Cloning of Two Rat PDIP1 Related Genes and Their Interactions with Proliferating Cell Nuclear Antigen

______http://www.paper.edu.cn

JOURNAL OF EXPERIMENTAL ZOOLOGY 303A:227–240 (2005)

Cloning of Two Rat PDIP1 Related Genes and Their Interactions With Proliferating Cell Nuclear Antigen JIANLIN ZHOU1, XIAOXIAO HU1, XIWEN XIONG1, XIN LIU, YUNHAI LIU, KAIQUN REN, TIESHAN JIANG, XIANG HU, AND JIAN ZHANGn Department of Biochemistry and Molecular Biology, College of Life Science, Hunan Normal University, Changsha, Hunan 410081, China

ABSTRACT Human polymerase delta-interacting protein 1 (PDIP1) is a tumor necrosis factor a and interleukin 6 inducible protein that interacts directly with proliferating cell nuclear antigen (PCNA) and the small subunit (p50) of DNA polymerase d. PDIP1 binds PCNA and p50 simultaneously and stimulates polymerase d activity in vitro in the presence, but not the absence, of PCNA. It has been suggested that PDIP1 provides a link between cytokine activation and DNA replication in eukaryotes. Here these authors report the cloning of two rat genes homologous to human PDIP1, termed rat PDIP1 and rat tumor necrosis factor-induced protein 1 (TNFAIP1). The rat PDIP1 is mapped to chromosome 1q36 cM region, spans approximately 18.7 kb, and is organized into six exons. The rat TNFAIP1 gene is mapped to chromosome 10q25 cM, spans approximately 12.9 kb, and is composed of seven exons. The deduced proteins of rat PDIP1 and rat TNFAIP1 share 63.1% sequence identity with each other and are highly conserved in the majority of the middle portion of the two proteins, which encode a BTB/POZ domain at the N-terminus and a PCNA binding motif (QTKV-EFP) at the C-terminus, respectively. The BTB / POZ domain and the PCNA binding motif are highly conserved during the evolution. Both rat PDIP1 and rat TNFAIP1 were demonstrated to interact with PCNA via BIAcore, GST pull-down, and co-immunoprecipitation assays. Like the human PDIP1, both rat PDIP1 and rat TNFAIP1 stimulate polymerase d activity in vitro in a PCNA-dependent way. J. Exp. Zool. 303A:227–240, 2005. r 2005 Wiley-Liss, Inc.

Human PDIP1 was ﬁrst identiﬁed as an inter- pol d core enzyme consists of two subunits, the acting partner of the small subunit (p50) of catalytic subunit of p125 and the small subunit of eukaryotic DNA polymerase d (pol d) in a yeast p50. It has been demonstrated that the processiv- two-hybrid screening of HepG2 cDNA library ity of the recombinant 125–kDa catalytic subunit using p50 as a bait (He et al., 2001). It was further of human pol d is not responsive to PCNA, shown that PDIP1 also interacts with proliferat- whereas coexpression of the catalytic (p125) and ing cell nuclear antigen (PCNA). DNA polymerase small (p50) subunits of pol d results in the d is one of the major polymerases to replicate formation of a recombinant heterodimer whose chromosomal DNA; its processivity is markedly processivity is markedly stimulated by PCNA increased in the presence of PCNA. PCNA is a (Zhou et al., ’96). A recent report from the multifunctional protein that plays roles in a laboratory of Antero G. So showed that PCNA variety of cellular processes, including DNA directly interacts with p50, not with p125 (Lu replication, DNA repair, and cell cycle control by et al., 2002). It has been proposed that the small interacting with proteins involved in these pro- subunit functions as a bridge connecting the cesses (Prelich et al., ’87; Gulbis et al., ’96; catalytic subunit and PCNA. The human PDIP1 Tsurimoto, ’99; Kedar et al., 2002). For example, binds simultaneously to PCNA and p50, and the interaction between PCNA and DNA pol d is essential for processive DNA synthesis during DNA replication. PCNA functions as a sliding Grant Sponsor: National Natural Science Foundation of China; Grant numbers: 30025009; 30270723 DNA clamp that forms a closed ring around duplex n Correspondence to: Jian Zhang, Department of Biochemistry and DNA. The binding of pol d to PCNA provides an Molecular Biology, College of Life Science, Hunan Normal University, elegant micromechanical solution to the biological Changsha, Hunan 410081, China. E-mail: [email protected]/ [email protected]. need to maintain an extraordinarily high level of Received 30 March 2004; Accepted 8 November 2004 processivity during the synthesis of chromosomal Published online in Wiley InterScience (www.interscience.wiley. com). DOI: 10.1002/jez.a.150 DNA (Prelich et al., ’87; Krishna et al., ’94). The 1These authors contributed equally to the work. r 2005 WILEY-LISS, INC. 中国科技论文在线______http://www.paper.edu.cn

