Insect N-Acetylglucosaminidases
Phylogenetic analysis of insect Af-acetylglu-cosaminidases Beta-N-acetylglucosaminidases (NAGs; EC 3.2.1.30) have been defined as enzymes that release -acetylglucosamine residues from the non-reducing end of chitooligosaccharides and from glycoproteins with terminal N-acetylglucosamines. Insect NAGs are members of the family-20 hexosaminidase super-family of the glycosylhydrolases of the Carbohydrate Active Enzymes database, CAZY (Coutinho and Henrissat, 1999; Cantarel., et al., 2009). These enzymes have been detected in the molting fluid, hemolymph, integument, and gut tissues of several species of insects (Kramer and Koga, 1986; Hogenkamp et al., 2008), and cooperate with CHTs to hydrolyze chitin to generate monomers of N-acetylglucosamine (Fukamizo and Kramer, 1985a, 1985b). Insect CHTs are unable to convert the chitin substrate completely to GlcNAc monomers. Therefore, NAG is the enzyme primarily responsible for the production of the monomer from chitooligosaccharides for recycling. Kinetic studies with M. sexta CHT (MsCHT5, group I CHT) have revealed that this enzyme is subject to substrate and/or product inhibition when chitooligosaccharides and/or colloidal chitin are utilized as substrates (Koga et al, 1982, 1983; Arakane et al, 2003). Therefore, one of the potential functions of NAGs may be to prevent the accumulation of chitooligosaccharides at concentrations that are high enough to interfere with efficient degradation of chitin by CHT (Kramer and Muthukrishnan, 2005).
cDNAs for epidermal -N-acetylglucosaminidases of B. mori, B. mandarina, T. ni, and M. sexta have been isolated and characterized (Nagamatsu et al., 1995; Zen et al., 1996; Goo et al., 1999; Hogenkamp et al., 2008). A NAG also has been detected in the gut of Ae. aegypti, where its activity increased dramatically upon blood feeding (Filho et al., 2002). A search of the D. melanogaster, A. gambiae, Ae. aegypti, Culex pipiens, A. mellifera, N. vitripennis, B. mori, and T. castaneum genome databases revealed the presence of multiple NAG genes, as well as the genes encoding ^-N-acetylhexosaminidases (HEXs) in these species (Hogenkamp et al., 2008). Phylogenetic analysis of NAGs from these insects indicates that NAGs can be classified into four distinct groups – NAG group I (NAG1), NAG group II (NAG2), N-glycan processing NAGs (FDL) (group III, Leonard et al., 2006), and HEX group IV – according to their amino acid sequences (Figure 8). To date, only a single gene representing each of the groups I, II, and III has been found in the various insect species, with the exception of C. pipiens, which appears to have three genes encoding NAG-like proteins closely related to group I NAGs. Group I is composed of the enzymatically well-characterized NAGs, including NAGs from M. sexta (MsNAG1) and B. mori (BmNAG1). DmHEXO2, which has been shown to have NAG activity (Mark et al., 2003; Leonard et al., 2006), was placed in group II. Group III is composed of the D. melanogas-ter fused lobes protein (DmFDL), along with the fused lobes (fdl) homologs of other insect species (Leonard et al. , 2006). All of the proteins belonging to this group possess a predicted transmembrane anchor and a signal anchor, except for a signal peptide that can be found in NAGs belonging to groups I, II, and IV. In T. castaneum, TcNAG3 could not be unambiguously assigned to any of the three subgroups. TcNAG3 is more closely related to TcFDL than to TcNAG1 and TcNAG2, but the TcFDL and TcNAG3 genes are present on different linkage groups (Figure 8, Hogenkamp et al., 2008).
