Since the Giardia lamblia cyst wall is necessary for survival in the environment and host infection, we tested the hypothesis that it contains proteins other than the three known cyst wall proteins. Serial analysis of gene expression during growth and encystation revealed a gene, “HCNCp” (High Cysteine Non-variant Cyst protein), that was upregulated late in encystation, and that resembled the classic Giardia variable surface proteins (VSPs) that cover the trophozoite plasmalemma. HCNCp is 13.9% cysteine, with many “CxxC” tetrapeptide motifs and a transmembrane sequence near the C-terminus. However, HCNCp has multiple “CxC” motifs rarely found in VSPs, and does not localize to the trophozoite plasmalemma. Moreover, the HCNCp C-terminus differed from the canonical VSP signature. Full-length epitope-tagged HCNCp expressed under its own promoter was upregulated during encystation with highest expression in cysts, including 42 and 21 kDa C-terminal fragments. Tagged HCNCp targeted to the nuclear envelope in trophozoites, and co-localized with cyst proteins to encystation-specific secretory vesicles during encystation. HCNCp defined a novel trafficking pathway as it localized to the wall and body of cysts, while the cyst proteins were exclusively in the wall. Unlike VSPs, HCNCp is expressed in at least five giardial strains and four WB subclones expressing different VSPs. Bioinformatics identified 60 additional large high cysteine membrane proteins (HCMp) containing ≥20 CxxC/CxC's lacking the VSP-specific C-terminal CRGKA. HCMp were absent or rare in other model or parasite genomes, except for Tetrahymena thermophila with 30. MEME analysis classified the 61 gHCMp genes into nine groups with similar internal motifs. Our data suggest that HCNCp is a novel invariant cyst protein belonging to a new HCMp family that is abundant in the Giardia genome. HCNCp and the other HCMp provide a rich source for developing parasite-specific diagnostic reagents, vaccine candidates, and subjects for further research into Giardia biology.
Citation: Davids BJ, Reiner DS, Birkeland SR, Preheim SP, Cipriano MJ, McArthur AG, et al. (2006) A New Family of Giardial Cysteine-Rich Non-VSP Protein Genes and a Novel Cyst Protein. PLoS ONE 1(1): e44. doi:10.1371/journal.pone.0000044
Academic Editor: Berend Snel, Utrecht University, Netherlands
Received: August 7, 2006; Accepted: October 10, 2006; Published: December 2006
Copyright: © 2006 Davids et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: BJD, DSR, and FDG were supported by NIH grants GM61896, AI51687, AI42488, and DK35108. AGM, SPP, SRB, and MJC were supported by NIH grant AI51089 and the Marine Biological Laboratory's Program in Global Infectious Diseases, funded by the Ellison Medical Foundation. Computational resources were provided by the Josephine Bay Paul Center for Comparative Molecular Biology and Evolution (Marine Biological Laboratory) through funds provided by the W.M. Keck Foundation and the G. Unger Vetlesen Foundation.
Competing interests: The authors have declared that no competing interests exist.
Giardiasis is a major contributor to the enormous burden of human diarrheal diseases, which are second only to respiratory infections as causes of mortality and morbidity worldwide –. Although diarrheal diseases rarely kill children in the USA, they remain important causes of infant death in less-developed countries , , . Unlike the rotaviruses, which affect infants, giardiasis can also hamper adults in their productive years. In the USA, G. lamblia has been one of the most frequently identified causes of water-borne intestinal disease, and causes diarrhea in daycare centers and among hikers and campers. Nonetheless, the basic biology and pathogenesis of this parasite remain incompletely understood. Although trophozoites do not invade or secrete any known toxin , , they can cause severe diarrhea that can cause failure of children to thrive. However, about half of infected people lack symptoms and the infection often resolves without treatment. Thus, both the duration and symptoms of giardiasis are highly variable in immunocompetent people for reasons that are not understood.
No giardial virulence factor is known and this protist's ability to survive in diverse, hostile environments may be a key to its pathophysiology. To survive within the small intestine, Giardia trophozoites are covered with a single layer of an unusual Type 1 integral membrane protein that protects them from proteolysis . Normally, each trophozoite expresses only one such surface protein from approximately 150 in the genome and switching of these variant-specific surface proteins (VSPs) – likely protects the trophozoite surface from both immunologic and environmental attack. VSP gene expression can spontaneously switch both in vitro and in vivo ,  every six to thirteen generations . Most isolates are mixed populations of trophozoites expressing different VSPs. Immune  or environmental  stress in vitro can select for or against expression of different VSPs, which are usually not immunologically cross-reactive , suggesting that a major function of antigenic variation is immune evasion. VSPs are extremely (~12%) cysteine-rich, with most Cys residues in a CxxC motif, where x can be any amino acid , . Although VSPs differ in size and sequence, they are characterized by a highly conserved C-terminal membrane spanning region, a hydrophilic cytoplasmic tail with a conserved five amino acid CRGKA signature sequence, and an extended polyadenylation signal , , .
If trophozoites are carried downstream in the intestine, they must encyst in order to survive outside the host . Encystation is characterized by construction of an extracellular cyst wall which protects the parasite from hypotonic lysis in the external environment and from gastric acid when infecting a new host , , . The Giardia' cyst wall is biologically important as a simple model extracellular matrix, with both structural and signal transduction functions , . Encystation contributes to immune evasion  by two distinct pathways: first, the cyst wall covers the trophozoite surface during encystation and sequesters it from host defenses and second, encystation leads to switching of VSPs , , . Despite its importance, the composition, architecture, and formation of the cyst wall are poorly understood. Only three structural cyst wall proteins (CWPs) have been identified –. CWPs 1–3, which do not undergo antigenic variation, have 14–16 Cys residues and similar leucine rich repeats (LRRs). In addition to CWPs, the cyst wall is also composed of fibrils of poly N-acetylgalactosamine –.
The VSPs and CWPs define distinct expression and trafficking patterns –, . A well-studied VSP called TSA 417 was highly expressed during vegetative growth and transported by a conventional ER pathway to the plasmalemma . TSA 417 was downregulated and the protein internalized during encystation. In contrast, CWPs are only expressed during encystation and are exclusively transported to the cyst wall within the matrix of novel regulated encystation secretory vesicles (ESVs) –. They do not localize to the plasma membrane. Neither the ESVs nor the cyst wall contains VSPs.
In theses studies, we tested the hypothesis that the cyst wall may contain additional proteins. Recently, our ongoing transcriptome analyses of the giardial life cycle revealed a gene encoding an unusual VSP-like protein that was upregulated during encystation. Functional and bioinformatic analyses showed that this gene encodes a divergent cysteine-rich protein (HCNCp) that is increasingly expressed in the cyst cell body and cyst wall during encystation and does not vary among giardial isolates. HCNCp may provide a functional link between immune avoidance and differentiation.
