Potential drug targets in cyst-wall biosynthesis by intestinal protozoa

Drug Resistance Updates 6 (2003) 239–246

Potential drug targets in cyst-wall biosynthesis by intestinal protozoa Edward L. Jarroll∗ , Keriman Sener ¸ Department of Biology, Northeastern University, 106 Egan Bldg., 360 Huntington Avenue, Boston, MA 02115, USA Received 8 July 2003; received in revised form 10 July 2003; accepted 14 July 2003

Abstract Given that resistance to antiprotozoal drugs exists and is likely to increase and given that currently no reliable treatments exist for some of these infections, efforts to find new targets for chemotherapy must be continued. In the case of cyst-forming pathogenic protozoa, one such target might be encystment pathways and cyst-wall assembly. Information is increasing on protozoan encystment and, as it does, we can begin to answer the question of whether targeting it for chemotherapy is a viable drug strategy. Currently, there are significant efforts to understand encystment in Giardia and Entamoeba, and potential targets are being discovered as work on their encystment mechanisms progress. We know with certainty now that Giardia and Entamoeba cyst walls contain unique proteins and polysaccharides which differ from those of their hosts and thus make them potentially interesting targets for a variety of chemotherapeutic attacks. Although we lack detailed information about the other protozoan cyst formers, enough evidence exists for Giardia and Entamoeba that it seems prudent to screen them with some of the antifungal drugs, especially those that target mannoproteins, chitin synthesis, and ␤ (1, 3) glucan synthesis to ascertain if they target elements in these protozoan pathways that are similar to those found in fungi. © 2003 Elsevier Ltd. All rights reserved. Keywords: Protozoa; Giardia; Entamoeba; Encystment; Cyst wall

1. Introduction Several types of protozoa infect the intestinal tract of humans; among these are flagellates, amoebae, ciliates, cyst-forming coccidia, and several species of microsporidia. While these organisms have been known to parasitologists for many years, many have only been recognized as potentially serious human pathogens since the onset of AIDS. Several of these protozoa exhibit rather simple asexual life cycles while others exhibit complex life cycles with a sexual phase. However, all of these protozoa have in common life-cycle strategies that involve leaving one host as an encysted form to infect another. The encysted form usually exits the host via the feces and enters the next host via the oral route. After ingestion, the encysted forms excyst at their preferred location along the host’s intestinal tract. Vegetative cells multiply within the host and establish an infection, and a specific stimulus (i) induces them to encyst before leaving for a new host. The protective walls of these encysted protozoa are often formed of proteins and polysaccharides (Gerwig et al., 2002; Eichinger, 1997, 2001; Arroyo-Begovich et al., 1980; Sterling and Arrowood, 1993) similar to those of fungal cell walls, have no equiva∗

Corresponding author. Tel.: +1-617-373-5828; fax: +1-617-373-7087. E-mail address: [email protected] (E.L. Jarroll).

1368-7646/$ – see front matter © 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S1368-7646(03)00065-7

lent structures in their mammalian hosts, and are needed by the protozoan for osmotic stability. However, unlike fungi which require their cell walls for osmotic stability inside or outside of the host, these parasitic protozoa do not need their protective walls for osmotic stability inside the hosts; rather, they need them only when the parasites pass through the external environment to get to the next host. Currently, it is unclear whether attacking the protozoan protective wall assembly or its precursor synthesis should be a primary strategy for attacking these parasites. Since relatively little is known of the details of protozoan encystment, the unknowns leave us wondering about the validity of such a strategy. The trigger for encystment of the vegetative trophozoites of Giardia intestinalis and Entamoeba invadens is often the depletion of some nutritional requirement. For Giardia, that requirement appears to be cholesterol and depriving them of it triggers encystment (Luján et al., 1996), while for E. invadens, the trigger can be the drastic dilution or removal of its carbon source (Eichinger, 1997). These triggers suggest that when the environment is not conducive to trophozoite survival, these protozoa encyst. Thus, disrupting the encystment process could lead to eventual cell death due to nutrient depletion. Furthermore, we do not know whether once triggered, encystment is irreversible, and we do not know whether encystment, if stopped in progress, can cause a cell to die. If encystment is irreversible and fatal

240

E.L. Jarroll, K. S¸ ener / Drug Resistance Updates 6 (2003) 239–246

when stopped in progress, then one could conceivably develop a chemotherapeutic agent that blocks encystment and produces a fatal outcome for the protozoan. Even if a fatal outcome does not occur directly, chemotherapy aimed at blocking encystment should be considered a potential target because (1) resistance to and discontinuation of currently used drugs has been reported for Giardia and Entamoeba (Upcroft and Upcroft, 2001), (2) there is a lack of effective treatment for some of the coccidia, and (3) blocking cyst formation could, at the very least, serve to reduce transmission even if it alone does not eradicate the infection. Giardia and Entamoeba cyst walls have been better studied than those of other protozoans. Their cyst walls are clearly composed of a combination of unique proteins and polysaccharides, ␤ (1, 3) (Gerwig et al., 2002) or ␤ (1, 4) (Arroyo-Begovich et al., 1980) hexosamine homopolymers, which offer distinct chemotherapeutic targets. Although reports on the detailed biochemical analyses of the encysted forms of ciliates, coccidia or microsporidia do not yet exist, Bigliardi et al. (1996) point to chitinous material in the exospore of Encephalitozoon hellum and Schottelius et al. (2000) detected chitinolytic activity in viable spores of Encephalitozoon cuniculi and Encephalitozoon intestinalis. Also, chitinous layers in the Cryptosporidium oocyst wall have been reported (Sterling and Arrowood, 1993), but as with the microsporidia, whether or not this is indeed chitin and the importance of the chitin, if present, to the wall’s integrity are unknown. Additionally, nothing is known of the cyst-wall composition of the ciliate, Balantidium coli. If proteins and chitin, or some other polysaccharide, are significant components of the Balantidium cyst wall, the Cryptosporidium oocyst wall or the microsporidial spore wall then these may offer unique targets similar to those of Giardia and Entamoeba. Unfortunately, the complexity of the coccidian and microsporidial life cycles and our inability to culture them successfully in vitro has made detailed biochemical analyses of these protozoan cyst walls impossible for the present. Thus, whether or not walls of all of these protozoa afford the same unique targets as those of Giardia and Entamoeba is a subject for future reports. It is likely, however, that they will offer analogous targets once their nature is elucidated.

