A Novel Human Striated Muscle RING Zinc Finger Protein, SMRZ, Interacts with SMT3b via Its RING Domain*

The RING domain is a conserved zinc finger motif, which serves as a protein-protein interaction interface. Searches of a human heart expressed sequence tag data base for genes encoding the RING domain identified a novel cDNA, named striated muscle RING zinc finger protein (SMRZ). The SMRZ cDNA is 1.9 kilobase pairs in length and encodes a polypeptide of 288 amino acid residues; analysis of the peptide sequence demonstrated an N-terminal RING domain. Fluorescence in situ hybridization localized SMRZ to chromosome 1p33–34. Northern blots demonstrated that SMRZ is expressed exclusively in striated muscle. In the cardiovascular system, SMRZ is more highly expressed in the fetal heart than in the adult heart (slightly higher expression in the ventricle than in the atrium), suggesting that SMRZ is developmentally regulated. SMRZ was found to interact with SMT3b, a ubiquitin-like protein, through the SMRZ-RING domain. This interaction was abolished by mutagenesis of conserved RING domain residues. Transient transfection of SMRZ into C2C12 myoblasts showed localization of SMRZ to the nucleus. These data suggest that SMRZ may play an important role in striated muscle cell embryonic development and perhaps in cell cycle regulation.

The zinc finger proteins (ZFPs) 1 are a group of proteins containing zinc finger domains. The finger-like structures are formed by the coordination of conserved cysteines (C) and histidines (H) with zinc ions (1)(2)(3). The discovery of the zinc finger domain in the Xenopus transcription factor IIIA (4), together with ongoing cDNA sequencing projects, has led to a significant acceleration in the discovery of ZFP genes. Combinational diversities among the conserved cysteines and histidines of the domain classifies ZFPs into several types: C 2 H 2 , C 2 C 2, C 2 HC, C 2 HC 4 C(HD), C 3 H, and C 3 HC 4 (3,(5)(6)(7)(8)(9)(10)(11)(12). These ZFP types have been described in a human heart EST data base survey as well as in an in silico Northern analysis of a total of 15 different tissues (13).
The C 3 HC 4 -type zinc finger domain, also termed RING domain, was first described in the sequence of the human ring 1 gene (14). This domain was subsequently found in a number of virus, yeast, plant, and animal proteins, which have various functions such as oncogenesis, signal transduction, peroxisome biogenesis, viral infection, development, transcriptional repression, and ubiquitination (15,16). The widespread presence of the RING domain in functionally distinct proteins and in many different species indicates that the domain is essential to basic cellular function and is conserved during evolution. With the aid of sequence alignment, the consensus sequence of this domain was defined as CX 2 CX (9 -39) CX (1)(2)(3) HX (2)(3) CX 2 CX (4 -48) CX 2 C, where X can be any amino acid. Unlike other zinc finger domains, this domain shows a unique cross-brace arrangement on the conserved residues that coordinate the two zinc ions (15,16).
Accumulated evidence suggests that the RING domain is mainly involved in protein-protein interactions (15,16), and the structural integrity of RING is reportedly essential for this type of interaction. For example, deletion of RING domain resulted in a loss of activity in Bmi-1 transformation (17) and in PML interaction with IE1 (18). The RING domain mutation was found in the tumor suppressor BRCA1 (19,20) and has been reported to cause impairments in bmi-1, which induces lymphomas (21); in c-Cbl, which desensitizes epidermal growth factor receptor (22); in MSL2, which forms protein complexes (23); in Pex12p, which maintains normal biological function of peroxisomes (24); in IE2, which blocks cell cycle progression in S phase (25); and in Vmw110, which interacts with PMLcontaining nuclear structures (26).
