**3.4 Peptidase screening in** *Crithidia*

Among the insect trypanosomatids, the genus *Crithidia* comprises monoxenic trypanosomatids of insects that were originally characterized by the presence of choanomastigote forms in their life cycles (Hoare and Wallace, 1966). The first studies employing zymograms in order to detect proteolytic activity in *Crithidia* spp. were performed by Frank and Ashall in 1990. In the two studies published in that year, the activity in *Crithidia fasciculata* extracts was compared to *T. cruzi*. In this sense, it is worth mentioning that *C. fasciculata*, among all non-pathogenic trypanosomatid species, has been considered an excellent model organism for many studies concerning trypanosomatids, because it can be cultivated in high yields and do not require specific bio-safety precautions (Vickerman, 1994).

In a first approach (Ashall, 1990), parasite extracts were made by the use of 0.5% Nonidet P-40 and mixed with SDS-PAGE sample buffer in non-denaturing conditions. After

Applications of Zymography (Substrate-SDS-PAGE)

reflecting their phylogenetic proximity.

wide range of trypanosomatids as well.

for Peptidase Screening in a Post-Genomic Era 277

species produced similar patterns of proteolysis at 28°C: two cysteine peptidase bands in the 80-110 kDa range and a minor cysteine peptidase activity detected at 45 kDa; and a metallopeptidase band detected in the 55-66 kDa range. When cells were lysed with the nonionic detergent Triton X-114, cysteine peptidases were detected in the aqueous phase, whereas the metallopeptidase partitioned into the detergent-rich phase, which suggested that the latter is membrane-associated. Interestingly, cell lysates of these species were also employed by our group in another comparative study (Santos et al. 2005), in which a single cysteine peptidase was found at 50 kDa and two metallopeptidases were detected at 70 kDa and at 90 kDa. In the latter study, peptidases were analyzed in 10% linear polyacrylamide gels containing gelatin, and the temperature of incubation after electrophoresis was 37°C. As explained by Martinez and Cazzulo (1992), the apparent molecular mass of each band varies depending on the experimental conditions, including acrylamide concentration and temperature of incubation, which may explain this discrepancy. Despite this fact, both studies highlighted the common proteolytic profile between these species, possibly

Unlike the similarities detected in the three *Crithidia* spp. described above, heterogeneous proteolytic profiles were observed in different members of this genus. For instance, *Crithidia acantocephali* produced 4 cysteine peptidases of 80, 75, 70 and 50 kDa, while *Crithidia harmosa*  presented 3 metallopeptidases at 63, 50 and 45 kDa (Santos et al. 2005). These data suggested the value of proteolytic enzymes in distinguishing between trypanosomatid species that cannot be differentiated on structural grounds (Santos et al. 2005). In order to study the distribution of metallopeptidases in trypanosomatids, our group also investigated cellassociated proteolytic activities in distinct species by gelatin-SDS-PAGE in conditions that favor the detection of this subgroup, specifically alkaline conditions (pH 9.0) and proteolytic inhibitors that putatively identified these enzymes, such as 1,10-phenanthroline (Santos et al. 2008). The analysis confirmed the previous results in all the species cited above, showing a great heterogeneity of expression of metallopeptidases not only in *Crithidia* spp. but in a

In a similar approach, our group described the differential expression of peptidases in endosymbiont-harboring *Crithidia* species in comparison to members of this genus that naturally lacks a bacterium in the cytoplasm (d'Avila-Levy et al. 2001). In this genus, the trypanosomatids *Crithidia deanei, Crithidia desouzai* and *Crithidia oncopelti* have been described to contain a bacterium symbiont in the cytoplasm, known as endosymbiont, which can be eliminated by the use of antibiotics, leading to the generation of cured strains (reviewed by De Souza and Motta, 1999). Gelatin-SDS-PAGE analysis was used to characterize the cell-associated and extracellular peptidases in these organisms, and our survey showed that a similar proteolytic profile was observed in cells of *C. desouzai* and in wild and cured strains of *C. deanei*: two cysteine peptidases migrating at 60-65 kDa and two metallopeptidases at 51-58 kDa. An additional cysteine peptidase was detected in wild strains at 100 kDa. A subsequent study from our group showed that, after Triton X-114 extraction preformed in *C. deanei* cells, a 65-kDa cysteine peptidase partitioned exclusively in the aqueous phase, possibly present in intracellular compartments, and a 51-kDa metallopeptidase was only detected in the detergent-rich phase (d'Avila-Levy et al. 2003). The remaining enzymes, at 60 kDa and at 58 kDa, which corresponds to a cysteine-type and to a metallo-type peptidase, respectively, were found in both aqueous and detergent-rich