Figure 8 Phylogenetic analysis of NAGs and hexoaminidases in Tribolium, other insects and metazoans. MEGA4.0 (Tamura et al., 2007) was used to construct the consensus phylogenetic tree using UPGMA. Bootstrap analyses of 1000 replications are shown. Protein sequences extracted from GenBank include: MsNAG, Manduca sexta (AY368703); BmNAG, Bombyx mori (genbank: AF326597); TnNAG, Trichoplusia ni (AY078172); AmNAG1, Apis mellifera (XM_624790); TcNAG1, Tribolium castaneum (EF592536); DmNAG1 (DmHEXO1), Drosophila melanogaster (NM_079200); AgNAG1, Anopheles gambiae (XP_315391); CqNAG1a, Culex quinquefasciatus (XP_001864406); AaNAG1, Aedes aegypti (EAT43909); CqNAG1b, Culex quinquefasciatus (XP_001864407); CqNAG1c, Culex quinquefasciatus (XP_001866097); TcNAG3, Tribolium castaneum (EF592538); TcFDL, Tribolium castaneum (EF592539); AmFDL, Apis mellifera (XP_394963); DmFDL, Drosophila melanogaster (NP_725178); AgFDL, Anopheles gambiae XP_308677); CqFDL, Culex quinquefasciatus (XP_001850423); AaFDL, Aedes aegypti (EAT36388); TcNAG2, Tribolium castaneum (EF592537); DmNAG2 (DmHEXO2), Drosophila melanogaster (NM_080342); AgNAG2, Anopheles gambiae (XM_307483); CqNAG2, Culex quinquefasciatus (XP_001842710); AaNAG2, Aedes aegypti (EAT40440); HsHEXA, Homo sapiens (NM_000520); HsHEXB, Homo sapiens (NM_000521); MsHEXA, Mus musculus (NM_010421); MsHEXB, Mus musculus (NM_010422); SfHEX1, Spodoptera frugiperda (DQ183187); SfHEX2, Spodoptera frugiperda (DQ249307); BmHEX, Bombyx mori (AY601817); TcHEX3, Tribolium castaneum (XM_970565); AmHEX, Apis mellifera (XM_001122538); TcHEX1, Tribolium castaneum (XM_970563); TcHEX2, Tribolium castaneum (XM_970567); CqHEX2, Culex quinquefasciatus (XP_001867058); AgHEX, Anopheles gambiae (XM_319210); CqFEX1, Culex quinquefasciatus (XP_001867057); and AaHEX, Aedes aegypti (EAT43655.
Expression and functional analysis of insect A-acetylglucosaminidases Hogenkamp and colleagues (2008) performed dsRNA-mediated post-transcriptional downregulation (RNAi) of transcripts for all four NAG genes from a single insect species (T. castaneum) to study the functions of insect NAGs. Injection of a dsRNA corresponding to any one TcNAG gene resulted in substantial downregulation of the target transcript without significantly affecting the levels of the other TcNAG transcripts. Depletion of transcripts for any one of the targeted genes produced lethal molting arrest phenotypes. However, some of the injected insects did succeed in completing each type of molt (larval-larval, larval-pupal, and pupal-adult). TcNAG1 appeared to be most critical in chitin catabolism during molting. Administration of dsRNA for TcNAGl resulted in developmental arrest, and more than 80% of the insects died at the time of the next molt (Figure 5). During each type of molt, larval-larval, larval-pupal, and pupal-adult, the insects were unable to completely shed their exoskeleton. The pupa-adult molting phenotype produced by injection of dsRNA for TcNAGl is strikingly similar to that obtained in RNAi studies with dsTcCHT5 (Figure 5; see section 7.4.1.3). Insects injected with dsRNA for TcCHT5 also failed to shed their old cuticle, and the new cuticle was visible underneath the old cuticle (Zhu et al., 2008c; Arakane and Muthukrishnan, 2010). It has been shown that in M. sexta, CHT is susceptible to oligosaccharide inhibition (Koga et al., 1982, 1983; Arakane et al., 2003). Injection of dsRNA for TcNAGl may result in the accumulation of chitiooligosaccharides in the molting fluid, and therefore it may cause inhibition of TcCHT5 activity, resulting in a phenotype similar to that observed in dsRNA for TcCHT5-treated insects. The high level of expression of TcNAGl, its phylogenetic relationship to other well-characterized molting-associated insect NAGs (Figure 8), and the phenotypic effect of knocking down TcNAGl transcripts suggest that, among all of the TcNAGs, TcNAG1 (group I NAG) is the enzyme primarily responsible for the efficient degradation of cuticular chitin, in concert with TcCHT5 (group I CHT), in T. castaneum, and that this may be the case in other insect species as well.
