HeLa (human cervical cancer cells) and HEK293 (human embryonic kidney cells) cell lines were obtained from Department of Infection Biology, Faculty of Medicine, Tsukuba University. The cells were maintained in Dulbecco’s modified eagle medium (DMEM); supplemented with 10% fetal bovine serum (FCS), 2 mM L-glutamine, 100 U/mL penicillin, and 100 μg/mL streptomycin; and cultured under the standard culture conditions (37°C, 5% CO₂, 95% relative humidity).

The Saccharomyces cerevisiae strain PJ69-4A was used to detect the proteins interacting with the influenza A virus PA protein by the yeast two-hybrid assay [Genotype: MATa trp1-901 leu2-3, 112 ura3-52 his3-200 gal4 (deletion) gal80 (deletion) LYS2::GAL1-HIS3 GAL2-ADE2 met2::GAL7-lacZ]. The cells were grown in a YPAD (yeast extract- peptone-dextrose plus adenine) rich medium at 30°C at 175 rpm shaking and stored at −80°C in a YPD (yeast extract-peptone-dextrose) medium containing 20% glycerin.

Gene cloning and amplification of plasmid DNAs were performed with the Escherichia coli Mach1 (Invitrogen) strain. The cells were cultured in Luria-Bertani (LB) broth and Luria-Bertani Agar (LBA) media.

The cells were infected with human influenza A/WSN/33-H1N1 (WSN) viruses. The viruses were grown in embryonated chicken eggs and/or MDCK cells. Virus titration was determined by a plaque forming assay (Turan et al., 1996).

The expression vectors encoding HA (YPYDVPDYA) tagged β-actin (hACTB) and β-actin variants with deleted regions corresponding to specific exons were generated by cloning the cDNA into the pCHA plasmid (Nagata et al., 1998). The plasmid DNA was digested with the EcoRV restriction enzyme (New England Biolabs, United States) and dephosphorylated with shrimp alkaline phosphatase (SAP) (Thermo Scientific, United States). β-actin open reading frame (ORF) was amplified with PCR over the cDNA library of HEK293 by using 5′-ATC​ATG​GAT​GAT​GAT​ATC​GCC​GC-3′ and 5′-ATC​TAG​AAG​CAT​TTG​CGG​TGG​AC-3′ primer pairs phosphorylated with T4 polynucleotide kinase (New England Biolabs, United States). PCR was performed as: 1 cycle “95°C, 3 min” initial denaturation; 30 cycles “95°C, 15 s; 56°C, 15 s; 68°C, 1 min”; and 68°C, 5 min final extension. The reaction was carried out with the high fidelity KOD DNA polymerase enzyme (Toyobo, Japan) which produces a blunt-end product. The PCR products were purified from agarose gel with a DNA agarose gel extraction kit (Qiagen, Germany) and ligated with the pCHA plasmid DNA in the frame of the HA-tag coding sequence by using T4 DNA ligase (Takara, Japan), and the resultant plasmid was designated as pCHA-ACTB. Plasmid vectors having a deletion of certain exons in the β-actin gene were derived from the pCHA-ACTB by inverse PCR. The primers were phosphorylated with T4-polynucleotide kinase. For the gene lacking the second and third exons (ACTBΔ2.3), ACTΔ2.3-For/Vec-Rev primers; for the gene lacking the third exon (ACTBΔ3), ACTΔ3-For/ACTΔ3-Rev primers; for the gene lacking the fifth and sixth exons (ACTBΔ5.6), Vec-For/ACTΔ5.6-Rev primers; and for the gene lacking the sixth exon (ACTBΔ6), Vec-For/ACTΔ6-Rev primers (Table 1) were used. The PCR was carried out with KOD DNA polymerase and conducted as: 1 cycle “95°C, 3 min” initial denaturation; 30 cycles “95°C, 20 s; 54°C, 15 s; 68°C, 5 min”; and 68°C, 10 min final extension. The PCR products were purified with a DNA agarose gel extraction kit and self-ligated with T4 DNA ligase. Ligated DNAs were transformed into competent E. coli cells, and the cells were grown on LBA (+amp) medium. Plasmid DNAs isolated from the colonies were checked by cutting with restriction endonuclease enzymes.

Yeast two-hybrid plasmid coding GAL4 binding domain (GAL4-BD)—viral PA fusion (BD-cPA) was constructed by cloning a 3′-half of the influenza A virus PA gene consisting of 1,074 nucleotides (cPA) into pGBD-C1 plasmid DNA (James et al., 1996). For this, the plasmid DNA was digested with EcoRI just after the GAL4-BD coding region, blunted with the Klenow enzyme, and then dephosphorylated with SAP. The cPA ORF carrying the linker sequence (ggaggatctgga/Gly-Gly-Ser-Gly) at 5′-end was amplified with PCR by using the pCAGGS-PA plasmid DNA (Pham et al., 2018) as a template and phosphorylated cPA-For/PA-Rev primers (Table 1) and purified with a DNA gel extraction kit. A pGBD-cPA bait plasmid was obtained with ligation of the cPA DNA fragment and linearized pGBD-C1 plasmid.