electrophoresis in SDS-PAGE, gels were overlaid with 0.75% agarose containing the chromogenic substrate Bz-Arg-pNA at 0.5 mM and at pH 8.0 and incubated for 4-6 h at 37°C, and then photographed using a blue filter to reveal yellow bands containing *p*nitroaniline. A single component with molecular mass >200 kDa that hydrolyzed this substrate was detected in *C. fasciculata* as well as in *T. cruzi* crude extracts. A modified procedure was also employed (Ashall et al. 1990), in which electrophoresis was followed by shaking the gels with 2% Triton X-100 and then the incubation of gels for 10 min in the presence of a range of amidomethylcoumarin substrates containing arginine adjacent to the amidomethylcoumarin moiety, at pH 8.0. Fluorescent bands were visualized in gels by ultraviolet light by the hydrolysis of each substrate. A single band of substrate hydrolysis occurred with all six substrates tested, in both *T. cruzi* and *C. fasciculata*, with the same electrophoretic mobility (150-200 kDa). Incubation of gel strips with various peptidase inhibitors showed that the enzyme was strongly inhibited by diisopropyl phosphorofluoridate (DFP), N--tosyl-L-lysinyl-chloromethylketone (TLCK), leupeptin and a peptidyldiazomethane containing lysine at P1, but not by E-64, PMSF, pepstatin A, 1,10 phenanthroline and a peptidyldiazomethane containing methionine at P1. This enzyme was characterized as an alkaline peptidase, probably from the serine-type, that cleaves peptide bonds on the carboxyl side of arginine residues at pH 8.0 (Ashall, 1990).

Following this set of experiments, Etges (1992) employed surface radioiodination of living cells, fractionation by Triton X-114 extraction and phase separation, and zymogram analysis by fibrinogen-SDS-PAGE in order to describe the presence of a surface metallopeptidase in *C. fasciculata* with biochemical similarities to the gp63 from *Leishmania* spp. This peptidase is one of major surface molecules in all *Leishmania* species and play vital roles in the different stages of *Leishmania* life cycle, being suggested its participation in many aspects of the infection inside the mammalian host (Yao, 2010). The presence of a similar neutral-toalkaline metallopeptidase at the surface of *C. fasciculata* led to the suggestion that gp63 should not be involved in the infection of the mammalian host by *Leishmania*, but rather contributes to the survival of the trypanosomatid inside the digestive tract of the insect (Santos et al. 2006).

The work of Etges (1992) opened the possibility to use the same technique in order to analyze the proteolytic profiles in different members of distinct trypanosomatid genera. With this task in mind, our group has analyzed the proteolytic profiles of a great number of species from 8 different genera of trypanosomatids by the use of SDS-PAGE containing 0.1% co-polymerized gelatin as substrate (Branquinha et al. 1996; Santos et al. 2005; 2008). In those studies, it became clear that two distinct proteolytic activities can be detected in total cell lysates: cysteine- and metallopeptidases. For detection of cysteine peptidase activity, the optimal conditions were established to be an acidic pH value (5.0-6.0) and the presence of a reducing agent, such as DTT, which was essential for detection of this activity. The use of specific inhibitors, such as E-64, prevented the development of cysteine peptidase activity bands. Metallopeptidases were consistently observed in a broad pH range (5.0-10.0), and the zinc-chelator 1,10-phenanthroline completely inhibited their activity.