Identification of HCNCp
A major goal of the giardial transcriptome analysis was to identify new genes whose expression patterns suggest an important role in differentiation. Correcting for sequencing error, we sampled 36,000–39,000 SAGE tags throughout the Giardia lamblia life cycle, except for 21 hours encystation, in which we were only able to sample 18,370 tags. 83.6% of all 15 bp SAGE tag sequences could be mapped to a single location in the 12 Mbp “giardia14” genome assembly, with an additional 9.0% mapping to multiple locations primarily due to duplicated genes. In addition to numerous upregulated transcripts, we identified ORF 40376 (originally annotated as a large VSP) as a gene significantly upregulated from 12 to 42 hr of encystation (Figure 1). Northern analyses confirmed an increase in the transcript for this gene during encystation (data not shown). ORF 40376 encodes a large gene of 3790 bp (1609 amino acids) and the encoded protein has a predicted molecular weight of 169,301.7 kDa. We named the gene product High Cysteine Non-variant Cyst protein (HCNCp) (GenBank accession DQ144994). Like the VSPs, HCNCp is a strikingly cysteine-rich (13.9%), acidic (theoretical pI 4.72), Type 1 integral membrane protein with multiple CxxC and GGCY motifs (Tables 1 and 3), although it did not have a Zn finger motif. It has a typical N-terminal signal peptide sequence whose predicted cleavage site is between amino acids 14 and 15. Also like the VSPs, HCNCp has a single C-terminal membrane spanning region and short cytoplasmic tail. However, the predicted C-terminal transmembrane (TM) region of HCNCp is divergent from those of the VSPs (1577-GAITGIAIAVIAVIGCAVGVLVW-1599) with a similar, but distinct amino acid tail, CRRSKAV for HCNCp versus CRGKA for the VSPs (Table 1). While VSPs utilize a unique extended polyadenylation signal , , the polyadenylation signal for HCNCp matches that of the majority of non-VSP genes within the G. lamblia genome (AGTAAAT). MEME analysis did not find a unique putative promoter sequence shared between HCNCp and the VSPs or CWPs, although HCNCp did share a typical AT-rich putative transcription initiation signal with the CWPs. There was no evidence of the C(T/A)ACAG promoter sequence associated with gMyb2 , a Myb-related protein involved in upregulation of CWPs and G6PI-B  during encystation.
SAGE revealed increased steady state abundance of HCNCp transcripts beginning at 12 hr encystation through 42 hr compared to trophozoite (“troph”) and 4 hr encystation.
SAGE data is presented as percentage of all tags sampled at a given time point.
HCNCp Protein Is Upregulated During Encystation and Does Not Traffic to the Plasma Membrane
Based on its degree of similarity to the VSPs, we tested the hypothesis that HCNCp is a large variable cysteine-rich protein increasingly expressed on the trophozoite surface during encystation. We expressed full-length (~170 kDa) HCNCp under its own promoter with a small C-terminal epitope-tag to evaluate protein expression profiles. Full-length HCNCp was faintly detected in trophozoites, but increasingly expressed at 21 and 42 hr of encystation and highest in water-resistant cysts (Figure 2). The protein expression profile of HCNCp is similar to its stable mRNA levels (Figure 1). Epitope-tagged (C-terminal) fragments of 42 and 21 kDa also increased during encystation. These are likely physiologically processed C-terminal fragments, as the antigens are prepared by trichloroacetic acid precipitation of live parasites in the presence of proteinase inhibitors to prevent degradation. Nonetheless, we cannot be certain that the native HCNCp processes the same way as epitope-tagged HCNCp.
Full length epitope-tagged HCNCp protein (arrow at 170 kDa) was detected beginning at 21 hr encystation (“21”) and was at peak levels at the cyst (“C”) stage.
Proteins having an AU1 tag and relative molecular masses of 42 and 21 kDa also increased during encystation (arrows).
Anti-AU1 antibodies did not react to non-transfected “control” trophozoites (“T”) or encysting cells (“21”, “42”, or “C”).
Size markers are indicated by dashes on the left side of figure in kDa and the taglin loading control is shown at the bottom of the figure (“LC”).
In vegetative trophozoites, HCNCp was primarily localized to the nuclei (Figure 3), as shown by DAPI staining. Double staining with antibody to protein disulfide isomerase 2 (PDI2; ; data not shown) confirmed that some HCNCp was located to the nuclear envelope, which is the first cisterna of the ER. Encystation is not synchronous and we observed this nuclei/ER staining in some encysting cells (not shown). We found that during encystation, HCNCp traffics through the encystation secretory vesicles (ESV), which export CWPs. Double staining with rabbit anti-cyst antibody, which co-localized with cwp-1 and –2 to the ESV and cyst wall (not shown), showed co-localization of cyst proteins with HCNCp in ESV (Figure 3). Double staining also showed HCNCp in the wall of mature, water-resistant cysts. However, in contrast to the cyst proteins that were exclusively in the cyst wall, most of the HCNCp localized to the cell body of the cyst that becomes the excyzoite (Figure 3). We have not observed HCNCp in the plasmalemma of vegetative or encysting trophozoites, even though our epitope-tag contains the C-terminal predicted TM sequence.
Differential interference contrast (DIC) merged with DAPI images are shown to the left of each panel, HCNCp in green, anti-cyst proteins in red, and nucleic acid in blue (DAPI).
In trophozoites, HCNCp localized to nuclei and nuclear envelope/ER.
During encystation HCNCp co-localized with cyst proteins in encystation secretory vesicles (ESV) and to the cyst wall of water-resistant cysts.
In mature cysts, most of the HCNCp was within the cell body.
Scale bar is 5 µM.
HCNCp is a Unique Non-Variant Cyst Protein
We searched the G. lamblia genome for sequences unique to HCNCp in order to develop HCNCp-specific internal PCR primers. We used these primers to detect the HCNCp gene and mRNA transcripts in different Giardia isolates. Isolate WB (clone C6) expresses the VSP called TSA 417 and 1F, A6, and E6 are subclones derived from WB C6 that do not express TSA 417 . Bris, RB, and LT are unrelated human isolates, while D3 was isolated from a dog (American Type Culture Collection; ATCC). All are in Assemblage A . We detected the HCNCp gene (Figure 4A) and transcript (Figure 4B) in trophozoites of all strains/subclones except GS/M, which belongs to Assemblage B and is considered the most distant Giardia strain infecting humans . Other primer sets we tested also did not detect HCNCp in GS/M. The GS/M strain either does not have the gene or it is too dissimilar to be detected with the primer sets used. These data suggest that HCNCp is not a classical giardial VSP, but a new invariant cyst protein.