2. Encystment In vivo, Giardia trophozoites encyst in the middle to lower jejunum. In vitro, Giardia trophozoites encyst when they are exposed to bile. Luján et al. (1996) showed that cholesterol deprivation in these lipid auxotrophs (Jarroll et al., 1981), brought on by bile, is the likely inducer of encystment rather than some component of the bile itself. Interestingly, metronidazole (MTZ), the drug most commonly used to treat giardiasis completely inhibits oxygen uptake (OU) and motility in the trophozoites; however, exposure of intact cysts to MTZ had no effect on their

OU or their ability to excyst (Paget et al., 1993). Approximately 10 h after encystment is induced, trophozoites lose their ability to take up exogenous glucose (Glc) and become resistant to MTZ both of which may reflect changes in metabolic activity, a change in membrane transport activity or both. During encystment, trophozoites express Golgi enzyme activities and develop a Golgi-like complex. Also, encysting and non-encysting Giardia trophozoites are sensitive to Brefeldrin A which inhibits their ability to transport proteins and supports Golgi involvement in vesicle formation. Furthermore, they express cyst antigens, and produce encystation-specific vesicles (ESVs) that contain material apparently destined to become cyst-wall filaments (Eichinger, 2001; McCaffery et al., 1994; McCaffery and Gillin, 1995; Luján et al., 1996; Luján and Touz, 2003). Unlike studies on Giardia encystment, which have been accomplished on the human pathogen grown and encysted axenically in vitro, many of the meaningful recent studies of Entamoeba encystment (Eichinger, 1997, 2001) have been performed using E. invadens, a reptilian intestinal parasite, rather than E. histolytica, the human pathogen. Just as there are clearly similarities between these two pathogens (Eichinger, 2001), there are clearly differences. For example, E. invadens (strain IP-1) encystment in vivo occurs in the large intestine and, at least in vitro, can be induced by either lowering the osmotic pressure or by removing the carbon source from its growth medium, but these same in vitro conditions will not induce encystment in E. histolytica (Eichinger, 1997, 2001). Despite that, significant progress in understanding encystment in Entamoeba is being made. Encystment of E. invadens requires either 5% bovine serum or defined galactose (Gal)-terminated ligands such as asialofetuin, GalBSA, or mucin in a strict concentration in the encystment medium (Coppi and Eichinger, 1999). Furthermore, catecholamines seem to play a role in the encystment signaling cascade somewhere downstream of the Gal lectin but upstream of adenylyl cyclase (Eichinger, 2001). Thus, Coppi and Eichinger (1999) speculated that a specific ratio of ligand to lectin may serve to convey an encystment signal. In the case of mucin, which amoebae dissolve to facilitate bacterial (food) availability, they also speculated that it could represent a quorum-sensing trigger in which a decreasing mucin concentration signals an increasing amoeba population. Increasing the numbers of amoebae while decreasing their food supply might be expected to signal amoebae to encyst and find a new host. Cryptosporidium and microsporidia have more complex life cycles than those of Giardia and Entamoeba, and triggers for these events have not been delineated to date.

3. Cyst-wall structure Giardia trophozoites are encased within a wall composed of an inner membranous and outer filamentous portion (Feeley et al., 1990). The outer portion, approximately

E.L. Jarroll, K. S¸ ener / Drug Resistance Updates 6 (2003) 239–246

0.3–0.5 ␮m thick (Feeley et al., 1990), is composed of filaments which measure 7–20 nm in diameter and which are arranged in a tightly packed meshwork (Erlandsen et al., 1989, 1996). Erlandsen et al. (1996) presented a detailed scanning electron microscopic (SEM) chronology of cyst-wall filament synthesis. The first extracellular sign of encystment is the appearance of approximately 15 nm (diameter) protrusions on the cell membrane at ca. 10 h after bile induction. The first cysts, which appear to be mature by virtue of the filament structure and water resistance, appeared at about 14 h and increased in abundance by 16 h. Arguello-Garcia et al. (2002) using carbohydrate-specific tags reported that polypeptide exposure on the cell surface precedes fibril patch assembly. The pattern for filament assembly in the Giardia cyst wall resembles that for microfibrils of the ␤ (1, 3) glucan in Candida albicans (Arguello-Garcia et al., 2002). During encystment in E. invadens, trophozoites aggregate and eventually round up and elaborate a cyst wall. SEM showed that the external surface displays a dense intertwined microfibrillar structure (Arroyo-Begovich et al., 1980; Arroyo-Begovich and Cárabez-Trejo, 1982). The Entamoeba cyst wall appears to be deposited outside the cell membrane by vesicular fusion with the membrane similar to that in Giardia, but this has not been as well documented (Eichinger, 1997, 2001).