In an effort to improve our understanding of the functional roles of ZFPs and the RING domain, we identified a novel RING domain ZFP gene, named SMRZ, through a human EST data base sequence similarity search. Northern expression analysis indicated that SMRZ shows restricted expression in striated muscle and its pattern of expression in heart tissue suggests that the gene is developmentally regulated. Fluorescence in situ hybridization located SMRZ on chromosome 1p33-34. A yeast two-hybrid screen demonstrated an interaction between SMRZ and SMT3b, a ubiquitin-like protein. This interaction was confirmed using in vitro interaction assays. Site-directed mutagenesis analysis further strengthens the hypothesis that RING domain plays an essential role in proteinprotein (SMRZ-SMT3b) interaction. Subcellular localization analysis showed that SMRZ is located in the nucleus.

EXPERIMENTAL PROCEDURES
Isolation of cDNA Clone-An EST clone, named striated muscle RING zinc finger protein (SMRZ), exhibiting an amino acid sequence similar to the conserved RING domain sequence, was isolated from a human heart EST data base (27)(28)(29) using BLAST searches (30 -32). The clone was subsequently excised in vivo from the ZAP Express vector using the ExAssist/XLOLR helper phage system (Stratagene). In brief, phagemid particles were excised by coinfecting XL1-BLUE MRFЈ cells with ExAssist helper phage. The excised pBluescript phagemids were used to infect Escherichia coli XLOLR cells, which lack the amber suppressor necessary for ExAssist phage replication. Infected XLOLR cells were selected using kanamycin resistance. Resultant colonies contained the double-stranded phagemid vector with the cloned cDNA insert. A single colony was grown overnight in LB-kanamycin, and DNA was purified using a plasmid purification kit (Qiagen).
Northern Blot Analysis-The expression pattern of SMRZ was analyzed using Northern blot hybridization. An ϳ1.2-kb SMRZ cDNA was amplified and labeled using DIG DNA labeling system (Roche Molecular Biochemicals). Two commercial human tissue Northern blots, Human 12-Lane MTN Blot (CLONTECH catalog no. 7780-1) and Human Cardiovascular System MTN Blot (CLONTECH catalog no. 7791-1), were hybridized with DIG-labeled probe in ExpressedHyb hybridization solution (CLONTECH). Blots were washed twice in 2ϫ SSC, 0.1% SDS at room temperature for 15 min each, followed by two washes in 0.1ϫ SSC and 0.1% SDS at 65°C for 20 min each. After these washes, blots were treated with CSPD according to the manufacturer's instructions (Roche Molecular Biochemicals). Signals were visualized by exposing the blots to Kodak X-Omat film. To normalize the hybridization signals, a commercial cDNA encoding ␤-actin probe (CLONTECH) was used as a control.
Full-length Nucleotide Sequencing and Data Base Comparisons-Phagemid DNA was sequenced using the Taq dye-deoxy terminator cycle sequencing kit for the Applied Biosystems 377 sequencing system (PerkinElmer Life Sciences). Both strands of the cDNA were then sequenced using commercial T3 or T7 primers and the specific nested forward and reverse primers: 5Ј-ATGGAGAACTTGGAGAAGCAGC-3Ј and 5Ј-ATTTACCACCCCGTATG TCTGG-3Ј. To obtain the full length of the SMRZ gene, 5Ј-rapid amplification of the cDNA ends (5Ј-RACE; Life Technologies, Inc.) was performed on human fetal heart total RNA. Using Trizol reagents (Life Technologies, Inc.), total RNA was isolated from pooled human fetal hearts (10 -12 weeks) and treated with DNase I. The purity and integrity of RNA was assessed by absorbance at 260/280 nm and by 1% formaldehyde/agarose gel electrophoresis. The first-strand cDNA was synthesized using 3 g of total RNA as template incubated with SuperScript™ II reverse transcriptase and an antisense primer located in the SMRZ coding region (5Ј-TCATTGGCACACTTC-CGGCACAGGTTGTGCTGGCACG GCAAGA-3Ј) at 42°C for 1 h. After treating with RNase Mix (RNase H and RNase T1) and performing 3Ј end tailing reaction, the 5Ј-RACE product was PCR-amplified according to the manufacturer's manual (Life Technologies, Inc.) and sequenced. Full-length SMRZ sequence comparisons against the nonredundant nucleotide and protein data bases were performed using BLASTN and BLASTX programs (30 -32), at the National Center for Biotechnology Information and the Swiss-Prot protein sequence data base. The RING domain amino acid sequence was aligned to the members of the RING domain-containing proteins to identify consensus sequence.
