Expression of Functional Schistosoma mansoni Smad4

Members of the transforming growth factor (TGF)-β superfamily play pivotal roles in cell migration, differentiation, adhesion, pattern formation, and apoptosis. The family of Smad proteins acts as intracellular signal transducers of TGF-β and related peptides. Smad4, a common mediator Smad (co-Smad), performs a central role in transmitting signals from TGF-β, BMP, and activins. Schistosoma mansoni receptor-regulated Smad1 and SmSmad2 were previously identified and shown to act in TGF-β signaling. Herein, we report the identification and characterization of a Smad4 homologue from S. mansoni and provide details about its role in mediation and down-regulation of TGF-β signaling in schistosomes. In order to identify the schistosome co-Smad, we designed degenerate primers based on the sequence of the conserved MH1/MH2 domains of Smad4 proteins, which were used in PCR to amplify a 137-bp PCR product. A S. mansoni adult worm pair cDNA library was screened resulting in the isolation of a cDNA clone that encodes a 738 amino acid protein (SmSmad4). SmSmad4 was shown to interact with schistosome R-Smads (SmSmad1 and SmSmad2) in vivo and in vitro. The interaction with SmSmad2 was dependent on the receptor-mediated phosphorylation of SmSmad2. In addition, several potential phosphorylation sites for Erk1/2 kinases were identified in the SmSmad4 linker region and shown to be phosphorylated in vitro by an active mutant of mammalian Erk2. Furthermore, Erk-mediated phosphorylation of SmSmad4 decreased its interaction with the receptor-activated form of SmSmad2, in vitro. SmSmad4 was shown to complement a human Smad4 deficiency through the restoration of TGF-β-responsiveness in MDA-MB-468 breast cancer cells.

Helminths of the genus Schistosoma are the causative agents of schistosomiasis, an endemic tropical disease affecting about 200 million people worldwide, which is a major cause of morbidity (www.who.int/tdr/diseases/schisto/diseaseinfo.htm). It is generally thought that the schistosome receives host signals that are utilized in parasite development. Thus, host-parasite interactions are likely to have co-evolved and to be selective for both the intermediate and the definitive host (1,2). In the definitive vertebrate host, the parasite migrates from the site of infection to the final destination in a complex journey that proceeds with concomitant development and maturation, likely guided by host factors (3,4). Furthermore, female worm sexual maturation and subsequent egg production, the major cause of the pathological consequences of schistosomiasis, is induced by mating with male worms. Female sexual development is likely to be stimulated by a set of chemical and/or mechanical signals from the male (5). Therefore, an investigation of signal transduction pathways in schistosomes may permit us to understand the mechanisms by which the parasite receives and responds both to self-and to host-derived signals. Such studies will yield insights into the host-parasite relationship and male-induced female maturation.
Prior investigations identified growth factor receptors and orthologues of components of signaling pathways in schistosomes (6 -12). Transforming growth factor ␤ (TGF-␤) 1 is an example of a signaling pathway that affects a wide variety of cell types and regulates different vital processes such as cell growth, differentiation, morphogenesis, and apoptosis (13)(14)(15). The intracellular signal transmission in TGF-␤ pathway is initiated through ligand-induced formation and activation of a heteromeric receptor complex of type II and type I serine/ threonine kinase receptors located at the cell surface. This process involves the binding of a TGF-␤ ligand to the constitutively phosphorylated type II kinase receptor, which triggers the interaction with and phosphorylation and activation of the corresponding type I receptor (16). The activated receptor complex in turn relays the signal to the downstream member, the receptor-regulated Smad (R-Smad). Activation of R-Smads occurs via direct phosphorylation of the two most C-terminal serine residues ((T/S)SXS motif) by the specific type I receptor (17,18). R-Smads are recruited to different activated receptor complexes, depending on the activating ligand. Smad1, Smad5, and Smad8 are activated and phosphorylated by receptors of the bone morphogenetic proteins (BMPs) subfamily (18 -21); while Smad2 and Smad3 receive the signal relayed from activated receptors of the TGF-␤ and activin subfamilies (17,19,22,23). Upon activation by the receptor complex, R-Smad forms a hetero-oligomeric complex with a common Smad (co-Smad). The co-Smad, called Smad4 in vertebrates, or Medea in Drosophila, acts as a shared partner for both BMP-specific and TGF-␤/activins-specific R-Smads (24,25). Smad4 was origi-* This work was supported by the United Nations Development Programme/World Bank/World Health Organization Special Programme for Research and Training in Tropical Diseases Grant A20357 and National Institutes of Health Grant AI46762. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) AY371484. nally identified as a tumor suppressor gene that is deleted or mutated in pancreatic carcinomas, and is also known as DPC4 (deleted in pancreatic carcinoma locus 4) (26,27). Although Smad4 shares many sequence similarities with R-Smads, it lacks the phosphorylation signature located at the C-terminal end of the MH2 domain of R-Smads and thus neither associates with the receptor complex nor acts as a substrate for receptor phosphorylation (17,28). The newly formed Smad4/R-Smad complex then translocates to the nucleus where it binds to promoter sequences and modulates the transcription of a subset of responsive genes required to exert the specific TGF-␤ ligand effect (29,30). In addition to the receptor complex and the corresponding R-Smad, signal specificity is determined by the nuclear partner that binds and directs Smad4/R-Smad complex to the target gene(s). Smad proteins are considered crucial members in the TGF-␤ signaling since they receive the signal from the activated receptor complex at the cell surface and relay it to the nucleus where the specific outcome/phenotype will be orchestrated.
Structurally, R-Smads and co-Smad consist of 2 conserved domains, located at the N and C termini of the proteins, called Mad homology domains 1 and 2 (MH1 and MH2), respectively, linked by a highly variable, proline-rich linker region (31). The MH1 domain mediates DNA binding (32)(33)(34) and negatively regulates the functions of the MH2 domain (35). The MH1 domain also contains a cluster of basic residues often followed by hydrophobic aliphatic and acidic residues, which serve as a nuclear localization signal (NLS) (36 -38). The MH2 domain is responsible for transcription activation (21,39,40) and interactions with other Smads, receptor kinases, and/or nuclear partners (41)(42)(43). The linker region of R-Smads contains PX(S/T)P phosphorylation motif(s), the consensus phosphorylation site for the Ras-activated mitogen-activated protein kinases Erk1 and Erk2 (44). Erk phosphorylation results in blocking the nuclear accumulation of R-Smads and consequently down-regulating TGF-␤ signaling (45)(46)(47). The N-terminal end of the linker region also contains a leucine-rich nuclear export signal (NES), which mediates the nuclear export of Smad4. The C terminus of the linker region of Smad4 contains a proline-rich transcription activation domain (Smad4 activation domain; SAD) that is not only necessary but also sufficient to activate maximal Smad-dependent transcriptional responses (35,48,49).