4. Cyst-wall proteins A unique polysaccharide accounts for approximately 63% of the Giardia cyst-wall filaments; the remaining ca. 37% of the Giardia cyst-wall filaments contain proteins. Three of these cyst-wall proteins (Cwps) are identified as Cwp1, Cwp2 (Luján et al., 1996, 1997) and Cwp3 (Sun et al., 2003) with molecular weights of 26, 39, and 27.3 kDa, respectively. Cwp1, Cwp2 share a 61% identity in their overlapping regions, and Cwp3 exhibits 36% identity to Cwp1 and 34% to Cwp2 in their leucine-rich repeat (LRR) regions. Cwp1 and Cwp2 have five tandem copies and Cwp3 has four complete and a possible fifth LRR. Furthermore, all of these proteins have a cysteine-rich region and all of the 14 cysteine residue positions in Cwp3 are conserved with corresponding cysteine residues in Cwp1 and Cwp2. All three of these proteins localize to ESVs as well as to the mature cyst wall. A single copy gene encodes for each Cwp, and transcripts for Cwp1 and Cwp2 increase ca. 140 times and those of Cwp3, 100 times compared to transcripts for these proteins in non-encysting trophozoites (Luján et al., 1995a,b, 1997; Luján and Touz, 2003; Sun et al., 2003). The concentration of Cwp1/Cwp2 complex at specific regions of rough endoplasmic reticulum may initiate secretory granule biogenesis (Luján and Touz, 2003). Also the chaperones, immunoglobulin heavy chain-binding protein and protein disulfide isomerases, are apparently upregulated to assist in folding cyst-wall proteins (Lujan et al., 1997).

241

During encystment, Giardia expressing Cwp1 with green fluorescent protein (Gfp) chimeras formed labeled dense granule-like vesicles and showed subsequent incorporation of Cwp1–Gfp into the cyst wall (Hehl et al., 2000). The N-terminal domain of Cwp1 was needed to target Gfp to the secretory compartments and central, LRRs were needed for association of the chimera with extracellular cyst-wall material (Hehl et al., 2000). In Cwp3, the N-terminal region and LRRs were needed to target AU1-tagged Cwp3 into secretory compartment and the C-terminal region was necessary for incorporation into cyst wall (Sun et al., 2003). Cwp1 was not observed in mature cysts leading Hehl et al. (2000) to speculate that there is a negative regulation of Cwp1 synthesis late in the encystment process. All three of the Cwps are exported as disulfide-bonded heterodimers in ESVs. Exposing encysting Giardia to dithiothreitol (DTT) reportedly inhibits the formation of ESVs (ca. 85%) and reduces the Cwps to monomers. Cwps reform as dimers when DTT is removed without a need for new protein synthesis (Reiner et al., 2001). Exposing Giardia to DTT did not stimulate an unfolded protein response as it typically does in higher eukaryotes suggesting that Giardia may have diverged evolutionally before this form of endoplasmic reticulum control evolved (Reiner et al., 2001). Entamoeba invadens cyst walls also contain proteins as well as carbohydrates (Eichinger, 1997, 2001). Frisardi et al. (2000) demonstrated a major cyst-wall protein, named Jacob, and suggested that it may help direct chitin deposition during cyst-wall formation. This is because Jacob exhibits cysteine-rich domains similar to those in insect proteins that bind chitin, and Jacob binds to chitin as well as to Entamoeba cyst wall filaments. In fungal cell walls, mannoproteins linked to their glucans constitute about 40% of the wall. These embedded matrix proteins are responsible for properties of the walls such as porosity, hydrophobicity, and immunogenicity (Georgopapadakou, 2001). Benanomicins and pradimicins are benzonaphthacene antibiotics from Actinomadura sp., which selectively interact with fungal mannoproteins producing lethal effects (Walsh and Giri, 1997). Currently, it is unproven that the cyst-wall proteins from Giardia and Entamoeba have similar functions but if they do then potentially these interactions could be drug targets in cyst-forming protozoa. Another potential point of protein attack for Giardia encystment comes from work by Touz et al. (2002a) who reported a novel 54 kDa encystment protein (Giardia granules-specific protein, gGsp). gGsp locates exclusively to ESVs, is induced during encystment, exhibits calcium binding characteristics, and exhibits no homology to any other protein in reference databases (Touz et al., 2002a). Inhibition of gGsp by antisense to gsp resulted in trophozoites that formed Cwp2 which was transported to granules as would be expected, but the granules failed to express their contents to the trophozoite exterior. gGsp may also play a role, along with chaperones, in keeping the Cwp1/Cwp2

242

E.L. Jarroll, K. S¸ ener / Drug Resistance Updates 6 (2003) 239–246

from assembling prematurely inside the ESV (Touz et al., 2002a; Luján and Touz, 2003). Recent studies also showed the importance of proteinases in Giardia encystment. Touz et al. (2002b) purified a ∼45 kDa cysteine proteinase that is induced during encystment and plays a role in the proteolytic cleavage and release of Cwp2 before cyst-wall formation. Proteinase inhibitors E64d (trans-epoxysuccinyl-l-leucylamido-(4-guanidino butane)), ALLM (acetyl leucyl leucyl methoinunal), ALLN (acetyl leucyl leucyl norleucinal) and bestatin against this cysteine proteinase and possibly other proteinases inhibited cyst production.

5. Cyst-wall carbohydrates Detailed carbohydrate analyses of purified Giardia cyst-wall filaments (Gerwig et al., 2002) showed that these filaments are composed, not of GlcNAc ␤ (1, 4) GlcNAc (GlcNAc: N-acetyl glucosamine) as originally believed (Jarroll et al., 1989), but of a unique GalNAc homopolymer, d-GalNAc ␤ (1, 3) d-GalNAc (GalNAc: N-acetyl galactosamine). Conformational analyses make it clear that the highly insoluble nature of these filaments is not due to the conformational properties of a single GalNAc polypeptide chain but is most probably due to strong interchain interactions. The potential covalent linkage (yet to be proven) between the polysaccharide and the protein may also play a role in the insolubility (Gerwig et al., 2002). The E. invadens cyst walls (presumably those of E. histolytica as well) reportedly contain chitin constituting ca. 25% of its dry weight. This claim is based mainly on work by Arroyo-Begovich et al. (1980) showing the presence of amino sugars and of X-ray diffraction patterns of purified cyst wall equated to an authentic chitin standard and on more recent work showing the activity of chitin synthase (Das and Gillin, 1991) and chitinases (de la Vega et al., 1997; Ghosh et al., 1999). Biochemical analyses of the Entamoeba cyst wall are lacking and it is unknown if other amino sugars

or polysaccharides are present, and potentially important, in these cyst walls.

