Anti-GFP and cyclin D1 antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Anti-acetyl-lysine antibody was purchased from Cell Signaling Technology (Danvers, MA, USA). Anti-tubulin was purchased from Sigma-Aldrich (St. Louis, MO, USA).

HeLa and 293T cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS) at 37°C and 5% CO₂ in a humid atmosphere.

To construct expression vectors for human ARD1 variants, ARD1 cDNA was amplified by PCR and sub-cloned into a GFP-tagged pCS2+ vector for cell expression, and a pGEX-4T vector for bacterial induction of the recombinant protein. Transfection was carried out using Lipofectamine (Life Technology, Carlsbad, CA, USA) or Polyfect (Qiagen, Valencia, CA, USA), according to the manufacturer’s instructions.

Cells were harvested and proteins were extracted using protein lysis buffer (10 mM HEPES at pH 7.9, 40 mM NaCl, 0.1 mM EDTA, 5% glycerol, 1 mM DTT and protease inhibitors). The concentration of extracted protein was measured using a BCA assay. Total cell lysates were resolved by SDS-PAGE and transferred onto nitrocellulose membranes (Amersham Pharmacia Bioscience, Piscataway, NJ, USA). The membrane was probed with a primary antibody followed by a secondary antibody conjugated to horseradish peroxidase, and protein was visualized using the ECL system (Intron Biotechnology, Gyeonggi-do, Korea).

Recombinants of GST-hARD1 variants were freshly prepared as previously described (21). ARD1 recombinants were incubated in reaction mixture (50 mM Tris-HCl at pH 8.0, 0.1 mM EDTA, 1 mM DTT, 10% glycerol and 10 mM acetyl-CoA) at 37°C for 1 h.

Cells were placed on cover slips then incubated with Hoechst 33342 (Molecular Probes, Eugene, OR, USA) for nucleus staining. Axiovert M200 microscopes (Carl Zeiss, Jena, Germany) were used for immunofluorescence imaging.

Total RNA was extracted using an RNA extraction kit (Invitrogen, Carlsbad, CA, USA). cDNA was synthesized from 2 μg of RNA using an oligo(dt) primer. Primers used for PCR reactions were as follows: human ARD1, 5′-ATGAACATCCGCAATGCGAG-3′ (forward) and 5′-CTCATATCATGGCTCGAGAGG-3′ (reverse); cyclin D1, 5′-CTGGCCATGAACTACCTGGA-3′ (forward) and 5′-GTC ACACTTGATCACTCTGG-3′ (reverse); GAPDH, 5′-ACCAC AGTCCATGCCATCAC-3′ (forward) and 5′-TCCACCACCCT GTTGCTGTA-3′ (reverse). The PCR reaction was performed for 25 cycles to allow ARD1, cyclin D1 and GAPDH amplification.

The rate of cell proliferation was measured using a Non-Radioactive Proliferation Assay kit (Promega) according to the manufacturer’s instructions. Briefly, cells were seeded into 96-well plates and cultured for three days. Subsequently, 20 μl of substrate solution was added and the cells were incubated for 1 h to allow color development. The absorbance at 492 nm was measured to determine the number of proliferating cells.

Results are presented as means ± SD and P-values were calculated by applying the two-tailed Student’s t-test to data derived from three independent experiments. Differences were considered statistically significant when P<0.05.

During the cloning of human ARD1 using mRNA prepared from HeLa cells, we observed that cDNA from one clone (no. 4) was shorter than cDNA from other clones (Fig. 1A). Sequence analysis revealed that the clone with the shorter cDNA sequence (clone no. 4) was hARD1¹³¹ (GenBank accession no. BC063377), whereas the other clones were all hARD1²³⁵ (GenBank accession no. NM_003491). As shown in Fig. 1B, the nucleotide sequence from residue 342 to 387, which is located in exon 6 and exon 7 of ARD1 mRNA, was deleted in hARD1¹³¹. Therefore, the nucleotide sequence of hARD1¹³¹ is 46 bp shorter than that of hARD1²³⁵.