Although TcNAG1 is most likely to be the principal NAG for catabolism of cuticle-associated chitin, the other three NAGs identified in T. castaneum also appear to play important and perhaps indispensable roles in cuticle turnover and development. Injection of dsRNA for TcNAG2 (encoding a group II NAG orthologous to DmHEXO2) prevents all types of molts, especially the pupal-adult molt. Like the phenotype produced by injection of dsRNA for TcNAGl (Figure 5), more than 75% of the animals treated with dsRNA for TcNAG2 were unable to fully shed the old pupal cuticle. Since injection of dsRNA for TcNAG2 did not change the level of TcNAGl transcripts, TcNAG1 could not compensate for the lack of TcNAG2 in adult eclosion in T. castaneum. In addition, TcNAG2 transcript level in the midgut is relatively higher than that in the carcass (whole body minus midgut), suggesting TcNAG2 as well as TcNAGl, which are highly expressed in both tissues, also play critical roles in the PM-associated chitin turnover.
Group III consists of the insect orthologs of the D. melanogaster fused lobes gene, DmFDL. The FDL proteins are predicted to be membrane-bound, with a single transmembrane helix located near the N-terminus. Furthermore, ultracentrifugation experiments on a lepi-dopteran protein from the culture media of Sf9 and Sf21 cells indicated that a major portion of the NAG activity resided in the membrane fraction (Altmann et al., 1995; Tomiya et al., 2006). This lepidopteran NAG was capable of effectively hydrolyzing chitotriose-PA (pyridylamino), while the recombinant DmFDL was unable to digest chitotriose (Leonard et al., 2006). The latter hydrolyzed only the GlcNAc residue attached to the a-1,3-linked mannose of the core pentasaccharide of N-glycans. No cleavage activity of any other GlcNAc residues was observed, including the GlcNAc residue attached to the a-1,6-linked mannose of the core pentasaccharide. Furthermore, DmFDL did not catalyze the endo-type hydrolysis of the N,N’-diacetylchitobiosyl unit in the high-mannose pentasaccharide core. A similar N-glycan substrate specificity for the terminal GlcNAc attached to the a-1,3-linked mannose was observed in membrane-bound P-N-acetylhexosaminidases from several lepidopteran insect cell lines, including Sf21, Bm-N, and Mb-0503 (Altmann et al., 1995; Tomiya et al, 2006). Taken together, FDLs may play a critical role in N-glycan processing.
Unlike RNAi for TcNAGl (group I NAG), injection of dsRNA for TcFDL exhibits a small percentage (10-20%) of lethal molting defect phenotypes at the larval-larval and larval-pupal molts (Hogenkamp et al., 2008). Much higher mortality (80%), however, was observed at the pupal-adult molting stage, indicating that TcFDL plays an essential role for adult eclosion. The transcript level of TcFDL in the midgut was relatively low compared to that of the carcass. Therefore, the observed lethal phenotype at the pharate adult stage may be a direct result of the knockdown of this transcript in the cuticular epidermal cells, rather than in the gut lining cells. If TcFDL does in fact play a role in chitin turnover in the cuticle, then this protein may be secreted and not membrane-bound. Indeed, Leonard and colleagues (2006) have observed that DmFDL is, to a large extent, secreted into the extracellular space. Whether there is another point of regulation at the level of release of membrane-bound FDLs is an interesting possibility.
Another T. castaneum NAG, TcNAG3, has not been unambiguously assigned to any of the three NAG groups (Figure 8). Similar to TcNAG2 (group II NAG), TcNAG3 is also expressed at a significantly higher level in the larval midgut than in the carcass (Hogenkamp et al., 2008). Furthermore, an analysis of the developmental pattern of expression of TcNAG3 indicated that it is primarily expressed during the larval stages. Unlike RNAi for the other three TcNAGs, injection of dsTcNAG3 did not consistently result in lethal phenotypes, and the majority of dsRNA-injected insects survived to adults with no visible phenotypic changes. However, a small number of individuals (approximately 20%) did exhibit a lethal larval phenotype similar to that of TcNAGl RNAi (Figure 5). In addition, a few insects (approximately 10%) exhibited a lethal pharate adult molting phenotype after dsRNA TcNAG3 injection. These insects were unable to fully shed their old pupal cuticle, similar to the phenotypes observed after dsRNA TcNAGl and dsRNA TcNAG2 injections. The TcNAG3 gene is expressed predominantly in the larval stages, with only trace levels of expression in the pupal and adult stages (Hogenkamp et al., 2008). In other insect species analyzed, only genes that can be classified into groups NAG1, NAG2, and FDL have been identified (Figure 8). Therefore, TcNAG3 appears to be unique, and its relatively high expression in the midgut compared to the carcass suggests that it may be specialized for the turnover of PM-associated chitin rather than cuticular chitin during larval stages.