6. Synthesis of cyst-wall saccharide units Many organisms including humans use the pathways discussed below and shown in Fig. 1 to synthesize amino sugars which in turn they use for a variety of cellular functions including polysaccharide synthesis and protein glycosylation. Thus, it becomes important to know if there are sufficient differences between the protozoan enzymes and those of their hosts if they are to be exploited as drug targets. Several potentially important differences in the pathway enzymes from various organisms have been discovered and clearly some are yet to be discovered. In encysting Giardia, UDP-GalNAc synthesis takes place by the activity of five nonsedimentable, encystment-induced enzymes (Macechko et al., 1992): glucosamine 6-phosphate isomerase (Gnp, EC 5.3.1.10), glucosamine 6-phosphate N-acetylase (Glcnpa, EC 2.3.1.4), phospho N-acetylglucosamine mutase (Pglcnacm, EC 2.7.5.2), uridine diphospho N-acetylglucosamine pyrophosphorylase (UDP-glcnacpp, EC 2.7.7.23), and uridine diphosphate N-acetylglucosamine 4 -epimerase (UDP-glcnace, EC 5.1.3.7). In human, fungal and bacterial systems, GlcN 6-P is synthesized from fructose 6-P by the activity of l-glutamine d-fructose 6-phosphate amidotransferase (Gf 6-pat, also called 2-amino-2-deoxy d-glucose-6-phosphate ketol-isomerase or Glcn synthase, EC 2.6.1.16) (Selitrennikoff and Nakata, 2003). Gf 6-pat activity was below the limits of detection in encysting or non-encysting Giardia (Macechko et al., 1992). Instead, Giardia uses a reversible Gnp with aminase and deaminase activities (Selitrennikoff and Nakata, 2003; Steimle et al., 1997). Gnp genes have been cloned and sequenced from Giardia strains MR4 (van Keulen et al., 1998) and WB (Knodler et al., 1999). In MR4, two genes were detected (gnp1 and gnp2) but only Gnp1 was expressed during encystment. In WB, Gnp-A has a short 5 untranslated region, [GlcNAc -β 1,4- GlcNAc]n

Fru 6-P + NH4Cl 1 GlcN 6-P 7 Fru 6-P + Asp or Glu

2

3 GlcNAc 6-P

4 GlcNAc 1-P

8 UDP-GlcNAc 5 UDP-GalNAc 6

[GalNAc- β1,3- GalNAc] n Fig. 1. Synthesis of ␤ (1, 4) GlcNAc and ␤ (1, 3) GalNAc polysaccharides. The pathway involved in ␤ (1, 3) GalNAc polysaccharide synthesis in Giardia is numbered 1–6. (1) Glucosamine 6-phosphate isomerase; (2) glucosamine 6-phosphate N-acetylase; (3) phosphoacetylglucosamine mutase; (4) uridine diphosphate N-acetylglucosamine pyrophosphorylase; (5) uridine diphosphate N-acetylglucosamine 4 -epimerase; (6) cyst-wall synthetase; (7) l-glutamine d-fructose 6-phosphate amidotransferase (glucosamine synthase); (8) chitin synthase.

E.L. Jarroll, K. S¸ ener / Drug Resistance Updates 6 (2003) 239–246

and it is expressed at a low level during vegetative growth and encystment, while Gnp-B has two transcripts: one expressed constitutively and a second is upregulated during encystment (Knodler et al., 1999). It is not clear what accounts for these differences in gene expression reported in these two studies but one suggestion is that it may be strain differences. Giardia’s Gnp1 exhibits only a 34% identity with the polypeptide subunit of the hexameric Nagb from Escherichia coli and with the single subunit Nag1 from C. albicans, but the entire active site, especially the Asp72 and His14 critical for catalytic activity of the E. coli isomerase, are conserved in both Candida’s Nag1 and in Giardia’s Gnp1 (van Keulen et al., 1998). Steimle et al. (1997) showed that Giardia’s Gnp1 exhibited a higher rate for the anabolic activity (aminase reaction) than for the catabolic activity (deaminase reaction). d-Glucose 6-P, d-mannose 6-P, and d-galactose 6-P did not substitute for d-fructose 6-P; l-glutamine and l-asparagine did not substitute for NH4 Cl. GlcNAc 6-P allosterically activates the E. coli and dog kidney isomerases (multimeric), but does not activate the purified Giardia or yeast isomerases (monomeric). Likewise, the Giardia Gnp was not activated or inhibited by GlcNAc 1-phosphate, UDP-GlcNAc, or UDP-GalNAc. However, 2-amino-2-deoxyglucitol-6-phosphate, an analogue of GlcN 6-P, strongly inhibited the Gnp (aminating (Ki = 2 × 10−8 M) and deaminating (Ki = 2.8 × 10−7 M)) activities (Steimle et al., 1997). In preliminary in vitro trials, this inhibitor reduced encystment from ca. 60–70% to ca. 2–3% at 1 mM (Jarroll, unpublished). Giardia’s Gnp1 conserves the same Asp72, His143, Arg172, and Lys208 that are critical for E. coli isomerase inhibition by 2-amino-2-deoxyglucitol-6-phosphate, and this inhibitor showed competitive inhibition with both native and recombinant Giardia enzymes (Steimle et al., 1997). Lopez et al. (2002) showed that Giardia’s Gnp is ubiquinated in the course of encystment suggesting a tight regulation of its expression. Excysted and non-encysting Giardia trophozoites may be prevented from premature cyst-wall formation by removal of the isomerase by ubiquination until they are triggered to increase production by transcriptional activation. Evidence that the encystment process is proteasome related in Giardia, as well as Entamoeba and other protozoa (Paugam et al., 2003), also may offer promising drug targets. Initial characterization of Glcnpa reveals a pH optimum of about 5. The activity of this enzyme is also increased in encysting cells compared to that of in non-encysting cells (Macechko et al., 1992). This acetylase has been reported in Saccharomyces cerevisiae, C. albicans and mice, but not in humans and the fungal enzymes have no apparent mammalian cognates (Selitrennikoff and Nakata, 2003). Whether or not the Giardia’s Glcnpa does remains to be seen. Partial purification (Lindmark and Schmidt, unpublished) and initial characterization of Pglcnacm showed a pH optimum of 8. The mutase requires Mg2+ and Glc 1,