In our first work (Branquinha et al. 1996), three *Crithidia* species were studied: *C. fasciculata, C. guilhermei* and *C. luciliae*. Cells were lysed by the addition of SDS-PAGE sample buffer in non-denaturing conditions, and peptidases were characterized by electrophoresis on 7-15% gradient SDS-PAGE with 0.1% gelatin co-polymerized as substrate. Cell lysates of the three

electrophoresis in SDS-PAGE, gels were overlaid with 0.75% agarose containing the chromogenic substrate Bz-Arg-pNA at 0.5 mM and at pH 8.0 and incubated for 4-6 h at 37°C, and then photographed using a blue filter to reveal yellow bands containing *p*nitroaniline. A single component with molecular mass >200 kDa that hydrolyzed this substrate was detected in *C. fasciculata* as well as in *T. cruzi* crude extracts. A modified procedure was also employed (Ashall et al. 1990), in which electrophoresis was followed by shaking the gels with 2% Triton X-100 and then the incubation of gels for 10 min in the presence of a range of amidomethylcoumarin substrates containing arginine adjacent to the amidomethylcoumarin moiety, at pH 8.0. Fluorescent bands were visualized in gels by ultraviolet light by the hydrolysis of each substrate. A single band of substrate hydrolysis occurred with all six substrates tested, in both *T. cruzi* and *C. fasciculata*, with the same electrophoretic mobility (150-200 kDa). Incubation of gel strips with various peptidase inhibitors showed that the enzyme was strongly inhibited by diisopropyl phosphorofluoridate (DFP), N--tosyl-L-lysinyl-chloromethylketone (TLCK), leupeptin and a peptidyldiazomethane containing lysine at P1, but not by E-64, PMSF, pepstatin A, 1,10 phenanthroline and a peptidyldiazomethane containing methionine at P1. This enzyme was characterized as an alkaline peptidase, probably from the serine-type, that cleaves peptide

Following this set of experiments, Etges (1992) employed surface radioiodination of living cells, fractionation by Triton X-114 extraction and phase separation, and zymogram analysis by fibrinogen-SDS-PAGE in order to describe the presence of a surface metallopeptidase in *C. fasciculata* with biochemical similarities to the gp63 from *Leishmania* spp. This peptidase is one of major surface molecules in all *Leishmania* species and play vital roles in the different stages of *Leishmania* life cycle, being suggested its participation in many aspects of the infection inside the mammalian host (Yao, 2010). The presence of a similar neutral-toalkaline metallopeptidase at the surface of *C. fasciculata* led to the suggestion that gp63 should not be involved in the infection of the mammalian host by *Leishmania*, but rather contributes to the survival of the trypanosomatid inside the digestive tract of the insect

The work of Etges (1992) opened the possibility to use the same technique in order to analyze the proteolytic profiles in different members of distinct trypanosomatid genera. With this task in mind, our group has analyzed the proteolytic profiles of a great number of species from 8 different genera of trypanosomatids by the use of SDS-PAGE containing 0.1% co-polymerized gelatin as substrate (Branquinha et al. 1996; Santos et al. 2005; 2008). In those studies, it became clear that two distinct proteolytic activities can be detected in total cell lysates: cysteine- and metallopeptidases. For detection of cysteine peptidase activity, the optimal conditions were established to be an acidic pH value (5.0-6.0) and the presence of a reducing agent, such as DTT, which was essential for detection of this activity. The use of specific inhibitors, such as E-64, prevented the development of cysteine peptidase activity bands. Metallopeptidases were consistently observed in a broad pH range (5.0-10.0), and the

In our first work (Branquinha et al. 1996), three *Crithidia* species were studied: *C. fasciculata, C. guilhermei* and *C. luciliae*. Cells were lysed by the addition of SDS-PAGE sample buffer in non-denaturing conditions, and peptidases were characterized by electrophoresis on 7-15% gradient SDS-PAGE with 0.1% gelatin co-polymerized as substrate. Cell lysates of the three

bonds on the carboxyl side of arginine residues at pH 8.0 (Ashall, 1990).

zinc-chelator 1,10-phenanthroline completely inhibited their activity.