Insect Chitin Deacetylases
Phylogenetic analysis and domain organization of chitin deacetylases The extracellular matrix (ECM) of the insect exoskeleton is modified in different ways to give the cuticle its proper physiological and mechanical properties – namely, rigidity and thickness, or flexibility and thinness (Kramer and Muthukrishnan, 2005). Chitin deacetylases (CDAs, EC 3.5.1.41) are secreted metalloproteins that belong to a family of extracellular chitin-modifying enzymes that catalyze the N-deacetylation of chitin to form chitosan, a polymer of ^-1,4-linked D-glucosamine residues with electrostatic properties very different from chitin. This modification might contribute to the affinity of chitosan for a variety of cuticular proteins distinct from those that bind specifically to chitin. CDAs have been well characterized in various fungi and bacteria (Caufrier et al., 2003), and belong to the carbohydrate esterase family 4 (CE4) of the CAZY database (www.cazy.org; Cantarel et al., 2009). CE4 esterases catalyze deacetylation of different carbohydrate substrates, such as chitin, acetylxylan, and bacterial peptidoglycan. Chitooligosaccharide deacetylases and NodB, a nodulation protein from Rhizobium, belong to this family, and possess a similar catalytic domain (John et al., 1993).
The first cDNA encoding an insect CDA-like protein (TnPM-P42, also referred to as TnCDA9) was characterized from the PM in the cabbage looper, Trichoplusia ni, only 5 years ago (Guo et al., 2005). Since then, several genes/cDNAs encoding insect CDAs have been identified from different species (Luschnig et al., 2006; Wang et al., 2006; Campbell et al., 2008; Dixit et al., 2008; Toprak et al., 2008; Jakubowska et al., 2010). A comparative analysis of CDA gene families in several insect species with fully sequenced genomes, including Diptera, Coleoptera, Hymenoptera, and Lepidoptera, revealed that the number of CDA genes varies with species. Based on amino acid sequence similarity, insect CDAs are classified into five groups, I to V (Figure 9; Dixit et al., 2008; Jakubowska et al., 2010).
Figure 9 A phylogenetic tree of putative CDAs from different insects. A consensus phylogenetic tree was constructed using neighbor-joining method in the software MEGA 4.0 (Tamura et al., 2007). Protein sequences obtained from GenBank as follows; NvCDAI, Nasonia vitripennis (XP_001604765); AmCDAI, Apis mellifera (XP_391915); TcCDAI, Tribolium castaneum (ABU2522); HaCDAI, Helicoverpa armigera (ADB43610); BmCDAI, Bombyx mori (BGIBMGA006213); DmCDAI, Drosophila melanogaster (NP_730444); AgCDAI, Anopheles gambiae (XP_320597); DmCDA2, Drosophila melanogaster (NP_001163469); AgCDA2, Anopheles gambiae (XP_320596); AmCDA2, Apis mellifera (XP_623723); TcCDA2, Tribolium castaneum (ABU25224); NvCDA2, Nasonia vitripennis (XP_001604838); BmCDA2, Bombyx mori (BGIBMGA006214); DmCDA3, Drosophila melanogaster (NP_609806); AgCDA3, Anopheles gambiae (XP_317336); BmCDA3, Bombyx mori (BGIBMGA008988); NvCDA3, Nasonia vitripennis (XP_001606617); AmCDA3, Apis mellifera (XP_001121246); TcCDA3, Tribolium castaneum (ABW74145); TcCDA4, Tribolium castaneum (ABW74146); AgCDA4, Anopheles gambiae (XP_310753); DmCDA4, Drosophila melanogaster (NP_728468); BmCDA4, Bombyx mori (BGIBMGA010573); AmCDA4, Apis mellifera (XP_001120478); NvCDA4, Nasonia vitripennis (XP_001607989); AmCDA5, Apis mellifera (XP_624655); NvCDA5, Nasonia vitripennis (XP_001603918); TcCDA5, Tribolium castaneum (ABW74147); BmCDA5, Bombyx mori (BGIBMGA002696); AgCDA5, Anopheles gambiae (XP_316929); DmCDA5, Drosophila melanogaster (NP_001097044); TcCDA6, Tribolium castaneum (ABW74149); TcCDA7, Tribolium castaneum (ABW74150); TcCDA8, Tribolium castaneum (ABW74151); TcCDA9, Tribolium castaneum (ABW74152); DmCDA9, Drosophila melanogaster (NP_611192); TnCDA9, Trichoplusia ni (AAY46199); BmCDA9-3, Bombyx mori (BGIBMGA013758); HaCDA5a, Helicoverpa armigera (ADB43611); HaCDA5b, Helicoverpa armigera (ADB43612); BmCDA9-1, Bombyx mori (BGIBMGA013756); BmCDA9-2, Bombyx mori (BGIBMGA013757).
