Portal de Periódicos da CAPES

Sorting of Encystation-specific Cysteine Protease to Lysosome-like Peripheral Vacuoles in Giardia lambliaRequires a Conserved Tyrosine-based Motif

2003; Elsevier BV; Volume: 278; Issue: 8 Linguagem: Inglês

10.1074/jbc.m208354200

1083-351X

Marı́a C. Touz, Hugo D. Luján, Stanley F. Hayes, Theodore E. Nash,

Legionella and Acanthamoeba research

Encystation-specific cysteine protease (ESCP) was the first membrane-associated protein described to be part of the lysosome-like peripheral vacuoles in the intestinal parasiteGiardia lamblia. ESCP is homologous to cathepsin C enzymes of higher eukaryotes, but is distinguished from other lysosomal cysteine proteases because it possesses a transmembrane domain and a short cytoplasmic tail. Tyrosine-based motifs within tails of membrane proteins are known to participate in endosomal/lysosomal protein sorting in higher eukaryotes. In this study, we show that a YRPI motif within the ESCP cytoplasmic tail is necessary and sufficient to mediate ESCP sorting to peripheral vacuoles in Giardia. Deletion and point mutation analysis demonstrated that the tyrosine residue is critical for ESCP sorting, whereas amino acids located at the Y+1 (Arg), Y+2 (Pro), and Y+3 (Ile) positions show minimal effect. Loss of the motif resulted in surface localization, whereas addition of the motif to a variant-specific surface protein resulted in lysosomal localization. Although Giardia trophozoites lack a morphologically discernible Golgi apparatus, our findings indicate that this parasite directs proteins to the lysosomes using a conserved sorting signal similar to that used by yeast and mammalian cells. Because Giardia is one of the earliest branching protist, these results demonstrate that sorting motifs for specific protein traffic developed very early during eukaryotic evolution. Encystation-specific cysteine protease (ESCP) was the first membrane-associated protein described to be part of the lysosome-like peripheral vacuoles in the intestinal parasiteGiardia lamblia. ESCP is homologous to cathepsin C enzymes of higher eukaryotes, but is distinguished from other lysosomal cysteine proteases because it possesses a transmembrane domain and a short cytoplasmic tail. Tyrosine-based motifs within tails of membrane proteins are known to participate in endosomal/lysosomal protein sorting in higher eukaryotes. In this study, we show that a YRPI motif within the ESCP cytoplasmic tail is necessary and sufficient to mediate ESCP sorting to peripheral vacuoles in Giardia. Deletion and point mutation analysis demonstrated that the tyrosine residue is critical for ESCP sorting, whereas amino acids located at the Y+1 (Arg), Y+2 (Pro), and Y+3 (Ile) positions show minimal effect. Loss of the motif resulted in surface localization, whereas addition of the motif to a variant-specific surface protein resulted in lysosomal localization. Although Giardia trophozoites lack a morphologically discernible Golgi apparatus, our findings indicate that this parasite directs proteins to the lysosomes using a conserved sorting signal similar to that used by yeast and mammalian cells. Because Giardia is one of the earliest branching protist, these results demonstrate that sorting motifs for specific protein traffic developed very early during eukaryotic evolution. Lysosomes are membrane-bound acidic organelles involved in degradation of endogenous and exogenous macromolecules via biosynthetic or endocytic pathways, respectively (1Schekman R. Orci L. Science. 1996; 271: 1526-1533Google Scholar, 2Hunziker W. Geuze H.J. Bioessays. 1996; 18: 379-389Google Scholar). In mammalian cells, trafficking between the trans-Golgi network (TGN), 1The abbreviations used are: TGN, trans-Golgi network; PV, peripheral vacuole; AP, adaptor protein; ESCP, encystation-specific cysteine protease; HA, hemagglutinin; VSP, variant-specific surface protein; mAb, monoclonal antibody; PBS, phosphate-buffered saline; CWP, cyst wall protein; ESVs, encystation-specific secretory vesicles 1The abbreviations used are: TGN, trans-Golgi network; PV, peripheral vacuole; AP, adaptor protein; ESCP, encystation-specific cysteine protease; HA, hemagglutinin; VSP, variant-specific surface protein; mAb, monoclonal antibody; PBS, phosphate-buffered saline; CWP, cyst wall protein; ESVs, encystation-specific secretory vesicles endosomes, and lysosomes involves several pathways. Mannose 6-phosphate receptors, TGN-38, furin, sortilin, and other proteins are cycled between the TGN and endosomes without ever reaching the lysosomes. Soluble hydrolases bind to mannose 6-phosphate receptors in the TGN by 6-phosphomannosyl residues and travel to endosomes, where they dissociate from their receptor and subsequently reach lysosomes (3Le Borgne R. Hoflack B. Biochim. Biophys. Acta. 1998; 1404: 195-209Google Scholar, 4Conibear E. Stevens T.H. Biochim. Biophys. Acta. 1998; 1404: 211-230Google Scholar). Structural lysosome-associated membrane proteins are sorted from the TGN to the lysosomes through endosomes by way of tyrosine-based motifs. In addition, other proteins are transported directly to the lysosomes without trafficking through endosomes (5Gu F. Crump C.M. Thomas G. Cell. Mol. Life Sci. 2001; 58: 1067-1084Google Scholar). Yeast, unlike mammalian cells, contains an endosomal or prevacuolar compartment and a large vacuole that functions like a lysosome. Carboxypeptidase Y is transported to the yeast vacuole by its receptor, Vps10p, which returns to the Golgi by the yeast retromer complex. Alternatively, yeast alkaline phosphatase is transported to the vacuole by a different mechanism that avoids the yeast prevacuolar compartment (5Gu F. Crump C.M. Thomas G. Cell. Mol. Life Sci. 2001; 58: 1067-1084Google Scholar). In yeast and mammalian cells, a clear distinction between early/late endosomes and lysosomes has been established. In contrast,Giardia lamblia possesses peripheral vacuoles (PVs) located underneath the plasma membrane that function as endosomes and lysosomes and are therefore considered a primitive endosomal/lysosomal complex (6Lanfredi-Rangel A. Attias M. de Carvalho T.M. Kattenbach W.M. De Souza W. J. Struct. Biol. 1998; 123: 225-235Google Scholar). These vacuoles, also called peripheral vesicles, are acidic organelles because they can be labeled with lysosomal markers like acridine orange (6Lanfredi-Rangel A. Attias M. de Carvalho T.M. Kattenbach W.M. De Souza W. J. Struct. Biol. 1998; 123: 225-235Google Scholar) and LysoSensor (7Touz M.C. Nores M.J. Slavin I. Carmona C. Conrad J.T. Mowatt M.R. Nash T.E. Coronel C.E. Luján H.D. J. Biol. Chem. 2002; 277: 8474-8481Google Scholar). PVs can take up horseradish peroxidase without delivering it to any other subcellular compartment, suggesting that PVs may be a unique endocytic compartment (6Lanfredi-Rangel A. Attias M. de Carvalho T.M. Kattenbach W.M. De Souza W. J. Struct. Biol. 1998; 123: 225-235Google Scholar). In addition, numerous soluble enzymes such as acid phosphatase, cathepsins B, and RNase are also present in these vacuoles, indicating their lysosomal nature (8Feely D.E. Dyer J.K. J. Protozool. 1987; 34: 80-83Google Scholar, 9Lindmark D.G. Exp. Parasitol. 1988; 65: 141-147Google Scholar, 10Ward W. Alvarado L. Rawlings N.D. Engel J.C. Franklin C. McKerrow J.H. Cell. 1997; 89: 437-444Google Scholar); nevertheless, no receptors involved in the sorting of any of these enzymes have yet been described. Interestingly, PVs are important not only for food degradation, but also for completion of the life cycle of this intestinal parasite.Giardia cycles between the disease-causing flagellated trophozoite and the environmentally resistant cyst, which is released with feces and is responsible for transmission of the disease (11Adam R.D. Clin. Microbiol. Rev. 2001; 14: 447-475Google Scholar). The participation of PVs has also been described to influence secretory granule discharge during cyst wall formation (12McCaffery J.M. Faubert G.M. Gillin F.D. Exp. Parasitol. 1994; 79: 236-249Google Scholar), and PVs act as secretory organelles that release cyst wall-disrupting enzymes during excystation (10Ward W. Alvarado L. Rawlings N.D. Engel J.C. Franklin C. McKerrow J.H. Cell. 1997; 89: 437-444Google Scholar, 13Slavin I. Saura A. Carranza P.G. Touz M.C. Nores M.J. Luján H.D. Mol. Biochem. Parasitol. 2002; 122: 95-98Google Scholar). In mammalian cells as well as in yeast, specific sorting signals direct transmembrane proteins to endosomes and/or lysosomes, either from the TGN or from the cell surface, and involve tyrosine-based motifs (YXXφ, where X is any amino acid and φ is an amino acid with a bulky hydrophobic side chain) and/or acidic cluster dileucine motifs located in their cytoplasmic tails. These motifs can be found in single or multiple copies and also in combination (14Dell'Angelica E.C. Payne G.S. Cell. 2001; 106: 395-398Google Scholar). The interaction of proteins carrying these motifs with adaptor proteins (APs) and GGA (Golgi-localized,gamma-ear-containing ADP-ribosylation factor-binding) proteins seems to be critical for endosomal/lysosomal protein transport (14Dell'Angelica E.C. Payne G.S. Cell. 2001; 106: 395-398Google Scholar, 15Bonifacino J.S. Dell'Angelica E.C. J. Cell Biol. 1999; 145: 923-926Google Scholar). We recently reported that a cysteine protease of the cathepsin C family is implicated in the processing of a cyst wall protein during encystation and that this enzyme localizes to the peripheral vacuoles of non-encysting Giardia trophozoites (7Touz