243

6-bisphosphate for its activity, and it is protected by DTT during purification. Even with DTT, the native mutase is unstable even in the cold. The activity of this mutase increased by from 8- to 20-fold by 20 h into encystment. The acetylase and the mutase have been cloned, sequenced, expressed in vitro as recombinant proteins (Lopez et al., 2003) and the assayed recombinant proteins showed their expected activity (Sener ¸ et al., unpublished). However, insufficient work has been done on these enzymes at the present for us to know if they are potential targets for inhibiting encystment specifically. Native UDP-glcnacpp activity has been purified from Giardia and exhibits a molecular mass of ca. 66 kDa (Bulik et al., 1998). UDP-glcnacpp exhibits no activity for other UDP-sugars as substrates, but it exhibits partial activity with GalNAc 1-P (58%) and Glc 1-P (53%). Interestingly, as little as 3 ␮M GlcN 6-P stimulated the activity of UDP-glcnacpp threefold to eightfold in the synthetic direction and only with UDP-GlcNAc 1-P as substrate (Bulik et al., 2000). At present, it appears that upregulation of GalNAc synthesis is tied to the increase in GlcN 6-P the product of GlcN 6-P isomerase aminating activity. GlcN 6-P appears to enhance the activity of a UDP-glcnacpp shifting the dynamics of the pathway from catabolic to anabolic. Western blot analysis suggested that this UDP-glcnacpp activity is present constitutively (Bulik et al., 1998), and may be a regulatory point in this pathway. Recently, a yeast-like UDP-glcnacpp gene has been detected in Giardia (Lopez et al., 2003) which produces a protein significantly smaller than the 66 kDa protein reported by Bulik et al. (1998) for the native UDP-glcnacpp. The specific activity of the recombinant yeast-like enzyme (Sener ¸ and Jarroll, unpublished) is apparently less than the native enzyme described by Bulik et al. (1998). Furthermore, Western blot analysis of the recombinant enzyme showed an increase in enzyme level during encystment. Studies are underway to assess the contribution of both of these activities to the process of encystment. Giardia’s cyst-wall synthesis requires that the UDPGlcNAc made during encystment be converted to UDPGalNAc by UDP-glcnace before it is incorporated into the polysaccharide. Giardia’s UDP-glcnace has been cloned and sequenced (Lopez et al., 2003), but it differs from the E. coli and mammalian UDP-glcnace in that these two can interconvert UDP-Glc to UDP-Gal as well as UDP-GlcNAc to UDP-GalNAc (Thoden et al., 2001) while Giardia’s can apparently only interconvert UDP-GlcNAc to UDP-GalNAc. This difference could represent a potentially unique chemotherapeutic site for attack in Giardia. It is currently unknown if this epimerase also occurs in Entamoeba.

7. Synthesis of cyst-wall polysaccharides Analogous to chitin synthase (Chs) in chitin-forming systems, there must be an enzyme that can fix the UDP-GalNAc

244

E.L. Jarroll, K. S¸ ener / Drug Resistance Updates 6 (2003) 239–246

into the ␤ (1, 3) GalNAc polymer for Giardia. An inducible, particle associated GalNAc transferase activity has been described (Jarroll et al., 2001; Das and Gillin, 1996) in encysting Giardia and has been tentatively named “cyst-wall synthetase” (Cws) (Jarroll et al., 2001). Cws activity is below the limits of detection in non-encysting trophozoites, but increases in encysting Giardia to detectable levels after about 8 h of encystment. Cws activity exhibits a requirement for Mg2+ or Ca2+ . Cws is specific for UDP-GalNAc. In light of the foregoing, it seems entirely likely that this Cws will turn out to be a unique ␤ (1, 3) GalNAc transferase which, in some fashion, is involved in the assembly of the Giardia cyst-wall filaments. The development of ␤ (1, 3) glucan synthase inhibitors (Georgopapadakou, 2001) is receiving attention in fungal systems because these inhibitors potentially target fungal cell wall structure. Among these drugs are the cyclic lipopeptides echinocandins and the liposaccharides papulacandins. No tests of these drugs on Giardia encystment have been reported yet, but these are experiments in which we are engaged. Such inhibitors exhibit some degree of efficacy against Pneumocystis and Candida which have ␤ (1, 3) glucans as cell wall components (Schmatz et al., 1990). In other systems, the UDP-GlcNAc is polymerized to chitin by Chs (Munro and Gow, 2001) and presumably this is similar to the events which occur in encysting Entamoeba. However, the details of the biochemical pathway leading to chitin synthesis have not been examined in Entamoeba, nor have the specifics of Entamoeba’s trophozoite or cyst-wall sugar composition for that matter. Even though chitin synthesis ends up with the polymerization of UDP-GlcNAc into the ␤ (1, 4) GlcNAc homopolymer, the details of how this happens can vary, and in the case of Entamoeba are unknown. For example, whether Entamoeba synthesizes UDP-GlcNAc via a pathway similar to that used by Giardia (Fig. 1) but stopping short of the epimerase, or one used by other microorganisms involving GlcN synthase (Fig. 1) is unknown. Also unknown is whether or not other amino sugars are made and/or are included to any great extent in the Entamoeba cyst wall. Also unknown is the nature of the Chs in Entamoeba. For example, Nikkomycin and Polyoxin D, GlcNAc analog drugs that inhibit Chs activity in some fungi, seem to have little if any effect on the Chs activity in E. invadens even though encystment is apparently inhibited (Das and Gillin, 1991). Due to transport problems in animal models, Nikkomycin and Polyoxin D apparently exhibit too little antifungal potential to be considered further for development as human drugs (Georgopapadakou, 2001) and thus it is unlikely that these antifungal agents will be developed for protozoan infections either. Parenthetically, Nikkomycin and Polyoxin D only partially inhibited Cws activity at high mM concentrations (Karr and Jarroll, unpublished). However, there are other Chs inhibitors (Sudoh et al., 2000; Hwang et al., 2000, 2002) which as of yet have not been tested against the Chs of Entamoeba or the Cws of Giardia.