228 J. ZHOU ET AL. stimulates pol d activity in the presence, but not in RT-PCR the absence, of PCNA (He et al., 2001). To confirm the sequences of two contigs The 1.6 kb cDNA of PDIP1 encodes a protein of assembled by overlapping EST, two pairs of 36 kDa that is highly homologous to one of the primers were designed and used to amplify the early response genes induced by TNF-a, the B12 rat liver cDNA. One pair of primers (Forward: (Wolf et al., ‘92), lately renamed as human GCGTGGAAGTGCTCACAG, Reverse: GAC- TNFAIP1. It has been shown that the expression TAGGTCCCAGAAGCAAG) correspond to 50–end of human PDIP1 can be induced by TNF-a and IL– sequence of BE110670 and 30–end sequence of 6 (He et al., 2001). TNF-a is a multifunctional AW141960, respectively. Another pair of primers cytokine which participates in wide biological (Forward: GGTGGTGAAGCTGCTGTAC, Re- activities, such as apoptosis, proliferation, B cell verse: CAAGGAATTGTCAGGGACTC) corre- activation, and some inflammatory responses spond to 50–end sequence of BF551628 and 30– (Fiers et al., ’95). Especially, TNF-a plays im- end sequence of AI598359. Amplification was portant roles during liver regeneration. Several conducted in the Mastercycler (Eppendorf, Ger- converging lines of evidence from recent work many) using HotStarTaq polymerase (Qiagen) have established that TNF-a and IL–6 are im- for 30 cycles of 941C for 30 sec, 541C/531C for 30 portant components of the early signaling path- sec, and 721C for 1 min. The products were cloned ways triggering liver regeneration after loss of into pMD18–T vector (TaKaLa, Dalian, China) hepatic tissues (reviewed in Michalopoulos and and sequenced. All primers used in the study DeFrances, ’97; Fauto, 2000). The findings that were designed by NetPremer (http://www. PDIP1 is inducible by TNF-a and IL–6, and that it premierbiosoft.com/netprimer). physically and functionally interacts with both PCNA and the small subunit of pol d when taken together, suggest that PDIP1 may play a role in the regulation of hepatocyte replication induced by TNF-a and IL–6, probably by providing a link 30 and 50 rapid amplification of cDNA ends between cytokine activation and DNA replication (RACE) for rat PDIP1 and TNFAIP1 (He et al., 2001). The rat homologue of human PDIP1 was cloned to determine whether and how After the sequences of two contigs were con- PDIP1 is involved in the liver regeneration, using firmed, 50–RACE and 30–RACE were performed to the rat partial hepatectomy model. These data extend the 50 upstream and 30 downstream describe the cloning of two rat homologues and sequences of the contigs using the SMART RACE show that the PDIP1 and related proteins are cDNA amplification kit (Clontech, Palo Alto, CA). highly conserved. It is further demonstrated that In brief, 1 mg of total RNA extracted from rat liver these two rat homologues interact directly with tissue was reverse-transcribed with the oligo (dT)- PCNA via BIAcore, pull-down, and co-immuno- anchor primer, and with the oligo (dT) primer and precipitation assays, and stimulate pol d activity in 50–anchor oligonucleotide provided in the kit for 30 vitro in a PCNA-dependent way. and 50 RACE, respectively. 50 RACE and 30 RACE were performed with the PDIP1 or TNFAIP1 gene–specific primer (Table 1) and the universal primer (UPM) provided in the kit, using HotStar- MATERIALS AND METHODS Taq polymerase. The PCR condition was as follows: a pre-incubation at 951C for 2 min, initial RNA preparation and reverse five cycles of 941C for 30 sec and 721C for 3 min, transcription followed by five cycles of 941C for 30 sec, 701C for Eight week-old male Wistar rats were sacrificed 30 sec, 721C for 3 min, and ended with 27 cycles of by carbon dioxide euthanasia. The total RNA was 941C for 30 sec, 681C for 30 sec, and 721C for 3 min. isolated from the liver tissue using UNIQ–10 Total To confirm the 30 and 50 RACE products, nested RNA Minipreps Super kit (Sangon, Shanghai, PCR was performed with the nested gene-specific China). The purified RNA was used for first- primer (Table 1) and nested universal primer strand cDNA synthesis and amplification of both (NUP) provided in the kit under the same 50 and 30 cDNA ends. The first strand of cDNA was conditions. The amplified products were purified synthesized by oligo-dT priming using Sensiscript with QIAquick Gel Extraction kit (Qiagen), cloned RT Kit (Qiagen, Valencia, CA). into pMD18–T vector and sequenced. 中国科技论文在线______http://www.paper.edu.cn

TWO RAT PDIP1 HOMOLOUGES INTERACT WITH PCNA 229

TABLE 1. Gene-speci¢c primers (GSPs) and nested gene-speci¢c TNFAIP1) were transformed into E. coli BL21. primers (NGSPs) used in RACE reactions The transformant containing pT7–PCNA was Primer Sequence (50 to 30) grown at 371C until an OD600 of 0.5–0.6 was reached. Expression was induced with 1 mM IPTG PDIP1-50GSP TGGTACTCTCAGGCAGTGGCACTGACC (final concentration) at 371C. The other two PDIP1-50NGSP GGCCACTCCGGTCAATCAGCACC 0 transformants (pGEX–4T–2–PDIP1 and pGEX– PDIP1-3 GSP CCATATCACCCATGATGAGCGTCCTCA 1 PDIP1-30NGSP GGCCAACAGATTGTCTTCAAGGACTGACC 4T–2–TNFAIP1) were grown at 37 C until an TNFAIP1-50GSP GGTGATCATCAGAGTTGCTGGTGTACGAGT OD600 of 1.0 was reached. A final concentration of TNFAIP1-50NGSP GCTCCTGTTGTACAGCAGCTTCACCACC 0.1 mM IPTG was then added to induce the TNFAIP1-30GSP GCAGAAGTGTGCTGCACCTCCATTGTG expression of GST-fusion proteins for 6 h at 171C. TNFAIP1-30NGSP CCCAGAGGCCCGTATCTATGAGGAGACA After induction, cells were lysed by sonication. The His-tag fusion protein and GST fusion proteins were purified using Chelating Sepharose Fast Flow and Glutathione Sepharose 4B (Amer- Plasmids construction sham Pharmacia), respectively. The purity of the proteins was analyzed by 10% SDS–PAGE. The expression plasmids pT7–PCNA were con- structed as follows: the coding sequence of PCNA was amplified from rat liver cDNA using primer pairs with an EcoR I site (underlined) at their 5’ ends (Forward: TTTGAATTCCCATGTTTGAGG- CA CGCCTGATC, Reverse: TTTGAATTCCTAA- BIAcore assay GATCCTTCTTCATCTTCGAT), digested with Binding kinetics of the PCNA to PDIP1 and EcoRI, and cloned in-frame with His-Tag into TNFAIP1 were monitored by surface plasmon the modified pT7 vector (kindly provided by Dr. resonance (SPR) using a Biacore X apparatus Antero G. So). The resulting constructs pT7– (Biacore AB Uppsala, Sweden) according to the PCNA were sequenced to confirm the orientation manufacturer’s instructions. The His-tag fusion and sequence. To construct plasmids expressing protein of PCNA was covalently immobilized onto rat PDIP1or TNFAIP1, two pairs of PCR primers the dextran modified surface of CM5 sensor chip (PDIP1 primers: TTTGTCGACTCCGCT- as below: 35 ml of a solution containing 50 mM N- CACTGGCATGT and GGGGTACCTCAGTCCTT- hydroxysuccinimide and 200 mM 1–ethyl–3–(3– GAAGACAATCTGT; TNFAIP1 primers: dimethylaminopropyl) carbodiimide were injected AAAGTCGACGATGTCAGGGGACACCTGT and to activate the dextran-modified surface. Then, the GGGGTACCTCAGTCACGATGAGTGGA) with chip was modified with PCNA protein (40 mg/ml in Sal I site at the 50 end of the forward primers 10 mM sodium acetate, pH 4.5). Finally, the and Kpn I site at the 50 end of reverse primers residual reacting sites were blocked with 35 ml were used to amplify the coding sequences of these solution of ethanolamine hydrochloride (pH8.6, two genes respectively from rat liver cDNA. The 1M). The above immobilization was conducted amplified fragments were digested and cloned in- with a flow rate of 5 ml/min. After immobilization, frame with HA-tag into pCMV-Myc (Clontech). this modified chip was ready for interaction The resulting constructs pCMV-Myc-PDIP1and measurement. All measurements were conducted pCMV-Myc-TNFAIP1 were sequenced to confirm in HBS buffer (10 mM HEPES, pH 7.4, 150 mM the reading frame and sequence. The coding NaCl, 3 mM EDTA and 0.005% Surfactant P20) sequences of PDIP1 and TNFAIP1 were excised with a flow rate of 10 ml/min. In all cases, 20 mlof subsequently from pCMV-Myc-PDIP1and pCMV- different concentrations of GST fusion proteins of Myc-TNFAIP1 with Sal I and Kpn I and subcloned PDIP1, dissolved in HBS buffer, was injected into into pGEX–4T–2 (Amersham Pharmacia, Frei- the BIAcore, and the kinetics of binding were burg, Germany), resulting in pGEX–4T–2–PDIP1 recorded by monitoring the increase in resonance and pGEX–4T–2–TNFAIP1. for 90 sec. Next, HBS buffer was injected and dissociation was followed for 50 sec. After each Expression and purification of measurement, the chip surface was regenerated recombinant proteins with 10 ml of 1 mM Gly-HCl, pH 2.0. GST was used The recombinant expressing plasmids (pT7– as a control protein to confirm the specificity of the PCNA, pGEX–4T–2–PDIP1 and pGEX–4T–2– interaction. 中国科技论文在线______http://www.paper.edu.cn