Recent studies showed that schistosomes contain several members of the TGF-␤ pathway. A TGF-␤ type I receptor homologue, (SmRK1 and called SmT␤R-I thereafter) was the first member to be identified and shown to be localized to the surface of the parasite (8). The identification of the schistosome type I receptor spurred the search for other members of TGF-␤ pathways in the parasite S. mansoni, which resulted in the identification of 2 members of R-Smad subfamily, SmSmad1 (10) and SmSmad2 (10,11). Interestingly, SmSmad2 was shown to interact with and act downstream of SmT␤R-I, where it serves as a substrate for receptor phosphorylation and translocates into the nucleus of mammalian cells after induction with human TGF-␤ (10, 11). The above data suggest the presence of a common Smad homologue in schistosomes and underscores the significance of pursuing the search for SmSmad4. In this study, we report the identification and characterization of SmSmad4, the common Smad homologue of S. mansoni, and we demonstrate its interaction with schistosome R-Smads.

EXPERIMENTAL PROCEDURES
Identification and Isolation of SmSmad4 cDNA-Protein sequences of Smad4 homologues were aligned using Pileup (Genetics Computer Group, GCG-Wisconsin Package, version 9.1) to define areas of high conservation in the MH1 and MH2 domains. Degenerate primers were synthesized, based on the amino acid sequence of the homologous regions, and were employed in different combinations in PCR using S. mansoni adult worm pair cDNA. One primer pair spanning a region in the MH1 domain (Fig. 1A) amplified a 137-bp PCR product, DPC4/F3 and R4 (5Ј-CCNCAYGTNRTNTAYGCNMGNHTNTGGMGNTGGCC-3Ј and 5Ј-CANACNCKYTCRTARTGRTANGGRTT-3Ј, respectively). Based on the DNA sequence of the 137-bp product, two specific primers (DPC-fwd and rev; bp 412-435 as forward primer and the complementary sequence of bp 467-489 as reverse primer, respectively), were designed downstream of F3 and R4 primer sequences, amplified a 78-bp-specific cDNA product. The later DNA fragment was radiolabeled with [␣-32 P]dCTP using MegaPrime random priming labeling kit (Amersham Biosciences) and used to screen 400,000 plaques from a nonamplified S. mansoni adult worm pair -ZapII cDNA library.
Cloning and Expression of SmSmad4 -DNA and protein sequence analyses were performed using programs in the GCG-Wisconsin Package, version 9.1. Sequence analysis of the SmSmad4 cDNA revealed the presence of a unique BglII site 742-bp downstream of the vector EcoRI cloning site. A 5Ј-primer (bp 47-72), in which the C 49 was replaced by an A creating an EcoRI site upstream of the start ATG codon, was used as a forward primer along with a reverse primer (complementary sequence of bp 752-774) to amplify a 728-bp product. The PCR product was cloned in pCR2.1-TOPO cloning vector (Invitrogen), sequenced to confirm the absence of any PCR-generated errors, digested with EcoRI and BglII and recloned into the parental pSmSmad4-BlueScript-SKϩ vector to generate a modified version of the cDNA clone in which the sequence upstream of the start ATG codon was removed and an EcoRI site was inserted just before and in-frame with the start codon. The modified vector, pSmSmad4-cod-BlueScript-SKϩ, was digested with EcoRI and XhoI to yield a SmSmad4 DNA fragment suitable for cloning in several vectors including pGEX-4T-1 (Amersham Biosciences; digested with EcoRI-XhoI), pMAL-c2x (New England Biolabs; digested with EcoRI-SalI) for prokaryotic expression; pADGAL4 -2.1 (GAL4-activation domain vector, Stratagene; digested with EcoRI-XhoI), pBDGAL4-cam (GAL4-DNA binding domain vector, Stratagene; digested with EcoRI-SalI) for yeast two-hybrid analysis, pcDNA3.1-His (Invitrogen; digested with EcoRI-XhoI) for expression in mammalian cells, and pCITE-4a (Novagen; digested with EcoRI-XhoI) for in vitro transcription/translation (see Table S1, online Supplemental Data).
Production of Anti-SmSmad4-specific Antisera-In addition to the full-length constructs, a DNA fragment encoding the SmSmad4-linker region (bp 401-1183; 261 amino acids) was also amplified, cloned in pCR2.1-TOPO, sequenced and cloned in pMAL-c2x vector for expression of a fusion protein that shares no homology to other schistosome Smad proteins. MBP-SmSmad4-Linker fusion protein was used to immunize a New Zealand rabbit and BALB/c mice. A dose of 200 g of the fusion protein in Freund's complete adjuvant was used to immunize the rabbit as a primary dose and 200 g in Freund's incomplete adjuvant were used for 2 booster doses at 4-week intervals. An activating dose (200 g in 1ϫ phosphate-buffered saline) was used 7 days prior to sacrificing the animal. Mice were immunized with the same reagents following same time frame except 20 g/animal were used for each dose. Sera were titered by ELISA and IgG fractions were affinity-purified over protein A-Sepharose resin (Amersham Biosciences) and quantified. Preimmune rabbit and mice sera were processed similarly to provide reagents for negative controls. Affinity-purified IgG was used for immunoprecipitation, immunofluorescence, and Western blot analyses.
Western Blotting and Immunofluorescence-Protein extracts were prepared from S. mansoni adult worm pairs as previously described (11). 20 g were loaded per lane on 4 -12% gradient SDS-polyacrylamide gels (Invitrogen) and size-separated. SDS gels were blotted onto polyvinylidene difluoride membranes (Immobilon P; Millipore), and the blots were probed with 0.75 g/ml of either preimmune IgG or anti-Smad4-linker IgG. Biotinylated secondary antibody reagents (goat antirabbit and goat anti-mouse IgG, 0.75 g/ml; Molecular Probes, Inc.) were used to probe the reactive primary antibodies and the immune complexes were detected by alkaline phosphatase-conjugated streptavidin (1 g/ml; Molecular Probes, Inc.) and alkaline phosphatase substrate kit (NBT/BCIP Vector kit, Vector Laboratories).