(A) Genomic PCR found that the gene for HCNCp is present in all giardial isolates except for GS/M (upper panel).
Lower panel shows tubulin controls for each isolate.
(B) HCNCp mRNA transcripts were detected in all strains/clones tested except for the most divergent known Giardia strain infecting humans, GS/M (upper panel).
Tubulin controls for rt-PCR without reverse transcriptase (“-“) or plus reverse transcriptase (“+”) are shown in the lower panel.
Sizing markers are indicated by dashes to the left side of the figure in kb.
Identification of other non-VSP cysteine-rich membrane proteins
We searched the G. lamblia genome for additional genes encoding a Type 1 integral membrane, high Cys containing proteins (HCMp) lacking the canonical VSP cytoplasmic tail sequence CRGKA (Table 1). Other search parameters were that the protein contain ≥400 amino acids, ≥10% Cys and ≥20 CxC and or CxxC. Of the 60 proteins we identified, none had specific sequence homology to HCNCp, but many contained the CxC motif that appears eight times in HCNCp, yet is rarely found in VSPs , . Only one HCMp (ORF 17328) used the extended VSP polyadenylation signal , , while all others utilized the conventional Giardia polyadenylation signal. SAGE data was available for a number of these HCMp (Table 2), but only five had evidence of changes in expression during encystation. ORFs 16318 and 25816 were upregulated in 4–12 hr encystation, while ORFs 11309, 14017, and 15317 were approximately equal in abundance in trophozoites and during encystation, but nearly absent at 42 hr encystation. Each of these five proteins is a large, VSP-like protein with epidermal growth factor (EGF)-like motifs (Table 3). We also found 26 ORFs (18 were <400 aa) that were ≥10% Cys that had a single TM domain and no CRGKA tails, but they did not have ≥20 CxC/CxxC motifs (Table S1).
We found that cysteine-rich (≥10%), Type 1 integral membrane proteins with ≥20 CxC and/or CxxC are absent or rare in most completed genomes (Table 4). For example, we found no representatives in the parasites Toxoplasma gondii, Trypanosoma brucei, Plasmodium falciparum, or Cryptosporidium parvum or in the yeast and plant model organisms Saccharomyces cerevisiae, Arabidopsis thaliana and Chlamydomonas rheinhardtii. In contrast, we found 6 HCMp's in the intestinal dwelling, cyst-forming human parasite Entamoeba histolytica, 4 each in Caenorhabditis elegans and Drosophila melanogaster, and 7 in humans. Surprisingly, the genome of the ciliated protozoan Tetrahymena thermophila has 30 HCMp. This number is still far fewer than Giardia if the genome size is considered. Based on predicted ORFs, T. thermophila has 5 times fewer HCMp than Giardia (Table 4). These 30 ORFs had high expectation values (from 0.0035 to 4.6e−08) to the TMK-like HMM from E. histolytica and half of them had between 1 and 25 PC6B motifs.
Bioinformatic analyses of HCMp
36 of the giardial HCMp are annotated as VSPs in the Giardia Genome Database (www.mbl.edu/Giardia ), nine as notch/EGF receptor-like , and 16 as E. histolytica receptor kinase-like proteins . We used comparative motif analyses with the programs MEME  and MAST  to determine whether any of the 61 G. lamblia HCMp (Table 1) contain repeats or known motifs or might be related to each other. Seven of the nine similarity groups contained regular repeat patterns (Figures S1 and S2). In addition to the repeated motifs, three groups (called VSP-like, EGF-like, and TMK-like) contained recognizable motif signatures. The proteins we placed in the VSP-like group had expectation values of e−54 to e−72 against the VSP domain HMM of the Pfam database (PF03302), while TSA 417 was e−308 (Table 3). The EGF-like group had high similarities to the EGF functional domain (PF00008). The TMK-like group scored highly to a HMM constructed using alignments from the extracellular domains from 6 of 8 E. histolytica transmembrane kinases . In addition, some members of the other groups had fewer instances of these motifs. Several members of the TMK-like, VSP-like, and Group 1, including HCNCp, also contained 1 or more PC6B domains (Table 3), which have been implicated in the possible stabilization or intracellular localization of a Kex-2 like endoproteinase .
The sequence of HCNCp is strikingly similar to the VSPs in many respects, however its expression and trafficking largely resemble that of the CWPs. Moreover, we discovered that the giardial genome contains a total of 60 additional high cysteine Type 1 integral membrane proteins (HCMp) that resemble HCNCp, and do not appear to be VSPs. The giardial isolate (“WB, clone C6”) used for genome analyses has about 150 VSPs in its repertoire. VSPs change under selection by antibodies in vitro and in vivo , . However, they also switch spontaneously , . VSP expression may also be influenced by physiological conditions, such as intestinal proteases or growth vs. differentiation ,  and may help in adaptation to growth in different host species . In contrast, HCNCp is not variable, as it was expressed in all the Assemblage A isolates and subclones we tested.
VSPs may differ greatly in length from about 20 to 200 kDa, frequently because of different numbers of repeated sequences , . They often belong to small families whose genes are recognized by the same probes or encode proteins with cross-reactive epitopes. Many VSP sequences have been used as to define different genetic assemblages , , . Despite their variability, the VSPs have strikingly constant features. They are all highly (~11–12%) cysteine-rich Type I integral membrane proteins, with a large proportion of the Cys residues in “CxxC” tetrapeptide motif, where x can be many amino acids . However, the CxC motif is very rare in VSPs , . CC sequences are rarely found in VSPs or HCMp. VSPs have a conventional N-terminal signal peptide and least one GGCY tetrapeptide motif near the C-terminus. Many have a Zn finger motif. VSPs are characterized by an extremely conserved C-terminal membrane spanning sequence and short hydrophilic cytoplasmic anchor sequence, CRGKA. The Cys is a substrate for palmitoylation , which may be a means to regulate cellular signaling. Virtually all of the HCMp also have Cys residues in their tails that may act as a substrate for this modification, although Giardia has relatively few palmitoylated proteins , . While most giardial genes have a typical consensus AGTUAAY polyadenylation signal, that of the VSPs is 15 nucleotides , . Most importantly, VSPs traffic to the plasma membrane via the ER in a constitutive secretory pathway , , . These features may be considered diagnostic of VSPs.
The three known cyst wall structural proteins (CWPs) are similar to each other with a signal peptide and leucine-rich repeats (LRR), but no TM regions . The 14 Cys residues they share are positionally conserved and they do not appear to vary among isolates. Expression of the CWPs is greatly upregulated in encystation and they are transported to the nascent, extracellular cyst wall by novel encystation secretory vesicles (ESV) that are specifically assembled during encystation .