Fluorescence in Situ Hybridization (FISH)-The SMRZ phagemid, used for human chromosome mapping, was prepared by dATP biotinylation with the BioNick labeling kit (Life Technologies, Inc.). Methods for preparation of cultured lymphocyte slide and FISH protocols have been described previously (33,34). Assignment of the FISH mapping data with chromosomal bands was achieved by superimposition of the FISH signals with 4,6-diamidino-2-phenylindole-banded chromosome. Detailed positioning of each of the assignments was determined based on the summary from 10 photographs.
Yeast Two-hybrid Screen-A yeast two-hybrid screen was conducted to determine the protein-protein interaction role of SMRZ-RING domain. Both RING domain region (nucleotides encoding amino acids 1-67) and non-RING domain region (nucleotides encoding amino acids 68 -288) were PCR-amplified from the SMRZ cDNA clone. The PCR primers were constructed to incorporate an EcoRI and a SalI site in forward and reverse primers, respectively. After digestion with restriction enzyme, these two regions of PCR products were subcloned into pAS2-1 vector (CLONTECH) in frame to generate pAS2-R1 (RING domain region) and pAS2-R2 (non-RING domain region) fused to a GAL4 DNA-binding domain as bait. After cloning, the expression plasmids were sequenced to verify that no mutation had been introduced during the PCR process. A human heart Matchmaker cDNA library (CLONTECH catalog no. HL4042AH), constructed in the pACT2 vector expressing GAL4 activation domain fusion protein, was used as prey. Using a lithium acetate method, bait and prey vectors were transformed sequentially into yeast strain Y190. Transformants were selected by the use of appropriate selective media lacking histidine, leucine, and tryptophan, but containing 25 mM 3-aminotriazole. Histi-dine prototrophs were assayed for ␤-galactosidase activity using a filter lift technique. Colonies remaining histidine-and ␤-galactosidase-positive were further characterized by sequencing and searched for gene sequence similarity on the GenBank™ data base using the BLAST program.
Expression Plasmids and Site-directed Mutagenesis-To generate epitope-tagged SMRZ and SMT3b, two prokaryotic expression vectors, pET27b(ϩ) and pET9a (Novagen), were used. The full open reading frame of SMRZ was amplified with PCR primers containing EcoRI and SalI enzyme digestion sites and ligated into the corresponding sites of pET27b(ϩ) vector in-frame to generate pET27b(ϩ)-SMRZ plasmid. For SMT3b, the full open reading frame was amplified with PCR primers containing BamHI and Bpu1102I enzyme digestion sites and inserted into the pET9a vector in-frame to generate pET9a-SMT3b plasmid. To further determine whether the RING domain of SMRZ is essential to protein-protein interaction, the RING domain was mutated using the QuickChange site-directed mutagenesis kit (Stratagene). Using pET27b(ϩ)-SMRZ plasmid as a template, one round of mutagenesis in which Cys 26 , His 28 , Cys 31 , and Cys 34 were mutated to alanine using sense primer (5Ј-ATC TTG CCG GCC CAG GCA AAC CTG GCC CGG AAG GCC GCC AAT GAC-3Ј) and antisense primer (5Ј-GTC ATT GGC GGC CTT CCG GGC CAG GTT TGC CTG GGC CGG CAA GAT-3Ј). The temperature cycling reaction was performed according to the manufacturer's instructions (Stratagene). After cloning, the sequence fidelity of all recombinants was verified by sequencing.