M.C. Nores M.J. Slavin I. Carmona C. Conrad J.T. Mowatt M.R. Nash T.E. Coronel C.E. Luján H.D. J. Biol. Chem. 2002; 277: 8474-8481Google Scholar). This encystation-specific cysteine protease (ESCP) possesses a transmembrane domain and a 12-amino acid cytoplasmic tail, unlike cathepsin C enzymes from higher eukaryotes. In the present study, we show that a tyrosine-based sorting signal (YRPI) within the ESCP cytoplasmic tail functions in the sorting of ESCP to lysosome-like peripheral vacuoles in Giardia. To constitutively express ESCP along with three influenza hemagglutinin (HA) epitopes (YPYDVPDYAYPYDVPDYAYPYDVPDYA) at the C terminus, the plasmid pTubH7pac carrying the variant-specific surface protein gene vsph7(16Elmendorf H.G. Singer S.M. Pierce J. Cowan J. Nash T.E. Mol. Biochem. Parasitol. 2001; 113: 157-169Google Scholar) was modified to introduce the tag just before the TGA stop codon and an ApaI site immediately following the vsph7ATG start codon. First, one round of PCR was performed using sense oligonucleotide 5′-ggtacgcgtacccctacgatgtaccagactatgcatagggatccgacttaggtagtaaacgtcatggt-3′ and antisense oligonucleotide 5′-tcgacgcgtaatcgggaacatcatacggataagcgtagtcaggcacatcatatggatagatatccgccttcc cgcggcagacgaacca-3′ (with the MluI site in boldface, the BamHI site underlined, and the EcoRV site in italics), which were restricted with MluI and ligated together. Another round of PCR using the same strategy allowed the insertion of an ApaI site using sense 5′-cctGGGCCCttaattaattgcctaatagcaagcact-3′ and antisense 5′-taaGGGCCCcatggttttatttccgcccgtccagact-3′ primers (with theApaI site in uppercase), resulting in pTubH7HApac. To exchange the escp gene for the vsph7 gene, pTubH7HApac was digested with ApaI and EcoRV to release vsph7. The DNA fragment corresponding to the entireescp coding region was amplified from Giardiagenomic DNA (isolate WB/clone 1267) by PCR using sense 5′-ctgGGGCCCcttttcatcttggcgctcctg-3′ and antisense 5′-taagatatctgcaattattggacggtattt-3′ primers carryingApaI and EcoRV sites, respectively; digested; and ligated into the restricted vector, resulting in pTubESCPHApac. A site-directed mutagenesis kit (QuikChange, Stratagene) was employed to construct deletion/mutations of ESCP inside pTubESCPHApac using two complementary mutagenic oligonucleotide primers based on escp sequence reported previously (GenBankTM/EBI accession numberAF293408) (7Touz M.C. Nores M.J. Slavin I. Carmona C. Conrad J.T. Mowatt M.R. Nash T.E. Coronel C.E. Luján H.D. J. Biol. Chem. 2002; 277: 8474-8481Google Scholar). For ΔK/A, a second EcoRV site was introduced at the end of the sequence coding for the ESCP transmembrane domain. The plasmid was restricted with EcoRV and ligated together, thereby eliminating the sequence corresponding to the ESCP cytoplasmic tail. For ΔK/K, two complementary primers were designed to omit bases 1588–1606. For ΔY/A, the same strategy was followed, introducing a deletion of bases 1606–1624 (see Fig. 2 A). For the ΔYRPI, ΔY, ΔRP, and ΔI point mutations, the corresponding amino acids were replaced with alanine residues (see Fig.4 A). ESCP variants were confirmed by sequencing using dye terminator cycle sequencing (Beckman Coulter).Figure 4The tyrosine-based motif in the ESCP cytoplasmic tail is essential for lysosomal localization. A, shown is an illustration of ESCP tyrosine-based motif point mutations. The tyrosine-based motif (YRPI) is show in red, and alanine substitutions are underlined. The signal peptide and the propeptide were omitted in these illustrations. 3xHA, three HA epitopes. B, point mutations of the tyrosine-based motif (ΔYRPI) and tyrosine (ΔY) changed the enzyme location to the plasma membrane. Replacement of residue Y+3 (ΔI) localized ESCP to the PVs and to the surface, suggesting a moderate effect on lysosomal sorting. Expression was determined using anti-HA mAb. 4,6-diamidino-2-phenylindole stained the Giardia nuclei. Magnification is ×630.View Large Image Figure ViewerDownload (PPT) VSPH7-HA, ΔH7, and ΔH7-AA were used as controls (see Fig. 5 A). The last two constructs were made following the same strategy described for ESCP deletions and mutations. The chimeras cH7TM and cH7CT (see Fig. 6 A) were generated by PCR using primers that have complementary sequences to the ESCP transmembrane domain and cytoplasmic tail, respectively, following the protocol described by Geiser et al. (17Geiser M. Cebe R. Drewello D. Schmitz R. BioTechniques. 