Although Giardia was reported to lack a conventional Golgi apparatus , we showed earlier that it could sort proteins to constitutive and regulated secretory pathways , . Specifically, in vegetative cells, the VSP TSA 417 is targeted to the plasmalemma via the constitutive pathway . During encystation, TSA 417 co-localizes with CWPs in the nuclear envelope and the ER, but does not enter the ESVs or traffic to the cyst wall. Conversely, CWPs are sequestered in the ESVs for transport to the cyst wall, but not found in the plasmalemma –, . Rather than a discrete targeting motif, most of the LRR regions of the CWPs seem to be required for entry into the ESV and incorporation into the cyst wall , , . Hehl's group also reported that chimeras containing the LRRs from a CWP  and the VSP  transmembrane region and tail localized to the plasmalemma in pre-encysting cells lacking ESV. However, during encystation, the chimera was targeted to the cyst wall via ESVs. This group has also localized homologs of Golgi proteins to ESVs ,  and proposes that they have post-ER Golgi-like sorting function.
In vegetatively growing cells, HCNCp localizes largely to the nuclei and nuclear envelope, as it co-localizes with DAPI staining and PDI-2 (Figure 3). Thus, it is within the innermost cisterna of the ER and in the constitutive protein secretory pathway . In contrast to the VSPs, it is not in the more distal ER or the plasmalemma. The implications of HCNCp within the nuclei remain to be elucidated. In encysting cells, HCNCp clearly co-localizes with cyst proteins within the ESV, which demarcate the regulated secretory pathway . In mature cysts, virtually all anti-cyst epitopes are in the extracellular cyst wall and little is detected within the cell body. However, only a fraction of HCNCp is in the cyst wall and most appears to be in vesicular structures throughout the cell body (Figure 3). We cannot tell if these are remnant ESVs that have discharged the CWPs, but retain at minimum the C-terminus of HCNCp. This is the first observation of a protein that localizes to both the cyst wall and cell body. Since the cell body is the precursor of the excyzoite, HCNCp may have a function during or after excystation. Therefore, HCNCp appears to follow a novel protein trafficking pathway, unlike those previously characterized in Giardia. We trace HCNCp with a C-terminal 6-amino acid epitope-tag. Thus, even the smallest fragment detected (~21 kDa) contains the hydrophobic putative TM region. Further studies are needed to determine the fate of the N-terminus of HCNCp and the function, if any of the proteolytic processing.
In addition to HCNCp, another cyst protein is proteolytically processed during encystation. Lujan et al (1995)  reported that CWP2 can be proteolytically cleaved from a 39 kDa protein to 26 kDa by an encystation specific cysteine proteinase (ESCP) found in the lysosomal compartment . We have not determined whether ESCP or other proteinases can cleave HCNCp, nor do we know if it can be self-cleaving. We have found a putative subtilisin-type proprotein convertase (SPC) that is also upregulated in encystation and localizes to the ESV (B. J. Davids unpublished data). Interestingly, there is an SPC cleavage site in HCNCp that would produce a C-terminal epitope-tagged fragment of v56 kDa, which may be one of the minor fragments seen in our immunoblot of cysts (Figure 2). The VSP TSA 417 also undergoes endogenous proteolysis , but the fragments are held together on the cell surface by disulfide bonding.
Although it is strikingly cysteine-rich, with 77 CxxC motifs, HCNCp clearly differs from the VSPs, especially in its C-terminus. HCNCp also has 6 CxC motifs while TSA 417, like most other VSPs, has none. We asked whether HCNCp is unique or resembles other proteins in Giardia. Genome-wide analyses revealed 173 giardial proteins of ≥400 amino acids with ≥10% Cys. 152 of these are Type 1 integral membrane proteins. Of these, 89 have a VSP CRKGA tail and have not been considered in our analyses. Aside from the VSPs, 61 ORFs including HCNCp have ≥20 instances of CxxC/CxC and satisfy all our criteria for a putative HCMp. The HCMp are difficult to classify and differ in various parameters, although they form small clusters with similar membrane spanning and or cytoplasmic sequences (Table 1), and the occurrence of GGCY motifs (Table 3). While not all groups had GGCY motifs, HCNCp had 5 and others within “group 1” had between 1 and 3. The significance of this motif in VSPs has not been fully examined, although Nash  reports that mutations in GGCY result in loss of VSP surface localization. We have not seen HCNCp localized at the surface membranes of trophozoites or encysting trophozoites, but it may have an important role at the surface of excyzoites during excystation.
The HCMp also differ from each other in whether they have a preponderance of CxxC or CxC motifs. The latter are very rare in VSPs . All have some CxxC motifs and many have no CxC, but others have from 1 to 28 CxC sequences. From a functional point of view , CxxC motifs are well-known in the active sites of redox proteins like PDIs and thioredoxins. The identity of the two intervening amino acids strongly affects the E0′ or redox potential. We showed earlier that in TSA 417 , all detectable Cys residues were in intramolecular disulfide bonds that were extremely resistant to reduction in the native protein. However, the vicinal disulfide bonds of three amino acid CxC rings are much less stable than those of four-member CxxC rings. The functional implications of the prevalence of CxxC vs CxC motifs in HCMp remain to be determined.
We asked whether extremely cysteine-rich membrane proteins are common in other organisms, especially those that live in a microaerophilic environment similar to that of Giardia. E. histolytica is a mucosal dwelling protozoan that also differentiates into cysts. Both parasites require high concentrations of Cys for growth in culture , . E. histolytica has 17 Type 1 integral membrane proteins with ≥10% Cys, but only 6 of these proteins have CxxC/CxC sequences that would qualify them as HCMp . Four of the other genomes we examined had fewer HCMp than Giardia although T. thermophila has 30 (Table 4). However, only 0.11% of the macronuclear T. thermophila genome is dedicated to HCMp , compared to 0.63% for Giardia. Currently, these two protozoans appear to have more HCMp than other organisms, although analyses of additional genomes could certainly reveal others. Genetic manipulations are not possible in Giardia, which is tetraploid . However, reverse genetics  in T. thermophila could provide clues or hypotheses on functions of HCMps.
Global comparisons of the HCMp are challenging as they group differently depending on the parameter used. Within the giardial HCMp, recognizable motifs (Figures S1 and S2) enabled us to classify them into subgroups (Table 3). These subcategories may represent specific classes of HCMp with various functions. It will be important to characterize different members of each MEME category to determine if their group similarity can predict possible expression, localization, and function during different stages of giardial differentiation. Presently, we can say that HCNCp has structural characteristics like that of VSPs, but is regulated and expressed similar to a CWP. It is interesting that the known giardial surface proteins are all cysteine-rich. This may help explain the high requirement of Giardia for Cys for growth and survival , . HCNCp is a member of a large group of cysteine-rich, non-VSP, Type 1 integral membrane proteins, which are rare to absent in the majority of genomes examined. The exact functions of HCNCp and other HCMp, their evolutionary origins, and their overall importance to the biology of G. lamblia remain to be determined.