To confirm the expression of hARD1¹³¹ in human cells, we performed RT-PCR using mRNA isolated from 293T and HeLa cells. As described in Table I and Fig. 1B, three kinds of primer sets named P1, P2 and P3 were designed for the PCR analysis: P1 and P2 contained the deleted mRNA region, whereas P3 did not. As predicted, when P1 and P2 were used in PCR, two bands corresponding to hARD1¹³¹ and hARD1²³⁵ were detected by polyacrylamide gel electrophoresis. However, only one band, corresponding to hARD1²³⁵, was detected following PCR with the P3 primer set. These data confirm the expression of the hARD1¹³¹ splice variant in human cells (Fig. 1C).

Next, we compared the coding sequences of hARD1¹³¹ and hARD1²³⁵. The coding sequence of hARD1²³⁵ terminates at amino acid 708, thus it encodes a 235 amino acid proteins. However, in hARD1¹³¹, deletion of 46 bp in hARD1¹³¹ results in a frame shift after amino acid 113, resulting in premature termination at amino acid 131 (Fig. 2A and B). ARD1 is predicted to have an acetyltransferase domain located between amino acid residues 45 and 130, in which an acetyl-CoA binding domain is positioned between amino acid residues 82 and 87. Compared with hARD1²³⁵, hARD1¹³¹ protein possesses a conserved acetyl-CoA binding domain. However, 20% of the acetyltransferase domain was deleted in hARD¹³¹ and it was found to have a different C-terminal region (Fig. 2B).

In a previous study, we reported that splice variants of mouse ARD1 have a different subcellular localization (22). Therefore, we speculated that hARD1¹³¹ might have a distinct localization compared to hARD1²³⁵. To investigate ARD1 location in human cells, we constructed GFP-tagged plasmids containing hARD1¹³¹ and hARD1²³⁵, which were subsequently transfected into HeLa and 293T cells. The localization of GFP-hARD1 variants and control GFP protein was analyzed by fluorescence microscopy. As shown in Fig. 3, GFP-hARD1²³⁵ was observed predominantly in the cytoplasm, despite hARD1 containing a putative nuclear localization sequence (NLS) (Fig. 2B). however, hARD1¹³¹ was specifically located in the cell nucleus. These results demonstrate that hARD1 variants have distinct subcellular localizations, and suggest that the roles of hARD1¹³¹ and hARD1²³⁵ might differ within the cell.

Several studies have reported that hARD1²³⁵ regulates the cell cycle and stimulates cell proliferation (3,21). Thus, we aimed to determine whether hARD1¹³¹ could promote cellular growth in a similar way to hARD1²³⁵. Consistent with previous studies, compared to control cells, cell proliferation was significantly accelerated in hARD1²³⁵-expressing HeLa cells. However, hARD1¹³¹ had no effect on cell growth, suggesting that hARD1 variants have different roles in the regulation of cell proliferation (Fig. 4A and B). In a previous study, cyclin D1 was found to mediate ARD1-induced cell growth, therefore we examined the expression levels of cyclin D1 mRNA and protein in hARD1²³⁵ and hARD1²³⁵ transfected cells (3,21). Consistent with enhanced cellular growth, cyclin D1 mRNA and protein expression were significantly increased in hARD1²³⁵, but not hARD1¹³¹ transfected cells (Fig. 4C and D). These results suggest that the diverse roles of hARD1 variants in the regulation of cell proliferation may be due to their different effects on cyclin D1 expression.

Previously, we showed that the autoacetylation activity of hARD1²³⁵ was required for enhanced cell proliferation (21). Thus, we compared the autoacetylation activities of hARD1¹³¹ and hARD1²³⁵ using an in vitro acetylation assay. While the hARD1²³⁵ recombinant acetylated itself, hARD1¹³¹ was not acetylated in vitro, indicating a lack of autoacetylation activity in this splice variant (Fig. 4E). These results suggest that, unlike hARD1²³⁵, hARD1¹³¹ has no autoacetylation activity, and is therefore unable to upregulate cyclin D1 expression and promote cell proliferation.