Isolation of a novel member of small G protein superfamily and its expression in colon cancer

Abstract

AIM: APMCF1 is a novel human gene whose transcripts are up-regulated in apoptotic MCF-7 cells. In order to learn more about this gene’s function in other tumors, we cloned its full length cDNA and prepared its polyclonal antibody to investigate its expression in colon cancers with immunohistochemistry.

METHODS: With the method of 5’ rapid amplification of cDNA end (RACE) and EST assembled in GenBank, we extended the length of APMCF1 at 5’ end. Then the sequence encoding the APMCF1 protein was amplified by RT-PCR from the total RNA of apoptotic MCF-7 cells and cloned into the prokaryotic expression vector pGEX-KG to construct recombinant expression vector pGEX-APMCF1. The GST-APMCF1 fusion protein was expressed in E. coli and used to immunize rabbits to get the rabbit anti-APMCF1 serum. The specificity of polyclonal anti-APMCF1 antibody was determined by Western blot. Then we investigated the expression of Apmcf1 in colon cancers and normal colonic mucosa with immunohistochemistry.

RESULTS: A cDNA fragment with a length of 1745 bp was obtained. APMCF1 was mapped to chromosome 3q22.2 and spanned at least 14.8 kb of genomic DNA with seven exons and six introns contained. Bioinformatic analysis showed the protein encoded by APMCF1 contained a small GTP-binding protein (G proteins) domain and was homologous to mouse signal recognition particle receptor β(SRβ). A coding region covering 816 bp was cloned and polyclonal anti-APMCF1 antibody was prepared successfully. The immunohistochemistry study showed that APMCF1 had a strong expression in colon cancer.

CONCLUSION: APMCF1 may be the gene coding human signal recognition particle receptor β and belongs to the small-G protein superfamily. Its strong expression pattern in colon cancer suggests it may play a role in colon cancer development.

Key Words: $[Keywords]

INTRODUCTION

Apoptosis or programmed cell death (PCD) is a form of cell death in human and other multicellular organisms[1-15]. It is a genetically controlled program of cellular selfdestruction, which is of central importance to the development and homeostasis of almost all animals, and also a highly regulated process involving a large number of genes that are conserved in all metazoans. A variety of genes have been found to take part in the process of apoptosis, such as the TNF/FasL family, the Bcl-2 family, the ICE/CED-3 family, oncogenes, tumor suppressor genes and the genes encoding heat shock proteins[16-24]. APMCF1 is a novel human gene fragment we cloned in 1999. Its transcript was found to be up-regulated in apoptotic MCF-7 cells, which might play a role in apoptosis of tumor cells[25]. As its 5’ end is not complete, the functional study of this gene has not been performed. In order to learn more about this gene’s biological functions, we extended its 5’ cDNA end and found it was a gene homologous to mouse SRβ. Then polyclonal antibody against this gene’s product was prepared and protein expression of APMCF1 in human colon cancer and normal colonic mucosa was examined with immunohistochemistry method.

MATERIALS AND METHODS

Main reagents

Reverse transcription kit was from TaKaRa. Pfu and Tf1 Taq DNA polymerase were from Promega. TRIzol Reagent was from Gibco. Retinoid acid (RA) was from Sigma.

Cell lines and specimens

MCF-7 cell line was kept by our lab oratory and maintained in DMEM medium supplemented with 100 mL/L fetal bovine serum. The cells were induced to apoptosis by RA with the method described before[25]. Fifty five tissue specimens of colon cancer and 5 tissue specimens of normal colonic mucosa were obtained from surgically resected tissues of patients with hepatocellular carcinoma in Xijing Hospital of Fourth Military Medical University. The tumor tissues consisted of 25 well-differentiated, 23 moderately differentiated, and 7 poorly differentiated colon cancer specimens, All the tissues were fixed in 100mL/L buffered formalin phosphate and embedded in paraffin wax.