Materials and Methods
Giardial isolates, cultivation and encystation
We used isolate WB clone C6 (ATCC 50803; ; human, Afghanistan) unless otherwise indicated. We tested expression of HCNCp in trophozoites of WB C6 subclones (1F, E6 ), A6 (ATCC 50583; ), and strains LT (; human, Puerto Rico), RB (isolated from a patient infected in Turkey), D3 (ATCC 203334, isolated from a dog Calgary, Alberta, Canada, 1986), Bris (, human, Australia), and GS clone H7 (GS/M; ATCC 50581; , human, Alaska). Parasites were grown in modified TYI-S33 medium and encysted as described .
Production of polyclonal anti-cyst antibody
Polyclonal antiserum against in vitro purified water-resistant cysts was raised in rabbits as described .
mRNA expression patterns
Serial Analysis of Gene Expression (SAGE) libraries were constructed using ~10 µg of total RNA isolated from TRIzol (Invitrogen, Carlsbad, CA) extractions of trophozoites and encysting cells (4, 12, 21, 42 hr) following the I-SAGE protocol (Invitrogen). Recombinant pZero1 clones produced by SAGE were purified using GeneMachines RevPrep Orbit (San Carlos, CA) and were sequenced on an ABI 3730xl DNA sequencer (Applied Biosystems, Foster City, CA). Sequences collected were analyzed with software created at the MBL specifically for Giardia SAGE analyses. The SAGE software extracts ditag sequences from the ABI 3730xl results according to the SAGE sequence grammar, parses out individual SAGE tags, excludes tags with sequence ambiguities, and reduces all SAGE tags to a look-up table of unique SAGE tag sequences and their observed frequencies among all of the Giardia SAGE libraries. SAGE tags not found more than once in at least one SAGE library were excluded from analyses as putative sequencing error, unless the tag sequence had a perfect match to available genome data. The unique tag sequences were mapped to all available Giardia “giardia14” genome assembly contigs (www.mbl.edu/Giardia) to determine the identity of expressed genes. The 15 bp SAGE tags were mapped to the assembly based on exact sequence matches.
HCNCp mRNA expression was verified by Northern analysis. 10 µg of total RNA was extracted from trophozoites and encysting cells as above and separated on a denaturing RNA gel, transferred to a ZetaProbe filter (Bio-Rad Laboratories, Hercules, CA), and hybridized as described . The strand-specific 244 bp probe for HCNCp was made using primers “Nprobe-for” (5′GCAGCATTCCTACCCGTTAT3′) and “Nprobe-rev” (5′TCCAAACATTGCGTTGATTC3′) in combination with the labeling kit Prime-it II (Stratagene, La Jolla, CA). The probed and washed filter was imaged on a Storm 860 phosphoimager (Molecular Dynamics, Sunnyvale, CA).
Selectable transformation and epitope tagging
The promoter and coding sequences for HCNCp were amplified from 200 ng of genomic DNA using the forward primer “HCNCp-for” (5′-ACAAGCTTTCACAGGATGTGGACGAT-3′) and the reverse primer “HCNCp-rev” (5′-AAATGGGCCCCACAGCTTTGCTCCTAC-3′) and the GeneAmp XL PCR kit (Applied Biosystems, Branchburg, NJ) by standard conditions at a final concentration of 1.2 mM Mg(OAc)2. Conditions for PCR were an initial denaturation at 95°C for 5 min, 4 cycles of: 95°C 50 sec, 50°C 1 min, 72°C 6 min; 31 cycles of: 95°C 50 sec, 55°C 1 min, 72°C 6 min; and a final extension for 10 min at 72°C (MJ Research, Waltham, MA). The ~4 kb product was gel purified using QIAquick gel extraction kit (Qiagen; Valencia, CA) and approximately 1 µg of each PCR product or AU1-epitope tagging Giardia vector  were first digested with Apa1 for 1 hr 37°C, cleaned with QIAquick for PCR products, digested with HindIII for 1 hr 37°C, and gel purified using QIAquick gel extraction kit. The vector was de-phosphorylated using shrimp alkaline phosphatase (Promega; Madison, WI) by standard methods and purified using QIAquick. The digested vector and PCR fragments were ligated (T4 DNA ligase; Roche, Nutley, NJ) overnight at 4°C, transformed into JM109 competent cells (Promega), and plasmids screened by size for automated sequencing of clones containing full length HCNCp (GenBank accession number DQ144994). WB clone C6 trophozoites were electroporated with 50 µg of plasmid containing HCNCp and were selected and maintained by puromycin resistance . Recombinant clones were purified using GeneMachines RevPrep Orbit, sequenced on an ABI 3730xl DNA sequencer (Applied Biosystems), and assembled using the computer program PHRAP .
Expression of HCNCp
Trophozoite, encysting cells, and cyst antigens from Giardia stably expressing the AU1-tagged HCNCp gene or non-transfected control cells were processed for Western analysis . Because of the large size of HCNCp, transfer buffer contained 20% MeOH/0.1% SDS. Blotted proteins were blocked overnight in PBST (PBS+0.05% Tween 20) containing 5% milk, probed in rabbit anti-AU1 antibody (1/5,000 in PBST/0.1% milk; Bethyl Laboratories, Montgomery, TX) or antibody to the constitutive protein taglin (1/700 in PBST/0.1% milk; loading control; ) for 1.5 hr, washed 3× over 30 min in PBST, incubated in Zymax goat anti-rabbit (1/8,000; Zymed Laboratories/Invitrogen) or Zymax goat anti-mouse (1/5,000; Zymed Laboratories/Invitrogen) antibodies for 1 hr, washed 3× in PBST over 30 min, and developed in ECL (Amersham Biosciences, Piscataway, NJ).