In Vitro Interaction Assay-To confirm the protein-protein interaction between SMRZ and SMT3b, in vitro interaction assay was performed as described previously (35). Five micrograms of wild-type HSV SMRZ and mutant HSV SMRZ MT fusion proteins were each immobilized on 100 l of His⅐Bind ® resin (Novagen) and washed with 60 mM imidazole, 500 mM NaCl, and 20 mM Tris⅐HCl (pH 7.9), to remove unbound HSV SMRZ and HSV SMRZ MT . Bacterial cell lysates from E. coli BL21 (DE3) expressing T7 SMT3b fusion protein were split into aliquots and incubated with His⅐Bind resin immobilized with HSV SMRZ or HSV SMR-Z MT in a final concentration of 1 mM ATP in 1.0 ml of lysis buffer (BugBuster protein extraction reagent; Novagen). Following incubation at 4°C overnight (with gentle rocking), the mixture was pelleted by centrifugation and washed three times with buffer containing 60 mM imidazole, 500 mM NaCl, and 20 mM Tris⅐HCl (pH 7.9). The bound proteins were eluted using buffer containing 1 M imidazole, 500 mM NaCl, and 20 mM Tris⅐HCl (pH 7.9). SDS sample buffer was added to the protein elution under reducing and non-reducing conditions for recombinant protein analysis using 12% SDS-polyacrylamide gel electrophoresis. Control experiments under reducing conditions were performed under conditions: 1) nontransformed bacterial lysates were used to incubate with His⅐Bind resin immobilized with HSV SMRZ or HSV SMR-Z MT and 2) bacterial lysates expressing T7 SMT3b were used to incubate with non-HSV SMRZ immobilized His⅐Bind resin.
Cell Culture, Transfection, and Confocal Microscopy Analysis-C2C12 myoblasts were purchased from the American Type Culture Collection (ATCC) and maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (Life Technologies, Inc.) at 37°C in a humidified 5% CO 2 atmosphere. Before transfections, the full open reading frame of SMRZ was amplified with PCR primers containing EcoRI and XhoI enzyme digestion sites and ligated into the corresponding sites of pcDNA3.1 vector (Invitrogen) in-frame with V5 epitope-tag to generate pcDNA3.1-SMRZ plasmid. Sequence fidelity of this recombinant was verified by sequencing. The cells were plated on multiwell chamber slides (Nunc) 1 day before transfection at a density of 5 ϫ 10 4 cells/well. Transfections were carried out using LipofectAMINE Plus transfection reagent according to the manufacturer's instructions (Life Technologies, Inc.). Briefly, 0.7 g of pcDNA3.1-SMRZ plasmid was precomplexed with 5 l of the Plus reagent in 50 l of serum-free DMEM for 15 min at 37°C was combined with 2 l of LipofectAMINE reagent in 50 l of serum-free DMEM and incubated for 15 min at 37°C. Complexes of DNA-plus-LipofectAMINE reagent were then added to each well. Three hours following transfection, medium containing the complexes was replaced with fresh medium. Forty-eight hours after incubation, the cells were fixed with 4% paraformaldehyde in PBS for 15 min, permeabilized with 0.2% Triton X-100 in PBS for 5 min, blocked with 1% bovine serum albumin in PBS (blocking buffer) for 30 min, incubated with a 1:200 diluted alkaline phosphatase-conjugated anti-V5 monoclonal antibody (Invitrogen) in blocking buffer for 1 h, and stained with Vector ® Red substrate (Vector Laboratories Inc.). After washing, the slides were analyzed using a confocal fluorescence microscope (Olympus). Cells transfected with pDNA3.1 vectors were used as control.
Western Blot Analysis-Proteins separated by SDS-polyacrylamide gel electrophoresis were transferred to polyvinylidene difluoride membranes (Helix) in transfer buffer (25 mM Tris-HCl, pH 8.3, 192 mM glycine, 20% methanol) with a Mini Trans-Blot apparatus (Bio-Rad) for 1 h at 100 V. Proteins were immunoblotted with specific antibodies. For example, membrane blotted with purified T7 SMT3b fusion protein and its co-immunoprecipitates was blocked with a 3% solution of nonfat milk powder in TBST buffer (10 mM Tris-HCl, pH 8.0, 150 mM NaCl, 0.1% Tween 20) and incubated with 1:5000 dilution of T7 tag horseradish peroxidase conjugate (Novagen) at room temperature. Membrane blotted with purified HSV SMRZ or HSV SMRZ MT fusion proteins and their co-immunoprecipitates was blocked with 3% solution of bovine serum albumin in TBST buffer, incubated with monoclonal antibody directed against HSV tag (dilution 1 into 10,000; Novagen) and then treated with 1:10,000 dilution of anti-mouse IgG alkaline phosphatase conjugate (Novagen). Unbound antibody was removed by washing with TBST for 5 ϫ 1 min. Bound antibody was detected using ECL Western blotting detecting reagents as described by the manufacturer's protocol (Novagen). The binding was visualized by exposing the blots to Kodak X-Omat film.