230 J. ZHOU ET AL.

GST pull-down assay DNA polymerase assay 2 mg of GST-PDIP1, GST-TNFAIP1, or GST was GST, GST-PDIP1, or GST-TNFAIP1 was pre- mixed with 2 mg of His-PCNA in binding buffer incubated with 0.3 unit of calf thymus pol d at 41C (25mM HEPES pH 7.5, 12.5 mM MgCl2, 5mM for 30 min, and then assayed for DNA polymerase DTT, 20mM ZnCl2, 150 mM KCl, 10% glycerol (vol/ activity. DNA polymerase d assay using poly(dA)/ vol), 0.1% NP–40 and protease inhibitors) respec- olig(dT) as template/primer was performed as tively. The reactions were incubated for 2 h at 41C. described by Tan et al. (’86). Then 20 ml of glutathione–Sepharose beads (Amer- sham Pharmacia) was added to each tube followed by incubation on a shaker for another 2 h at 41C. RESULTS The tubes were centrifuged at 500 g for 2 min, and the pellets were washed three times for 15 min at cDNA cloning of rat PDIP1 homologues 41C by resuspending and centrifuging them in To find the rat homologue of human PDIP1, a bead-binding buffer without glycerol and Triton series of sequence similarity searches was con- X–100. After the fourth wash, the beads were ducted in the GenBank EST database. This study resuspended with 20 ml of loading buffer and found two groups of overlapping EST sequences heated at 1051C for 5 min, then resolved on SDS/ that shared significant homologies with human PAGE. Proteins were detected by Western analy- PDIP1. One group included the EST sequences sis using the anti-PCNA monoclonal antibody with the GenBank accession numbers BE110670 (Sant Cruz). and AW141960; another group included BF551628 and AI598359. To confirm the sequences of the above two contigs assembled from overlapping Cell culture, transfection, and co- EST sequences, two pairs of primers were de- immunoprecipitation signed and used to amplify the rat liver cDNA. The NIH3T3 cells were grown in DMEM (Invitro- amplification products were cloned into pMD18–T gen/Gibco, Carlsbad, CA) supplemented with 10% vector and sequenced. Based on the sequence of newborn calf serum, L–glutamine (2 mM), peni- these two contigs, 50– and 30– RACE primers were cillin (100 U/ml), and streptomycin (100 g/ml) at designed; the full-length cDNAs of two rat homo- 371C in a 5% CO2 incubator (Shellab, USA). Cells logue genes of human PDIP1 by RACE-PCR were were co-transfected with pT7–PCNA and pCMV- then successfully cloned. Myc-PDIP1, or pCMV-Myc-TNFAIP1 using Lipo- One of the rat cDNA sequences consists of 1716 fectamine 2000 (Invitrogen). After 48 h, cells were bases with an ATG codon at nucleotides 178–180. harvested and lysed with RIPA buffer (50mM Tris, An in-frame stop codon upstream is found at pH 7.2, 150mM NaCl, 1% Triton X–100, 1% positions 55–57. The open reading frame is sodium deoxycholate,0.1% SDS, and 1mM PMSF) followed by a 549–base 30–untranslated region containing a cocktail of protease inhibitors (Sigma, with a polyadenylation signal (AATAAA) at posi- Germany) by freezing at 701C for 30 min and tions 1670–1675. The deduced amino acid se- then thawing at 371C for 3 min. The freeze-thaw quence predicts a protein of 329 amino acids cycle was repeated 3 times. The cell extracts were with a calculated molecular mass of 36.4 kDa. incubated for 2 h at room temperature with anti- Using the BLAST program (www.ncbi.nlm.nih.- Myc antibody (BD Biosciences Pharmingen, San gov/blast), the nucleotide and deduced amino acid Diego, CA). Sepharose coupled protein-A/G (Santa sequences of this cDNA were shown to share a Cruz) was then added for a further 3 h incubation. significantly high identity with human PDIP1 The beads were washed four times with 1ml RIPA (GenBank accession no: AF401315), in particular, and the proteins associated with the beads were 72.0% identity for nucleotides and 92.7% identity heated at 1051C for 5 min, centrifuged, and then for amino acids. These authors believe this rat resolved on 10% SDS–PAGE. Proteins were cDNA is a homlogue of human PDIP1, and detected by Western blot analysis using the anti- therefore named this homologue as rat PDIP1. PCNA monoclonal antibody. Conversely, the ex- Another cDNA sequence consists of 2123 bases periment was performed by using anti-PCNA with an ATG codon at nucleotides 172–174. An in- antibody in immunoprecipitation reactions, and frame stop codon upstream is found at positions using anti-Myc antibody in the Western-blot 61–63. The open reading frame ends at position analysis. 1122 followed by a 1002–base 30–untranslated 中国科技论文在线______http://www.paper.edu.cn

