INTRODUCTION

Liver regeneration is an ideal model for study of cell proliferation1,2. In intact adult liver the majority of hepatocytes stay in quiescent and highly differentiated status. In the condition of cell loss by surgical resection, infectious, or toxic injury, hepatocytes almost initiate immediately replication and a wave of cell proliferation appears.

In 70% partial hepatectomy (PHx) liver regeneration model, the majority of cells in the remnant liver rapidly reenter cell cycle accompanied by the induced expression of a large number of growth-response genes. Several immediate early-phase genes related to cell proliferation can be induced within 2 h. These genes include c-fos, c-jun, etc3. The onset of S phase of cell cycle emerges at about 12 h and DNA synthesis reached the peak at 24 h4. The remainder liver will double its volume after 48 h and restore the mass of liver before hepatectomy within day 7 to 10 when proliferation ceases5.

Liver-specific genes play important roles in liver functions3,6. lfire-1 gene (GenBank accession number: AF168954) was isolated by our lab and has 97.6% identity with hfrep-17, but a main discrepancy at their 5′ untranslated region (manuscript submitted). Since both of them encode the same polypeptide, we designated it as hfrep-1/lfire-1 here. HFREP-1/LFIRE-1 belongs to fibrinogen superfamily, whose members share a common fibrinogen-like domain at their carboxy-termini7. They might act as 'molecular facilitator' participating in the proliferation and adhesion of cells, angiogenesis and carcinogenesis8, 9, 10. Northern blot assay revealed that hfrep-1/lfire-1 was exclusively expressed in normal rat liver and two human hepatoma cell lines, HepG2 and PLC/PRF/57. However, the functions of hfrep-1/lfire-1 remain elusive.

Here we are the first to report the isolation and characterization of mouse ortholog mfrep-1 gene to human hfrep-1/lfre-1, by in silico cloning. The full-length mfrep-1 cDNA sequence was confirmed by RT-PCR and sequencing. This gene encodes a 314 amino-acid protein exhibiting 80.4% similarity to the human HFREP-1/LFIRE-1. We investigated the expression of mfrep-1 in mouse multiple tissues by Northern blot analysis. To explore its potential function, we employed liver regeneration model by PHx. After 70% hepatectomy of mouse liver, mfrep-1 was induced at 2 h after PHx and the second peak arrived after 24 h. The expression of mfrep-1 maintained high until 72 h and declined to the basal level thereafter. The results suggested that mfrep-1, a liver specific gene, may function as a regulator of cell growth in liver regeneration.

MATERIALS AND METHODS

Animals

Male C57Bl/6J mice (body weight, 20-25 g) were ether-anesthetized and subjected to midventral laparotomy with 70% liver hepatectomy as described by Higgins and Anderson11. For sham operations, mice were anesthetized, an abdominal incision was made, and the liver was manipulated but not removed. For cycloheximide-treated sample, mice were pretreated 15 min prior to laparotomy with cycloheximide (50 mg/kg of body weight). All mice were maintained in temperature-controlled rooms under 12-h dark/light cycles to synchronize feeding.

In silico cloning of mfrep-1

Based on the hfrep-1/lfire-1 sequence, we found its mouse ortholog by EST database searching. We used Blast algorithms through the National Center of Biotechnology Information and downloaded the mouse UniGene cluster Mm. 28340. All of the EST sequences were auto-transformed into the recognizable format for DNASTAR software package (DNASTAR. Inc.) by a self-made software, called SMART-DNA12. These EST sequences were assembled by Seqman program (DNASTAR) and the clone with the longest insertion was purchased. We checked the nucleotide sequence by sequencing the clone in two directions. The primers 5′GAGGCTCTGTGTGGATGGACTG3′ and 5′TTACAGAAAGGAGATCCCAATGAGG3′ were designed according to the sequence above and used in RT-PCR assay with PfuTurbo® DNA polymerase (Stratagene) and mouse liver cDNA as template. The full-length MFREP-1 cDNA was isolated, confirmed by sequencing and submitted to the GenBank.