RESULTS
Isolation and Sequence Analysis of SMRZ cDNA-An EST clone (SMRZ; GenBank™ accession no. AF361946) exhibiting (approximately) 39% amino acid sequence similarity to the RING domain of several known ZFPs was isolated and sequenced. To obtain the full length of the cDNA, 5Ј-RACE was performed. The full-length nucleotide and predicted amino acid sequences of SMRZ are shown in Fig. 1. The 1903-bp cDNA contains an 864-bp open reading frame extending from 984 to 1847, which corresponds to an encoded protein of 288 amino acid residues with a predicted molecular mass of 32.3 kDa. The sequence around the initiation ATG codon at nucleotide 984 -986 was matched with the Kozak consensus ((A/G)CCATGG) (36,37). The coding region contained an N-terminal RING domain between residues 2-67. In the 3Ј-untranslated region, a potential polyadenylation signal AATAAA (38) was detected 26 bases upstream of a poly(A) tail. A search of the EST data base, disclosed 10 ESTs (GenBank™ accession no. F35868, F21990, AI351421, F36292, F32077, F25523, F18452, C02645, F34432, and BE894237) showing identity to portions of the SMRZ coding region. Fifty ESTs also matched an Alu sequence containing portion (nucleotides 220 -295) of the 5Ј untranslated region (5Ј-UTR). No gene sequence deposited in public data bases was found to match to SMRZ suggesting that SMRZ is a novel gene. A comparison of the amino acid sequence of the RING domain regions of SMRZ and 26 known RING domaincontaining ZFPs is shown in Fig. 2. The alignment shows that the cysteine and histidine residues for constructing the RING finger (i.e. C 3 HC 4 ) are 100% conserved (shown in bold). Six additional highly conserved (Ͼ50% conservation) amino acids within the sequences are also shown in the consensus sequence (CONS).
Expression of SMRZ mRNA in Human Tissues-To determine the expression of the SMRZ transcript in various human tissues, both human multiple-tissue and cardiovascular system Northern blot analyses were probed using DIG-labeled SMRZ cDNA. For the human 12-lane multiple-tissue blot, a single transcript of ϳ2.1 kb was detected exclusively in heart and skeletal muscle lanes (Fig. 3A). A second blot (human cardiovascular system) probed in the same way showed a high expression of an ϳ2.1-kb transcript in fetal heart, intermediate expression in ventricle, and low expression in atrium and adult heart (Fig. 3B).
Chromosomal Mapping of SMRZ Gene-Chromosomal localization of SMRZ was performed using FISH (Fig. 4). 4,6-Diamidino-2-phenylindole staining and detailed position analysis allowed for assignment of the signal to 1p33-34. No additional loci were detected under the conditions applied.