(Santos et al. 2006).

species produced similar patterns of proteolysis at 28°C: two cysteine peptidase bands in the 80-110 kDa range and a minor cysteine peptidase activity detected at 45 kDa; and a metallopeptidase band detected in the 55-66 kDa range. When cells were lysed with the nonionic detergent Triton X-114, cysteine peptidases were detected in the aqueous phase, whereas the metallopeptidase partitioned into the detergent-rich phase, which suggested that the latter is membrane-associated. Interestingly, cell lysates of these species were also employed by our group in another comparative study (Santos et al. 2005), in which a single cysteine peptidase was found at 50 kDa and two metallopeptidases were detected at 70 kDa and at 90 kDa. In the latter study, peptidases were analyzed in 10% linear polyacrylamide gels containing gelatin, and the temperature of incubation after electrophoresis was 37°C. As explained by Martinez and Cazzulo (1992), the apparent molecular mass of each band varies depending on the experimental conditions, including acrylamide concentration and temperature of incubation, which may explain this discrepancy. Despite this fact, both studies highlighted the common proteolytic profile between these species, possibly reflecting their phylogenetic proximity.

Unlike the similarities detected in the three *Crithidia* spp. described above, heterogeneous proteolytic profiles were observed in different members of this genus. For instance, *Crithidia acantocephali* produced 4 cysteine peptidases of 80, 75, 70 and 50 kDa, while *Crithidia harmosa*  presented 3 metallopeptidases at 63, 50 and 45 kDa (Santos et al. 2005). These data suggested the value of proteolytic enzymes in distinguishing between trypanosomatid species that cannot be differentiated on structural grounds (Santos et al. 2005). In order to study the distribution of metallopeptidases in trypanosomatids, our group also investigated cellassociated proteolytic activities in distinct species by gelatin-SDS-PAGE in conditions that favor the detection of this subgroup, specifically alkaline conditions (pH 9.0) and proteolytic inhibitors that putatively identified these enzymes, such as 1,10-phenanthroline (Santos et al. 2008). The analysis confirmed the previous results in all the species cited above, showing a great heterogeneity of expression of metallopeptidases not only in *Crithidia* spp. but in a wide range of trypanosomatids as well.

In a similar approach, our group described the differential expression of peptidases in endosymbiont-harboring *Crithidia* species in comparison to members of this genus that naturally lacks a bacterium in the cytoplasm (d'Avila-Levy et al. 2001). In this genus, the trypanosomatids *Crithidia deanei, Crithidia desouzai* and *Crithidia oncopelti* have been described to contain a bacterium symbiont in the cytoplasm, known as endosymbiont, which can be eliminated by the use of antibiotics, leading to the generation of cured strains (reviewed by De Souza and Motta, 1999). Gelatin-SDS-PAGE analysis was used to characterize the cell-associated and extracellular peptidases in these organisms, and our survey showed that a similar proteolytic profile was observed in cells of *C. desouzai* and in wild and cured strains of *C. deanei*: two cysteine peptidases migrating at 60-65 kDa and two metallopeptidases at 51-58 kDa. An additional cysteine peptidase was detected in wild strains at 100 kDa. A subsequent study from our group showed that, after Triton X-114 extraction preformed in *C. deanei* cells, a 65-kDa cysteine peptidase partitioned exclusively in the aqueous phase, possibly present in intracellular compartments, and a 51-kDa metallopeptidase was only detected in the detergent-rich phase (d'Avila-Levy et al. 2003). The remaining enzymes, at 60 kDa and at 58 kDa, which corresponds to a cysteine-type and to a metallo-type peptidase, respectively, were found in both aqueous and detergent-rich

Applications of Zymography (Substrate-SDS-PAGE)

parasite.

