996; Brune, 1998; Dillon and Dillon, 2004).
Microorganisms play a limited role in digestion, but they may enable phytophagous insects to overcome biochemical barriers to herbivory — for example, detoxifying flavonoid alkaloids and the phenolic aglycones of plant glycosides. They may also provide complex-B vitamins for blood-feeders and essential amino acids for phloem feeders, produce pheromone components, or withstand the colonization of the gut by non-indigenous species (including pathogens) (Dillon and Dillon, 2004; Genta et al., 2006a).
Midgut Conditions Affecting Enzyme Activity
The pH of the contents of the midgut is one of the important internal environmental properties that affect digestive enzymes. Although midgut pH is hypothesized to result from adaptation of an ancestral insect to a particular diet, its descendants may diverge, feeding on different diets, while still retaining the ancestral midgut pH condition. Thus, there is not necessarily a correlation between midgut pH and diet. In fact, midgut pH correlates well with insect phylogeny (Terra and Ferreira, 1994; Clark, 1999).
The pH of insect midgut contents is usually in the 6-7.5 range. Major exceptions are the very alkaline midgut contents (pH 9-12) of Lepidoptera, scarab beetles, and nematoceran Diptera larvae; the very acid (pH 3.1-3.4) middle region of the midgut of cyclorrhaphous Diptera; and the acid posterior region of the midgut of heterop-teran Hemiptera (Terra and Ferreira, 1994; Clark, 1999). pH values may not be equally buffered along the midgut. Thus, midgut contents are acidic in the anterior midgut and nearly neutral or alkaline in the posterior midgut in Dictyoptera, Orthoptera, and most families of Coleop-tera. Cyclorrhaphan Diptera midguts have nearly neutral contents in the anterior and posterior regions, whereas in middle midgut the contents are very acid (Terra and Ferreira, 1994).
A pH in the ectoperitrophic space lower than in the midgut lumen was reported in some lepidopteran larvae. This is an artefact caused by a halt in alkaline secretion by the isolated midgut tissue (Grigorten et al., 1993). Nevertheless, the pH in the immediate neighborhood of the negatively-charged microvillar glycocalyx is expected to be lower than in the bulk solution, because of proton retention (Quina et al., 1980).
The high alkanity of lepidopteran midgut contents is thought to allow these insects to feed on plant material rich in tannins, which bind to proteins at lower pH, reducing the efficiency of digestion (Berenbaum, 1980). This explanation may also hold for scarab beetles and for detritus-feeding nematoceran Diptera larvae that usually feed on refractory materials such as humus. Nevertheless, mechanisms other than high gut pH must account for the resistance to tannin displayed by some locusts (Bernays et al., 1981) and beetles (Fox and Macauley, 1977). One possibility is the effect of surfactants such as lysolecithin, which is formed in insect fluids due to the action of phospholipase A on cell membranes (Figure 2), and which occurs widely in insect digestive fluids (De Veau and Schultz, 1992). Surfactants are known to prevent the precipitation of proteins by tannins even at pH as low as 6.5 (Martin and Martin, 1984). Present knowledge is not sufficient to relate midgut detergency to diet or phylogeny, or to both.
Tannins may have deleterious effects other than precipitating proteins. Tannic acid is frequently oxidized in the midgut lumen, generating peroxides, including hydrogen peroxide, which readily diffuses across cell membranes and is a powerful cytotoxin. In some insects (e.g., Orgyia leucostigma), tannic acid oxidation and the generation of peroxides are suppressed by the presence of high concentrations of ascorbate and glutathione in the midgut lumen (Barbehenn et al., 2003). Dihydroxy phenolics in an alkaline medium are converted to quinones that react with lysine E-amino groups. This leads to protein aggregation and a decrease in lysine availability for the insect. Other compounds (e.g., oleuropein, alkylate lysine residues in proteins) cause the same problems as dihydroxy pheno-lics. These phenomena are inhibited in larvae of several lepidopteran species by secreting glycine into the midgut lumen. Glycine competes with lysine residues in the dena-turating reaction (Konno et al., 2001). In some insects, tannins reduce the overall efficiency of conversion of ingested matter to body mass by an unknown mechanism. Nevertheless, the performance of these insects remains unchanged, because of compensatory feeding (Barbehenn et al., 2009).