8. Regulation of cyst-wall synthesis At present, little is known of the regulation of cyst-wall synthesis. However, information is beginning to appear (Luján and Touz, 2003) and as it does this area too may lend itself to the development of new drug targets. For Giardia, Luján et al. (1996) proposed that since cholesterol deprivation can induce encystment in Giardia, then perhaps cholesterol-induced changes in the trophozoite’s plasma membrane fluidity triggers a second messenger pathway and cholesterol might mediate transcription (Luján et al., 1996) regulation. Ellis et al. (2003) identified Giardia homologues of two members of the mitogen-activated protein kinase (Mapk) family, Erk1 and Erk2. Mapks are a major signaling system for eukaryotic cells to transfer extracellular stimuli to intracellular events. Both Erk1 and Erk2 exhibited a significantly reduced kinase activity, reaching a minimum during 24 h in encystment (Ellis et al., 2003). The decrease in activities was correlated to the decrease in phosphorylation which is known to control the function of these proteins. In the same work, it is suggested that reduced cholesterol starvation in encysting cells may be a factor that indirectly prevents Erk1 and Erk2 activation. The Myb family of transcriptional factors is important in regulating developmental processes in a variety of eukaryotic organisms. In two independent studies a Giardia homologue of myb, gmyb2 was identified (Sun et al., 2002; Yang et al., 2003). Expression of gMyb2 increased during encystment and it bound to the promoter region of the encystment-induced genes: gmyb2, to all three cwps, and to gnp. Deletion of putative gMyb2 binding sites from the promoter region of gnp reduced its encystment-specific activity and fusion of the gmyb2 binding site to a constitutive promoter or to gnp with a deleted gMyb2 binding site induced encystment-specific expression. These results reveal that gMyb2 is the first identified transcriptional factor in Giardia and may play an important role in regulation of encystment genes. The sedimentable enzymes in the Giardia UDP-GalNAc synthesizing pathway are transcriptionally activated (Lopez et al., 2003). Because the particulate Cws activity appears in a time-related manner during encystment like all of the other pathway enzymes, it seems likely that Cws production is also controlled transcriptionally, but that remains to be proven. 9. Conclusions and future directions Less than 20 years ago, the discussion of cyst-wall formation in protozoa would have taken a paragraph or two at most and certainly there was little thought given to the idea that it may eventually become a drug target. Today, significant information is beginning to appear on the encystment of Giardia, and more slowly, but just as surely, on Entamoeba.

E.L. Jarroll, K. S¸ ener / Drug Resistance Updates 6 (2003) 239–246

Efforts with both of these protozoa have been made possible by our ability to cultivate them axenically and to encyst them in vitro. However, the axenic in vitro cultivation of Giardia and Entamoeba, parasites with relatively simple life cycles, did not come quickly. Therefore, we must not expect that the cultivation of either coccidia or microsporidia will be accomplished quickly given their complex life cycles which involve a need for host cells during their development. Nonetheless, work on such cultivation of these later parasites must continue, as must the pursuit of the encystment process and cell wall assembly as drug targets in all protozoans. This is true especially for those that can be cultivated currently because of the limited number of available antiprotozoal agents and the emerging resistance to the few that are available. Although we are unable to study encystment in each of these parasite species individually at present, we can hypothesize that there are likely some commonalties in the process among them. Thus, the details of encystment in the more primitive protozoa, Giardia and Entamoeba, might shed light on encystment mechanisms, and thus drug targets, in the coccidia and the microsporidia as well. Furthermore, it is likely that unique drug targets will be found for the coccidia and microsporidia once we are able to study their oocyst/spore wall formation in greater detail. References Arguello-Garcia, R., Arguello-Lopez, C., Gonzalez-Robles, A., et al., 2002. Sequential exposure and assembly of cyst wall filaments on the surface of encysting Giardia duodenalis. Parasitology 125, 209–219. Arroyo-Begovich, A., Cárabez-Trejo, A., 1982. Location of chitin in the cyst wall of Entamoeba invadens with colloidal gold tracers. J. Parasitol. 68 (2), 253–258. Arroyo-Begovich, A., Cárabez-Trejo, A., Ru´ız-Herrera, J., 1980. Identification of the structural component in the cyst wall of Entamoeba invadens. J. Parasitol. 66 (5), 735–741. Bigliardi, E., Seinfi, M., Lupetti, P., et al., 1996. Microsporidian spore wall: ultrastructural findings on Encephalitozoon hellem exospore. J. Eukaryot. Microbiol. 43, 181–186. Bulik, D., Lindmark, D., Jarroll, E., 1998. Purification and characterization of UDP-N-acetylglucosamine pyrophosphorylase from encysting Giardia. Mol. Biochem. Parasitol. 95, 135–139. Bulik, D., van Ophem, P., Manning, J., et al., 2000. UDP-N-acetylglucosamine pyrophosphorylase: a key enzyme in encysting Giardia is allosterically regulated. J. Biol. Chem. 275, 14722–14728. Coppi, A., Eichinger, D., 1999. Regulation of Entamoeba invadens encystation and gene expression with galactose and N-acetylglucosamine. Mol. Biochem. Parasitol. 102, 67–77. Das, S., Gillin, F.D., 1991. Chitin synthase in encysting Entamoeba invadens. Biochem. J. 280, 641–647. Das, D., Gillin, F.D., 1996. Giardia lamblia: increased UDP-N-acetyl-dglucosamine and UDP-N-acetyl-d-galactosamine transferase activities during encystation. Exp. Parasitol. 83, 19–29. de la Vega, H., Specht, C.A., Semino, C., et al., 1997. Cloning and expression of chitinases of Entamoeba. Mol. Biochem. Parasitol. 85, 139–147. Eichinger, D., 1997. Encystation of Entamoeba parasites. Bioassays 19 (7), 633–639. Eichinger, D., 2001. Encystation in parasitic protozoa. Curr. Opin. Microbiol. 4, 421–426.