Group I CDAs (CDA1s and CDA2s) consist of D. melanogaster Serpentine (DmSerp) and Vermiform (DmVerm) (referred to as DmCDAl and DmCDA2, respectively) and their orthologs (CDAs l and 2) from each species. All group I CDAs have a chitin-binding peritrophin-A domain (CBD), a low-density lipoprotein receptor class A domain (LDLa), and a CDA catalytic domain. There are two to four transcript variants produced by alternative splicing and/or exon skipping from the CDA2 pre-mRNAs (Dixit et al, 2008). Group II, III, and IV families are represented by only one CDA in each species, namely CDA3, CDA4, and CDA5, respectively. Although, like group I CDAs, CDA3s also possess a single copy of each of the three domains, the overall amino acid sequence identity is only about 38% with CDA1s and CDA2s (amino acid sequence identity between CDA1s and CDA2s is about 60%). Group III enzymes (CDA4s) have a single copy of the CBD and the CDA catalytic domain, but lack an LDLa domain. Group IV CDAs (CDA5s), like CDA4s, each possess a single CBD and a single CDA catalytic domain. These two domains, however, are connected by a long Ser/Thr/Pro/Gln-rich linker (e.g., about 2400 amino acids in AgCDA5), which results in CDA5s being the largest CDA proteins. At least three insect species, D. melanogaster, A. mellifera, and T. castaneum, have more than one isoform of CDA5 due to alternative splicing and/or exon skipping during the processing of pre-mRNA for these genes. Group V consists of two subgroups. One subgroup includes the CDA9s. Two CDAs (HaCDA5a and HaCDA5b), identified recently by proteomic analysis and EST sequence analysis of the PM of the cotton bollworm Helicoverpa armigera (Campbell et al., 2008; Jakubowska et al., 2010), also belong to this CDA9 subgroup of group V (Figure 9). Interestingly, two lepidopterans, B. mori and H. armigera, appear to have multiple genes related to CDA9. The other subgroup of group V consists of paralogs from T. castaneum only (TcCDAs 6, 7, and 8), and not from other insect species. All the proteins belonging to this group have only a CDA catalytic domain, and no CBD or LDLa domains.
Functional analysis of insect chitin deacety-lases Developmental patterns and tissue-specific expression of different CDA genes in the same species suggest that the chitin deacetylases may have specific functions. In D. melanogaster, the two group I genes, DmSerp (DmCDAl) and DmVerm (DmCDA2), are required for normal tracheal tube development and morphology (Lusching et al., 2006; Wang et al., 2006). D. melanogaster mutants lacking either serp or verm exhibited excessively long and tortuous embryonic tracheal tubes. In T. castaneum, injection of dsRNA for TcCDAl or TcCDA2, which are predominantly expressed in epidermis and tracheae, prevented all types of molts, including larval-larval, larval-pupal, and pupal-adult (Figure 5; Arakane et al., 2009). Furthermore, alternative exon-specific RNAi for TcCDA2 (TcCDA2a and TcCDA2b) revealed functional specialization of the isoforms for this CDA. Unlike exon non-specific RNAi for TcCDA2, injection of dsRNAs specific for either one of alternative exons did not prevent any molts, suggesting that the proteins TcCDA2a and TcCDA2b could compensate for each other. However, the resulting adults exhibited different abnormal phenotypes. RNAi for TcCDA2a affected only femoral-tibial joint movement, while dsRNA for TcCDA2b resulted in elytra with crinkled and rough dorsal surfaces (Arakane et al., 2009). These results suggest that group I CDAs play critical roles in maintaining the structural integrity of the cuticlular chitin laminae and chitin fibers of the tracheal tube. It is possible that there are unique cuticular proteins that preferentially bind to deacetylated portions of chitin, whereas others preferentially bind to fully acetylated chitin. These proteins may help to organize the chitinous cuticular layers and provide the proper rigidity and/or flexibility in different regions of the cuticle.