2001; 31: 88-90Google Scholar). The correct sequences of all constructs were verified by sequencing.Figure 6VSPH7 carrying the ESCP cytoplasmic tail is displaced from the surface to the peripheral vesicles. A, shown is an illustration of ESCP and tagged VSPH7 chimeras.Bars denote different domains of ESCP and VSPH7-HA. The tyrosine-based motif (YRPI) of ESCP is show in red. 3xHA, three HA epitopes. B, the localization of cH7TM, in which VSPH7 has the transmembrane domain of ESCP, remained on the surface, covering the entire parasite surface including the flagella. C, cH7CT, in which VSPH7-HA carries the ESCP cytoplasmic tail, relocated to the PVs close to the plasma membrane. Chimeras were detected using anti-HA mAb. PC, phase-contrast. Magnification is ×630.View Large Image Figure ViewerDownload (PPT) Trophozoites of isolate WB/clone 1267 (18Nash T.E. Aggarwal A. Adam R.D. Conrad J.T. Merritt Jr., J.W. J. Immunol. 1988; 141: 636-641Google Scholar) were cultured as described (19Keister D.B. Trans. R Soc. Trop. Med. Hyg. 1983; 77: 487-488Google Scholar). Encystation of trophozoite monolayers was accomplished following the method described by Boucher and Gillin (20Boucher S.E. Gillin F.D. Infect. Immun. 1990; 58: 3516-3522Google Scholar). Trophozoites were transfected with the constructs by electroporation and selected by puromycin as previously described (16Elmendorf H.G. Singer S.M. Pierce J. Cowan J. Nash T.E. Mol. Biochem. Parasitol. 2001; 113: 157-169Google Scholar, 21Singer S.M. Yee J. Nash T.E. Mol. Biochem. Parasitol. 1998; 92: 59-69Google Scholar, 22Yee J. Nash T.E. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 5615-5619Google Scholar). For fixed cells, trophozoites cultured in growth medium were harvested and processed as described previously (23Luján H.D. Mowatt M.R. Conrad J.T. Bowers B. Nash T.E. J. Biol. Chem. 1995; 270: 29307-29313Google Scholar). Primary anti-HA mAb (Sigma) was used to detect ESCP and ESCP variants, and anti-HA mAb or VSPH7-specific mAb G10/4 (27Nash T.E. Conrad J.T. Mowatt M.R. J. Eukaryotic Microbiol. 1995; 42: 604-609Google Scholar) was used to detect VSPH7 and VSPH7 variants. For assays of viable trophozoites, the cells were washed twice with PBS and 0.1% growth medium and incubated with the specific mAb G10/4 for VSPH7 surface localization. For CWP2 localization, mAb 7D2 (30Mowatt M.R. Luján H.D. Cotten D.B. Bowers B. Yee J. Nash T.E. Stibbs H.H. Mol. Microbiol. 1995; 15: 955-963Google Scholar) was directly labeled with Texas Red (Molecular Probes, Inc., Eugene, OR) following the manufacturer's instructions and used in encysting trophozoites transfected with pTubESCPHApac. The specimens were examined with a Zeiss Axioplan fluorescence microscope and/or a Leica TCS-NT/SP confocal microscope. Controls included omission of primary antibody and staining of untransfected cells. For PV localization,Giardia trophozoites were incubated at 37 °C for 3 h in growth medium containing 70 nm LysoTrackerTMRed DND-99 (Molecular Probes, Inc.) as suggested by the manufacturer, chilled, washed twice with PBS and 0.1% growth medium, and attached to coverslips for 30 min at 37 °C. After 40 min of fixation with 4% formaldehyde, the cells were blocked with 10% normal goat serum in PBS containing 0.1% Triton X-100 for 30 min and then incubated for 1 h with anti-HA antibody in PBS containing 3% normal goat serum and 0.1% Triton X-100. Fluorescein-conjugated anti-mouse IgG secondary antibody (Cappel, West Chester, PA) was used to reveal the labeling patterns. Western blotting was performed as previously reported (23Luján H.D. Mowatt M.R. Conrad J.T. Bowers B. Nash T.E. J. Biol. Chem. 1995; 270: 29307-29313Google Scholar). Briefly, 10 μg of total protein/lane from transfected non-encysting trophozoites were resuspended in 30 μl of Laemmli sample buffer (Bio-Rad) with 2-mercaptoethanol, boiled for 5 min, and electrophoresed on a 4–12% Tris/glycine-polyacrylamide gel. The proteins were transblotted onto polyvinylidene difluoride membranes (Invitrogen) and probed with anti-HA antibody (1:2000 dilution). Encysting Giardiatrophozoites were rinsed twice with PBS and 0.1% growth medium; chilled; attached to Thermanox coverslips (Nunc, Naperville, IL); and processed as described previously (24Hayman J.R. Hayes S.F. Amon J. Nash T.E. Infect. Immun. 2001; 69: 7057-7066Google Scholar), except that the primary antibody for tagged ESCP was anti-HA mAb diluted 1:1000. HA-tagged ESCP was constitutively expressed in WB/1267 trophozoites. With anti-HA mAb, ESCP showed a PV localization pattern and colocalized with LysoTracker (Fig.1 and Supplemental Fig. 1), a probe for acidic organelles in living cells (25Wubbolts R. Fernandez-Borja M. Oomen L. Verwoerd D. Janssen H. Calafat J. Tulp A. Dusseljee S. Neefjes J. J. Cell Biol. 1996; 135: 611-622Google Scholar). To verify that the HA tag does not interfere with ESCP sorting, tagged variants of ESCP carrying V5 and FLAG epitopes as well as green fluorescent protein were expressed and localized to the PVs by immunofluorescence assays using specific mAbs (data not shown). In addition, acid