Affinity-purified IgG (5 g/ml) was also used to localize the native SmSmad4 in adult worm cryosections, probed with biotinylated secondary antibody reagents (5 g/ml; Molecular Probes, Inc.) and detected with AlexaFluor 647-conjugated streptavidin (5 g/ml; Molecular Probes, Inc.). AlexaFluor 647 is a far-red fluorochrome that emits at a maximum wavelength of 647 nm. At this wavelength, auto-fluorescence produced by phenolic compounds in schistosome sections would not be visualized. Probed sections were evaluated using a Bio-Rad MRC-1024 confocal microscope equipped with Krypton-Argon laser and 522 nm and 680 nm filters.
RT-PCR Analysis-Analysis of SmSmad4 mRNA levels in different developmental stages was performed by semiquantitative RT-PCR. Total RNA was extracted from different developmental stages, representing growth in both mammalian and molluscan hosts, using TRIzol reagent (Invitrogen). All RNA samples were treated with RNase-free DNaseI (RQ1 DNase; Promega) and reverse transcribed using a random hexamer, and SuperScript Reverse Transcriptase II (SSRTaseII; Invitrogen) following the vendor's recommended conditions. Reverse-transcribed cDNA samples were used as templates in PCR reactions using specific primer pairs. PCR reactions were separated by electrophoresis in 2% agarose gels, ethidium bromide-stained, and analyzed using a gel-documentation system (GelDoc1000; Bio-Rad) and quantified using Quantity One software (version 4.2.3; Bio-Rad). Control PCR reactions using reverse transcription reaction mix lacking SSRTaseII were also performed to evaluate the efficiency of DNase treatment. Specific primers for S. mansoni ␣-tubulin gene (GenBank TM accession no.: M80214; bp 424 -444 and the complementary sequence of bp 777-801 as forward and reverse primers, respectively, yielding a 378-bp PCR product) were used to amplify a PCR product that served as a constitutively transcribed control (50). This primer pair amplifies a DNA fragment of about 410 bp when S. mansoni genomic DNA is used as a template. Therefore, in addition to its role as an amplification control to which PCR products for other genes were compared, ␣-tubulin served also as an additional control to assess the efficiency of DNase treatment. In order to compare SmSmad4 mRNA levels in different developmental stages with those of S. mansoni R-Smads, specific primer pairs representing C-terminal end of SmSmad4 (bp 1866 -1888 and the complementary sequence of bp 2240 -2269 as forward and reverse primers, respectively, yielding a 404-bp PCR product), linker region of SmSmad1 (GenBank TM accession no. AF215933; bp 454 -476 and the complementary sequence of bp 761-780 as forward and reverse primers, respectively, yielding a 339-bp PCR product) and part of the MH2 domain of SmSmad2 (GenBank TM accession no. AF232025; bp 1614 -1626 and the complementary sequence of bp 1902-1926 as forward and reverse primers, respectively, yielding a 313-bp PCR product) were used to amplify the specific PCR products which were separated by electrophoresis and processed as described above. Due to differences in abundance of cDNA species included in this assay, amounts of input template were varied according to the stage and/or the gene under study. Also, in order to ensure that the amplification products were analyzed in the exponential phase and below saturation limits (PCR plateau), the number of PCR cycles was also varied. 26 cycles were used for ␣-tubulin, 27 cycles for SmSmad4 and SmSmad2 while SmSmad1 was cycled for 29 times. All variables were considered and compensated for in data analysis. SmSmad4 Interactions with Schistosome R-Smads-Similar to the situation in higher eukaryotes, schistosome Smad4 was expected to interact with SmSmad1 and SmSmad2. In order to study these interaction patterns, a series of expression vectors (prokaryotic, yeast, and in vitro transcription/translation) of SmSmad4, SmSmad2, SmSmad1, and SmT␤R-I were constructed and used.
(i) Vectors- Table S1 (see online Supplemental Data) lists the plasmid vectors, insert description, and the interaction assay(s) in which they were used. SmSmad4 was expected to interact with SmSmad2, for example, upon receptor activation; therefore, SmT␤R-I was included in the experimental design of the in vivo and in vitro assays. The yeast expression vector (pYES2/NT-C, Invitrogen), which carries Ura3 yeast selection marker and the expression of recombinant gene is derived by a galactose-inducible GAL1 promoter (P GAL1 ), was modified to replace the P GAL1 with the glucose-inducible promoter, alcohol dehydrogenase (P ADH1 ), which is used in the GAL4 yeast two-hybrid vectors resulting in pADH-YES. The vector pYES2/NT-C was digested with AgeI, which removed most of the P GAL1 and the multiple cloning sites, then bluntended by treatment with T4 DNA polymerase in the presence of dNTP. A DNA fragment spanning part of the Saccharomyces cerevisiae ADH1 promoter (433 bp) was amplified from the vector pBDGAL4-cam using high fidelity Platinum TaqDNA polymerase (Invitrogen) with a primer pair representing bases 1-21 and the complementary sequence of bases 407-433 as forward and reverse primers, respectively. In addition to a HindIII site normally present at the 3Ј-end of the reverse primer, AgeI site was also added at its 5Ј-end. PCR product was cloned into the blunt ended vector using Fast Link ligation kit (EpiCentre) and the orientation was confirmed. The new vector was then digested with HindIII and AgeI, which removed the sequence of the ADH reverse primer, and a 220-bp HindIII-AgeI fragment, excised from the parent pYES2/NT-C, was ligated to regenerate the multiple cloning sites back into the new vector. SmT␤RI-wt and SmT␤RI-Q/D were cloned into this modified vector. Both vectors were then used in the yeast interaction assays to provide a third member (three-hybrid system) in which, expression was under control of the same promoter (P ADH1 ), and selection was provided by Ura3.
(ii) In Vivo Interaction, Yeast Two-hybrid and Three-hybrid Assays-In the yeast two-hybrid assays, yeast host strains AH109 (BD Biosciences-Clontech) and YRG2 (Stratagene) were used. The positive interactions were judged by activation of HIS3/ADE2 reporter genes for AH109, as determined by the ability of transformed cells to grow on selective medium (synthetic-dextrose; S.D.) lacking leucine, tryptophan, histidine, and adenine in the presence of 2.5 mm 3-amino 1,2,4 triazole (an inhibitor of HIS3 gene in the host cells; 3-AT) (S.D. -Leu, -Trp, -His, -Ade, 2.5 mm 3-AT). Positive interactions in YRG2 cells were assessed based on the activation of HIS3 reporter gene, as indicated by growth on the selective medium S.D. -Leu, -Trp, -His. In both strains, when activated, the LacZ reporter gene yielded blue colonies in the LacZ colony-(filter-) lift assay. The yeast strains PJ69 -4a (51) and Y187 (BD Biosciences-Clontech) were used for the yeast three-hybrid assay and a positive interaction was assessed by activation of HIS3 and ADE2 reporter genes, permitting growth on the selective medium S.D.