Injection of a mixture of dsRNAs for T. castaneum group V CDAs, TcCDAs 6, 7, 8, and 9, which are all predominantly expressed in the gut, significantly reduced the transcript levels of individual CDAs. However, no adverse effects on the appearance, behavior, or survival of these dsRNA-treated insects were observed (Arakane et al., 2009). Interestingly, Jakubowska et al. (2010) observed that one of the group V (CDA9 subgroup) CDA genes from H. armigera (HaCDA5a) was downregulated by bac-ulovirus infection in larvae. Like TnCDA9, HaCDA5a had a strong binding affinity for chitin, although it lacks any predicted chitin-binding domain. Incubation of the PM from S. frugiperda with recombinant HaCDA5a increased PM permeability in a concentration-dependent manner. Infection of insects with a recombinant baculovi-rus carrying this gene significantly increased the speed of kill for S. frugiperda and S. exigua. Together, these observations indicate that the group V CDA, HaCDA5a, may have a role in determining PM structure/morphology or permeability. For instance, downregulation of transcripts for this gene after pathogen attack resulted in reduced PM permeability, presumably to avoid pathogen infection. Additional studies in the future may reveal the physiological functions of the many CDAs belonging to groups II, III, and IV.
Chitin-Binding Proteins
Chitin is almost always found in association with numerous proteins that influence the overall mechanical and physicochemical properties of the chitin-protein matrix, which can range from very rigid (e.g., head capsule and mouth parts) to fully flexible (e.g., larval body and wing cuticle). Since chitin is an extracellular matrix polysac-charide, the proteins that have an affinity for chitin are expected to be extracellularly secreted proteins. This is generally true, with the constraint that some CBPs can be in vesicles or storage granules between the time they are synthesized and when they are secreted or released into the extracellular space by exocytosis.
There are three broad groups of insect proteins containing sequence motifs that have been associated with chitin-binding ability. The first group consists of a very large number of insect cuticular proteins, belonging to the CPR family, containing a consensus sequence(s) known as the extended Rebers & Riddiford Consensus (R&R Consensus) of a stretch of about 70 amino acids that defines pfam 00379.The second group of proteins contains an amino acid sequence motif known as the "peritrophin A" motif (Tellam et al., 1999). To avoid confusion about its biological role(s), this motif will be referred to as the ChtBD2 domain in this topic, because it is found not only in the group of proteins extracted from the peritrophic matrix, but also in proteins extracted from (or expressed in) cuticle-forming tissues. Proteins with the ChtBD2 motif are further subdivided intro three groups: peritrophic matrix proteins (with 1-19 ChtBD2 domains, determined to date); cuticular proteins analogous to peritrophins-3 (with 3 ChtBD2 domains); and cuticular proteins analogous to peritrophins-1 (with 1 ChtBD2 domain) (Jasrapuria et al., 2010). This domain consists of a linear sequence of about 60 amino acids with 6 cysteines and conserved spacings between successive cysteine residues. The ChtBD2 domain defines family 14 of carbohydrate-binding proteins with chitin-binding ability (CBM14; pfam01607; SMART 00494). The second group also includes enzymes of chitin metabolism (chitinases, chitin deacetylases, and a protease) that have one or more ChtBD2 domains in addition to their catalytic domains. The third group of chitin-binding proteins consists of the family of antimicrobial peptides related to tachystatins from horseshoe crab (denoted as A1, A2, B1, B2, and C subfamilies), as well as the calcium channel antagonists, agatoxins from spider venom. Tachys-tatins are expressed in hemocytes, where they are stored in the form of small granules and are released into the hemolymph upon an immune stimulus. This group of proteins with six cysteines and a high affinity for chitin has a triple-stranded P-sheet structure with an inhibitory cysteine knot motif (Fujitani et al., 2007). This structure is quite different from the peritrophin A motif and tachys-tatin (see below), and belongs to pfam 11478. They are not associated with cuticle or the PM, but they do play a major role in immune defense against bacteria, fungi, and other pathogens.