245

Ellis, J.H., Davila, M., Chakrabarti, R., 2003. Potential involvement of extracellular signal-regulated kinase 1 and 2 in encystations of a primitive eukaryote, Giardia lamblia. J. Biol. Chem. 278 (3), 1936– 1945. Erlandsen, S.L., Bemrick, W., Pawley, J., 1989. High resolution electron microscopic evidence for the filamentous structure of the cyst wall in Giardia muris and Giardia duodenalis. J. Parasitol. 75, 787– 797. Erlandsen, S.L., Macechko, P.T., van Keulen, H., et al., 1996. Formation of the Giardia cyst wall: studies on extracellular assembly using immunogold labeling and high resolution field emission SEM. J. Eukaryot. Microbiol. 43, 416–429. Feeley, D., Holbertson, D., Erlandsen, S., 1990. The biology of Giardia. In: Meyer, E.A. (Ed.), Giardiasis. Elsevier, New York, pp. 11–50. Frisardi, M., Ghosh, S.K., Field, J., et al., 2000. The most abundant glycoprotein of amoebic cyst walls (Jacob) is a lectin with five Cys-rich, chitin-binding domains. Infect. Immun. 68, 4217–4224. Georgopapadakou, N.H., 2001. Update on antifungals targeted to the cell wall: focus on ␤-1, 3 glucan synthase inhibitors. Exp. Opin. Invest. Drugs 10 (2), 269–280. Gerwig, G.J., van Kuik, J.A., Leeflang, B.R., et al., 2002. The Giardia intestinalis filamentous cyst wall contains a novel ␤ (1-3)-N-acetyld-galactosamine polymer: a structural and conformational study. Glycobiology 12 (8), 499–505. Ghosh, S.K., Field, J., Frisardi, M., et al., 1999. Chitinase secretion by encysting Entamoeba invadens and transfected Entamoeba histolytica trophozoites: localization of secretory vesicles, endoplasmic reticulum and Golgi apparatus. Infect. Immun. 67 (6), 3073–3081. Hehl, A.B., Marti, M., Kohler, P., 2000. Stage-specific expression and targeting of cyst wall protein-green fluorescent protein chimeras in Giardia. Mol. Biol. Cell. 11, 1789–1800. Hwang, E.-I., Yun, B.-S., Kim, Y.-K., et al., 2000. Phellinsin A, a novel chitin synthase inhibitor produced by Phellinus sp. PL3. J. Antibiot. 53 (9), 903–911. Hwang, E.-I., Kwon, B.-M., Lee, S.-H., et al., 2002. Obovatols, new chitin synthase 2 inhibitors of Saccharomyces cerevisiae from Magnolia obovata. J. Antimicrob. Chemother. 49, 95–101. Jarroll, E.L., Muller, P.J., Meyer, E.A., et al., 1981. Lipid and carbohydrate metabolism of Giardia lamblia. Mol. Biochem. Parasitol. 2, 187–196. Jarroll, E.L., Manning, P., Lindmark, D.G., et al., 1989. Giardia cyst wall-specific carbohydrate: evidence for the presence of galactosamine. Mol. Biochem. Parasitol. 32, 121–132. Jarroll, E.L., Macechko, P.T., Steimle, P.A., et al., 2001. Regulation of carbohydrate metabolism during Giardia encystment. J. Eukaryot. Microbiol. 48 (1), 22–26. Knodler, L.A., Svärdm, S.G., Silberman, J.D., et al., 1999. Developmental gene regulation in Giardia lamblia: first evidence for an encystationspecific promoter and differential 5 mRNA processing. Mol. Microbiol. 34 (2), 327–340. Lopez, A.B., Hossain, M.T., van Keulen, H., et al., 2002. Giardia intestinalis glucosamine 6-phosphate isomerase: the key enzyme to encystment appears to be controlled by ubiquitin attachment. J. Eukaryot. Microbiol. 49 (2), 134–136. Lopez, A.B., Sener, K., Jarroll, E.L., et al., 2003. Transcription regulation is demonstrated for five key enzymes in Giardia intestinalis cyst wall polysaccharide synthesis. Mol. Biochem. Parasitol. 128, 51–57. Luján, H.D., Touz, M.C., 2003. Protein trafficking in Giardia lamblia. Cell. Microbiol. 5 (7), 427–434. Luján, H.D., Merotta, A., Mowatt, M., et al., 1995a. Developmental induction of Golgi structure and function in the primitive eukaryote Giardia lamblia. J. Biol. Chem. 270, 4612–4618. Luján, H.D., Mowatt, M.R., Conrad, J.T., et al., 1995b. Identification of a novel Giardia lamblia cyst wall protein with leucine-rich repeats. J. Biol. Chem. 270, 29307–29313. Luján, H.D., Mowatt, M.R., Byrd, L.G., et al., 1996. Cholesterol starvation induces differentiation of the intestinal parasite Giardia lamblia. Proc. Natl. Acad. Sci. U.S.A. 93, 7628–7633.