TWO RAT PDIP1 HOMOLOUGES INTERACT WITH PCNA 231 sequence. The cDNA does not contain a polyade- there are seven casein kinase II sites (positions 0 nylation signal in the 3 – untranslated sequence. S22,T60,S110,S155,S186,T250,T315), eight protein However, there is a poly A sequence in the rat kinase C sites (positions S32,T55,S71,S89,S110, 0 genomic sequence downstream of 3 untranslated S155,T166,T245) and one cAMP-dependent protein region. This cDNA encodes a protein of 316 amino kinase site (position T247), whereas in the se- acids with a calculated molecular mass of 36.0 quence of rat TNFAIP1 there are nine casein kDa. The sequence shares lower identity with kinase II sites (positions T47,T65,S97,S142,S173, human PDIP1, 43.8% for nucleotide and 63.0% for T237,T265,T278,T301), nine protein kinase C sites amino acids, respectively. However, it shares (positions T42,S58,T65,S142,S153,S190,T232,T258, higher identity with another human gene – T270) and three cAMP-dependent protein kinase TNFAIP1 (GenBank accession no: NM_021137). sites (positions S127,T234,S298). Analysis of The identity of the nucleotide between the two hydrophilicity of the two rat proteins revealed cDNAs is 61.2% while the amino acid identity is that the rat PDIP1 is more hydrophobic than the 96.2%. These authors conclude that this cDNA is a rat TNFAIP1. The hydrophobic amino acid re- rat homologue of human TNFAIP1 and name this sidues (L, G, A, V, I, F, M, W) in the rat PDIP1 homologue rat TNFAIP1. Human TNFAIP1 is a represent 41.6% of the total protein while those of previously identified gene induced by TNF-a (Wolf the rat TNFIAP1 (L, G, V, I, A, F, M, W) represent et al., ’92), and its function has yet to be described. 37.3%. The highly hydrophilic amino acids (E, R, The sequence data of rat PDIP1 and rat TNFAIP1 K, D) in the rat PDIP1 represent 24.9% of the total are available in GenBank with accession no#: protein, whereas those of the TNFAIP1 (R, D, E, AY305029 and AY336949, respectively. K) represent 27.2%. The overall profile of hydrophilicity of the two proteins shows that the N- terminal portion of the protein is more hydro- Amino acids sequence analysis of rat phobic while the C-terminal portion is more PDIP1 and TNFAIP1 and their hydrophilic (Fig. 1). Secondary structures pre- conservation in animal species dicted by Garnier-Robson in the C-terminus of Through a Protein pattern search of rat PDIP1 both proteins are a potential helix-turn-helix motif and TNFAIP1 based on the PROSITE database, a that could form a DNA binding motif (Fig. 1). number of potential phosphorylation sites have The deduced proteins of PDIP1 and TNFAIP1 been identified. In the sequence of rat PDIP1, share 63.1% sequence identity with each other and

Fig. 1. Schematic representation and secondary structure of rat PDIP1 (A) and rat TNFAIP1 (B). Secondary structure prediction was made using Garnier-Robson algorithm in the DNAstar program. Hydrophilicity plot was made according to Kyte and Doolittle with a widow size of 10 residues. 中国科技论文在线______http://www.paper.edu.cn

232 J. ZHOU ET AL. are highly conserved in the majority of the middle human. A spacer of 14 amino acids between the portion of the two proteins. In the N-terminus of two basic clusters has been conserved from the middle portion, there is a highly conserved protozoan parasite to human and the amino acid region corresponding to amino acid residues 41– sequence of the spacer has also been almost 138 of the rat PDIP1 and 28–125 of rat TNFAIP1 perfectly conserved from insect to human. Phylo- that was identified as a consensus sequence of genetic tree analysis and pairwise comparisons BTB/POZ domain (Harrison and Travers, ’90; showed the evolutionary relationship of the PDIP1 Chen et al., ’93; Godt et al., ’93; Kerckaert et al., and related proteins (Fig. 2c). The PDIP1 proteins ’93) (Fig. 2a). In the C-terminus of the middle from human, rat, and mouse share 92.7%–98.8% portion, there is another highly conserved se- identity, and TNFAIP1 proteins from those quence corresponding to residues 168–283 of rat species share 96.2%–98.4% identity. Sequences PDIP1 and 155–270 of rat TNFAIP1. A consensus from zebra fish (AAH44362) and from insects PCNA binding motif (QTKV-EFP) previously (NP_610165, XP_316233) are closer to the identified in the human PDIP1 (He et al., 2001) TNFAIP1. The sequence from Schistosoma japo- is present in this region, corresponding to residues nicum is also closer to TNFAIP1, though it is 249–255 of rat PDIP1 and 236–242 of rat TNFAIP evolutionarily farther away. (Fig 2b). In addition, highly conserved basic residues RK are noted at positions 231–232 Genomic structure and chromosomal (corresponding to rat PDIP1) followed by another mapping of rat PDIP1 and TNFAIP1 highly conserved basic motif KKQTK at positions 247–251 (corresponding to rat PDIP1); the second To determine the genomic structure of rat basic cluster partially overlaps with the PCNA PDIP1 and rat TNFAIP1, rat PDIP1 and rat binding motif (Fig 2b). Sequences with these TNFAIP1 cDNAs were compared against the rat characteristics have been implicated as recogni- genome database. The rat PDIP1 is mapped to tion signals for receptor-mediated transport into chromosome 1q36 cM region, spans approximately the nucleus through nuclear pores (Robbins et al., 18.7 kb, and is organized into six exons. The first ’91). These sequence motifs represent a potential exon of rat PDIP1 contained 50– untranslated nuclear localization signal (NLS). Secondary region and translation initiation site, the last exon structure predicted by Garnier-Robson in this encodes the stop codon and the 30–untranslated region is a potential helix-turn-helix motif that long sequence. The polyadenylation signal (AA- could form a DNA binding motif (Fig.1). Both TAAA) was identified in the 30–untranslated nuclear localization signal and PCNA binding region of the last exon. The rat TNFAIP1 gene is motif are located in the second helix. mapped to chromosome 10q25 cM, spans approxi- Homology searches of the rat PDIP1 and rat mately 12.9 kb, and is composed of seven exons. TNFAIP1 were performed using BLAST program. The first exon does not encode proteins, and the A BLASTP search revealed strong similarity translation initiation site is located in the second among the PDIP1 and related proteins distributed exon. The first exon and the second exon are in various animals including protozoan parasites separated by a large intron (6.2 kb). All exon- (Schistosoma japonicum), insects (eg. Anopheles intron junctions of both genes accord with the gambiae, Drosophila melanogaster), zebra fish (eg. GT/AG splice donor/receptor rule (Shapiro and Danio rerio), and mammals (Fig. 2). Proteins from Senapathy, ’87). all these species seem to be highly conserved, in particular in the majority of the middle portion of Expression and purification of rat PDIP1 the proteins. The PCNA binding motif (QTKV- and rat TNFAIP1 EFP) is totally conserved from fish to mammals, while two insect sequences have an H at the The rat PDIP1 and rat TNFAIP1 genes were position Q as in the other sequences. However, the cloned into the pGEX–4T–2 vector and expressed residues KVEFP are identical in the sequences as a GST fusion protein, respectively. The rat from all species in which residues V and F have PCNA gene was cloned in-frame with His-Tag into been shown to be critical for PCNA binding (He pT7 vector. After induction with IPTG, the et al., 2001). The potential nuclear localization induced recombinant proteins were purified using signal is highly conserved from zebra fish to affinity chromatography and analyzed by SDS– human (Fig. 2b). The sequences of the two basic PAGE (Fig. 3). SDS–PAGE analysis indicated that clusters are identical in proteins from zebra fish to the fusion proteins of rat PDIP1 and rat TNFAIP1 中国科技论文在线______http://www.paper.edu.cn