Yeast Two-hybrid Screening Reveals an SMRZ RING Domain-SMT3b Interaction-To determine whether the SMRZ-RING domain is the region involved in protein-protein interactions and to identify its possible interactive partners, a yeast two-hybrid screen was performed using two baits: the RING domain region of SMRZ and the non-RING domain region (a region outside of the RING domain) of SMRZ. Each bait was fused to GAL4 DNA-binding domain serving to screen a commercial human heart cDNA library fused to a GAL4 activation domain in yeast Y190. After screening 1 ϫ 10 6 transformants, 126 yeast colonies fitting the criteria for the interaction between bait and prey proteins (i.e. positive for both histidine growth and ␤-galactosidase activity) were obtained from the bait fused with RING domain. No colony was found to grow on the histidine dropout medium from the bait fused with non-RING domain region (Table I), suggesting that RING domain is the region of SMRZ responsible for protein-protein interaction. Nucleotide sequencing was carried out with 12 randomly selected cDNA clones fused with activating domain isolated from yeast His and LacZ positive colonies. A sequence similarity search through the GenBank™ data base revealed 12 selected positive clones: 1) one H ϩ -ATP synthase subunit b (GenBank™ accession no. X60221), 2) one NADH dehydrogenase subunit 2 (GenBank™ accession no. AF014892), 3) one ferritin light chain (GenBank™ accession no. M11147), 4) one overlapping cosmid cSRL72 g7-140b8 (GenBank™ accession no. AC002037), 5) two cytochrome oxidase subunit IV (GenBank™ accession no. X54802), 6) two SMT3b (GenBank™ accession no. X99585), 7) two ribosomal proteins L27 (GenBank™ accession no. L19527), and 8) two KAT11391 clones (GenBank™ acces-sion no. AK026634). Previous evidence has shown that many RING ZFPs were involved in either ubiquitination (39) or in interacting with ubiquitin-like protein such as the interactions between RING ZFPs (PML (Refs. 40 -42) and Mdm2 (Ref. 43)) and ubiquitin-like protein (sentrin-1/PIC1/SUMO-1). It is interesting that 2 of the 12 selected positive clones were identical to SMT3b, a ubiquitin-like protein. Thus, SMT3b was selected for further study as an interaction partner of SMRZ.
In Vitro Demonstration of Interaction between SMRZ and SMT3b-Yeast two-hybrid results demonstrated that RING domain region is a protein-protein interface. This raises the question as to whether the full-length SMRZ protein can interact with SMT3b. Using bacterial cell lysate expressing T7 SMT3b incubated with HSV SMRZ-immobilized His⅐Bind resin, bound proteins were eluted and analyzed under reducing (Fig. 5A) and non-reducing conditions (Fig. 5B). Under reducing conditions, the eluted proteins were detected as two bands corresponding to SMRZ (Fig. 5A, lane 4; top panel) and SMT3b (Fig. 5A, lane 4; bottom panel) confirming the presence of interaction between HSV SMRZ and T7 SMT3b. The specificity of this interaction was strengthened by control experiments. For example, no T7 SMT3b could be eluted with HSV SMRZ from nontransformed bacterial lysates incubated with His⅐Bind resin immobilized with HSV SMRZ (Fig. 5A, lane 3) suggesting a specific interaction between HSV SMRZ and T7 SMT3b. Furthermore, no HSV SMRZ and T7 SMT3b could be eluted from bacterial lysates expressing T7 SMT3b incubated with non-HSV SMRZ immobilized His⅐Bind resin (Fig. 5A, lane 2), suggesting that the interaction between HSV SMRZ and T7 SMT3b is not an artifact due to the binding of T7 SMT3b to the His⅐Bind resin. Since specific HSV SMRZ-T7 SMT3b interaction has been demonstrated under reducing conditions, we examined the presence of this interaction under non-reducing conditions (Fig. 5B). A slight increase in the molecular masses of epitope-tagged proteins as compared with the predicted molecular weight of the protein, was due to the extra amino acids derived from the expression vectors, which primarily served as epitope tags (Fig. 5B, lanes  1 and 4). For example, the predicted molecular mass of SMRZ is 32.3 kDa. As compared with the predicted molecular weight of the protein, the slight increase in the molecular masses of epitope-tagged proteins was due to the extra amino acids (ϳ7.26 kDa) derived from the expression vectors, which primarily served as epitope tags.
