TcRho1: a Farnesylated Rho family homologue from Trypanosoma cruzi CLONING, TRANS-SPLICING AND PRENYLATION STUDIES

: Rho GTPases are members of the Ras superfamily and are involved in signal transduction pathways including maintenance of cell morphology and motility, cell cycle progression, and transcriptional activation. We report the molecular identification in trypanosomatids ( Trypanosoma cruzi ) of the first member of the Rho family. The cloned Rho protein, TcRho1 shares ~40% homology with other members of the Rho family. Southern blot analysis reveals that TcRho1 is a single copy gene per haploid genome, and northern blot assays shows a transcript of 1200 nt in length. Mapping the 5’ untranslated region of TcRho1 transcripts revealed at least five different transcripts derived from differential trans-splicing. Three of the five transcripts contain the trans-splicing site within the coding region of the TcRho1 gene. TcRho1 also contains the C-terminus sequence CQLF (CaaX motif) which is predicted to direct post-translation prenylation of the cysteine residue. A synthetic peptide containing this C-terminus motif, when tested against Q-Sepharose chromatography fractions from T. cruzi cytosol, was shown to be efficiently farnesylated but not geranylgeranylated, despite the fact that the CaaX motif with X = F specifies geranylgeranylation by mammalian protein geranylgeranyltransferase-I. Furthermore, immunoblot analyses of epimastigote protein with anti-S-farnesyl-cysteine methyl ester and anti-TcRho1 antisera strongly suggest that TcRho1 is farnesylated in vivo. The farnesylation of proteins such as Rho GTPases could be the basis for the selective cytotoxic action of protein farnesyltransferase inhibitors on trypanosomatids versus mammalian cells.


Introduction
for each hybridization, and plugs containing positive plaques in both of them were selected and submitted to secondary and tertiary screens. After three rounds of selection, three phage clones giving positive hybridization signals were selected. One of them, named TcRho1, was selected for a characterization of TcRho1.
Subcloning of the TcRho1 coding region. The TcRho1 clone was amplified in the LE392 E. coli strain. DNA from this clone was extracted and purified as described elsewhere (36) and was submitted to Southern blot analysis as described above. A 4.0 Kb EcoRI fragment was selected as an initial target for cloning. Ligation of EcoRI digested genomic cloned DNA with EcoRI digested and 5' dephosphorylated pKS-II+ Bluescript (Stratagene) followed by transformation of XL1-Blue E. coli, generated several recombinant clones. Plasmid DNA from these clones was extracted as described elsewhere (36), and those containing cloned fragments in the range of 4.0 Kb were selected, and their 5' and 3' ends were sequenced. As the TcRho1 coding region was not fully contained in the EcoRI fragment, we further subcloned a 300 bp KpnI fragment from λTcRho1 containing part of the 3'coding region of TcRho1. This clone was radiolabeled and used as a probe to clone a 1.3 Kb EcoRI/PstI fragment from λTcRho1 encompassing the remaining TcRho1 sequence.
Sequence analysis of TcRho1. Subcloned fragments in pKS-II+ were sequenced using different primers by the dideoxy chain termination method (39) using the T7 Sequencing kit (Amershan Pharmacia Biotech). We used the commercially available T3 and T7 sequencing primers and also three sequencing primers based on the TcRho1 sequence: G2 5' CGGAATTCCCGGTACCCCGCC 3' , G4: 5' GCGTGGGATGACCC 3' , G5: Mapping the 5'UTR of TcRho1. For mapping the TcRho1 5'UTR and for locating the transsplicing acceptor sites, we carried out mini-exon, semi-nested RT-PCR against T. cruzi RNA.