Representative members of each of the three groups of chitin-binding proteins have been extracted from the cuticle or the PM, or isolated from hemocytes. They have also been expressed in bacterial or other hosts, and some of the purified proteins have been shown to have chitin-binding ability. Several proteins belonging to the first and second groups of chitin-binding proteins are only predicted from known cDNA or genomic sequences and have not been biochemically characterized, largely as a result of difficulties associated with extracting them from highly sclero-tized cuticular preparations or exuviae. The following sections will focus on the proteins of the second group of proteins with ChtBD2 motifs, and also include a limited discussion of group 3 chitin-binding proteins. A discussion on the first group of cuticular proteins with the R&R or other consensus motifs is kept to a minimum.
Chitin-Binding Proteins with the R&R Consensus
The CPR family of cuticular proteins is generally rich in histidines and devoid of cysteines. The absence of cyste-ines has been regarded as a defining characteristic of this group of proteins, with rare exceptions. The number of cuticular proteins belonging to the CPR subfamily in different insects varies widely, ranging from 32 in A. mellifera to >150 in A. gambiae,indicating a genus-specific expansion of specific families of cuticular proteins. Among the many families of cuticular proteins in insects, only some members of the CPR family with the R&R Consensus have been unequivocally shown to bind to chitin.A member each of the Tweedle family from B. mori (Tang et al., 2010) and one protein of the CPAP family (see below) have also been shown to possess chitin-binding ability. Modeling studies using the 65-aa long R&R Consensus have led to the notion that this region assumes a half-barrel structure into which a liner chain of N-acetylglucosamines can be fitted using van der Waals interactions between the sugar oligo-mer and the hydrophobic rings of conserved aromatic amino acids in this consensus (Iconomidou et al., 2005). In an interesting study, Rebers and Willis (2001) demonstrated that the addition of this consensus sequence alone to glutathione-S-transferase resulted in acquisition of an affinity for chitin by this chimeric protein.
Peritrophic Matrix Proteins
The second group of proteins with the ChtBD2 motif is the family of proteins known as "peritrophins" that can be extracted from the PM using strong denaturing/ chaotropic reagents, such as 6-M urea or 6-M guanidine hydrochloride (Tellam et al, 1999). The extracted PMPs or recombinantly expressed PMPs have chitin-binding activity (Elvin et al., 1996; Wijffels et al., 2001; Wang et al. , 2004). This motif was shown to be responsible for binding to chitin by expressing a single ChtBD2 domain of Trichoplusia ni peritrophin, CBP1, in an insect cell line, and demonstrating its chitin-binding ability (Wang et al., 2004). Proteins with multiple ChtBD2 domains are commonly found strongly associated with the PM. Not all of them are actually extractable, even with strong chao-tropic agents. Some require extraction with strong organic solvents, such as anhydrous trifluoromethanesulfonic acid, which also deglycosylates O-linked glycoproteins (Campbell et al., 2008).