Computer-aided analysis of mouse MFREP-1

The deduced amino acid sequence of MFREP-1 was aligned against the GenBank databases (nucleotide, EST and protein) at the National Center for Biotechnology Information, using BLAST to search for sequence matches. The alignment of proteins was done using the MegAlign program (DNASTAR. Inc.). The signal peptide fragment was determined by SignalP program V1.1 (http://www.cbs.dtu.dk/services/signalP)13.

RNA extraction and Northern blot assay

Total RNA was isolated by Trizol (GIBCO/BRL). Northern blot assay was performed as described previously14. In brief, 30 μg of total RNA was separated in 1.0% agarose formaldehyde gel and transferred to nylon membrane Hybond XL (Pharmacia) and fixed to the membrane by baking at 80°C for 2 h. Membranes were probed with full-length mfrep-1 cDNA labeled with [α-32P]-dCTP (Amersham) by random-primed labeling system (Promega) at 68 °C vernight followed by washing with 0.2 × SSC/ 0.1% SDS at room temperature for 15 min and then at 60°C or 15 min twice. Then membranes were subjected to phosphor screen with a FLA-3000A plate/Fluorescent Image Analyzer.

Quantitation of mfrep-1 expression level during liver regeneration

The intensity of mfrep-1 bands during liver regeneration was measured and analyzed by ImageGauge program Ver. 3.12 for band quantification (Fuji Photo Film Co. LTD). The value of mfrep-1 expression was determined by calculating the ratio of the expression level at each time during liver regeneration and that in the intact normal liver tissue, each of which was normalized for the corresponding 18S rRNA expression level. Reproducibility was confirmed by three independent Northern blot assays.

Immunohistochemistry

Immunohistochemistry was performed as described previously15. In brief, paraffin-embedded tissues were sectioned at a thickness of 4 to 6 μm sections and were microwaved for 5 min for antigen retrieval. After the specific pretreatment procedures, all samples were incubated with 3% H2O2 in methanol solution for 12 min at room temperature to block endogenous peroxidases. Then the samples were incubated with human HFREP-1/LFIRE-1 polyclonal antibody (prepared in our lab) to localize the MFREP-1 proteins in the cells, and visualized by AEC (Maxin Biotech. Inc.). Sections were blocked and counterstained with hematoxylin. Sections incubated in rabbit serum instead of the corresponding primary antibody were used for comparison as negative control.

RESULTS

Cloning and characterization of the Full-length mfrep-1 cDNA

In order to obtain the mouse ortholog of human HFREP-1/LFIRE-1, we screened the GenBank EST database using HFREP-1/LFIRE-1 cDNA sequence as template. We found that two ESTs (5′ EST: AI181881 and 3′ EST: AI173236), both of which belong to a clone IMAGE: 1450566 from Washington University school of Medicine, contained the longest insert among 57 EST entries in mouse UniGene cluster Mm. 28340. We purchased this clone and sequenced the insert.

To examine the authenticity of the sequence of insert, we designed the primers according to the sequence of 5′ EST (AI181881) and 3′ EST (AI173236) and amplified the ORF of this gene by RT-PCR with mouse adult liver cDNA as template. Sequencing of the resultant PCR product confirmed the identity of the cDNA from IMAGE: 1450566, indicating that we had isolated a novel cDNA of 1127 -bp transcript. This cDNA sequence has a complete Kozak sequence with AGAATGG around the second ATG triplet (Fig 1A)16. As illustrated in Fig 1A, the ORF extends from the ATG (at position 98) codon to the TAA (at position 1042), and encodes 314 amino acids with a predicted molecular mass of 36 kDa. According to the results from Northern blot analysis, we inferred that we had obtained the full-length 1.2 kb cDNA. We designated it as mfrep-1 (mouse fibrinogen-related protein-1).