The SMRZ RING Domain Is Necessary for Interaction with SMT3b-To further confirm that the RING domain is essential for HSV SMRZ-T7 SMT3b interaction, site-directed mutagenesis of the RING domain was used for in vitro interaction analysis. The RING domain is known to form a unique cross-brace arrangement, with the first and the third pairs of conserved residues forming the first zinc binding site, and the second and the fourth pairs forming the second (15,16). We substituted Cys 26 , His 28 , Cys 31 , and Cys 34 of the RING domain with alanines, which disrupt the structural integrity of the RING domain. Under reducing conditions, no T7 SMT3b could be detected (Fig. 6A, lane 4) suggesting that the interaction was abolished. This explanation was further strengthened by the observation that no higher molecular weight bands could be visualized by immunoblotting with anti-HSV antibody (Fig. 6B,  lane 2) and anti-T7 antibody (Fig. 6B, lane 3) under nonreducing conditions, indicating that the large protein complexes ( HSV SMRZ-T7 SMT3b) were not formed. The purified mutated RING domain SMRZ ( HSV SMRZ MT ) used as a control (Fig. 6, A (lane 1) and B (lane 1)) demonstrated that HSV SMR-Z MT was expressed as a stable protein in order to exclude the possibility that loss of the interaction in in vitro assay was due to the instability of the mutated protein. The interaction between SMRZ and SMT3b was completely abolished by the mutated SMRZ RING domain, indicating that the RING domain is essential to protein-protein interaction.
Subcellular Localization of SMRZ-To facilitate our under-standing of the putative role of SMRZ, the subcellular localization of SMRZ was determined using a plasmid expression construct encoding a V5-tagged SMRZ ( V5 SMRZ) fusion protein transiently transfected into C2C12 myoblasts. Staining for V5 SMRZ in transfected C2C12 cells was performed using alkaline phosphatase-labeled antibody against the V5 epitope tag and Vector Red substrate, which produced red fluorescent precipitate. Using confocal fluorescence microscope, red fluorescent spots were observed on the nucleus (Fig. 7A). In contrast, no such red spot was seen in control cells (Fig. 7B). It is thus clear that the expressed SMRZ fusion proteins are present in the nuclei.

DISCUSSION
The ZFPs are a group of proteins involved in a wide spectrum of cellular regulatory functions (13). Since protein-protein interaction plays an important role in cell regulatory networks, identification of novel ZFP genes and characterization of their protein-protein interactions will facilitate our understanding of ZFP cellular regulatory mechanisms. Previously, we have established a cardiovascular ZFP profile (cvbZFP; Ref. 13). To expand this cardiovascular ZFP profile, a sequence similarity analysis search of a human heart EST data base was carried out in order to identify novel ZFP genes. An EST clone (SMRZ) exhibiting an amino acid sequence similar to the RING domain, a protein-protein interaction RING domain (15,16) was isolated and sequenced. With the aid of 5Ј-RACE, the full-length sequence of SMRZ was obtained. Analysis of the full-length sequence indicated that SMRZ contains an Alu sequence in the yeast two-hybrid system Baits were constructed by Gal4 binding domain fused with in frame RING domain (pAS2-R1) and non-RING domain regions (pAS2-R2) of SMRZ. Specificity of the interaction between bait and prey was determined by growth (ϩ) on Trp Ϫ Leu Ϫ His Ϫ plates and blue color development (ϩ) in a filter lift ␤-galactosidase assay in the yeast two-hybrid screen analysis. Ϫ indicates no growth on plates or no color development in filter lift assay.

Construct
Plates with selective media LacZ expression 5Ј-UTR and a RING domain at the N terminus of the coding region. Examples of Alu sequence in the 5Ј-UTR has been shown by a survey of GenBank™ sequence data base, which indicated that 14% of the Alu-containing human cDNAs have Alu sequence located in the 5Ј-UTR (44). The role of Alu sequence in 5Ј-UTR is currently unknown although it is believed that the presence of Alu sequence may influence translation. The presence of the RING domain is of functional importance since this domain has been shown to be mainly involved in protein-protein interactions (15,16).