5’ RACE

A human EST named AGENCOURT_8343738 was initially identified with APMCF1 cDNA sequence(AF141882) by running a BLASTN search against the public database of ESTs (dbEST). It shared 98% identification with APMCF1 but had more than 325 nucleotide acids at the 5’ end. Then we designed a reverse transcription primer against this sequence. The total RNA extracted from the apoptotic MCF-7 cells with TRIzol reagent according to the manufacturer’s instructions was used as reverse transcription template. While transcription, 5’ cap sequence was added according to the protocol of SMART^TM PCR cDNA synthesis kit (Clontech). RT-PCR was performed using the cap sequence as a sense primer, reverse transcription primer as a antisense primer. The sequences of the primers were as follows: sense primer (cap sequence), 5’TACGGCTGCGAGAAGACGACAGAA3’, and antisense primer, 5’CAGGCAATTTTAGCCAGCCATTT3’ (reverse transcription primer). A total of 30 cycles of PCR reactions were performed: at 95 °C for 15 s, at 68 °C for 5 min. Then 5 μL PCR products was extracted as a template for the next 30 cycles with the same PCR parameters. The resulting PCR fragment (810 bp) was subcloned into the pEGM-T easy vector (Promega) for sequencing.

Bioinformatics

The fragment of 810 bp was used as a bait to search the Genbank database and a homologous sequence FLJ14619 was found. Then all these ESTs were retrieved and a coontig was assembled from overlapping ESTs. A full length cDNA sequence of APMCF1 was obtained and identified in human genomic database. The putative cleavage site of the signal peptide was predicted by using Signal P server (http://www. cbs.dtu.dk/services/Signal P). The gene localization of APMCF1 on the chromosome was analyzed with the HTGS database (http://www.ncbi.nlm.nih.gov).

Construction of recombinant pGEX-APMCF1

The entire APMCF1 coding region was amplified by PCR by using upstream and downstream primers, which introduce a BamHI and Sal I site respectively according to the conjunct sequence. The sequences of the primers were as follows: sense primer, 5’AGGATCCTCATCCATGGCTTCCG3’, and antisense primer, 5’ACGCGTCGACCTGCCTCTCAGGCAAT3’. The PCR product was cloned into the pGEM-T easy vector for sequencing. Then the fragment containing the ORF of APMCF1 was released from the recombinant plasmid by cutting with BamHI and Sal I, and subcloned into the prokaryotic expression vector pGEX-KG. The recombinant plasmid was termed pGEX-APMCF1.

Preparation of polyclonal antibody to APMCF1

The GST-APMCF1 fusion protein was expressed in E. coli DH5α and purified on glutathione-Sepharose 4B according to the manufacturer’s instructions. After the purified protein was separated by 100 g/L sodium dodecylsulfate-poluacrylamide gel electrophoresis (SDS-PAGE), it was then transferred from the gel to nitrocellulose. The specific strip corresponding to the antigenic protein was excised and cut into very small pieces for immunizing rabbits. The specificity of the anti-APMCF1 antibodies was identified by Western blot.

Immunohistochemical staining

All the colon cancer specimens were fixed in 100 mL/L buffered formalin and embedded in paraffin. ABC technique was adopted for immunohistochemical staining. Briefly, deparaffinized sections were treated with 30 mL/L H₂O₂ in methanol for 30 minutes to abolish the endogenous peroxidase activity. Sections were further blocked with 10 mL/L normal goat serum for 0.5 hour, followed by incubation with polyclonal rabbit anti-APMCF1 serum (1:100) at 4 °C overnight. The sections were then washed in phosphate buffer solution (PBS) (0.01 mol/L, pH 7.2) and sequentially incubated with biotinylated goat anti-rabbit IgG followed by avidin-biotin horseradish peroxidase complex following the manufacturer’s instructions (ABC kit, Sigma. USA). Staining was developed by immersing slides in 0.5 g/L diaminobenzidine tetrahydrochloride (DAB) with 3.3 g/L hydrogen peroxide. All slides were counterstained with haematoxylin, dehydrated and mounted. PBS substituting for the primary antibody was used as the negative control.