Figure 1
figure 1

A. The nucleotide and amino acid sequence of mouse MFREP-1. One putative signal peptide predicated is underlined. Four conserved cysteine residues are enclosed in circle. The asterisk indicates a termination codon. One possible polyadenylation signal is boxed. B. Predicted hydrophobic plot of mouse MFREP-1 by Protean program (DNAStar package). Negative values indicate hydrophilic regions and positive values indicate hydrophobic region.

Sequence comparison

The deduced protein from mfrep-1 cDNA has a 230-a.a. fibrinogen-related domain at its carboxy-terminus, which contains four conserved cysteine residues (Fig 1A, 2B). The N-terminal sequence consisting of 22 a.a. was assigned as the signal sequence based on its hydrophobic nature (Fig 1B)13. After comparing MFREP-1 with HFREP-1/LFIRE-1, we found the predicted a.a. sequence of MFREP-1 has 80.4% similarity with its human ortholog HFREP-1/LFIRE-1. The MFREP-1 protein resembled human HFREP-1/LFIRE-1 closely except 2 more unconserved a.a. at N-terminal end and 59 a.a. residue variations between these two proteins, most of which are with the same hydrophathy (Fig 2A).

Figure 2
figure 2

Comparison of MFREP-1 with other members of fibrinogen superfamily. A. Comparison of deduced amino acid sequences of mouse MFREP-1 and human HFREP-1/LFIRE-1. Shaded areas indicate identical amino acids. B. Amino acid sequences alignment of fibrinogen-related domain of MFREP-1 with those of seven other human members of fibrinogen superfamily: HFREP-1/LFIRE-1; angiopoietin-1; ficolin-1; fibrinogen beta; fibrinogen gamma; human MAP, and tenascin. Dark shaded areas indicate the residues conserved in all eight sequences, gray shaded areas indicate residues conserved at least in six sequences.

We compared the deduced a.a. sequence of MFREP-1 with those of other members of the fibrinogen superfamily (Fig 2B). The similarity and divergence is shown in the table (Fig 3A), using Pro-tean program (DNASTAR, Inc.). From these alignments using Megalign, the phylogenetic tree was established (Fig 3B). Our data indicate that MFREP-1 belongs to fibrinogen superfamily.

Figure 3
figure 3

A. Similarity and divergence of MFREP-1 among members of fibrinogen superfamily. B. Phylogenetic tree established with the data from Megalign.

Tissue-specific expression of mfrep-1

Analysis of the tissue distribution of mfrep-1 transcript was carried out using total RNA isolated from mouse uterus, ovary, bladder, pancreas, placenta, oesophagus, brain, heart, lung, stomach, liver, intestine, spleen, kidney, testis, muscle. Our data showed that the mfrep-1 transcript appears as a band of 1.2 kb mRNA only in mouse liver (Fig 4). There existed no detectable signals in other tissues we tested.

Figure 4
figure 4

Northern blot analysis of mfrep-1 mRNA expression in multiple mouse tissue. Total RNA was extracted from 16 various mouse tissues and liver tissues 12 h after partial hepatectomy (12 h reg) and 12 h after sham operation (12 h sham). 30 μg of total RNA was electrophoresed on a 1.0% agarose/formaldehyde gel, transferred to a nylon membrane and hybridized with [α-32P] dCTP-labeled mfrep-1 cDNA probe. rRNA is shown as loading control. The multiple tissue sources of the mRNA in each lane are indicated at top of the blot.