RING was first described in the sequence of the human ring 1 gene (14) and was later found in a large number of proteins through sequence comparison analysis (15,16). These proteins are involved in a variety of functions such as oncogenesis, signal transduction, peroxisome biogenesis, viral infection, development, transcriptional repression, and ubiquitination (15,16). Since the RING domain serves as a protein-protein interface, we hypothesized that determination of the RING domain interaction partner may facilitate our understanding of the SMRZ regulatory role. Using a yeast two-hybrid screen, we found that the RING domain is the only region of SMRZ that is involved in protein-protein interaction, and we identified SMT3b as an interaction partner of SMRZ. The specificity of interaction between SMRZ and SMT3b and the essential role of SMRZ-RING domain in this interaction were further confirmed in vitro using bacterially expressed full-length SMT3b and wild-type or mutated SMRZ fusion proteins. These results are supported by previous studies demonstrating that the structural integrity of RING domain is critical for the protein-protein interactions of RING ZFPs (17)(18)(19)(20)(21)(22)(23)(24)(25)(26).
The restricted expression of SMRZ mRNA in the human heart and skeletal muscle suggests that SMRZ plays an important role in striated muscle cells. SMRZ transcript is more highly expressed in human fetal heart than in adult heart, suggesting that SMRZ may be involved in the cardiovascular developmental process. It is interesting to note that ESTs matching to portions of SMRZ were mainly found in heart, muscle, and whole embryo libraries. A search of UniGene disclosed that these ESTs were located in UniGene cluster Hs.157119T, which was assigned to chromosome 1. This chromosomal assignment was supported by FISH analysis, which mapped SMRZ to chromosome 1p33-34, a region associated with cell proliferation and differentiation (45)(46)(47)(48)(49)(50)(51)(52)(53)(54). This predicted putative role of SMRZ based on the chromosomal assignment was supported by the subcellular localization experiment, which showed that SMRZ is located in the nucleus using C2C12 myoblasts transfected with SMRZ.
To facilitate our understanding of the putative functional role of SMRZ, it is necessary to understand the biological activities of its interactive partner SMT3b, a member of the ubiquitin-like gene family. SMT3b was originally cloned as a human homolog to yeast SMT3, first isolated as a high copy mutation suppressor in MIF2 (55). MIF2 is an essential centromere protein binding to the AϩT-rich CDEII region of centromere DNA (55). Loss of MIF2 function results in blocking of DNA replication and nuclear division (56), chromosome missegregation, mitotic delay, and aberrant microtubule morphologies (57). Although the functions of most ubiquitin-like proteins are still largely unknown, it is assumed that since members of the family share a significant degree of sequence similarity they may have similar functional roles. Putative roles for a few ubiquitin-like proteins have been proposed. For example, sentrin-1/PIC1/SUMO-1, a ubiquitin-like protein with high sequence homology to SMT3b (58), has been shown to interact with RING domain of PML, a tumor suppressor associated with the pathogenesis of acute promyelocytic leukemia (41,42). Overexpression of Sentrin-1/PIC1/SUMO-1 has also been shown to protect cells from Fas/APO-and TNF-induced apoptosis (59). The interaction between Sentrin-1/PIC1/ SUMO-1 and PML has been shown to be associated with cell cycle regulation (60 -62). One recent study indicated that SMT3b may play a role in cellular responses to environmental stress, as SMT3b was shown to be up-regulated by many protein-damaging stimuli (63). In addition, SMT3lP1, a novel isopeptidase 1 specifically bound to SMT3b, was found to be located in the nucleolus at interphase (64). These data, together with our Northern blot analysis, the chromosomal assignment, and the subcellular localization of SMRZ, leads us to believe that SMRZ-SMT3b interaction may be involved in cell cycle regulation occurring during the developmental process of striated muscle cells. Further study is needed to develop this interpretation.
In conclusion, we discovered a novel RING ZFP (SMRZ), analyzed its full-length sequence, and mapped its expression patterns, chromosomal, and subcellular locations. The identification of an interaction between SMRZ RING domain and SMT3b, as demonstrated in our yeast two-hybrid screen, in vitro interaction analysis, and site-directed mutagenesis, provides evidence indicating that SMRZ-RING domain plays an important role in protein-protein interaction and suggests that this interaction may be involved in the cell cycle regulatory process of striated muscle cells.