TWO RAT PDIP1 HOMOLOUGES INTERACT WITH PCNA 233

Fig. 2. Highly conserved regions (A, B) and phylogenetic tree (C) of PDIP1 related proteins: rp (rat PDIP1), rt (rat TNFAIP1), mp (mouse PDIP1), mt (mouse TNFAIP1), hp (human PDIP1), ht (human TNFAIP1), dr (accession no: AAH44362, Danio rerio), dm (accession no: NP_610165, Drosophila melanogaster), ag (accession no: XP_316233, Anopheles gambiae), sj (accession no:AAP05924, Schistosoma japonicum). Amino acids potentially important in nuclear transport are boxed. Identical amino acids are shaded in black or grey. Numbers indicate the amino acid position relative to the N-terminus of each protein. 中国科技论文在线______http://www.paper.edu.cn

234 J. ZHOU ET AL.

Fig.3. Electrophoresis of puriﬁed proteins expressed in bacteria. Lane1, GST-PDIP1; lane 2, GST-TNFAIP1; lane 3, GST; lane 4, His-PCNA.

1 migrated at molecular weight of about 63 kDa, Equilibrium binding (affinity) constants (KA,M ) which were the same with the molecular weights were evaluated from the association and dissocia- predicated from primary amino acid sequence. tion phases of the sensorgrams using BIA evalua- tion Software (Biacore AB, Uppsala, Sweden). Interaction of rat PDIP1 and TNFAIP1 Each analyte was tested with four different with PCNA by BIAcore assay concentrations (4.50 mM, 3.00 mM, 2.25mM, 1.50 The human PDIP1 was demonstrated to be mM). Each concentration was tested at least three capable of interacting with PCNA by GST-pull- times and the average value obtained. 4.50 mM down assay (He et al., 2001). As mentioned above, GST was used to check the specificity of the both rat PDIP1 and TNFAIP1 contain common interaction, and a non-detectable binding signal PCNA-binding motifs, such as human PDIP1, was obtained, confirming the specificity of the suggesting that rat PDIP1 and TNFAIP1 should interaction between PCNA and PDIP1 or be capable of binding to PCNA. To confirm their TNFAIP1. From Fig. 4, it is obvious that PDIP1 interactions, binding interaction experiments and TNFAIP1 are capable of binding to PCNA, were first conducted in the Biacore X, a system and the average binding signals were similar while for real-time biomolecular interaction analysis using the same concentration. The estimated (BIA) using surface plasmon resonance technology equilibrium binding constants of PDIP1 and 7 6 (reviewed by Fivash et al., ’98). In the Biacore, TNFAIP1 were also similar, (1.6 0.7) 10 KA, 1 7 6 1 His-tagged PCNA was immobilized on the surface M and (2.0 0.7) 10 KA,M , respectively. of a sensor chip, while analyte (GST-tagged PDIP1 The interactions of rat PDIP1 and rat or TNFAIP1) is injected in the flow over the TNFAIP1 with PCNA by GST pull-down surface. If the analyte binds the PCNA protein, and co-immunoprecipitation then a signal is generated in real time and displaced on the sensorgram (Resonance Units vs To further confirm the interaction of rat PDIP1 time). A solution of each analyte was injected into and rat TNFAIP1 with PCNA, in vitro GST the system, and association and dissociation pull-down and co-immunoprecipitation (co-IP) phases of the interaction were recorded (Fig. 4). assays were carried out. For the GST pull-down 中国科技论文在线______http://www.paper.edu.cn

TWO RAT PDIP1 HOMOLOUGES INTERACT WITH PCNA 235

Fig. 4. Interaction of rat PDIP1 and TNFAIP1 with PCNA in BIAcore assay. Puriﬁed rat PCNA was immobilized on the sensor chip and probed with different concentrations of GST-rat PDIP1 (A), GST-rat TNFAIP1 (B). GST was used at the highest concentration (4.50 mM) in the test to monitor unspeciﬁc background (C). Region I represented baseline, region II represented association phase and region III represented dissociation phase.

experiment, PCNA was tagged with histidine (His- alone, and glutathione–Sepharose beads were then PCNA), whereas rat PDIP1 and rat TNFAIP1 added to each mixture, followed by incubation and were fused to GST, respectively. His-PCNA was wash. The bound proteins were detected by mixed with GST-PDIP1, GST-TNFAIP1, or GST Western analysis using anti-PCNA antibody. The 中国科技论文在线______http://www.paper.edu.cn

236 J. ZHOU ET AL. results showed that His-PCNA binds to GST- PDIP1 and TNFAIP1 but fails to bind to GST alone (Fig. 5). In vivo evidence for this association was obtained in a co-IP assay. Two types of plasmids expressing Myc-tagged PDIP1 (or TNFAIP1) and HA-tagged PCNA were used to co-transfect NIH3T3 cells. In one set of experiments, the anti-Myc antibody was used in precipitation reactions, and the anti-PCNA antibody was used in the Western-blot analysis. Conversely, the anti-PCNA antibody was used in precipitation reactions, and the anti-Myc antibody was used in the Western-blot analysis. When cell lysate was precipitated with anti-Myc antibody, PCNA was detected in Myc-tagged PDIP1 or Myc-tagged TNFAIP1 immune complexes (Fig 6). No PCNA was detected when pre-immune serum was used in the precipitation reaction. Conversely, when cell lysate was precipitated with anti-PCNA antibody, Myc-tagged PDIP1 or Myc-tagged TNFAIP1 was detected in PCNA immune complex (Fig 6). Myc- tagged PDIP1 or Myc-tagged TNFAIP1 was not detected when pre-immune serum was used in the precipitation reaction.