for Peptidase Screening in a Post-Genomic Era 279

possibly reflecting the adaptations of the parasite to the different environments it might confront during its life cycle. Besides the gelatinolytic activity, the 60-, 67- and 80-kDa bands were also able to degrade casein incorporated into SDS-PAGE, but with minor activity, and no proteolytic activity was detected when bovine serum albumin incorporated into the gel. A distinct pattern of degradation was observed when hemoglobin was used as substrate: a 43-kDa metallopeptidase was exclusively detected in these conditions. These hemoglobinases are possibly involved in supplying exogenous iron and heme for the

Besides the characterization of these peptidases when *C. guilhermei* cells were grown in yeast extract-peptone-sucrose medium, log-phase cells grown in different culture medium composition were obtained and analyzed. The proteolytic zymograms displayed no qualitative difference, only quantitative variations. In this sense, the replacement of sucrose by glucose enhanced the proteolytic activity of the four bands, while either the replacement of sucrose by glycerol or the cultivation of cells in BHI decreased the proteolytic detection. These results pointed out to the influence of the culture medium composition in the

**4. Two-dimensional zymography coupled to peptidase identification through** 

For decades, one-dimensional (1D) zymographic gel systems have been broadly used for the analysis and characterization of proteolytic activities in several organisms. Especially in protozoa parasites, this technique has been extensively useful to detect and identify peptidases involved in virulence of pathogenic protozoa (North and Coombs 1981; Coombs and North 1983; Lockwood et al. 1987; Williams and Coombs 1995; Cuervo et al. 2006; De Jesus et al, 2009). Also, through this technique, crucial roles of these enzymes during the cell cycle of parasites have been revealed (Brooks et al. 2001; De Jesus et al. 2007). In the postgenomic era, this methodology is shedding light on the biochemical traits of organisms of unknown genomes (Santos et al. 2005; Pereira et al. 2009; d'Avila-Levy et al. 2001), and has the potential of increasing the functional annotation of the genome for those organisms yet sequenced. However, information regarding on isoforms of proteolytic enzymes, isoelectric point of peptidases, and even a higher resolution of complex proteolytic profiles cannot be obtained by 1D zymographic systems. In superior eukaryotes, a broader analysis of functional peptidases has been achieved by combining zymographic techniques with proteomic technologies, specifically two-dimensional electrophoresis (2D) and mass spectrometry that enable a better resolution of peptidase arranges and the direct identification of peptidase species (Ong and Chang 1997; Park et al. 2002; Zhao and Russell 2003; Wilkesman and Schröder 2007; Lee et al. 2011). Nevertheless, this combined approach

production of extracellular peptidases in this microorganism (Melo et al. 2002).

**mass spectrometry: Possibilities and technical difficulties** 

has been little used in the study of protozoan parasites (De Jesus et al. 2009).

Proteomic approaches intend to produce the widest possible resolution of individual proteins from a protein mixture, followed by protein identification by mass spectrometry (MS). The fractionation of complex cellular extracts by 2D is attained by combining two independent electrophoretic separations, the isoelectric focusing (IEF) in the first dimension and SDS-PAGE in the second dimension (MacGillivray and Rickwood 1974; O'Farrell 1975). After, protein spots are excised from the gel, submitted to enzymatic digestion and the resulting peptides are analyzed by MS. The developments of soft ionization sources for

phases. In cells of *C. oncopelti*, two metallopeptidases were detected in 59-63 kDa range (d'Avila-Levy et al. 2001) (Figure 1).