-Leu, -Trp, -His, -Ade, -Uracil (-Ura), 2.5 mm 3-AT for PJ69 -4a strain and by activation of LacZ reporter gene in colonies grew on the selective medium S.D. -Leu, -Trp, -Ura for yeast strain Y187. In this assay, the use of the recombinant pADH-YES plasmids allowed growth in absence of uracil. Activation of the reporter genes was assessed in absence or the presence of uracil to provide a comparative measure to determine the effect of the expression of the third protein on the interaction of the GAL4-AD and -BD fusion proteins. The degree of interaction was quantified by performing liquid LacZ assay on the transformed colonies of both strains, in addition to growth on selective medium (for PJ69 -4a cells) and the intensity of color of the LacZ filter lift assay (for Y187 cells). Procedures for yeast culture media preparations and yeast cell manipulations were performed following the recommended instructions of the suppliers (BD Biosciences-Clontech and Stratagene).
(iii) In Vitro Interaction, GST Pull-down, and Co-immunoprecipitation-R-SmSmads/SmSmad4 interactions were evaluated by immunoprecipitation in the presence or absence of wild type or constitutively activated SmT␤R-I. The first step was to assess the interaction status of the 2 forms of SmT␤R-I with SmSmads. S-tagged, unlabeled SmT␤R-I-wt or SmT␤R-I-Q/D (10 l) was allowed to interact with 35 S-labeled, non-S-tagged SmSmads (5 l) for 1 h at room temperature, and complexes were precipitated by the addition of S-protein agarose beads (30 l; Novagen). Protein-bound beads were washed, resuspended into SDS gel loading buffer, and size-separated in 4 -12% gradient SDS-polyacrylamide gels. The gels were treated with Amplify (Amersham Biosciences), dried, and exposed to x-ray film.
GST pull-down assays were performed to evaluate the interaction of SmSmad4 with SmSmad2, as described previously (11). Briefly, [ 35 S]methionine-labeled SmSmad4 (5 l) was allowed to interact with GST-SmSmad2 or GST-SmSmad2-3Ј-bound glutathione-Sepharose beads (ϳ2 g), in the presence or absence of SmT␤RI-wt or SmT␤RI-Q/D synthesized by in vitro translation (10 l). Binding reactions were incubated overnight at 4°C with constant agitation. The beads were washed and processed as above. GST-bound beads and GST-SmSmad2-Linker-bound beads were similarly processed to serve as negative controls.
Co-immunoprecipitation assays were conducted using unlabeled in vitro translated SmSmad4 (10 l), with [ 35 S]methionine-labeled SmSmad2-MH2, SmSmad2-MH2/AAA or full-length SmSmad1 (5 l), in the presence or absence of in vitro translated, unlabeled SmT␤RI-wt or SmT␤RI-Q/D (10 l). Reactions were incubated for 1 h at room temperature, then anti-MBP-SmSmad4-Linker rabbit IgG (3 g) was added, and incubation was extended for an additional 30 min at room temperature. Protein complexes were precipitated with rProtein A Sepharose beads (Amersham Biosciences, 20 l). The reactive beads were washed and processed as described above.
Erk-mediated in Vitro Phosphorylation of SmSmad4 -An activated form of the recombinant rat skeletal muscle MAP kinase (Erk2; Calbiochem) was used to evaluate the ability of Erk kinase to phosphorylate MBP-SmSmad4 or MBP-SmSmad4-Linker in vitro. Kinase assays were performed using 1 g of purified MBP, MBP-SmSmad4 or MBP-SmSmad4-Linker with 50 units of activated Erk2 in the presence of 20 mm HEPES, 1 mm MnCl 2 , 1 mm MgCl 2 , 1 mm dithiothreitol, 1 mm sodium vanadate, and 10 Ci of [␥-32 P]ATP, in a 20-l reaction volume. Reactions were incubated at 37°C for 30 min, and 1 l was mixed with 10 l of SDS-loading buffer and separated onto a 4 -12% SDS gradient polyacrylamide gels. The gels were dried and exposed to x-ray film. Activated Erk2 was also used to treat the in vitro translated, unlabeled SmSmad4 to determine the effect of Erk phosphorylation on the interaction of SmSmad4 with R-Smads. The co-immunoprecipitation assay described above was repeated in the presence or absence of 50 units of activated Erk2, and samples were processed as described before.
TGF-␤1 Responsiveness of SmSmad4 Protein-To determine the re-sponsiveness of SmSmad4 to human TGF-␤ ligands, SmSmad4 was cloned in the mammalian expression vector, pcDNA3.1-His (Invitrogen). Mink lung epithelial cells, Mv1Lu (ATCC CCL-64) and the Smad4-deficient human breast cancer cell line, MDA-MB-468 (ATCC HTB-132) were used and maintained according to ATCC recommended instructions. Double CsCl-purified pSmSmad4-cDNA-His was used to transfect the two cell lines along with the TGF-␤-responsive reporter plasmid vector, p800-Luc, (kindly provided by Dr. D. Loskutoff) in which 800 bp of the promoter region of the TGF-␤-positively regulated gene, plasminogen activator inhibitor 1 (PAI-1) was cloned upstream of firefly luciferase reporter gene (52). SmSmad4 was also transfected along with the pCAL2 reporter vector (Kindly provided by Dr. Rik Derynck), which contains the luciferase reporter gene under control of the human cyclin A promoter, a TGF-␤-negatively regulated gene (53). LipofectAMINE 2000 reagent (Invitrogen) was used to transfect the mammalian cells following the recommended manufacturer's instructions. Transfected cells were incubated with transfection mix, in triplicate, after 6 h, the mix was changed to complete culture media containing 10% fetal bovine serum and incubated for an additional 18 h. The cells were serum-deprived overnight in culture media containing 0.1% fetal bovine serum and either left untreated or treated with 0.5 nM recombinant human TGF-␤1 (rhTGF-␤1; R&D Systems) for 4 h. Cells were washed, harvested in lysis buffer, and assayed for luciferase activity. The constitutively expressed reporter construct, pRL-CMV (Promega), which uses the immediate early CMV promoter to drive the expression of sea pansy (Renilla reniformis) luciferase gene, was co-transfected in all cells to serve as an internal control for normalizing the differences in luciferase activity due to transfection efficiency. The dual luciferase assay kit (Promega) was used to determine the luciferase activities in transfected cells following the manufacturer's instructions employing an Orion MPL2 microplate luminometer (Berthold Detection Systems) to measure the luminescence in transfected samples.