246

E.L. Jarroll, K. S¸ ener / Drug Resistance Updates 6 (2003) 239–246

Luján, H.D., Mowatt, M.R., Nash, T.E., et al., 1997. Mechanism of Giardia lamblia differentiation into cysts. Microbiol. Mol. Rev. 61, 294–304. Macechko, P.T., Steimle, P.A., Lindmark, D., et al., 1992. Galactosamine synthesizing enzymes are induced when Giardia encyst. Mol. Biochem. Parasitol. 56, 301–310. McCaffery, M., Gillin, F.D., 1995. Giardia lamblia: ultrastructural basis of protein transport during growth and encystation. Exp. Parasitol. 79, 220–235. McCaffery, M., Faubert, G.M., Gillin, F.D., 1994. Giardia lamblia: traffic of a trophozoite variant surface protein and a major cyst wall epitope during growth, encystation and antigenic switching. Exp. Parasitol. 79, 236–249. Munro, C.A., Gow, N.A.R., 2001. Chitin synthesis in human pathogenic fungi. Med. Mycol. 39 (Suppl. 1), 41–53. Paget, T.A., Manning, P., Jarroll, E., 1993. Oxygen uptake in cysts and trophozoites of Giardia lamblia. J. Eukaryot. Microbiol. 40, 246–250. Paugam, A., Bulteau, A.-L., Dupouy-Camet, J., et al., 2003. Characterization and role of protozoan parasites proteasomes. Trend Parasitol. 19 (2), 55–59. Reiner, D.S., McCaffery, J.M., Gillin, F.D., 2001. Reversible interruption of Giardia lamblia cyst wall protein transport in a novel regulated secretory pathway. Cell Microbiol. 3 (7), 459–472. Schmatz, D.M., Romancheck, M.A., Pittarelli, L.A., et al., 1990. Treatment of Pneumocystis carinii pneumonia with 1, 3-␤-glucan synthesis inhibitors. Proc. Natl. Acad. Sci. U.S.A. 87, 5950–5954. Schottelius, J., Hünger, F., Schüler, T., et al., 2000. Chitinolytic activity in viable spores of Encephalitozoon species. Mem. Inst. Oswaldo Cruz 95 (5), 701–705. Selitrennikoff, C.P., Nakata, M., 2003. New cell wall targets for antifungal drugs. Curr. Opin. Invest. Drugs 4 (2), 200–205. Steimle, P.A., Lindmark, D., Jarroll, E.L., 1997. Purification and characterization of encystment-induced glucosamine 6-phosphate isomerase in Giardia. Mol. Biochem. Parasitol. 84, 149–153.

Sterling, C., Arrowood, M., 1993. Cryptosporidia. In: Kreier, J. (Ed.), Parasitic Protozoa. Academic Press, New York, pp. 159–225. Sudoh, M., Yamazaki, T., Masubuchi, K., et al., 2000. Identification of a novel inhibitor specific to the fungal chitin synthase. J. Biol. Chem. 275, 32901–32905. Sun, C.H., Palm, D., McArthur, A.G., et al., 2002. A novel Myb-related protein involved in transcriptional activation of encystation genes in Giardia lamblia. Mol. Microbiol. 46 (4), 971–984. Sun, C.-H., McCaffery, J.M., Reiner, D.S., et al., 2003. Mining the Giardia lamblia genome for new cyst wall proteins. J. Biol. Chem. 278 (24), 21701–21708. Thoden, J.B., Wohlers, T.M., Fridovich-Keil, J.L., et al., 2001. Human UDP-galactose 4-epimerase. J. Biol. Chem. 276 (18), 15131–15136. Touz, M.C., Gottig, N., Nash, T.E., et al., 2002a. Identification and characterization of a novel secretory granule calcium-binding protein from the early branching eukaryote Giardia lamblia. J. Biol. Chem. 277 (52), 50557–50563. Touz, M.C., Nores, M.J., Slavin, I., et al., 2002b. The activity of a developmentally regulated cysteine proteinase is required for cyst wall formation in the primitive eukaryote Giardia lamblia. J. Biol. Chem. 277 (10), 8474–8481. Upcroft, P., Upcroft, J.A., 2001. Drug targets and mechanisms of resistance in the anaerobic protozoa. Clin. Microbiol. Rev. 14 (1), 150–164. van Keulen, H., Steimle, P.A., Bulik, D., et al., 1998. Cloning of two Giardia glucosamine 6-phosphate isomerase genes only one of which is transcriptionally activated during encystment. J. Eukaryot. Microbiol. 45, 637–642. Walsh, T.J., Giri, N., 1997. Pradimicins: a novel class of broad-spectrum antifungal compounds. Eur. J. Clin. Microbiol. Infect. Dis. 16, 93– 97. Yang, H., Chung, H.J., Yong, T., et al., 2003. Identification of an encystation-specific transcriptional factor, Myb protein in Giardia lamblia. Mol. Biochem. Parasitol. 128, 167–174.