The analysis of the spent culture medium showed a similar profile among the abovementioned species: *C. desouzai* and both strains of *C. deanei* displayed an 80-kDa cysteine peptidase and a 60-kDa metallopeptidase, and *C. oncopelti* showed four bands of protein degradation migrating at 101 kDa, 92 kDa, 76 kDa and 59 kDa, all belonging to the metallopeptidase class. For comparison, *C. fasciculata* displayed a more complex extracellular profile, comprising five metallopeptidases migrating at 101 kDa, 92 kDa, 76 kDa, 60 kDa and 43 kDa (d'Avila-Levy et al. 2001). In summary, the proteolytic profiles of *C. deanei* and *C. desouzai* are identical, and distinct from *C. oncopelti*, which is in accordance to a revision in *Crithidia* taxonomy proposed previously by Brandão et al. (2000) and d'Avila-Levy et al. (2004) and recently confirmed by molecular phylogenetic analyses (Teixeira et al. 2011). In this sense, this genus must be subdivided into three groups: the first one (*Angomonas*) must include *C. deanei* and *C. desouzai*, the second one (designated as *Strigomonas*) must include *C. oncopelti* and the remaining *Crithidia* spp. would remain in the originally described genus.

In the same work (d'Avila-Levy et al. 2001), the availability of *C. deanei* wild and cured strains allowed us to study whether the presence of the endosymbiont induces any alteration in the proteolytic profile. The absence of the cell-associated 100-kDa cysteine peptidase in the cured strain was the only qualitative difference found, and may possibly be related to the absence of the endosymbiont. In addition, the activity of extracellular peptidases was enhanced in the cured strain, which provides evidence that the presence of the endosymbiont diminishes the secretion of proteolytic enzymes, mainly the metallopeptidase (d'Avila-Levy et al. 2001).

Extracellular peptidases were also the focus of studies in some species belonging to the genus *Crithidia*. Unlike cell-associated enzymes, qualitative differences were observed when extracellular proteolytic enzymes were analyzed. In all the species tested, only metallopeptidases were detected, and 3 bands in the 60-80 kDa range were common to *C. fasciculata, C. guilhermei* and *C. luciliae*. Nevertheless, bands with lower molecular mass (30- 40 kDa) were found exclusively in *C. fasciculata*, while higher molecular mass bands (90-100 kDa) were only detected in *C. fasciculata* and *C. guilhermei* (d'Avila-Levy et al. 2001; Santos et al. 2005). Interestingly, the extracellular proteolytic profile of *C. luciliae* was also analyzed by Jaffe and Dwyer (2003), but only two metallopeptidases were detected at 97 kDa and at 50 kDa, which could be explained by the smallest amount of spent culture medium employed as well as by the reduced incubation period for proteolysis development.

Melo et al. (2002) characterized the extracellular peptidases from *C. guilhermei* through the incorporation of different protein substrates into SDS-PAGE. When cells were grown in yeast extract-peptone-sucrose medium, the extracellular proteolytic zymogram comprised four bands with gelatinolytic activity migrating at 80 kDa, 67 kDa, 60 kDa and 55 kDa. All bands were inhibited by 1,10-phenanthroline, which classified these enzymes as metallopeptidases, and these gelatinases remained active over a broad pH range, being the maximum activity reached at pH 5.0, which is in accordance to their proper activity in the insect gut. Interestingly, these enzymes were mainly detected at 37°C; when gels were incubated at 28°C, which corresponds to the room temperature and to the expected value in the insect gut, the proteolytic activity was reduced and the 55-kDa band was not detected,

phases. In cells of *C. oncopelti*, two metallopeptidases were detected in 59-63 kDa range

The analysis of the spent culture medium showed a similar profile among the abovementioned species: *C. desouzai* and both strains of *C. deanei* displayed an 80-kDa cysteine peptidase and a 60-kDa metallopeptidase, and *C. oncopelti* showed four bands of protein degradation migrating at 101 kDa, 92 kDa, 76 kDa and 59 kDa, all belonging to the metallopeptidase class. For comparison, *C. fasciculata* displayed a more complex extracellular profile, comprising five metallopeptidases migrating at 101 kDa, 92 kDa, 76 kDa, 60 kDa and 43 kDa (d'Avila-Levy et al. 2001). In summary, the proteolytic profiles of *C. deanei* and *C. desouzai* are identical, and distinct from *C. oncopelti*, which is in accordance to a revision in *Crithidia* taxonomy proposed previously by Brandão et al. (2000) and d'Avila-Levy et al. (2004) and recently confirmed by molecular phylogenetic analyses (Teixeira et al. 2011). In this sense, this genus must be subdivided into three groups: the first one (*Angomonas*) must include *C. deanei* and *C. desouzai*, the second one (designated as *Strigomonas*) must include *C. oncopelti* and the remaining *Crithidia* spp. would remain in the