RESULTS
Identification and Isolation of SmSmad4 cDNA-Mad homology regions (MH1 and MH2) of Smad4 homologues from different species, were aligned to define areas of high conservation. Several degenerate primers were designed and used in PCR with S. mansoni adult worm cDNA yielding a PCR product that shared homology to Smad4. A primer pair located downstream of the original degenerate primers amplified a 78-bp DNA fragment, which was used as a probe to screen an aliquot of the original, non-amplified ZapII cDNA library (4 ϫ 10 5 independent clones), yielding 4 clones, 2 of which were found to contain the entire coding sequence of S. mansoni homologue of Smad4 (SmSmad4). The isolated cDNA clone is 3146 bp in length and encodes a 738 amino acid protein. Like its Drosophila counterpart, the schistosome Smad4 protein has about 200 amino acids more than Smad4 homologues from most other species. An NCBI BLASTP search showed that SmSmad4 exhibits the highest homology with Drosophila Me-dea for the MH1 domain (homology score: 215), while its MH2 domain exhibited the highest homology hit to the mammalian counterparts, mouse and human (homology scores: 320 and 318, respectively) ( Table 1). Analysis of the protein sequence revealed the conservation of the Smad4 structural features, such as the NLS and DNA-binding motif in the MH1 domain and the NES in the N-terminal end of the linker region (Fig.  1A). The NLS as well as the NES are well conserved in SmSmad4 with the retention of all basic and leucine residues constituting the cores of the NLS and NES domains, respectively (Fig. 1A). The corresponding SAD sequence in SmSmad4 (Fig. 1B) although showing very little sequence conservation, retains a proline-rich pattern. Interestingly, the SmSmad4 linker region was found to contain 3 Erk1/2 phosphorylation motifs (PX(S/T)P) (Fig. 1B) compared with one or two motifs present in vertebrate homologues and none in Drosophila Medea. None of schistosome R-Smads, SmSmad1 (GenBank TM accession no. AF215933), SmSmad2 (GenBank TM accession no. AF232025), and the recently identified Smad8/9 homologue (54) contain Erk phosphorylation motifs.
Western Blotting and Immunohistochemistry-Affinity-purified IgG fractions of mice and rabbit sera, raised against the SmSmad4 linker region, were used to detect the native protein in extracts and in cryosections of adult worms. The apparent molecular size of the native SmSmad4 is ϳ78 kDa as determined by SDS-polyacrylamide gel and Western blot analyses (data not shown). The native protein was localized to the epithelial tissues surrounding the gut and vitelline lobules in female worms (Fig. 2E) and in the subtegumental tissues and muscle layers of male worms (Fig. 2G). IgG fractions of preimmune sera were used as a negative control (Fig. 2, A-C). As shown in the negative control sections, auto-fluorescence, observed in females using a 522-nm filter, attributed to the phenolic compounds present in vitelline cells (Fig. 2B) was not observed at the far-red wavelength (680-nm filter) used to visualize the AlexaFluor-647 streptavidin conjugated to the reactive antigens (Fig. 2C).
RT-PCR Analysis-SmSmad4 mRNA levels were determined in different stages in the parasite life cycle comprising both mammalian and molluscan hosts. Compared with ␣-tubulin mRNA, SmSmad4 mRNA levels exhibited little variation throughout development of the parasite (Fig. 3C and Fig. 4A) including sexually immature parasites (single-sex female and male worms) (Fig. 3, B and C, lanes 16 -18, and Fig. 4, A and B,  lanes 7 and 8). In contrast, in the infected snail stage, which represents secondary sporocysts and to some extent the freeliving infective cercarial stage, SmSmad4 showed a relatively  low level (Fig. 3, B and C, lane 4 and Fig. 4, A and B, lane 3). R-Smads (SmSmad1 and 2), however, exhibited comparable profiles, although at different levels, only in mammalian host stages (Fig. 4, A and B, lanes 4 -8). In parasite eggs as well as stages representing development in molluscan host (primary sporocysts and infected snail stages), the mRNA levels of SmSmad1 and SmSmad2 were ϳ25-50% of their average mRNA levels in mammalian host stages (Fig. 4, A and B, lanes 1-3).
In Vivo Interaction of SmSmad4 with Schistosome R-Smads-SmSmad4 interaction with SmSmad1 and 2 was assessed in vivo by using the yeast two-hybrid and yeast three-hybrid assays. In the yeast two-hybrid assay, interaction was evaluated in absence of SmT␤R-I, whereas in the three-hybrid assay SmT␤R-I, either wild type (wt) or the constitutively activated form (Q/D), was included. Table 2, left column shows that, SmSmad2 and its non-phosphorylatable form (SmSmad2-3Ј) weakly interact with SmSmad4. Weak interaction was judged by the ability of few, but not all cells, to grow under selective conditions as compared with the number of colonies grown on the control plate (-Leu, -Trp). Those colonies also produced a faint blue-green color in the LacZ assay that required incubation for up to 3 days. The MH2 domain of SmSmad2, its AAAmutant form, and SmSmad1 showed greater interaction with SmSmad4, as assessed by the number of colonies grown under selective conditions and the intensity of blue color produced in the LacZ assay. When SmT␤R-I was included in the assay, it significantly affected the SmSmad4 interaction pattern with R-Smads. In the presence of the wild-type form of receptor I ( Table 2, middle column), none of SmSmad2 constructs showed detectable interaction with SmSmad4, while SmSmad1 interaction was still apparent at a decreased level. Thus, wild-type SmT␤R-I squelches the interaction between SmSmad4 and all forms of SmSmad2. In contrast, inclusion of the constitutively active form of receptor I, SmT␤R-I-Q/D, significantly elevated SmSmad4 interaction with either wild-type full-length SmSmad2 or its wild-type MH2 domain. Neither of the nonphosphorylatable constructs, SmSmad2-3Ј and MH2-AAA, showed detectable interaction with SmSmad4 in the presence of SmT␤R-I-Q/D ( Table 2, right column). These results demonstrate that phosphorylation of SmSmad2 by T␤R-I-Q/D stimulates binding to SmSmad4. In contrast, the SmSmad1 interaction with SmSmad4 was significantly inhibited in the presence of SmT␤R-I-Q/D.