phosphatase (GenBankTM/EBI accession number AAK97085) and the variant-specific surface protein VSPH7 tagged with HA localized to the PVs 2M. C. Touz and T. E. Nash, unpublished data. and the plasma membrane, respectively, indicating that the HA tag at the C terminus does not influence protein trafficking (see below). To analyze whether the YRPI motif located in the cytoplasmic tail of ESCP determines its localization to the PVs, we constructed a series of variants of this enzyme by deletions and mutations (see "Experimental Procedures" and Figs. 2 A and4 A). The sorting of these variants was examined by immunofluorescence assays using anti-HA mAb. As was previously documented, ESCP localizes to the PVs in Giardia (7Touz M.C. Nores M.J. Slavin I. Carmona C. Conrad J.T. Mowatt M.R. Nash T.E. Coronel C.E. Luján H.D. J. Biol. Chem. 2002; 277: 8474-8481Google Scholar), but the truncated version ΔK/A showed surface localization indicated by staining of the trophozoite surface and flagella (Fig. 2 B and Supplemental Fig. 2). Expression of ΔK/K, which still has the YXXφ motif, resulted in no change in ESCP localization (Fig. 2 B). In contrast, ΔY/A, which lacks the YRPIIA sequence, relocated the enzyme to the plasma membrane (Fig.2 B). Western blot analysis of total protein extracted from transfected trophozoites confirmed the expression of ESCP and its variants. In every case, 65- and 45-kDa bands corresponding to the immature and mature forms of ESCP, respectively, were observed (Fig.3) (7Touz M.C. Nores M.J. Slavin I. Carmona C. Conrad J.T. Mowatt M.R. Nash T.E. Coronel C.E. Luján H.D. J. Biol. Chem. 2002; 277: 8474-8481Google Scholar). These results prompted us to perform a more detailed analysis of the YRPI sorting signal because only the construct lacking this motif failed to localize ESCP to peripheral vacuoles. Mutation of YRPI (ΔYRPI) to alanine residues resulted in missorting of the protein to the plasma membrane, showing that this motif is essential for ESCP localization (Fig. 4,A and B). Exchanging tyrosine (ΔY) with alanine localized ΔY to the surface, whereas replacement of the residue that follow tyrosine at position +3 (ΔI) had an intermediate effect because the enzyme was detected both at the surface and in the PVs (Fig. 4, A and B). However, when Arg and Pro were replaced, the enzyme remained in the PVs (data not shown). These results indicate that the tyrosine within the cytoplasmic tail is critical for ESCP peripheral vacuole localization. VSPH7 is a variant-specific surface protein ofGiardia clone GS/M-H7 that possesses a single transmembrane domain and a conserved CRGKA cytoplasmic tail and that covers the entire cell surface, including the flagella (26Nash T.E. Mowatt M.R. Mol. Biochem. Parasitol. 1992; 51: 219-227Google Scholar). VSPH7 is not expressed in Giardia clone WB/1267, allowing detection of VSPH7 at the surface of transfected WB trophozoites using VSPH7-specific mAb G10/4 (16Elmendorf H.G. Singer S.M. Pierce J. Cowan J. Nash T.E. Mol. Biochem. Parasitol. 2001; 113: 157-169Google Scholar, 27Nash T.E. Conrad J.T. Mowatt M.R. J. Eukaryotic Microbiol. 1995; 42: 604-609Google Scholar). First, it was important to determine whether VSPH7 does have also a sorting motif for its plasma membrane localization. In this way, expression of VSPH7-HA (VSPH7 with an HA tag at its C terminus), ΔH7 (tagged VSPH7 without its cytoplasmic tail), and ΔH7-AA (tagged VSPH7 with the amino acids in its tail changed to alanine residues) showed the same localization profile on the surface of trophozoites compared with expression of native VSPH7, as determined using either anti-HA mAb (Fig. 5,A and B) or mAb G10/4 (data not shown). Thus, these result shows that the HA epitope does not affect VSPH7 localization and that the conserved CRGKA cytoplasmic tail is not involved in VSPH7 plasma membrane sorting. To analyze whether the ESCP cytoplasmic tail can modify VSPH7 sorting, two different chimeras were expressed in WB/1267 trophozoites: VSPH7-HA possessing the transmembrane domain of ESCP (cH7TM) and VSPH7-HA possessing the ESCP cytoplasmic tail instead of its own conserved tail (cH7CT) (Fig. 6 A). Expression of cH7TM resulted in no change in localization because the chimera remained in the plasma membrane (Fig. 6 B). Immunofluorescence assay using G10/4, a mAb that recognizes the VSPH7 extracellular domain, confirmed the surface localization of cH7TM in viable trophozoites (Supplemental Fig. 3). These results indicate that the 27-amino acid transmembrane domain of VSPH7 does not contain a specific signal for surface localization and that a protein carrying the 24-amino acid ESCP transmembrane domain is transported to the plasma membrane. In contrast, cH7CT was localized to the PVs, the same