The number of ChtBD2 domains in insect PMPs varies from 1 to as many as 19 in the bertha armyworm Mam-estra configurata (Shi et al., 2004; Dinglasan et al., 2009; Venancio et al., 2009; Jasrapuria et al., 2010; Toprak et al., 2010). Some PMPs have multiple ChtBD2 repeats in a tandem arrangement with short spacers rich in P, S, and T residues. Some of these linkers are potential sites of O-glycosylation. Other PMPs have mucin domains interspersed between ChtBD2 domains in various patterns of alternating ChtBD2 and mucin domains (Wang et al., 2004; Venancio et al, 2009). PMPs with only one or two ChtBD2 domains have also been reported (Jasrapuria et al., 2010; Toprak et al., 2010). The number of PMPs in different species is variable. Both Ae. aegypti and D. melanogaster have been predicted to have about 65 PMPs, though many of these may not be components of the PM (Venancio et al., 2009). Detailed expression studies of all proteins with ChtBD2 domains in T. castnaeum have demonstrated that there are only 11 bona fide PMPs in this beetle (Jasrapuria et al., 2010). Direct proteomic analysis of >200 proteins extracted from PMs dissected from adult A. gambiae females fed a protein-free diet has revealed the presence of only 12 PMPs, with the number of ChtBD2 repeats ranging from 1 to 4. It is likely that the total number of PMPs in insects is in the range of 10-20, although it can’t be ruled out that additional PMP genes are expressed in the gut. However, their conceptual protein products were not detected in proteomic analyses because they were still in the insoluble pellet after extraction with detergents used in an extensive study (Dinglasan et al., 2009). Interestingly, different PMP genes of T. cas-taneum were not expressed uniformly through the length of the midgut, with some PMPs being expressed in the anterior midgut, whereas others coding for proteins with multiple ChtBD2 domains were expressed in the posterior midgut (Jasrapuria et al., 2010). Whether this differential spatial expression results in altered permeability of the PM along the length of the midgut remains to be investigated.
Cuticular Proteins Analogous to Peritrophins (CPAPs)
In addition to the PMP genes, which are expressed exclusively in the midgut lining cells, there are other genes encoding proteins with ChtBD2 domains, which are expressed in tissues other than the midgut. All of these proteins are predicted to have a cleavable signal peptide, and are expected to be capable of interacting with extracellular chitin. These genes are expressed predominantly in epidermal tissue as well as in other cuticle-forming tissues, including tracheae, elytra, hindwings, and hind-gut. These genes have been subdivided into two groups, CPAPPl and CPAP3, to reflect the fact that they encode proteins with one or three ChtBD2 domains, respectively (Jasrapuria et al., 2010). CPAP3 is the new name given to the orthologs of the previously characterized D. melano-gaster "obstructor" or "gasp" gene family.
Mutants of the D. melanogaster CPAP3-C gene are embryo-lethal, and have been reported to exhibit cuticular defects (Barry et al, 1999; Behr and Hoch, 2005). In D. melanogaster there are 10 genes encoding CPAP3 proteins, which can be further subdivided into two groups of 5 genes each. Only orthologs for the first group (CPAP3-A, CPAP3-B, CPAP3-C, CPAP3-D, and CPAP-E) are present in insects other than Drosophila species. There are significant variations in the expression profiles of these genes in different cuticle-forming tissues and/or developmental stages, suggesting functional differences among the CPAP3 proteins. RNA interference studies carried out in T. castaneum are consistent with such specialized functions of individual CPAP3 proteins (Jasrapuria et al., unpublished data).
While it is expected that the CPAP3 proteins with three ChtBD2s will bind to chitin strongly, this has been demonstrated for only one recombinant protein from the spruce budworm Choristoneura fumiferana, which was expressed in E. coli (Nisole et al., 2010). However, only a minor percentage of the His-tagged protein bound to the chitin, with the major portion appearing in the flow-through fraction, perhaps indicating that not all molecules of this recombinant protein had folded properly to exhibit strong chitin-binding activity. So far, there is no report of expression of this class of proteins in an insect cell system that may overcome the problem of misfolding as demonstrated for two PMP proteins with 10 and 12 repeats of ChtBD2 domains (Wang et al., 2004).
A second group of genes encoding proteins with one ChtBD2 domain, referred to as the CPAPl family proteins, has been characterized extensively using a bioinfor-matics analysis of the T. castaneum genome (Jasrapuria et al., 2010). These proteins vary extensively in size, and in the location of the ChtBD2 domain. Like CPAP3, they are also expressed in cuticle-forming tissues and have putative cleavable signal sequences consistent with a role involving interactions with chitin. So far, there are no reports on the chitin-binding ability of these proteins. Only some of these proteins have orthologs in D. melano-gaster, casting doubt on whether these proteins are ubiquitous in insects. However, RNAi studies have produced lethal phenotypes when transcripts for 3 of the 10 genes encoding CPAP1 proteins were depleted in T. castaneum (Jasrapuria, unpublished data).