Up-regulated expression of mfrep-1 mRNA and protein during liver regeneration

To explore the function of mfrep-1 in vivo, we analyzed the kinetics of mfrep-1 gene expression during liver regeneration after 70% PHx. In the PHx mice, the liver weight was restored to 96% of the original weight at 168 h after PHx, which indicated a standard time frame of PHx regeneration. As shown in Fig 4, the expression of mfrep-1 was obviously elevated in 12 h regenerative liver compared with normal tissues and 12 h sham operation tissue. Then, the kinetic change of mfrep-1 expression levels during liver regeneration was examined by Northern blot assay. The level of mfrep-1 mRNA remained almost unchanged at 0.5 h and increased to 2.13-fold above quiescent level at 2 h in the PHx mice and maintained thereafter (Fig 5A and B). Interestingly, mfrep-1 expression continued to elevate at 6 h. The expression level reached at 3.42-folds higher than control level 24 h after PHx (Fig 5A and B). Then, the expression level of mfrep-1 RNA slowly declined from day 3 to d 7 around 3-fold above quiescent level, and reduced to 2-fold at d 14 (Fig 5A and C). At d 21, its expression level reached the control level. Pretreatment with protein synthesis inhibitor cycloheximide down-regulated mfrep-1 expression level at 2 h after PHx, indicating the up-regulated expression of mfrep-1 needs prior protein synthesis of induction (Fig 5A).

Figure 5
figure 5

Expression kinetics of mfrep-1 during liver regeneration after PHx. A. The expression of mfrep-1 mRNAs during liver regeneration were determined by Northern blot assay. Total mRNAs was obtained from liver tissues 0.5 h (Lane 2), 1 h (Lane 3), 2 h (Lane 4 and 19), 4 h (Lane 5), 6 h (Lane 6), 12 h (Lane 7), 1 d (Lane 8), 2 days (Lane 9), 3 days (Lane 10), 4 days (Lane 11), 5 d (Lane 12), 6 d (Lane 13), 7 d (Lane 14), 8 d (Lane 15), 12 d (Lane 16), 14 d (Lane 17), 21 d (Lane 18) after PHx, before PHx (Lane 9) as well as RNA from 2 h after 70% hepatectomy with cycloheximide (50 mg/kg of body weight) pretreatment (lane 20). 30 μg of mRNAs was separated in 1.0% agarose gel containing formaldehyde. EB staining of 18S and 28S rRNA was used as loading control. B. Graphic representation of mfrep-1 mRNA level in the growth period of liver regeneration. C. Graphic representation of mfrep-1 mRNA level in the long-term period of liver regeneration. The quantity of mfrep-1 determined by densitometries analysis as shown as increase over that of untreated controls after normalization against signals of 18s control. Each point represents the average of three independent experiments.

To check whether the up-regulated mRNA expression of mfrep-1 was accompanied by the enhanced expression of MFREP-1 protein in mouse liver regeneration, we performed immunohistochemistry assessment using the polyclonal antibody against human HFREP-1/LFIRE-1. Since this polyclonal antibody is prepared by using HFREP-1/LFIRE-1 fragment (a.a. 22-312) as antigen, it can also recognize the evolutionarily conserved MFREP-1. Immunohistochemsitry results revealed that the signals were seen from liver parenchyma cells. The signals were not very strong in normal intact mouse liver and sham-operation mouse liver, but intensive in mouse remnant liver 12 h after PHx, which coincided with the enhanced expression of mfrep-1 mRNA (Fig 6).

Figure 6
figure 6

Immunohistochemical assessment of MFREP-1 protein expression during mouse liver regeneration after PHx. Paraffin-embedded sections (4 to 6 μm) were pretreated and immunodetected by anti-HFREP-1/LFIRE-1 polyclonal antibody. The samples were visualized by AEC (Maxin Biotech. Inc.). Sections were counterstained with hematoxylin. A. mouse intact adult liver, B. mouse liver 12 h after PHx, C. mouse liver 12 h after sham operation. Magnitude, ×200.

DISCUSSION

The isolation and characterization of mouse ortholog gene can offer important information on the structural conservation and potential function of a specific gene. The concept of comparative gene identification (CGI) has been previously used by many laboratories to search for orthologous genes once a particular gene of interest has been identified in other species. Nowadays, with the development of EST database, molecular cloning in silico comes to be an efficient method to isolate the full-length cDNA sequence17,18.