subcellular localization as ESCP (Fig. 6 C). Moreover, in viable cells, mAb G10/4 failed to detect cH7CT at the surface of the trophozoite (Supplemental Fig. 3). Taken together, these results show that the cytoplasmic tail of ESCP has all the information necessary to direct proteins to Giardia peripheral vesicles. Furthermore, these findings suggest that a long transmembrane domain is essential for VSPs to be transported to the plasma membrane (see "Discussion"). ESCP expression and activity increase during encystation and are involved in the processing of one of the proteins forming the cyst wall (CWP2) (7Touz M.C. Nores M.J. Slavin I. Carmona C. Conrad J.T. Mowatt M.R. Nash T.E. Coronel C.E. Luján H.D. J. Biol. Chem. 2002; 277: 8474-8481Google Scholar). Although CWP2 and ESCP colocalize in encysting cells, how and where this interaction takes place are unknown (7Touz M.C. Nores M.J. Slavin I. Carmona C. Conrad J.T. Mowatt M.R. Nash T.E. Coronel C.E. Luján H.D. J. Biol. Chem. 2002; 277: 8474-8481Google Scholar). Here, we analyzed ESCP/CWP2 interaction during encystation by immunofluorescence and ESCP subcellular localization by immunoelectron microscopy in encysting trophozoites. Using directly labeled anti-CWP2 mAb and anti-HA mAb for the detection of ESCP, we found that, at the beginning of encystation, ESCP was in the PVs close to the encysting trophozoite plasma membrane, whereas CWP2 was detected in the ESVs (Fig.7 A). In contrast, during cyst wall formation, both colocalized in the developing cyst wall and in the cyst wall in mature cysts (Fig. 7 A). Immunoelectron microscopy showed ESCP in ESVs as well as on the surface and in the PVs of encysting cells (Fig. 7 B and Supplemental Fig. 4). The fact that ESCP could be found inside ESVs close to peripheral vacuoles and its localization at the surface suggest that ESVs interact with PVs during ESV discharge and/or at the time of release onto the surface of encysting trophozoites. In higher eukaryotes, the endoplasmic reticulum and the Golgi complex play a central role in the correct protein folding and transport. Proteins transported to endosomes or lysosomes are generally sorted away from the trafficking pathway taken by secretory proteins and are instead targeted to the endocytic compartments (28Traub L.M. Kornfeld S. Curr. Opin. Cell Biol. 1997; 9: 527-533Google Scholar).Giardia does have an endomembranous system that differs from that of higher eukaryotes (29Adam R.D. Int. J. Parasitol. 2000; 30: 475-484Google Scholar). Giardia lacks organelles that resemble early and late endosomes and instead has peripheral vacuoles with hydrolytic activity. The property of these organelles to accumulate macromolecules and, at the same time, the presence of lysosome-like soluble hydrolases suggest that this parasite possesses an endosomal/lysosomal system represented in this single organelle (6Lanfredi-Rangel A. Attias M. de Carvalho T.M. Kattenbach W.M. De Souza W. J. Struct. Biol. 1998; 123: 225-235Google Scholar). In addition, these vacuoles seem to perform multiple cellular functions because they also act as secretory organelles at certain points of theGiardia life cycle (7Touz M.C. Nores M.J. Slavin I. Carmona C. Conrad J.T. Mowatt M.R. Nash T.E. Coronel C.E. Luján H.D. J. Biol. Chem. 2002; 277: 8474-8481Google Scholar, 13Slavin I. Saura A. Carranza P.G. Touz M.C. Nores M.J. Luján H.D. Mol. Biochem. Parasitol. 2002; 122: 95-98Google Scholar). Despite these differences, in a number of ways, protein transport in Giardia resembles that in higher eukaryotes. One example is the constitutive secretion of VSPs (26Nash T.E. Mowatt M.R. Mol. Biochem. Parasitol. 1992; 51: 219-227Google Scholar, 27Nash T.E. Conrad J.T. Mowatt M.R. J. Eukaryotic Microbiol. 1995; 42: 604-609Google Scholar) and the regulated secretion of CWPs (23Luján H.D. Mowatt M.R. Conrad J.T. Bowers B. Nash T.E. J. Biol. Chem. 1995; 270: 29307-29313Google Scholar, 30Mowatt M.R. Luján H.D. Cotten D.B. Bowers B. Yee J. Nash T.E. Stibbs H.H. Mol. Microbiol. 1995; 15: 955-963Google Scholar, 31Luján H.D. Mowatt M.R. Nash T.E. Microbiol. Mol. Biol. Rev. 1997; 61: 294-304Google Scholar). In addition, signal peptides target VSPs and CWPs through the secretory pathway inGiardia (26Nash T.E. Mowatt M.R. Mol. Biochem. Parasitol. 1992; 51: 219-227Google Scholar, 27Nash T.E. Conrad J.T. Mowatt M.R. J. Eukaryotic Microbiol. 