The intensity of staining for APMCF1 in colon cancer tissues and normal colonic mucosa tissues was scored for each specimen on a scale of 0 to 3, in which 0 = negative staining, 1 = weakly positive staining, 2 = moderately positive staining, and 3 = strongly positive staining. The percentage of positive cells in each specimen was estimated and scored on a scale of 0 to 4, in which 0 = negative, 1 = positive staining in 1% to 25% of cells counted, 2 in 26% to 50%, 3 in 51% to 75%, and 4 in 76% to 100%. Each section was evaluated for the sum of these two parameters. Group means of expression scores of APMCF1 antigens were compared by using ANOVA followed by Fisher’s protected t test. P < 0.05 was considered statistically significant.

RESULTS

Identification of human APMCF1

With 5’ RACE and assembled homologous ESTs, the 5’ end of original APMCF1 cDNA sequence was extended successfully. The length of new sequence was 1745 bp. It was mapped to chromosome 3q22.2 and spanned at least 14.8 kb of genomic DNA with seven exons and six introns contained. Sequence analysis revealed that APMCF1 encoded a protein of 271 aa, with an ATG initiation codon at nt 17-19. The deduced APMCF1 protein contained no putative N-glycosylation site, and no typical signal cleavage site was revealed by searching on the signal P server. There was a transmembrane helice from 37 aa to 54 aa predicted by the method of TMpred. It contained a small GTP-binding protein (G proteins) domain and it was homologous to mouse signal recognition particle receptor β(SRβ) both in nucleic acid and amino acid sequence by 86% and 89% respectively. Further analysis showed that APMCF1 protein was homologous to gene products of SRβ from species ranging from mice to yeast. This indicated that APMCF1 protein was a human SRβ gene and also was a well conserved protein (Figure 1).

Figure 1 Nucleotide and Predicted Amino Acid Sequence of the APMCF1 cDNA and sequence alignment of APMCF1 wth the protein from other species. A: The full-length gene sequence and amino acids sequence of APMCF1; B: Alignment was done by using the Align software in Vector NTI. Dark gray background shows identity with at lease five aligned residues in different species. The gray region represents the sequence conserved among different species.

Identification of polyclonal antibody to APMCF1

To detect APMCF1 protein, we generated anti-human APMCF1 antibody by immunizing rabbit with the purified GST-APMCF1 fusion protein. On immunoblots, this anti-APMCF1 antibody specifically recognized GST-APMCF1 fusion protein, the dominant band was the expected 56-ku molecular size. No band was evident in the blot of the same samples using a preimmune rabbit serum. When the antibody was preabsorbed with GST-APMCF1 fusion protein, the corresponding band was no longer observed, confirming the antibody specificity (Figure 2).

Figure 2 Immunoblotting of APMCF1 in GST-APMCF1 (56 ku) using antiserum, normal rabbit serum before immunization and antiserum preabsorbed with GST-APMCF1. M: Protein marker; 1: antiserum; 2: rabbit serum before immune from GST-APMCF1; 3: antiserum absorbed in the peptides used for im-munization (control antibody).

Expression of APMCF1 in colon cancer

Fifty five cases of colon cancer tissues and 5 cases of normal colonic mucosa were examined for expression of APMCF1 by using immunohistochemistry. Specific cytoplasmic staining for APMCF1 was observed in 44 out of 55 cases of carcinomas (80%) and 4 out of 5 normal mucosa. All of the positive samples had cytoplasmic staining, suggesting that APMCF1 might be a cytoplasmic protein. The positive ratios of APMCF1 in well-differentiated, moderately differentiated, and poorly differentiated carcinomas were 80%, 74% and 86% respectively. A significantly high expression level of APMCF1 was found in poorly and moderately differentiated colon cancer, compared with that of normal tissue (Figures 3). There were also differences between different histological grades of colon cancer, but these differences did not reach statistical significance (Table 1).