In the same work (d'Avila-Levy et al. 2001), the availability of *C. deanei* wild and cured strains allowed us to study whether the presence of the endosymbiont induces any alteration in the proteolytic profile. The absence of the cell-associated 100-kDa cysteine peptidase in the cured strain was the only qualitative difference found, and may possibly be related to the absence of the endosymbiont. In addition, the activity of extracellular peptidases was enhanced in the cured strain, which provides evidence that the presence of the endosymbiont diminishes the secretion of proteolytic enzymes, mainly the

Extracellular peptidases were also the focus of studies in some species belonging to the genus *Crithidia*. Unlike cell-associated enzymes, qualitative differences were observed when extracellular proteolytic enzymes were analyzed. In all the species tested, only metallopeptidases were detected, and 3 bands in the 60-80 kDa range were common to *C. fasciculata, C. guilhermei* and *C. luciliae*. Nevertheless, bands with lower molecular mass (30- 40 kDa) were found exclusively in *C. fasciculata*, while higher molecular mass bands (90-100 kDa) were only detected in *C. fasciculata* and *C. guilhermei* (d'Avila-Levy et al. 2001; Santos et al. 2005). Interestingly, the extracellular proteolytic profile of *C. luciliae* was also analyzed by Jaffe and Dwyer (2003), but only two metallopeptidases were detected at 97 kDa and at 50 kDa, which could be explained by the smallest amount of spent culture medium employed

Melo et al. (2002) characterized the extracellular peptidases from *C. guilhermei* through the incorporation of different protein substrates into SDS-PAGE. When cells were grown in yeast extract-peptone-sucrose medium, the extracellular proteolytic zymogram comprised four bands with gelatinolytic activity migrating at 80 kDa, 67 kDa, 60 kDa and 55 kDa. All bands were inhibited by 1,10-phenanthroline, which classified these enzymes as metallopeptidases, and these gelatinases remained active over a broad pH range, being the maximum activity reached at pH 5.0, which is in accordance to their proper activity in the insect gut. Interestingly, these enzymes were mainly detected at 37°C; when gels were incubated at 28°C, which corresponds to the room temperature and to the expected value in the insect gut, the proteolytic activity was reduced and the 55-kDa band was not detected,

as well as by the reduced incubation period for proteolysis development.

(d'Avila-Levy et al. 2001) (Figure 1).

originally described genus.

metallopeptidase (d'Avila-Levy et al. 2001).

possibly reflecting the adaptations of the parasite to the different environments it might confront during its life cycle. Besides the gelatinolytic activity, the 60-, 67- and 80-kDa bands were also able to degrade casein incorporated into SDS-PAGE, but with minor activity, and no proteolytic activity was detected when bovine serum albumin incorporated into the gel. A distinct pattern of degradation was observed when hemoglobin was used as substrate: a 43-kDa metallopeptidase was exclusively detected in these conditions. These hemoglobinases are possibly involved in supplying exogenous iron and heme for the parasite.

Besides the characterization of these peptidases when *C. guilhermei* cells were grown in yeast extract-peptone-sucrose medium, log-phase cells grown in different culture medium composition were obtained and analyzed. The proteolytic zymograms displayed no qualitative difference, only quantitative variations. In this sense, the replacement of sucrose by glucose enhanced the proteolytic activity of the four bands, while either the replacement of sucrose by glycerol or the cultivation of cells in BHI decreased the proteolytic detection. These results pointed out to the influence of the culture medium composition in the production of extracellular peptidases in this microorganism (Melo et al. 2002).