1995; 42: 604-609Google Scholar), and conserved motifs such as the BiP chaperone/endoplasmic reticulum retention motif (KDEL) are present in this parasite (32Luján H.D. Mowatt M.R. Conrad J.T. Nash T.E. Biol. Cell. 1996; 86: 11-18Google Scholar). The tyrosine-based motif (YRPI) involved in ESCP transport to the PVs in Giardia is another example of similarity of secretory mechanisms to more evolved cells. In higher eukaryotes, a tyrosine-based signal defines a motif that has the consensus YXXφ or NPXY (33Marks M.S. Woodruff L. Ohno H. Bonifacino J.S. J. Cell Biol. 1996; 135: 341-354Google Scholar). In vitro analyses have shown that this motif interacts with the μ-subunits of almost all APs described so far (14Dell'Angelica E.C. Payne G.S. Cell. 2001; 106: 395-398Google Scholar, 15Bonifacino J.S. Dell'Angelica E.C. J. Cell Biol. 1999; 145: 923-926Google Scholar). Despite that each μ-subunit has a preference for 1 amino acid at the X position favoring a nonpolar, an arginine-rich, and an acidic amino acid for AP1, AP2, and AP3, respectively, there is also an overlapping specificity (YIPL) among them (34Ohno H. Aguilar R.C. Yeh D. Taura D. Saito T. Bonifacino J.S. J. Biol. Chem. 1998; 273: 25915-25921Google Scholar). Furthermore, AP μ-subunits also have predilections for the φ position, preferring leucine and isoleucine over other hydrophobic amino acids (34Ohno H. Aguilar R.C. Yeh D. Taura D. Saito T. Bonifacino J.S. J. Biol. Chem. 1998; 273: 25915-25921Google Scholar). In the case ofGiardia, the YRPI motif within the ESCP cytoplasmic tail appears to be a putative adaptor-binding domain because it has a proline and an isoleucine at the Y+2 and Y+3 positions, respectively. The exchange of YRPI for alanine residues altered the localization of ESCP from the PVs to the plasma membrane. Moreover, recognition by tyrosine plays a major role in ESCP localization because its replacement was sufficient to relocate the enzyme to other cellular organelles. Point mutation of residues Y+1/Y+2 and Y+3 showed that only the isoleucine at position +3 has a moderate effect on ESCP subcellular localization. It is possible that, as was described for other proteins (15Bonifacino J.S. Dell'Angelica E.C. J. Cell Biol. 1999; 145: 923-926Google Scholar), residues Y+2 and Y+3 may help expose the tyrosine residue to the adaptor subunit, rather than being involved in adaptor recognition. To better understand protein sorting signals in the primitive eukaryoteG. lamblia, we performed additional experiments using a type I membrane protein, the variant-specific surface protein VSPH7. When VSPH7 (VSP of Giardia clone GS/M) is expressed inGiardia clone WB/1267, it shows a surface pattern (16Elmendorf H.G. Singer S.M. Pierce J. Cowan J. Nash T.E. Mol. Biochem. Parasitol. 2001; 113: 157-169Google Scholar). Like all VSPs described so far (26Nash T.E. Mowatt M.R. Mol. Biochem. Parasitol. 1992; 51: 219-227Google Scholar, 27Nash T.E. Conrad J.T. Mowatt M.R. J. Eukaryotic Microbiol. 1995; 42: 604-609Google Scholar, 35Papanastasiou P. Bruderer T. Li Y. Bommeli C. Kohler P. Mol. Biochem. Parasitol. 1997; 86: 13-27Google Scholar, 36Papanastasiou P. McConville M.J. Ralton J. Kohler P. Biochem. J. 1997; 322: 49-56Google Scholar, 37Luján H.D. Mowatt M.R. Wu J.J. Lu Y. Lees A. Chance M.R. Nash T.E. J. Biol. Chem. 1995; 270: 13807-13813Google Scholar, 38Nash T.E. Luján H.T. Mowatt M.R. Conrad J.T. Infect. Immun. 2001; 69: 1922-1923Google Scholar), VSPH7 possesses a 27-amino acid transmembrane domain and a conserved 5-amino acid cytoplasmic tail (CRGKA). When the 24-residues ESCP transmembrane segment replaced the VSPH7 transmembrane domain, the localization of the VSPH7 chimera remained unchanged. In addition, when VSPH7 lacking the cytoplasmic tail was expressed, the protein appeared on the surface, similar to when the CRGKA tail of VSPH7 was exchanged for 5 alanine residues. These results are consistent with the hypothesis that the length of the transmembrane domain is critical for protein localization, supporting the model wherein short transmembrane domains (≤17 residues) direct proteins to the endoplasmic reticulum and cis-Golgi, and proteins with long transmembrane domains (≥23 residues) direct proteins to the plasma membrane (39Honsho M. Mitoma J.Y. Ito A. J. Biol. Chem. 1998; 273: 20860-20866Google Scholar, 40Yang M. Ellenberg J. Bonifacino J.S. Weissman A.M. J. Biol. Chem. 1997; 272: 1970-1975Google Scholar, 41Itin C. Schindler R. Hauri H.P. J. Cell Biol. 1995; 131: 57-67Google Scholar, 42Itin C. Foguet M. Kappeler F. Klumperman J. Hauri H.P. Biochem. Soc. Trans. 1995; 23: 541-544Google Scholar, 43Watson R.T. Pessin J.E. Am. J. Physiol. 2001; 281: C215-C223Google Scholar). Transmembrane proteins inGiardia seem to follow the same criteria. Similar to VSPs, other proteins such as dipeptidyl peptidase IV (44Touz M.C. Nores M.J. Slavin I. Piacenza L. Acosta D. Carmona C. Luján H.D. Biochem. J. 2002; 364: 703-710Google Scholar) and syntaxin-1 (GenBankTM/EBI accession numberAF293409), 2M. C. Touz and T. E. Nash, unpublished data., 3M. C. Touz, M. J. Nores, N. Gottig, and H. D. Luján, unpublished data. which have long transmembrane domains, also localized to the plasma membrane, whereas syntaxin-2 (accession number AF293410), a protein with a short transmembrane domain, localized at the Golgi of encysting trophozoites. 