Table 1 Expression of APMCF1 in colon cancer and normal colon.

Histological type	n	Expression of APMCF1	Positive rate (%)
A) Well differentiated	25	4.20 ± 1.02	80
B) Moderately differentiated	23	4.78 ± 1.10	74
C) Poorly differentiated	7	5.18 ± 1.89	86
D) Normal	5	3.84 ± 1.73	80

Expression score (mean ± SE) in different histological grades of colon cancer and normal colon tissue. Statistically significant differences: B versus D, P < 0.05; C versus D, P < 0.01.

Figure 3 Expression of AMPCF1 in normal colon tissue and colon cancer ( × 200). A. Normal colon tissue; B. Colon cancer.

DISCUSSION

We have successfully extended the 5’ end of APMCF1. The new sequence contains an open-reading frame of 816 bp that encodes a protein with a predicted molecular mass of 30ku. By searching the GenBank, it was found that the amino acid sequence of APMCF1 was homologous to mouse signal recognition particle receptor β(SRβ) and conserved in several other species, including Caenorhabditis elegans, Saccharomyces cerevisia and Drosophila melanogaster. This indicates that APMCF1 is likely a human SRβ gene and also a well conserved protein. The conservation of APMCF1 implies that it may have a very important function, since it remains stable during the evolution of species[26,27].

The mammalian SRP receptor, also known as docking protein, is composed of two subunits, SRα and SRβ. Based on the study results in mouse and canine, both of them are required to target nascent secretory and membrane proteins to the membrane of endoplasmic reticulum (ER). SRβ anchors alpha subunit to the ER membrane during this course[26,28-30]. The result of conserved domain searching showed that APMCF1 containd a GTPase domain closely related to ADP-ribosylation factor family (ARF) and Sar1p-like members of the Ras-family of small GTPases, suggesting it belongs to the small GTP-binding protein (G proteins) superfamily[31-36].

Over 100 members of the Ras superfamily of GTPases have been found so far. Based on both their sequence homology and function, they have been subdivided into at least six families: Ras, Rho, Rab, Arf, Ran, and Rad/Gem[37-42]. Each family, in turn, is comprised of several members with distinct expression, cellular localization and biological activities. Among these members, both Ras and Rho GTPases can mediate key cellular processes in response to diverse stimuli, such as cell growth, apoptosis, lipid metabolism, cytoarchitecture, membrane trafficking, and transcriptional regulation. The Rab and Sar1/Arf families regulate vesicle trafficking, and the Ran family regulates nucleocytoplasmic transport and microtubule organization. However, the negative aspect of these multifunctional proteins arises in the context of scenarios that cause their constitutive activation (i.e. point mutations or overexpression) and render them insensitive to regulatory signals. In this case, these GTPases trigger specific signals that lead to uncontrolled cell growth, enhanced angiogenesis, inhibition of apoptosis, and genetic instability, all of which result in tumor development[43-48].

From the analysis of bioinfomatics we can see that APMCF1 might be a new member of small-G protein superfamily. It may have the potential function to regulate the survival, proliferation, differentiation and apoptosis of cells as well as other small-G protein members such as Ras and Rho. Preliminary studies also showed that transcripts of APMCF1 were up-regulated in apoptotic MCF-7 cells, indicating it might play a role in apoptosis of tumor cells. So it is necessary to know expression pattern of APMCF1 in various tissues, especially in tumors. In the present study, APMCF1 was expressed highly in poorly and moderately differentiated colon cancers compared with that in normal colon mucosa, perhaps indicating its possible biological function in tumor development. It is early to say there are some relationship between the expression of APMCF1 and tumorigenesis before further study has been done. But it will be very interesting and worthy to further study these aspects in the future.