In Vitro Interaction among Schistosome Smads-In order to further evaluate the interaction of SmSmad4 with SmSmad1 and SmSmad2, and the effect of SmT␤R-I on this interaction, an in vitro approach was designed. Initially, the interaction of SmT␤R-I with SmSmads was evaluated by co-precipitation. Both the wild-type and the constitutively activated form of SmT␤R-I interacted with all R-Smad constructs, but not with SmSmad4 (Fig. 5A). Wild-type SmSmad2 and its MH2 domain interacted with SmT␤R-I-wt more than SmT␤R-I-Q/D (Fig. 5A,  panels I and II, lanes 3, 4, 7, and 8), while SmSmad1 showed stronger interaction with SmT␤R-I-Q/D (Fig. 5A, panels I and  II, lanes 9 and 10). SmSmad2-MH2/AAA exhibited slight preference toward the Q/D form (Fig. 5A, panels I and II, lanes 1  and 2).
The direct interaction of SmSmad4 with SmSmad2 was evaluated by a GST pull-down assay. GST alone and the SmSmad2linker control exhibited minimal binding with SmSmad4 in the presence or absence of SmT␤R-I (Fig. 5B, panels I and II). SmSmad2-GST showed basal level of interaction with SmSmad4 (Fig. 5B, panel III, lane 1), which was significantly enhanced in the presence of SmT␤R-I-Q/D (Fig. 5B, panel III,  lane 3), while SmT␤R-I-wt had less effect (Fig. 5B, panel III,  lane 2). However, addition of SmT␤R-I completely abolished the already weak interaction of the SmSmad2-3Ј with SmSmad4 (Fig. 5B, panel IV).

FIG. 1-continued
Similar results were also observed in co-immunoprecipitation assays. SmSmad2-MH2 and its non-phosphorylatable AAA-mutant form showed a comparable basal level of interaction with SmSmad4 (Fig. 5C, panels I and II, lane 3). In the presence of SmT␤R-I-Q/D, the interaction of SmSmad4 with SmSmad2-MH2 was significantly enhanced, while it exhibited a minor decrease with SmSmad2-MH2/AAA (Fig. 5C, panels I  and II, lane 5). On the other hand, the addition of wild-type receptor I (SmT␤R-I-wt) produced a modest change in the interaction of SmSmad4 with SmSmad2-MH2 and with MH2/ AAA (Fig. 5C, panels I and II, lane 4). SmSmad1 readily interacted with SmSmad4 (Fig. 5C, panel III, lane 3). However, both forms of SmT␤R-I inhibited the SmSmad1 interaction with SmSmad4 (Fig. 5C, panel III, lanes 4 and 5). This can be attributed to the observed interaction between SmSmad1 and both forms of type I receptor (Fig. 5A, lanes 9 and 10).
Erk-mediated in Vitro Phosphorylation of SmSmad4 -Analysis of the SmSmad4 protein sequence revealed the presence of 3 possible Erk1/2 phosphorylation sites in the linker region and extending to beginning of the MH2 domain. MBP-fusion proteins of both full-length SmSmad4 and its linker region were tested as substrates for the activated Erk2 kinase in vitro (Fig.  6A, lanes 5 and 8). The effect of Erk2 phosphorylation on the interaction of SmSmad4 with R-Smads was also tested. Two sets of reactions were assayed in the absence (Fig. 6B, lanes [3][4][5] or in the presence of Erk2 (Fig. 6B, lanes 6 -8). Effect of Erk2 treatment was assessed by comparing similar samples from the 2 sets. Interestingly, although Erk2 treatment has no effect on the interaction of SmSmad4 with SmSmad2 or its MH2 domain in absence of SmT␤R-I (Fig. 6B, panels I and II,  lanes 3 and 6), activated Erk2 significantly inhibited SmSmad4 interaction with SmSmad2-wt and SmSmad2-MH2 in the presence of the SmT␤R-I-Q/D (Fig. 6B, panels I and II, lanes 5 and  8). In case of SmSmad1, Erk2 phosphorylation of SmSmad4 had little effect on SmSmad1 interaction with SmSmad4 in the presence or absence of SmT␤R-I (Fig. 6B, panel III).
SmSmad4 Participates in TGF-␤ Signaling in Mammalian Cells-The ability of SmSmad4 to restore TGF-␤ responsiveness in Smad4-deficient human breast cancer cell line, MDA-MB-468 was tested. In absence of TGF-␤1 treatment, MDA-MB-468 cells transfected with pCAL2 and p800-luc reporters exhibited background luciferase activity. TGF-␤1 treatment of these cells resulted in a minor change in the luciferase activities, which was expected because of the TGF-␤ non-responsiveness of this cell line. Expression of SmSmad4 in these cells moderately decreased the luciferase activity in pCAL2-transfected cells, and produced a comparable increase in cells transfected with p800-luc. TGF-␤1 treatment of the cells transfected with SmSmad4 sharply down regulated the luciferase activity in cells transfected with pCAL2 reporter, while it significantly increased the luciferase activity of the p800-luc-transfected cells (Fig. 7A). This demonstrates that SmSmad4 complements the human Smad4 deficiency and restores TGF-␤-responsiveness in these cells.
Mink lung epithelial cells, Mv1Lu, transfected with p800-luc reporter, were tested to further evaluate the activity of schistosome Smad proteins in heterologous cells. Cells transfected with the reporter construct alone showed a modest change in luciferase activity upon addition of TGF-␤1. TGF-␤1 treatment of cells transfected with either SmSmad2 or SmSmad4 alone exhibited a comparable change in luciferase activity. When Mv1Lu cells were transfected with both SmSmad2 and SmSmad4, the expression of luciferase reporter gene was elevated in the absence of TGF-␤ and further enhanced in response to TGF-␤ in a dose-dependent manner. The increase in luciferase activity significantly surpassed the effect of the Smad plasmids transfected individually (Fig. 7B, left panel).
A similar experiment was conducted on the Smad4-deficient cell line, MDA-MB-468. TGF-␤1 treatment of cells transfected either with p800-luc reporter alone or co-transfected with SmSmad2 failed to boost the luciferase activity above the background level. Co-transfection with SmSmad4 significantly enhanced the luciferase activity in the absence or presence of TGF-␤1. Cells co-transfected with both SmSmad4 and SmSmad2 exhibited a further increase in luciferase activity in the absence or presence of TGF-␤1 (Fig. 7B, right panel). DISCUSSION The current investigation focuses on signaling pathways in the parasite Schistosoma, as an important step to identify the molecular mechanisms of biological events mediated by self or host signals. Such studies contribute to strategies aimed toward the development of therapies and vaccines for control of schistosomiasis. In higher organisms, TGF-␤ signaling controls a diverse set of cellular processes ranging from cell proliferation to apoptosis. Smad4 is a central mediator that plays an essential role in most TGF-␤ mediated pathways. SmSmad4 exhibits the basic features characteristic of other co-Smads. The high degree of conservation in the MH1 and MH2 domains provided the means to identify the schistosome homologue. As is the case with other transcription factors, the intracellular distribution of Smad4 determines to a great extent its function. . Single sex stages were obtained by perfusion of hamsters infected with cercariae produced from a B. glabrata snail infected with a single miracidium. Parasite eggs were obtained from liver of infected hamsters. Primary sporocysts were prepared by in vitro transformation of miracidia. The hepato-pancreas regions of infected snails (30-days postinfection), cercariae-producing snails, representing different stages of daughter sporocysts, as well as from uninfected snails were also used. pattern (48) rather than strict sequence conservation. This sequence variation may reflect structural differences that have evolved to accommodate multiple interacting partners involved in transcription activation (49).