Rat PDIP1 and rat TNFAIP1 stimulates pol d activity in the presence of PCNA The human PDIP1 was demonstrated to be capable of stimulating pol d activity in the presence of PCNA (He et al., 2001). To identify whether rat PDIP1 and rat TNFAIP1 have the

Fig. 6. Interaction of rat PDIP1 and rat TNFAIP1 with PCNA in co-immunoprecipitation assay. (a) NIH3T3 cells were co-transfected with PCNA and Myc-tagged PDIP1 or (b) with PCNA and Myc-tagged TNFAIP1. Lysates from transfected NIH3T3 were used for immunoprecipitation. The levels of starting PCNA, PDIP1 and TNFAIP1 in the transfected cells were determined by using same antibody for both immunoprecipitation and Western blot. About 500 mgof protein lysate was used for each co-immunoprecipitation. IP, immunoprecipitation.

same function as human PDIP1, the effects of rat Fig. 5. Interaction of rat PDIP1 and rat TNFAIP1 with PDIP1 and rat TNFAIP1 on the DNA polymerase PCNA in GST-pull down assay. GST-PDIP1 (lane 1) GST- activity of pol d were examined in the presence TNFAIP1 (lane 2), and GST (lane 3) was mixed with His- and absence of PCNA. In the absence of PCNA, PCNA, then glutathione–Sepharose beads was added to each mixture followed by incubation and wash. The bound proteins GST-PDIP1 and GST-TNFAIP1 had little or no were detected by Western blot analysis using anti-PCNA effect on the activity of pol d. However, in the antibody. presence of PCNA, GST-PDIP1 and GST-TNFAIP1 中国科技论文在线______http://www.paper.edu.cn

TWO RAT PDIP1 HOMOLOUGES INTERACT WITH PCNA 237

Fig. 7. Rat PDIP1 and rat TNFAIP1 stimulates PCNA-dependent pol d activity. GST, GST-PDIP1, or GST-TNFAIP1 was preincubated with 0.3 unit of calf thymus pol d at 41C for 30 min, and then assayed for DNA polymerase activity. White bars, activity in the absence of PCNA; gray bars, activity in the presence of PCNA. Data shown are mean7SD of triplicate experiments.

stimulated the activity 3.1– and 3.9–fold, respec- shown that a synthetic peptide containing this tively (Fig. 7). PCNA-binding motif was shown to bind PCNA. A double mutated peptide (V-A, F-A) within the DISCUSSION motif could not bind to PCNA, suggesting the specificity of this PCNA-binding motif (He et al., This study reported the cloning of two rat 2001). This PCNA-binding motif is present PDIP1 related genes, termed rat PDIP1 and rat in two rat proteins and has been conserved TNFAIP1. The rat PDIP1 protein shares 92.7% through the evolution, from protozoan parasite amino acid identity with human PDIP1 protein, to human (Fig 2). This finding strongly suggests whereas the rat TNFAIP1 shares high amino acid that the function involving in DNA synthesis identity (96.2%) with another human PDIP1– (DNA replication and/or repair) of the PDIP1 related protein, the human TNFAIP1. The PDIP1 and related genes has been highly conserved family of proteins was quite conserved and through the evolution. Though the precise role of distributed in various animals including protozoan these two proteins in DNA replication or repair is parasites (Schistosoma japonicum), insects (eg. not known at present, one of the possible Anopheles gambiae, Drosophila melanogaster), mechanisms to increase the DNA synthesis is to fish (eg. Danio rerio) and mammals (Fig 2). stabilize the association of pol d and PCNA. A Previous study has shown that the human PDIP1 number of potential phosphoryzation sites, of interacts with PCNA and stimulates DNA pol casein kinase II, protein kinase C, and protein d activity in presence of PCNA (He et al., 2001). kinase A in rat PDIP1 and rat TNFAIP1 proteins This report also showed that both rat PDIP1 and were identified. Primary data indicated that the TNFAIP1 genes encode functional proteins. Like two proteins can be phosphorylated by casein the human PDIP1 protein, they interact kinase II (data not shown). The activity of the with PCNA and stimulate pol d activity 3–4 fold PDIP1 and related proteins is quite possibly in a PCNA-dependent way. A putative PCNA- regulated by phosphorylation and dephosphoryla- binding motif was previously identified within tion mechanisms. These experiments are in pro- the C-terminus of human PDIP1. It has been gress in our laboratory. 中国科技论文在线______http://www.paper.edu.cn