One primer was directed to the mini-exon sequence (ME: temperatures, a touch-down PCR program was used by decreasing the annealing temperature from 75 0 C to 57 0 C in 1 0 C steps. Then, 20 additional cycles were performed at 93 0 C for denaturing, 55 0 C for annealing and 72 0 C for extension followed by a 10 minutes extension step. On tenth of this first reaction was used as template in the second reaction with the same conditions, but with a conventional thermocycler program (30 cycles consisting of 93 0 C for denaturing, 55 0 C for annealing and 72 0 C for extension, followed by a 10 min extension step). Products of both reactions were resolved on a 2.5% agarose gel. Sites for EcoRI were present in the mini-exon and G2 primers in order to allow ligation of products from the second reaction into pKS-II+. Ligated products were introduced in E. coli XL1-Blue, amplified and sequenced, allowing an accurate mapping of TcRho1 trans-splicing acceptor sites. sec pulses interrupted by cooling on ice). The fusion protein was recovered from inclusion bodies using the urea solubilization protocol previous described for the Ras protein (40).
Purification of the fusion protein was accomplished by glutathione-sepharose chromatography according to the manufacturer's suggested procedure. The purified protein was cleaved with factor Xa (Amersham Pharmacia Biotech) at 10 U/mg fusion protein to release TcRho1 protein from the GST tag. Protein yield was measured by the Bradford quantification method (41). Preparations were analyzed for purity by 12% SDS-PAGE.
T. cruzi transfection. As described in the Results Section, we desired a strain of T. cruzi that overexpresses a mutant TcRho1 that cannot be prenylated. We had available a mutant of TcRho1 in which the C-terminal CQLF is replaced with FNFFDFA, and this mutant DNA fragment was used to construct the vector for overexpression of mutant TcRho1 in T. cruzi.
Overexpression was performed using the T. cruzi expression vector pBS:IL2-CnFc (42). The IL2 encoding insert was excised from the vector with BamHI and replaced with the TcRho1 ORF flanked by BamHI sites. A clone with properly oriented insert was identified, and 5 µg of supercoiled DNA was electroporated into Tulahuen epimastigotes as previously described (42). Transfectants were selected and expanded in 500 µg/ml of G418.
Immunoblotting. Antiserum to TcRho1 was raised in a rabbit against the KLH-conjugated peptide NDNGVVDTSNKQSIEL, present in the C-terminal hypervariable region.
Antiserum was submitted to affinity purification using the resin prepared by reacting the same peptide used for immunization with CNBr-activated Sepharose 4 Fast Flow (Amersham  of radiolabeled prenylated peptide was quantified using the avidin-agarose method as described (44). Fractions were also assayed for PFT activity with 5 µM of recombinant RAS1-CVIM (a generous gift from Dr. C. Omer, Merck) and for PGGT-I activity with 5 µM Ras-CVLL (a generous gift from Prof. G. James, University of Texas), using the glass fiber method (45).

Results
Genomic Organization of TcRho1. We have characterized several cloned RT-PCR fragments amplified from T. cruzi RNA which share homology in their predicted peptide sequences with several Ras superfamily genes (unpublished results). These products were obtained by degenerated RT-PCR, using a primer directed to the mini-exon sequence and a degenerated primer directed to the G3 conserved GTPase domain (DTAGQE). One of the obtained fragments, named pTcrho, shares about 40% similarity with several members of the Rho family proteins. We used pTcrho as a homologous probe in order to characterize TcRho1, a Rho family gene from T. cruzi.
Genomic DNA from Dm28c epimastigotes was digested with several restriction enzymes and probed with the pTcrho labeled fragment. Southern blot analysis revealed single bands, suggesting that TcRho1 is present as a single copy gene in the Dm28c haploid genome ( Figure 1A). This pattern resembles other characterized trypanosome GTPase genes, reinforcing the hypothesis of preferential organization of small GTPases in trypanosomes as single copy genes (18).  Table 1.