Immunolocalization of HCNCp and CWP
Transfected parasites stably expressing AU1-tagged HCNCp or non-transfected control cells were prepared for immunofluorescence assay (IFA) by methods similar to Abel et al., 2001 . Giardial trophozoites or encysting cells were allowed to attach to glass coverslips coated with 0.1% polyethylenimine (Sigma, St. Louis, MO) in PBS on a slide warmer at 37°C. Water-resistant cysts were allowed to dry onto the coverslip and then placed in IFA block as described below. Attached cells were fixed in 100% methanol (−20°C) for 10 min, dried, and stored overnight at room temperature or permeabilized immediately in 0.5% Triton-X 100 for 10 min, and placed in IFA block (10% goat serum, 1% glycerol, 0.1% bovine serum albumin, 0.1% fish gelatin, and 0.04% sodium azide in PBS) for 1 hr. Cells were incubated for 1 hr in primary antibody (anti-AU1 mAb diluted 1/500 in IFA block; Covance, Berkeley, CA; or rabbit polyclonal anti-cyst (1/250 in IFA block)), washed 4× 5 min in PBS, and incubated in goat anti-mouse FITC-labeled (1/100 in IFA block; Pierce Biotechnology, Inc., Rockford, IL) or goat-anti rabbit TRITC-labeled (1/100 Zymed Laboratories/Invitrogen) secondary antibodies for 1 hr, washed 4× 5 min in PBS, post-fixed in 4% PFA for 7 min, rinsed 2× in PBS, and mounted with ProLong Gold with DAPI (Invitrogen). Cells were screened on an E800 Nikon (New York City, NY) microscope with an EXFO (Vanier, Canada) X-cite fluorescent 120W metal halide illuminator and imaged with a DMX 1200F Nikon fluorescence sensitive digital camera.
RtPCR and genomic PCR
cDNA was prepared directly from 104 log-phase trophozoites (SuperScriptTM III CellsDirect cDNA synthesis system, Invitrogen). HCNCp-specific internal primers were: 5′-GGAAAACGGTGACCTTATTGTG-3′ (forward) and 5′-GTCTACCAGGACTGAAAGTAT-3′ (reverse); and control primers for alpha-2-tubulin were: 5′-CGTTCAACACGTTCTTCTCG-3′ (forward) and 5′-ATGTCGTACATGGCCTCGTT-3′ (reverse). 3 µl of the cDNA template in combination with 0.2 mM dNTP mix, 2 mM MgSO4, 0.25 µM each primer, and 1 U Platinum Taq high fidelity polymerase (Invitrogen) was used in a standard PCR reaction. Conditions for PCR were an initial denaturation at 95°C for 3 min, 36 cycles of: 95°C 1 min, 50°C 1 min, 72°C 2 min, and a final extension for 7 min at 72°C.
Genomic DNA was prepared from all strains and subclones using DNA STAT-60TM (Tel-Test, Inc., Friendswood, TX) by manufacturer protocols. Conditions for genomic PCR were as described above using ~200 ng of DNA as template.
A profile Hidden Markov Model (HMM) of the Entamoeba histolytica transmembrane kinase (TMK) extracellular domain  was created with HMMER version 2.3.2 . The HMM's for VSP (PF03302) and EGF (PF00008) were from the Pfam database . Conservative nucleotide or amino acid motifs were detected using Multiple Em for Motif Elicitation (MEME)  and Motif Alignment & Search Tool (MAST) , predication of signal peptides using SignalP 3.0 , and prediction of transmembrane domains using transmembrane Hidden Markov Model (TMHMM) 2.0 . A custom PERL script was used to search genomes for cysteine-rich protein sequences containing CxC and/or CxxC motifs (A. G. McArthur). The PC6B motifs present in HCMp's were found using a regular expression constructed from the published consensus sequence: Cx2Cx3Cx2Cx5-7Cx2Cx10-15Cx3C .
Cysteine-Rich Proteins Not Classified As HCMp Because They Are Too Short And/OR Have < 20 CxxC + CxC's
(0.04 MB PDF)
Diagram Of MEME/MAST Analyses Of The 9 HCMp Groupings
(0.29 MB PDF)
Motif Sequences Found By MEME Analysis For The 9 HCMp Groupings
(0.06 MB PDF)
We thank Michael Gilbert for his initial input. We thank Sarah Williams for her skilled technical support. We thank Rodney Adam (University of Arizona) for examining our HCMp ORF list to be sure we did not classify a true VSP as HCMp. We are grateful to Mitchell Sogin and Hilary Morrison for their interest in HCMp's and for providing genomic resources.
Conceived and designed the experiments: AM FG BD DR. Performed the experiments: BD DR SB SP. Analyzed the data: AM FG BD DR SB SP MC. Contributed reagents/materials/analysis tools: AM. Wrote the paper: AM FG BD. Other: Provided grant funding: FG AM.
- 1. Wolfe MS (1992) Giardiasis. Clin Microbiol Rev 5: 93–100.
- 2. Adam RD (1991) The biology of Giardia spp. Microbiological Reviews 55: 706–732.
- 3. Savioli L, Smith H, Thompson A (2006) Giardia and Cryptosporidium join the ‘Neglected Diseases Initiative’. Trends Parasitol 22: 203–208.
- 4. Marshall MM, Naumovitz D, Ortega Y, Sterling CR (1997) Waterborne protozoan pathogens. Clin Microbiol Rev 10: 67–85.
- 5. Thompson RC, Hopkins RM, Homan WL (2000) Nomenclature and genetic groupings of Giardia infecting mammals. Parasitol Today 16: 210–213.
- 6. Smith PD, Gillin FD, Spira WM, Nash TE (1982) Chronic giardiasis: studies on drug sensitivity, toxin production, and host immune response. Gastroenterology 83: 797–803.
- 7. Nash TE (2002) Surface antigenic variation in Giardia lamblia. Mol Microbiol 45: 585–590.
- 8. Adam RD (2001) Biology of Giardia lamblia. Clin Microbiol Rev 14: 447–475.
- 9. Adam RD, Aggarwal A, Lal AA, de La Cruz VF, McCutchan T, et al. (1988) Antigenic variation of a cysteine-rich protein in Giardia lamblia. J Exp Med 167: 109–118.
- 10. Gillin FD, Hagblom P, Harwood J, Aley SB, Reiner DS, et al. (1990) Isolation and expression of the gene for a major surface protein of Giardia lamblia. Proc Natl Acad Sci U S A 87: 4463–4467.
- 11. Nash TE, Aggarwal A, Adam RD, Conrad JT, Merritt JW Jr. (1988) Antigenic variation in Giardia lamblia. J Immunol 141: 636–641.
- 12. Nash TE (1989) Antigenic variation in Giardia lamblia. Exp Parasitol 68: 238–241.
- 13. Nash TE, Banks SM, Alling DW, Merritt JW Jr., Conrad JT (1990) Frequency of variant antigens in Giardia lamblia. Exp Parasitol 71: 415–421.
- 14. Nash TE, Merritt JW Jr., Conrad JT (1991) Isolate and epitope variability in susceptibility of Giardia lamblia to intestinal proteases. Infect Immun 59: 1334–1340.
- 15. Nash T (1992) Surface antigen variability and variation in Giardia lamblia. Parasitol Today 8: 229–234.