238 J. ZHOU ET AL.

The PDIP1 and TNFAIP1 proteins are highly addition, in several transcription factors, the BTB/ conserved in the majority of the middle portion of POZ domain was also shown to be involved in the two proteins. The two proteins share in interactions with various co-repressor proteins common the presence of the near N-terminal (Hong et al., ’97; Huynh and Bardwell, ’98; Lin BTB/POZ domain. The BTB/POZ domain was et al., 98). The hydrophilic interactions with co- originally identified in a group of transcription repressors were proposed to occur through a factors such as Bcl6 (Kerchaert et al., ’93), PLZF cluster of conserved residues that lie at a surface (Chen et al., ’93), Drosophila Tramtrack (Harrison exposed groove formed at the dimmer interface. and Travers, ’90) and bric-a-brac proteins (Godt Analysis with DNAStar software showed several et al., ’93). It is a highly conserved structural motif hydrophilic regions which are rich in charged that has been found in proteins involved in a amino acids, such as Lys, Arg, Asp, and Glu within variety of biological processes from the organiza- the BTB/POZ domain. It is possible that these tion of the cytoskeleton and the nuclear matrix, or regions of the PDIP1 family proteins may be the regulation of transcription to the formation of involved in interacting with other proteins. ion-selective channels (reviewed by Aravind and It has been shown that human PDIP1 is Koonin, ’99). The BTB/POZ domain is most inducible by TNF-a and IL–6 (He et al., 2001), frequently found in combination with other and human TNFAIP1 is also inducible by TNF-a domains such as zinc finger, kelch repeats, ankryn in endothelial cells (Wolf et al, ’92). This finding repeats, or transmembrane domains. Analysis impliesanintriguingpossibilityofaconservedTNFa using PredictProtein softwares (http://maple.bioc.- induced regulatory mechanism mediated through columbia.edu), has not identified any of the typical the PDIP1 and related proteins. Giving the structural domains mentioned above. However, in evidence that one of TNF-a’s functions is to the C-terminal portion of the two proteins, there initiate the liver regeneration (Michalopoulos are several highly conserved cystine residues and DeFrance, ’97; Fausto, 2000), it is quite forming a consensus structure of CX12C72– possible that the PDIP1 family may play roles in 74CCX12CC which is different from the typical liver regeneration by linking the TNF-a signal and zinc finger motif of C2H2,C3HorC4. Whether this DNA synthesis. In fact, in a lipopolysaccharide structure can also chelate zinc ion(s) and form a (LPS)-induced liver injury mouse model, the domain that binds to DNA needs to be further TNFAIP1 protein, was significantly up-regulated determined. In addition, there are a several highly at 6 hours after the LPS treatment (Liu et al, conserved leucine residues at the C-terminal 2004). Our preliminary results in a rat partial portion of the two proteins, where every 5–7 hepatectomy model also revealed up-regulation of amino acids, instead of every seventh amino acid, transcription of PDIP1 at 12 hours and TNFAIP1 is a leucine, as identified in the typical leucine at 10 hours after surgery (unpublished data). zipper motif. Whether these leucine residues could These results suggest the potential role of PDIP1. form a leucine zipper needs to be further tested. The expression of PDIP1 is not restricted to The BTB/POZ domain mediates protein-protein liver tissue and the PDIP1 gene has been detected interactions. In several proteins, the BTB/POZ in every tissue tested including liver, colon, domain has been found to participate in the thymus, kidney, small intestine, placenta, lung, formation of homo-oligmers (Bardwell and Treus- and leukocyte (He et al., 2001). The expression of man, ’94; Li et al., ’97; Espinas et al., ’99) and it rat PDIP1 and rat TNFIAP1 was also detected in was also proposed that BTB/POZ domains could every rat tissue by RT PCR including brain, liver, form hetero-oligomers (Bardwell and Treusman, colon, thymus, kidney, small intestine, placenta, ’94). The BTB/POZ domain of the human pro- lung, spleen, and skin (data not shown), suggest- myelocytic leukaemia zinc finger (PLZF) protein ing that PDIP1 and related proteins may play was crystallized and its structure was resolved roles not only in liver regeneration but also in (Ahmad et al., ’98). The crystal structure shows a developmental or disease processes of other tightly interwound dimmer and, therefore, pro- tissues. Recently, Link et al. (2003) engineered vides a general structural basis for the participa- transgenic C. elegans animals to express the tion of BTB/POZ domain in homomeric human amyloid peptide (Ab). In this transgenic interactions. Dimerization occurs through an C. elegans model, expression of C. elegans extensive hydrophobic interface that involves TNFAIP1 was found to be robustly induced. To approximately one quarter of the monomer acces- ask whether this Ab-induced gene is also up- sible surface (Ahmad et al., ’98; Li et al., ’99). In regulated in Alzheimer’s disease (AD) brains, 中国科技论文在线______http://www.paper.edu.cn

TWO RAT PDIP1 HOMOLOUGES INTERACT WITH PCNA 239 quantitative RT-PCR was used to assay tissue Bardwell VJ, Treusman R. 1994. The POZ domain: a samples from specific brain regions of AD for conserved protein-protein interaction motif. Genes Dev TNFAIP1 expression. Up-regulation was found for 8:1664–1677. Barger SW, Horster D, Furukawa K, Goodman Y, Krieglstein TNFAIP1 in the superior frontal gyrus and the J, Mattson MP. 1995. Tumor necrosis factors alpha and beta cerebellum, but not in the hippocampus. The protect neurons against beta-peptide toxicity: evidence for pattern of TNFAIP1 RNA up-regulation observed involvement of a kappa B-binding factor and attenuation of in the AD brain samples did not parallel the peroxide and Ca2þ accumulation. Proc Natl Acad Sci USA typical pattern of brain pathology severity ob- 92:9328–9332. served in the brain regions assayed (i.e. hippo- Chen Z, Brand NJ, Chen A, Chen SJ, Tong JH, Wang ZY, Waxman S, Zelent A. 1993. Fusion between a novel Kruppel- campus4superior frontal gyrus4cerebellum). It like zinc finger gene and the retinoic acid receptor-alpha has been suggested that the TNFAIP1 plays locus due to a variant t (11:17) translocation associated with protective roles during the process of developing acute promyelocytic leukaemia. EMBO J 12:1161–1167. AD, and thus brain regions with the least Espinas ML, Jimenez-Garcia E, Vaquero A, Canudas S, pathology would have the highest expression of Bernues J, Azorin F. 1999. The N-terminal POZ domain of GAGA mediates the formation of oligomers that bind TNFAIP1 (Link et al., 2003). Alternatively, it is DNA with high affinity and specificity. J Biol Chem possible that TNAIP1 is up-regulated in cells very 274:16461–16469. early during the disease process and hippocampus Fausto N. 2000. Liver regeneration. J Hepatol 32 suppl: neurons expressing TNFAIP1 have been lost in 19–31. the course of the disease. One of the molecular Fiers W, Beyaert R, Boone E, Cornelis S, Declercq W, Decoster mechanisms involving Ab toxicity is to cause an E, Denecker G, Depuydt B, De Valck D, De Wilde G, Goossens V, Grooten J, Haegeman G, Heyninck K, Penning early DNA strand breakage. The extensive DNA L, Plaisance S, Vancompernolle K, Van Criekinge W, damage provoked by Ab constitutes an early event Vandenabeele P, Vanden Berghe W, Van de Craen M, in the pathogenic cascade leading to neuronal Vandevoorde V, Vercammen D. 1995. TNF-induced intra- death, which could contribute to the neuropatho- cellular signaling leading to gene induction or to cytotoxicity genesis of AD (Santiard-Baron et al., 2001). It has by necrosis or by apoptosis. J Inflamm 47:67–75. Fivash M, Towler EM, Fisher RJ. 1998. BIAcore for macro- been previously reported that TNFa treatment molecular interaction. Curr Opin Biotech 9:97–101. protects hippocampal neurons from Ab toxicity Godt D, Couderc JL, Cramton SE, Laski FA. 1993. Pattern (Barger et al., ‘95). Together, it is possible that formation in the limbs of Drosophila: bric a brac is expressed TNFAIP1 functions as a bridge between the TNF- in both a gradient and a wave-like pattern and is required a signal and the neuron survival, probably by for specification and proper segmentation of the tarsus. repairing the damaged DNA caused by Ab in Development 119:799–812. Gulbis JM, Kelman Z, Hurwitz J, O’Donnell M, Kuriyan J. neurons. 1996. Structure of the C-terminal region region of p21 In conclusion, the PDIP1 and TNFAIP1 may be WAF1/CIP1 complexed with human PCNA. Cell 87:297–306. a general link between cytokine signals and Harrison SD, Travers AA. 1990. The tramtrack gene encodes a subsequent processes involving DNA synthesis/ Drosophila finger protein that interacts with the ftz repair, and they fulfill their roles in these transcriptional regulatory region and shows a novel embryonic expression pattern. EMBO J 9:207–216. processes through interacting with a number of He H, Tan CK, Downey KM, So AG. 2001. A tumor necrosis components in DNA synthesis and repair. factor - and interleukin 6–inducible protein that interacts with the small subunit of DNA polymerase and proliferating ACKNOWLEDGEMENTS cell nuclear antigen. Proc Natl Acad Sci USA 98:11979– 11984. The authors are grateful to Prof. Songping Hong SH, David G, Wong CW, Dejean A, Privalski ML. 1997. Liang and Prof. Mei Han (College of Life Science, SMRT corepressor interacts with PLZF and with PML- retinoic acid receptor alpha (RAR alpha) and PLZF-RAR Hunan Normal University, China) for their help alpha oncoproteins associated with acute promyelocytic in BIAcore assay and critical reading of the leukemia. Proc Natl Acad Sci USA 94:9028–9033. manuscript, respectively. Huynh KD, Bardwell VJ. 1998. The BCL–6 POZ domain and other POZ domains interact with the co-repressors N-coR LITERATURE CITED and SMAT. Oncogene 17:2473–2484. Kedar PS, Kim SJ, Robertson A, Hou E, Prasad R, Horton JK, Ahmad KF, Engel CK, Prive GG. 1998. Crystal structure of Wilson SH. 2002. Direct interaction between mammalian the BTB domain from PLZF. Proc Natl Acad Sci USA DNA polymerase b and proliferating cell nuclear antigen. 95:12123–12128. J Biol Chem 277:31115–31123. Aravind L, Koonin EV. 1999. Fold prediction and evolutionary Kerckaert JP, Deweindt C, Tilly H, Quief S, Lecocq G, Bastard analysis of the POZ domain: structural and evolutionary C. 1993. LAZ3, a novel zinc-finger encoding gene, is relationship with the potassium channel tetramerization disrupted by recurring chromosome 3q27 translocations in domain. J Mol Biol 285:1353–1361. human lymphomas. Nat Genet 5:66–70. 中国科技论文在线______http://www.paper.edu.cn