Sequence database searching using the tBLASTn algorithm (49) against ESTs from diverse organisms detected a T. cruzi EST (28j18 clone, accession number AI075525) sharing 100% identity with TcRho1. This clone spans part of the 3' region of TcRho1. A multiple alignment of TcRho1 and its closest homologues was performed using the CLUSTAL X program (50) ( Figure 3). As noted for many cloned sequences from trypanosomatids (27, 33), We mapped the 5'untranslated region of TcRho1 mRNA using a mini-exon seminested RT-PCR approach with a sense oligonucleotide directed to the mini-exon and antisense oligonucleotides directed next to the G3 region (in the first reaction) and to the G2 region (in the second reaction). Electrophoresis of PCR products revealed three major bands, corresponding in size to the two splice-leader acceptor sites in the 5'UTR of TcRho1 mRNA ( Figure 5A). These bands were not seen in a negative control using epimastigote RNA not submitted to reverse transcription (not shown). White arrows indicate the specific products generated in the reaction. PCR products were further cloned and sequenced, showing that the two largest fragments indeed correspond to the acceptor sites located at -85 and -39 in the In order to examine the type of prenyl group attached to TcRho1 in vivo, we used a recently described antiserum that recognizes S-farnesyl-cysteine methyl ester but fails to recognize the S-geranylgeranylated compound (43). This immunological method was used because of the impracticality of obtaining sufficient amounts of native TcRho1 from T. cruzi for direct prenyl group structure determination by radiometric or mass spectrometric methods (54) (RNA analysis described above suggests that TcRho1 is present at low levels in epimastigotes). For these experiments, we prepared a stable T. cruzi transfectant that overexpresses a mutant TcRho1 that cannot be prenylated (CQLF replaced with FNFFDFA, already available in our lab as described in Methods). This mutant protein serves as a gel position marker of TcRho1 from whole parasites and also serves to confirm the specificity of the anti-S-farnesyl-cysteine methyl ester for the farnesyl portion of TcRho1.
As shown in Figure 7, the immunoblot using anti-TcRho1 antiserum detects a protein band from whole parasites that co-migrates with recombinant TcRho1 produced in E. coli.
The observed apparent MW for these bands is ~39-kDa (predicted MW 31-kDa). The immunoblot from parasites that overexpress the TcRho1 mutant ( Figure 7) shows a ~10-fold increase compared to non-transfected parasites in amount of protein detected at the ~39-kDa position, thus supporting the assignment of this band as TcRho1. As shown in Figure 7, the immunoblot analysis of non-transfected parasites with the anti-S-farnesyl-cysteine methyl ester antiserum clearly shows a band at ~39-kDa, which co-migrates with the band detected with anti-TcRho1 antiserum. The intensity of this band did not increase when transfected parasites were analyzed with the anti-S-farnesyl-cysteine methyl ester antiserum (Figure 7). This latter result shows that the farnesyl group but not the protein component of wild type, endogenous TcRho1 is being detected. These immunological studies strongly support the farnesylation of TcRho1 in vivo, which is consistent with the in vitro data with T. cruzi PFT.
As shown in Figure 8,

Discussion
TcRho1 is the first Rho family member from trypanosomatids to be identified, albeit some of Ras superfamily genes have been cloned in these organisms (18). TcRho1 has conserved GTPase motifs and a C-terminus CaaX motif that is a target for post-translational prenylation. Phylogenetic analysis shows that Tchro1 clearly belongs to the Rho family clade of GTP binding proteins, however, it does not seem to branch within Rho or Rac/Cdc42 subgroups, apparently having diverged from the clade before the division between Rho and Rac/Cdc42 proteins. As trypanosomatids are believed to have branched early on eukaryotic evolution, this GTPase may be an ancestral Rho family member of higher organisms.
Another monomeric GTPase protein, the Ras/Rap protein found in T. brucei, also branched in a similar way (27).
Interestingly, five trans-splicing sites were mapped in TcRho1 mRNA, three of them lying inside the coding region ( Figure 5). As far as we know, these are the first naturally 22 The results in this study could explain why PFT inhibitors are highly cytotoxic to trypanosomatids (16,32). In fact, Rho family proteins are important regulators of mammalian cell growth and morphology, and it has been shown that mammalian cell growth is much more sensitive to PGGT-I inhibitors than to PFT inhibitors (67)