The distribution pattern of SmSmad4 mRNA as well as the native protein indicates that SmSmad4 is likely to be involved in diverse developmental processes in different tissues and different stages throughout the parasite life cycle. As is the case of SmSmad2 (11), the SmSmad4 native protein was localized to the female reproductive tissues as well as muscle layers and gut epithelia indicating a more generalized role. SmSmad4 also demonstrated a relatively constant mRNA level in the stages representing the parasite development in the mammalian host, similar to the schistosome R-Smads, SmSmad2 and SmSmad1 (10,11). Furthermore, these data show that SmSmad4 exhibits an elevated mRNA level in eggs and primary sporocysts, the early developmental stage in molluscan host, as compared with those displayed by SmSmad1, SmSmad2, and SmT␤R-I (data not shown). This suggests that SmSmad4 may play an additional role in intramolluscan development independent of SmT␤R-I, SmSmad1, and SmSmad2. This may also suggest the presence of another type I receptor and a downstream R-Smad member(s), which synergize with SmSmad4 to orchestrate the biological events induced by a TGF-␤-like ligand(s) in these stages of the parasite life cycle. This hypothesis is supported by the genomic analysis of the related nematode, Caenorhabditis elegans, which revealed the presence of 4 TGF-␤-like ligands that signal in two main pathways via two type-I receptors. These ligands initiate the signaling cascades by binding only one TGF-␤ type II receptor (Daf-4), which phosphorylates and activates both type-I receptors (Daf-1 and Sma-6). The later receptors in turn interact with and transduce the signal to 6 members of the Smad family of proteins (Sma-2, Sma-3, Sma-4, Daf-8, Daf-14, and Daf-3) (59).
SmSmad4 was found to interact with SmSmad1 and SmSmad2. Both in vivo and in vitro analyses demonstrate that SmSmad4 associates with SmSmad2 upon the phosphorylation and activation of the later by SmT␤R-I. The inhibition of the interaction between SmSmad2 and SmSmad4 in the presence of wild-type receptor I could be attributed to the SmT␤R-I binding of SmSmad2, thus preventing SmSmad2 binding to SmSmad4. Similarly, the interaction between the non-phosphorylatable forms of SmSmad2 with SmSmad4 is reduced in the presence of any form of receptor I. In contrast, in the presence of the activated form of receptor I, SmT␤R-I-Q/D, the SmT␤R-I/SmSmad2 complex dissociates upon SmSmad2 phosphorylation permitting a functional association of SmSmad2 with SmSmad4. These results are in agreement with previous reports, which demonstrated that agonist-induced activation of human Smad1 and Smad2 leads to their association with Smad4 (18,24). Furthermore, mutation of the C-terminal serine residues of the receptor-phosphorylation motif in R-Smads prevents the subsequent activation events such as association with Smad4, accumulation in the nucleus and gain of transcriptional activity (18). SmSmad1 associates with SmSmad4 in absence of type I receptor and this interaction was diminished in the presence of either form of receptor I. Our data and a previous report (60) show that SmSmad1 binds to the type I receptor, preferably to the constitutively active form. Further-

TABLE II SmSmad4 in vivo interaction with schistosome R-Smads
Yeast two-hybrid assay, in which SmSmad4 fused to GAL4 DNAbinding domain (SmSmad4-DBD) was co-transformed with different schistosome R-Smads constructs fused to GAL4 transcription activation domain (AD constructs). The presented data are collected from assays performed on AH109, YRG2, PJ69 -4A, and Y187 yeast strains. Yeast three-hybrid assay, in which SmSmad4-DBD was co-transformed with any of the AD constructs of R-SmSmads and either wild type or the constitutively active version of receptor 1-Ura3 constructs, selected for with uracil removal (Ura3). Yeast host strains PJ69 -4A and Y187 were used in the three-hybrid assay. The cumulative data presented in this table considered at least 2 of the following parameters: Number of colonies present on selective media as compared to control plates (number on control plates were between 200 -500 colonies); the intensity and duration of development of the blue color in LacZ-filter assay; ␤-galactosidase units calculated from liquid LacZ assay using ONPG substrate. Negative sign (Ϫ) stands for undetectable interaction; (ϩ/Ϫ), weak interactions (1-4 colonies grew on selective media, a faint blue-green color produced in LacZ-filter assay that requires incubation of about 3 days; and/or Ͻ0.5 ␤-galactosidase units in liquid LacZ assay; (ϩ), positive interactions (5-10 colonies, greenish-blue color in LacZ-filter assay developed in 2-3 days, and/or 0.5-1.0 ␤-galactosidase units in liquid LacZ assay); (ϩϩ), stronger positive interactions (11-25 colonies, blue color in LacZ-filter assay developed in 1-2 days and/or 1.0 -2.0 ␤-galactosidase U); (ϩϩϩ), strong interactions (Ͼ25 colonies, blue color developed overnight in LacZ-filter assay and/or Ͼ2.0 ␤-galactosidase units in liquid LacZ assay).