4M. C. Touz, H. D. Luján, and T. E. Nash, unpublished data. In addition, a recent report showed that when the VSPH7 transmembrane domain and cytoplasmic tail were added to the extracellular domain of a membrane-anchored SAG1 protein of Toxoplasma gondii, the protein was localized to the surface, including the flagella ofGiardia trophozoites (45Marti M. Li Y. Kohler P. Hehl A.B. Infect. Immun. 2002; 70: 1014-1016Google Scholar). This agrees with the idea that the transmembrane domain, but not a motif or special structure inside the VSPH7 extracellular domain, is critical for its localization. Furthermore, we found that the VSPH7 cytoplasmic tail does not contain a trafficking motif, but, because it is highly conserved, may have an additional unknown function. When the cytoplasmic tail of VSPH7 was substituted for the ESCP counterpart, the localization of VSPH7 changed from the cell surface to the PVs. These findings suggest that the transmembrane domain directs the transport of membrane-associated proteins in Giardiaunless they have a sorting signal that specifically routes the protein to another organelle. There are at least two different mechanisms involved in protein trafficking to the PVs in Giardia. This study shows that a conserved tyrosine-based motif in the cytoplasmic tail of ESCP is critical for ESCP localization to the PVs. In contrast, soluble PV proteins such as acid phosphatase and cathepsins B do not contain a tyrosine-based motif. It is possible that these soluble lysosomal proteins, similar to those in the mammalian system, require a receptor-mediated sorting process that involves mannose 6-phosphate receptor- and adaptor-like proteins. Although mannose 6-phosphate receptor-like proteins have not been reported in Giardia, proteins with some homology to the α-subunit (α-adaptin gene, GenBankTM/EBI accession number AF486293) and γ-subunit (γ-adaptin gene, accession number AF486294) as well as the μ-subunit (GiMuA, accession number AAL82729; and GiMuB, accession number AY078978) of putative APs have been identified, supporting this idea. It is also well known that the Giardia secretory system undergoes radical changes during encystation (23Luján H.D. Mowatt M.R. Conrad J.T. Bowers B. Nash T.E. J. Biol. Chem. 1995; 270: 29307-29313Google Scholar, 30Mowatt M.R. Luján H.D. Cotten D.B. Bowers B. Yee J. Nash T.E. Stibbs H.H. Mol. Microbiol. 1995; 15: 955-963Google Scholar, 46Luján H.D. Marotta A. Mowatt M.R. Sciaky N. Lippincott-Schwartz J. Nash T.E. J. Biol. Chem. 1995; 270: 4612-4618Google Scholar, 47Reiner D.S. McCaffery M. Gillin F.D. Eur. J. Cell Biol. 1990; 53: 142-153Google Scholar, 48Reiner D.S. McCaffery J.M. Gillin F.D. Cell Microbiol. 2001; 3: 459-472Google Scholar). The most remarkable events are the presence of a well defined Golgi apparatus and the biogenesis of ESVs that transport newly synthesized CWPs to the plasma membrane for release and cyst wall formation. In the present study, during encystation, ESCP was seen inside ESVs and in the plasma membrane in addition to the PVs. Because ESCP is involved in the processing of CWP2 during cyst wall formation, it is possible that PVs fuse with ESVs where interaction between the enzyme ESCP and the substrate CWP2 takes place. After CWP2 processing, both proteins could then be released by exocytosis in a way that involves a calcium-dependent process (49Touz M.C. Gottig N. Nash T.E. Luján H.D. J. Biol. Chem. 2002; 277: 50557-50563Google Scholar). The findings presented here confirm the previous suggestion that ESVs interact with PVs during the latter stages of encystation (31Luján H.D. Mowatt M.R. Nash T.E. Microbiol. Mol. Biol. Rev. 1997; 61: 294-304Google Scholar). As an early diverging protist, Giardia seems to have a relatively elementary subcellular organization (11Adam R.D. Clin. Microbiol. Rev. 2001; 14: 447-475Google Scholar). In particular, it has one of the most basic systems for protein transport and degradation (11Adam R.D. Clin. Microbiol. Rev. 2001; 14: 447-475Google Scholar). However, it also shares many characteristics with higher eukaryotes, as in the case of conserved sorting motifs. Further studies regarding different sorting signals in Giardia and the molecules interacting with them will provide new insight to better understand the evolution of intracellular protein transport and subcellular organization in eukaryotes and also contribute to defining new targets for therapeutic intervention. We thank Drs. Dennis M. Dwyer and Cecilia Arighi for helpful discussion and John T. Conrad and Liudmila Kulakova for technical suggestions.