- 16. Aley SB, Gillin FD (1993) Giardia lamblia: post-translational processing and status of exposed cysteine residues in TSA 417, a variable surface antigen. Exp Parasitol 77: 295–305.
- 17. Aley SB, Gillin FD (1995) Specialized surface adaptations of Giardia lamblia. Infect Agents Dis 4: 161–166.
- 18. Svard SG, Meng TC, Hetsko ML, McCaffery JM, Gillin FD (1998) Differentiation-associated surface antigen variation in the ancient eukaryote Giardia lamblia. Mol Microbiol 30: 979–989.
- 19. Gillin FD, Reiner DS, McCaffery JM (1996) Cell biology of the primitive eukaryote Giardia lamblia. Annu Rev Microbiol 50: 679–705.
- 20. Eichinger D (2001) Encystation in parasitic protozoa. Curr Opin Microbiol 4: 421–426.
- 21. Abel ES, Davids BJ, Robles LD, Loflin CE, Gillin FD, et al. (2001) Possible roles of protein kinase A in cell motility and excystation of the early diverging eukaryote Giardia lamblia. J Biol Chem 276: 10320–10329.
- 22. Reiner DS, Hetsko ML, Meszaros JG, Sun CH, Morrison HG, et al. (2003) Calcium signaling in excystation of the early diverging eukaryote, Giardia lamblia. J Biol Chem 278: 2533–2540.
- 23. Larocque R, Nakagaki K, Lee P, Abdul-Wahid A, Faubert GM (2003) Oral immunization of BALB/c mice with Giardia duodenalis recombinant cyst wall protein inhibits shedding of cysts. Infect Immun 71: 5662–5669.
- 24. Meng TC, Hetsko ML, Gillin FD (1993) Antigenic switching of TSA 417, a trophozoite variable surface protein, following completion of the life cycle of Giardia lamblia. Infect Immun 61: 5394–5397.
- 25. Nash TE, Lujan HT, Mowatt MR, Conrad JT (2001) Variant-specific surface protein switching in Giardia lamblia. Infect Immun 69: 1922–1923.
- 26. Lujan HD, Mowatt MR, Conrad JT, Bowers B, Nash TE (1995) Identification of a novel Giardia lamblia cyst wall protein with leucine-rich repeats. Implications for secretory granule formation and protein assembly into the cyst wall. J Biol Chem 270: 29307–29313.
- 27. Mowatt MR, Lujan HD, Cotten DB, Bowers B, Yee J, et al. (1995) Developmentally regulated expression of a Giardia lamblia cyst wall protein gene. Mol Microbiol 15: 955–963.
- 28. Sun CH, McCaffery JM, Reiner DS, Gillin FD (2003) Mining the Giardia lamblia genome for new cyst wall proteins. J Biol Chem 278: 21701–21708.
- 29. Ortega-Barria E, Ward HD, Evans JE, Pereira ME (1990) N-acetyl-D-glucosamine is present in cysts and trophozoites of Giardia lamblia and serves as receptor for wheatgerm agglutinin. Mol Biochem Parasitol 43: 151–165.
- 30. Jarroll EL, Manning P, Lindmark DG, Coggins JR, Erlandsen SL (1989) Giardia cyst wall-specific carbohydrate: evidence for the presence of galactosamine. Mol Biochem Parasitol 32: 121–131.
- 31. Karr CD, Jarroll EL (2004) Cyst wall synthase: N-acetylgalactosaminyltransferase activity is induced to form the novel N-acetylgalactosamine polysaccharide in the Giardia cyst wall. Microbiology 150: 1237–1243.
- 32. McCaffery JM, Faubert GM, Gillin FD (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.
- 33. Svard SG, Hagblom P, Palm JE (2003) Giardia lamblia – a model organism for eukaryotic cell differentiation. FEMS Microbiol Lett 218: 3–7.
- 34. Reiner DS, McCaffery M, Gillin FD (1990) Sorting of cyst wall proteins to a regulated secretory pathway during differentiation of the primitive eukaryote, Giardia lamblia. Eur J Cell Biol 53: 142–153.
- 35. Hehl AB, Marti M (2004) Secretory protein trafficking in Giardia intestinalis. Mol Microbiol 53: 19–28.
- 36. Gottig N, Elias EV, Quiroga R, Nores MJ, Solari AJ, et al. (2006) Active and passive mechanisms drive secretory granule biogenesis during differentiation of the intestinal parasite Giardia lamblia. J Biol Chem 281: 18156–18166.
- 37. Sun CH, Palm D, McArthur AG, Svard SG, Gillin FD (2002) A novel Myb-related protein involved in transcriptional activation of encystation genes in Giardia lamblia. Mol Microbiol 46: 971–984.
- 38. Knodler LA, Svard SG, Silberman JD, Davids BJ, Gillin FD (1999) Developmental gene regulation in Giardia lamblia: first evidence for an encystation-specific promoter and differential 5′ mRNA processing. Mol Microbiol 34: 327–340.
- 39. Knodler LA, Noiva R, Mehta K, McCaffery JM, Aley SB, et al. (1999) Novel protein-disulfide isomerases from the early-diverging protist Giardia lamblia. J Biol Chem 274: 29805–29811.
- 40. Adam RD (2000) The Giardia lamblia genome. Int J Parasitol 30: 475–484.
- 41. McArthur AG, Morrison HG, Nixon JE, Passamaneck NQ, Kim U, et al. (2000) The Giardia genome project database. FEMS Microbiol Lett 189: 271–273.
- 42. Baron M (2003) An overview of the Notch signalling pathway. Semin Cell Dev Biol 14: 113–119.
- 43. Beck DL, Boettner DR, Dragulev B, Ready K, Nozaki T, et al. (2005) Identification and gene expression analysis of a large family of transmembrane kinases related to the Gal/GalNAc lectin in Entamoeba histolytica. Eukaryot Cell 4: 722–732.
- 44. Bailey TL, Elkan C (1994) Fitting a mixture model by expectation maximization to discover motifs in biopolymers. Proc Int Conf Intell Syst Mol Biol 2: 28–36.
- 45. Bailey TL, Gribskov M (1998) Combining evidence using p-values: application to sequence homology searches. Bioinformatics 14: 48–54.
- 46. Nakagawa T, Murakami K, Nakayama K (1993) Identification of an isoform with an extremely large Cys-rich region of PC6, a Kex2-like processing endoprotease. FEBS Lett 327: 165–171.
- 47. Bienz M, Siles-Lucas M, Wittwer P, Muller N (2001) vsp gene expression by Giardia lamblia clone GS/M-83-H7 during antigenic variation in vivo and in vitro. Infect Immun 69: 5278–5285.