We described here for the first time a novel mouse liver-specific gene for a protein designated mfrep-1, an ortholog gene of human hfrep-1/lfire-1 that has been identified as a liver cancer associated gene. Both of mfrep-1 and hfrep-1/lfire-1 belong to fibrinogen superfamily, with a 220∼250-a.a. fibrinogen-like domain at carboxy-terminus. Fibrinogen β and γ is composed of two each of three homologous polypeptide chains (α2, β2, γ2) and is an essential protein for blood clotting. Tenascins take responsibility for cell adhesion, whereas angiopoietins can stimulate angiogenesis9. The precise functions of the fibrinogen-like domain-containing proteins have not been fully elucidated.

mfrep-1, like hfrep-1/lfire-1, is a liver-specific gene that may play an essential role in liver functions. To investigate the potential role of MFREP-1 in cell proliferation regulation, we used mouse liver regeneration model to analyze the expression kinetics of mfrep-1. The processes of liver regeneration reflect very important events in the cell cycle such as initiation of cell division, termination of cell growth and cell differentiation19,20.

The expression of early genes in PHx can be categorized into two kinds, immediate-early genes and delayed-early genes. Immediate-early genes are induced in a protein synthesis-independent manner in the transition from the normal quiescent state of the liver (G0) to the growth phase (G1). The expression of c-fos and c-jun proto-oncogenes, two well-known immediate-early genes, was induced within 1h after PHx and can be superinduced by pretreatment with cycloheximide. The protein products of these nuclear proto-oncogenes are expressed abundantly in most proliferating cells4. At the cellular level, proteins encoded by immediate-early genes may help control progression through G1 phase. The induced expression of delayed-early genes occurs several hours later and requires new protein synthesis. Delayed-early genes are also important modulators of the regenerative response. The temporal profiles of Bcl-x mRNA expression, which is regarded as delayed-early gene, displays two distinct peaks at 4 hours and 48 to 72 h, suggesting a cell-cycle-dependent regulation21.

In PHx model, the level of mfrep-1 mRNA was found to reach its first higher level at 2 h (2.13 times higher than the basal level) after 70% PHx. Our data showed that after treatment of cycloheximide, a protein synthesis inhibitor, mfrep-1 was reduced rather superinduced 2 h after PHx. mfrep-1 can thus be characterized as a delayed early gene during liver regeneration in light of its expression during G1 phase and its requirement of protein synthesis for induction.

It has been demonstrated that the peak rate of DNA synthesis occurs at about 24 h after PHx and then a minor peak of DNA synthesis appears at about day 35. After the shoulder peak during 2-4 h, the expression level of mfrep-1 mRNA reached nearly 3.42-fold at 24 h and was slightly increased to 3.73-fold at day 3. These two peaks of expression corresponded to the first and second peak of DNA synthesis in hepatocytes. Thus, the results suggest that the mfrep-1 gene is closely associated with DNA synthesis in liver regeneration, and imply that it might be involved in the enhancement of cell replication.

Furthermore, the expression pattern of mfrep-1 gene resembles those of liver-specific genes, such as c/ebp α which shows maximal expression after the growth period of liver regeneration3,4. The expression level of c/ebp α reaches a broad peak between 60 and 216 h after PHx, correlating with the time when the liver cease growing22. This is consistent with the fact that c/ebpα is related with the differentiated state, and in some cells c/ebp α can arrest the proliferative phase23,24. Moreover, the expression of some growth-inhibitory genes, or example, tgf-β and activin a, are also up-regulated and kept high during this period in liver regeneration25, 26, 27, 28, 29. Here we demonstrated that mfrep-1 maintained at more than 2.5-fold the control level after the wave of DNA synthesis. Therefore, we conclude that this gene plays an important role in growth regulation of liver regeneration, but, the mechanism and the significance of enhanced expression of mfrep-1 after day 7 still remains to be defined.

In conclusion, our results revealed that liver-specific gene mfrep-1 was enhanced in liver regeneration after PHx. Its expression pattern in liver regeneration suggested that it might play an essential role in cell proliferation.