240 J. ZHOU ET AL.

Krishna TS, Kong XP, Gary S, Burgers PMJ, Kuriyan J. 1994. Prelich G, Tan CK, Kostura M, Mathews MB, So AG, Downey Crystal structure of the eukaryotic DNA polymerase KM, Stillman B. 1987. Functional identity of proliferating processivity factor PCNA. Cell 87:1233–1243. cell nuclear antigen and a DNA polymerase auxiliary Lin RJ, Nagy L, Inoue S, Shao W, Miller WH Jr, Evans RM. protein. Nature 326:517–520. 1998. Role of the histone deacetylase complex in acute Robbins J, Dilworth SM, Laskey RA, Dingwall C. 1991.Two promyelocytic leukaemia. Nature 391:811–814 interdependent basic domains in nucleoplasmin nuclear Li X, Lopez-Guisa JM, Ninan N, Weiner EJ, Rauscher III FJ, targeting sequence: identification of a class of bipartite Marmorstein R. 1997. Overexpression, purification, char- nuclear targeting sequence. Cell 64:615–623. acterization, and crystallization of the BTB/POZ domain Santiard-Baron D, Lacoste A, Ellouk-Achard S, Soulie C, from the PLZF oncoprotein. J Biol Chem 272:27324–27329. Nicole A, Sarasin A, Ceballos-Picot I. 2001. The amyloid Li X, Peng H, Schultz DC, Lopez-Guisa JM, Rauscher FJ 3rd, peptide induces early genotoxic damages in human pre- Marmonstein R. 1999. Structure-function studies of the neuron NT2. Mutat Res 479 (1–2):113–120. POZ/BTB transcriptional repression domain from the Shapiro MB, Senapathy P. 1987. RNA splice junctions of promyelocytic leukemia zinc finger oncoprotein. Cancer different classes of eukaryotes: sequence statistics and Res 59:5275–5282. functional implications in gene expression. Nucl Acid Res Link CD, Taft A, Kapulkin V, Duke K, Kim S, Fei Q, Wood DE, 15:7155–7174. Sahagan BG. 2003. Gene expression analysis in a transgenic Tan CK, Castillo C, So AG, Downey KM. 1986. An auxiliary Caenorhabditis elegens Alzheimer’s disease model. Neurol- protein for DNA polymerase-delta from fetal calf thymus. ogy of Aging 24:297–413. J Biol Chem 261:12310–12316. Liu XW, Lu FG, Zhaang GS, Wu XP, You Y, Ouyang CH, Yang Tsurimoto T. 1999. PCNA binding proteins. Front Biosci DY. 2004. Proteomics to display tissue repair opposing 4:849–858. injury response to LPS-induced liver injury. World J Wolf FW, Marks RM, Sarma V, Byers MG, Katz RW, Shows Gastroenterol 10:2701–2705. TB, Dixit VM. 1992. Characterization of tumor necrosis Lu X, Tan CK, Zhou JQ, You M, Carastro LM, Downey KM, So factor-a-induced protein endothelial primary response gene. AG. 2002. Direct interaction of proliferating cell nuclear J Biol Chem 267:1317–1326. antigen with the small subunit of DNA polymerase d. J Biol Zhou JQ, Tan CK, So AG, Downey KM. 1996. Purification and Chem 277:24340–24345. characterization of the human catalytic subunit of human Michalopoulos GK, DeFrances MC. 1997. Liver regeneration. DNA polymerase d expressed in baculovirus-infected insect Science 276:60–66. cells. J Biol Chem 271:29740–29745.