- 48. Singer SM, Elmendorf HG, Conrad JT, Nash TE (2001) Biological selection of variant-specific surface proteins in Giardia lamblia. J Infect Dis 183: 119–124.
- 49. Yang YM, Adam RD (1995) Analysis of a repeat-containing family of Giardia lamblia variant-specific surface protein genes: diversity through gene duplication and divergence. J Eukaryot Microbiol 42: 439–444.
- 50. Mowatt MR, Nguyen BY, Conrad JT, Adam RD, Nash TE (1994) Size heterogeneity among antigenically related Giardia lamblia variant-specific surface proteins is due to differences in tandem repeat copy number. Infect Immun 62: 1213–1218.
- 51. Nash TE, Keister DB (1985) Differences in excretory-secretory products and surface antigens among 19 isolates of Giardia. J Infect Dis 152: 1166–1171.
- 52. Touz MC, Conrad JT, Nash TE (2005) A novel palmitoyl acyl transferase controls surface protein palmitoylation and cytotoxicity in Giardia lamblia. Mol Microbiol 58: 999–1011.
- 53. Hiltpold A, Frey M, Hulsmeier A, Kohler P (2000) Glycosylation and palmitoylation are common modifications of giardia variant surface proteins. Mol Biochem Parasitol 109: 61–65.
- 54. Das S, Traynor-Kaplan A, Reiner DS, Meng TC, Gillin FD (1991) A surface antigen of Giardia lamblia with a glycosylphosphatidylinositol anchor. J Biol Chem 266: 21318–21325.
- 55. Marti M, Hehl AB (2003) Encystation-specific vesicles in Giardia: a primordial Golgi or just another secretory compartment? Trends Parasitol 19: 440–446.
- 56. Hehl AB, 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.
- 57. Marti M, Li Y, Schraner EM, Wild P, Kohler P, et al. (2003) The secretory apparatus of an ancient eukaryote: protein sorting to separate export pathways occurs before formation of transient Golgi-like compartments. Mol Biol Cell 14: 1433–1447.
- 58. Marti M, Regos A, Li Y, Schraner EM, Wild P, et al. (2003) An ancestral secretory apparatus in the protozoan parasite Giardia intestinalis. J Biol Chem 278: 24837–24848.
- 59. Stefanic S, Palm D, Svard SG, Hehl AB (2006) Organelle proteomics reveals cargo maturation mechanisms associated with Golgi-like encystation vesicles in the early-diverged protozoan Giardia lamblia. J Biol Chem 281: 7595–7604.
- 60. Touz MC, Nores MJ, Slavin I, Carmona C, Conrad JT, et al. (2002) The activity of a developmentally regulated cysteine proteinase is required for cyst wall formation in the primitive eukaryote Giardia lamblia. J Biol Chem 277: 8474–8481.
- 61. Woycechowsky KJ, Raines RT (2003) The CXC motif: a functional mimic of protein disulfide isomerase. Biochemistry 42: 5387–5394.
- 62. Lujan HD, Nash TE (1994) The uptake and metabolism of cysteine by Giardia lamblia trophozoites. J Eukaryot Microbiol 41: 169–175.
- 63. Gillin FD, Diamond LS (1981) Entamoeba histolytica and Giardia lamblia: growth responses to reducing agents. Exp Parasitol 51: 382–391.
- 64. Loftus B, Anderson I, Davies R, Alsmark UC, Samuelson J, et al. (2005) The genome of the protist parasite Entamoeba histolytica. Nature 433: 865–868.
- 65. Eisen JA, Coyne RS, Wu M, Wu D, Thiagarajan M, et al. (2006) Macronuclear Genome Sequence of the Ciliate Tetrahymena thermophila, a Model Eukaryote. PLoS Biol 4:
- 66. Bernander R, Palm JE, Svard SG (2001) Genome ploidy in different stages of the Giardia lamblia life cycle. Cell Microbiol 3: 55–62.
- 67. Howard-Till RA, Yao MC (2006) Induction of gene silencing by hairpin RNA expression in Tetrahymena thermophila reveals a second small RNA pathway. Mol Cell Biol. In Press.
- 68. Gillin FD, Diamond LS (1981) Entamoeba histolytica and Giardia lamblia: effects of cysteine and oxygen tension on trophozoite attachment to glass and survival in culture media. Exp Parasitol 52: 9–17.
- 69. Aggarwal A, Nash TE (1988) Antigenic variation of Giardia lamblia in vivo. Infect Immun 56: 1420–1423.
- 70. Smith PD, Gillin FD, Kaushal NA, Nash TE (1982) Antigenic analysis of Giardia lamblia from Afghanistan, Puerto Rico, Ecuador, and Oregon. Infect Immun 36: 714–719.
- 71. Capon AG, Upcroft JA, Boreham PF, Cottis LE, Bundesen PG (1989) Similarities of Giardia antigens derived from human and animal sources. Int J Parasitol 19: 91–98.
- 72. Aggarwal A, Merritt JW Jr., Nash TE (1989) Cysteine-rich variant surface proteins of Giardia lamblia. Mol Biochem Parasitol 32: 39–47.
- 73. Gillin FD, Reiner DS, Gault MJ, Douglas H, Das S, et al. (1987) Encystation and expression of cyst antigens by Giardia lamblia in vitro. Science 235: 1040–1043.
- 74. Weiland ME, Palm JE, Griffiths WJ, McCaffery JM, Svard SG (2003) Characterisation of alpha-1 giardin: an immunodominant Giardia lamblia annexin with glycosaminoglycan-binding activity. Int J Parasitol 33: 1341–1351.
- 75. Gordon D, Abajian C, Green P (1998) CONSED: a graphical tool for sequence finishing. Genome Res 8: 195–202.
- 76. Davids BJ, Mehta K, Fesus L, McCaffery JM, Gillin FD (2004) Dependence of Giardia lamblia encystation on novel transglutaminase activity. Mol Biochem Parasitol 136: 173–180.
- 77. Ward HD, Keusch GT, Pereira ME (1990) Induction of a phosphomannosyl binding lectin activity in Giardia. Bioessays 12: 211–215.
- 78. Eddy SR (1998) Profile hidden Markov models. Bioinformatics 14: 755–763.
- 79. Bateman A, Coin L, Durbin R, Finn RD, Hollich V, et al. (2004) The Pfam protein families database. Nucleic Acids Res 32 Database issue D138–141.
- 80. Bendtsen JD, Nielsen H, von Heijne G, Brunak S (2004) Improved prediction of signal peptides: SignalP 3.0. J Mol Biol 340: 783–795.
- 81. Krogh A, Larsson B, von Heijne G, Sonnhammer EL (2001) Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes. J Mol Biol 305: 567–580.