A high midgut pH may also be of importance, in addition to its role in preventing tannin binding to proteins, in freeing hemicelluloses from plant cell walls ingested by insects. Hemicelluloses are usually extracted in alkaline solutions for analytical purposes (Blake et al., 1971), and insects such as the caterpillar Erinnyis ello are able to digest hemicelluloses efficiently without affecting the cellulose from the leaves they ingest to any degree (Terra, 1988). This explanation is better than the previous one in accounting for the very high pH observed in several insects, since a pH of about 8 is sufficient to prevent tannin binding to proteins (Terra, 1988).
The acid region in the cyclorrhaphous Diptera mid-gut is assumed to be involved in the process of killing and digesting bacteria, which may be an important food for maggots. This region is retained in Muscidae that have not diverged from the putative ancestral bacteria-feeding habit, as well as in the flesh-feeding Calli-phoridae and the fruit-feeding Tephritidae (Terra and Ferreira, 1994). The acid posterior midgut of Hemiptera may be related to their lysosome-like digestive enzymes (cysteine and aspartic proteinase) (see sections 11.5.3 and 11.5.4).
Few papers have dealt with midgut pH buffering mechanisms. Dow (1992) described a carbonate secretion system, which may be responsible for the high pH found in Lepidoptera midguts (Figure 3). Phosphorus NMR microscopy has been used to show that valinomycin leads to a loss of alkalinization in the midgut of Spodoptera litura (Skibbe et al., 1996). As valinomycin is known to transport K+ down its concentration gradient, this result gives further support to the model described in Figure 3. Midgut alkalinization in nematoceran Diptera occurs by mechanisms similar to those of lepidopteran larvae (Okech et al., 2008), whereas no data are available for scarab beetles. Terra and Regel (1995) determined pH values and concentrations of ammonia, chloride, and phosphate in the presence or absence of ouabain and vanadate in Musca domestica midguts. From the results, they proposed that middle midgut acidification is accomplished by a proton pump of mammalian-like oxyntic cells, whereas the neutralization of posterior midgut contents depends on ammonia secretion (Figure 4).
Figure 3 A model for generation of high gut pH by the goblet cells of lepidopteran larvae. Carbonic anhydrase (CA) produces carbonic acid that dissociates into bicarbonate and a proton. The proton is pumped by a V-ATPase into the goblet cell cavity, from where it is removed in exchange with K+ that eventually diffuses into lumen. Bicarbonate is secreted in exchange with chloride and loses a proton due to the intense field near the membrane, forming carbonate and raising the gut pH. Data from Dow (1992).
Redox conditions in the midgut are regulated and may be the result of phylogeny, although data are scarce. Reducing conditions are observed in clothes moth, sphinx moths, owlet moths, and dermestid beetles (Appel and Martin, 1990), and in Hemiptera (Silva and Terra, 1994). Reducing conditions are important to open disulfide bonds in keratin ingested by some insects (clothes moths, dermestid beetles) (Appel and Martin, 1990), to maintain the activity of the major proteinase in Hemiptera (see section 11.5.3), and to reduce the impact of some plant allelochemicals, such as phenol, in some herbivores (Appel and Martin, 1990). In spite of this, the artificial lowering of in vivo redox potentials did not significantly impact digestive efficiency of the herbivore Helicoverpa zea, although the reducing agent used (dithiothreitol) inhibited some proteinases in vitro (Johnson and Felton, 2000). Midgut antioxidant enzymes in Spodoptera litto-ralis are upregulated in response to increased oxidative stress caused by oxidizable allelochemicals (Krishnan and Kodrik, 2006).
Although several allelochemicals other than phenols may be present in the insect gut lumen, including alkaloids, terpene aldehydes, saponins, and hydroxamic acids (Appel, 1994), data on their effect on digestion are lacking.
Figure 4 Diagrammatic representation of ion movements, proposed as being responsible for maintenance of pH in the larval midgut contents of Musca domestica. Carbonic anhydrase (CA) in cup-shaped oxyntic cells in the middle of the midgut (A) produces carbonic acid which dissociates into bicarbonate and a proton. Bicarbonate is transported into the hemolymph, whereas the proton is actively translocated into the midgut lumen acidifying its contents to pH 3.2. Chloride ions follow the movement of protons. NH3 diffuses from anterior and posterior midgut cells (B) into the midgut lumen, becoming protonated and neutralizing their contents to pH 6.1-6.8. NH4+ is then exchanged for Na+ by a microvillar Na+/K+-ATPase. Inside the cells, NH4+ forms NH3, which diffuses into midgut lumen, and proton that is transferred into the hemolymph.