In vitro interaction of SmSmad4 with different schistosome R-Smads. A, co-precipitation of different SmSmads with SmT␤R-I-wt (wt) or SmT␤R-I-Q/D (Q/D). In vitro translated, unlabeled-S-tagged wt and Q/D were incubated with 35 S-labeled, in vitro translation products of each of the following constructs: SmSmad2-MH2/AAA, SmSmad2-MH2, SmSmad4, SmSmad2 and SmSmad1. Binding reactions were purified using S-protein agarose beads, and the products were separated onto SDS-polyacrylamide gels (SDS-PAG) and subjected to autofluorography (panel I). The bar graph (panel II) shows the binding of each construct to each of SmT␤R-I. Values are mean values obtained from two independent assays. B, GST pull-down analysis of SmSmad4 binding to wt and C-terminally tagged SmSmad2. GST (I) and GST fusion proteins of SmSmad2-linker (II), wild-type SmSmad2 (III), and C-terminally-tagged SmSmad2 (SmSmad2-3Ј) (IV) were expressed in bacteria, affinity-purified over glutathione-Sepharose beads and the protein-bound beads were allowed to interact with 35 S-labeled, in vitro translated SmSmad4 in the presence or absence of in vitro translated SmT␤R-I wt and Q/D. Binding reactions were separated by SDS-PAGE and subjected to autofluorography. Lanes are labeled according to the absence or the presence of receptor I. C, co-immunoprecipitation of SmSmad2-MH2 and SmSmad1 with SmSmad4. 35 S-labeled, in vitro translated products of SmSmad2-MH2 (I), SmSmad2-MH2/ AAA (II), and SmSmad1 (III) were immunoprecipitated using ␣-SmSmad4-linker IgG in the presence or absence of SmSmad4 and SmT␤R-I (wt or Q/D). Immunoprecipitation products were separated by SDS-PAGE and subjected to autofluorography. Lanes are labeled to specify the input components of each reaction. In vitro translated products (20% of input in each reaction) are shown in the left lane of each panel. Bkg, background. more, Beall and Pearce (60) reported that SmSmad1 was not phosphorylated by the Q/D form of receptor I nor could it be induced to stimulate luciferase activity in response to TGF-␤ when co-transfected with a human T␤RI/SmT␤R-I chimera (extracellular, transmembrane, and juxtamembrane of human T␤RI and intracellular kinase domain of SmT␤R-I). Taken together, the above data demonstrate that SmSmad1 does not serve as a substrate for SmT␤R-I, yet noticeably interacts with it. This could explain why SmT␤R-I does not stimulate the receptor-dependent activation and association of SmSmad1 with SmSmad4.
Protein sequence analysis revealed that SmSmad4 contains three possible Erk1/2 phosphorylation sites, while none were found in either SmSmad1 or SmSmad2. An activated form of rat muscle Erk2 phosphorylated SmSmad4 and this phosphorylation was found to exert a substantial inhibitory effect on the interaction of SmSmad4 with the receptor-phosphorylated form of R-Smad (SmSmad2) but not on the non-activated form (SmSmad1 and SmSmad2), in vitro. In the human model, previous studies demonstrated that cells expressing an oncogenic hyperactive Ras mutant typically acquire loss of TGF-␤ antiproliferative responses. This effect is primarily attributed to the inhibition of the nuclear translocation of the Smad complex upon Erk-mediated phosphorylation of R-Smads, Smad1 (46) and Smad 2 and 3 (45,47). An alternative mechanism of down-regulation of TGF-␤ signaling by oncogenic Ras was reported by Saha et al. (61) who showed that hyperactive Ras induced proteasome-dependent degradation of Smad4. Our results are consistent with the Rasmediated regulation of TGF-␤ signaling as hypothesized by Saha et al. (61), as only SmSmad4 contains Erk-phosphorylation sites, while none of the currently identified schistosome R-Smads (SmSmad1, SmSmad2, and Smad8/9 homologue) contain Erk phosphorylation motifs. Since all Smad4 homologues, except the Drosophila Medea, contain at least one Erk-phosphorylation site, it is plausible that Erk-phosphorylation of Smad4 could induce the proteasome-mediated degradation and that such degradation is a secondary event to Erk phosphorylation. Therefore, the broad spectrum of the inhibitory effects of the oncogenic hyperactive Ras on TGF-␤ signaling could be due to a combined effect of Erk phosphorylation of both R-Smads and co-Smad and that the later could induce a proteasomedependent proteolysis of co-Smad (61). However, this hypothesis awaits further investigation.
In MDA-MB-468 breast cancer cells, which are deficient in Smad4 (62), Smad4 deficiency accounts for the non-responsiveness to TGF-␤-antiproliferative effects and contributes to malignancy (24,63). Transfection of wild-type human Smad4 into these cells strongly increased the basal level expression of luciferase reporter gene in the TGF-␤ inducible construct and conferred responsiveness to both activin and TGF-␤ (24). Expression of SmSmad4 was found to complement the loss of human Smad4 in this cell line and to restore cellular responsiveness to TGF-␤. This demonstrates that SmSmad4 is able to function with the cellular Smad proteins, which indicates the integration of schistosome protein in this heterologous system. When both SmSmad2 and SmSmad4 were co-expressed a further enhancement of transcription was observed. Since SmSmad2 had no effect on its own, the further increase in activity is likely to be due to cooperation between the schistosome Smads in the host cells. Previous studies reported that SmSmad2 stimulated the luciferase activity of TGF-␤-responsive luciferase reporter construct in response to TGF-␤ when overexpressed with either human T␤RI (10) or with the activated form of schistosome receptor I, SmT␤R-I-Q/D (60) in R1B mink lung epithelial cells, a T␤RI-deficient cell line. SmSmad2 was also shown to respond to human TGF-␤ induction by increasing its nuclear localization (10,11), an additional evidence of integration of SmSmad2 in mammalian systems represented by mammalian cell lines. Herein, SmSmad4 demonstrates a synergism with SmSmad2 in response to human TGF-␤ induction, which indicates that SmSmad4 is functionally integrated in this system. The conservation of Smad4 functions in heterologous systems was previously demonstrated. In COS cells, Drosophila Medea was shown to interact with human Smad1 in the presence of activated BMP receptor (BMPR-IB) and to interact with human Smad2 in the presence of activated TGF-␤ type I receptor (T␤R-I) (64). In Xenopus embryos, human Smad4 could act as a ventral mesoderm inducer, mimicking the effect of low activin concentrations (24).
The identification of SmSmad4 raises many questions to be addressed such as the possibility of another receptor-I and an associated R-Smad, which might work upstream of SmSmad4 in the snail host. Another issue is the identification of the ligands, whether host or self in origin, and the different pathways mediated by SmSmad1, SmSmad2, or other yet to be identified R-Smad. TGF-␤ signaling pathways cross-talk via SmSmad4, which participates in heteromeric Smad complexes with different R-Smads then the newly formed complex moves into the nucleus where it modulates target genes transcription. After being imported to the nucleus, the Smad complex binds to a nuclear partner, which directs the complex to the target gene(s) in order to elicit the specific effects/phenotypes that fit the tissue and/or the developmental stage. These downstream nuclear partner(s), and the target gene(s) are more issues that await answers. On the other hand, SmSmad4 identification has also addressed unanswered questions about the mechanism of interaction of schistosome R-Smads with co-Smad in response to an activating signal and how the phosphorylation status of the Smad protein regulates its interaction and consequently its function. In conclusion, the identification of Smad4 homologue in S. mansoni, a central TGF-␤ signal transducer, is an important step toward the characterization of the TGF-␤ pathways in schistosomes.