A pCAGGS-GSTnls plasmid (Sugiyama et al., 2015) was used to obtain the plasmid vectors encoding GST-tagged β-actin (GST.ACTB) and the carboxy-terminal half of the influenza PA (GST.cPA) fusion proteins. The pCAGGS-GSTnls was linearized by inverse PCR using the pCA-Vec-For/GST (GGSG)-Rev primers (Table 1) and ligated with PCR-amplified β-actin or cPA cDNA. The resultant plasmids were named pCAGGS-GST.ACTB and pCAGGS-GST.cPA.

The plasmid vectors used in the study and the proteins encoded by these vectors are given in Table 2. The nucleotide sequences of all genes cloned into the plasmids were confirmed by the Sanger DNA sequencing method.

The cDNA library constructed by cloning human embryonic kidney cells (HEK293) cDNA immediately after the sequence of yeast GAL4 activation domain (GAL4-AD) in pACT2 plasmid (pACT2-cDNAn) was purchased as transformed in E. coli cells (Clontech #638826). The cells were grown on LB agar (+amp) plates according to the manufacturer’s instructions, and the plasmid DNA was isolated.

The yeast two-hybrid (Y2H) assay was performed as previously described (Senbas-Akyazi et al., 2020; Pirincal and Turan, 2021), taking into account the manufacturer’s instructions. Briefly, a small-scale yeast culture (5 mL) was prepared in the YPAD medium and transformed with the pGBD-cPA plasmid DNA coding the bait protein (GAl4-BD.cPA) by the lithium acetate/polyethylene glycol (LiAc/PEG) method (Gietz and Schiestl, 2007). Transformants were selected on a yeast synthetic drop-out (SD) agar medium (without Trp). One of the colonies harboring the bait plasmid was grown in the YPAD medium, transformed with the Y2H Matchmaker cDNA library, and seeded on SD agar plates (without Her, Leu, Ade, and Trp) for selection of positive colonies. The relative β-galactosidase activities of the yeast colonies grown on the selective media were measured by using an o-nitrophenyl-β-D-galactopyranoside (ONPG) substrate (Senbas-Akyazi et al., 2020). The plasmids carrying cDNA were isolated from the colonies having high β-galactosidase activity with a yeast plasmid DNA extraction kit (Bio Basic, Canada). The cDNAs carried on the plasmids were amplified with PCR using the oligonucleotide primers 5′-AAT​ACC​ACT​ACA​ATG​GAT​GAT​GT-3′ and 5′-CCA​AGA​TTG​AAA​CTT​AGA​GGA​GT-3’. The PCR products were purified with a gel extraction kit, and the sequence of the cDNAs was defined with Sanger sequencing. The cDNAs were identified with the basic local alignment search tool (BLAST).

HEK293 and HeLa cells were transiently transfected with the plasmid DNA using polyethyleneimine (PEI) (Longo et al., 2013). The cells were seeded in a DMEM complete medium in multi-well plates (1 × 10⁵ HEK293 cells/well or 5 × 10⁴ HeLa cells/well in 24-well plate; 2 × 10⁵ HEK293 cells/well or 1 × 10⁵ HeLa cells/well in 12-well plate) or a 6 cm Petri dish (1 × 10⁶ HEK293 cells) and incubated overnight under standard culture conditions. An average of 0.25–0.5 µg/well (in 24-well plate), 0.5–1 µg/well (in 12-well plate), or 2.5–5 µg of plasmid DNA (6 cm Petri dish) were used for transfection of the cells. Plasmid DNAs were diluted to 15 ng/μL in OPTİ-MEM (Thermo Scientific, Gibbco, United States). The 20 μg/μL PEI solution in deionized distilled water was also diluted in OPTİ-MEM to a 40–50 ng/μL final concentration. Diluted DNA and PEI solutions were mixed in equal volumes, incubated at room temperature for 5–10 min, and added to the cell cultures. The cultures were incubated for 10 h under standard culture conditions, and then the media were replaced with fresh media. The cells were harvested 45–48 h after transfection.

The proteins were expressed in transfected and/or infected cells or precipitated with glutathione-Sepharose beads (GE-Healthcare, CL-4B) and protein A-Sepharose (GE-Healthcare, CL-4B) and were analyzed by immunoblotting. The proteins were separated with sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to PVDF membranes. The membranes were blocked in 5% skim milk in tris-buffered saline (TBS) (137 mM NaCl, 20 mM Tris, pH: 7.4) with 0.1% Tween 20 (TBS-T) at room temperature for 60 min and subsequently treated with mouse anti-HA (Santa Cruz, sc-7392) (for HA-tagged β-actin proteins), mouse anti-actin (Abcam, ab11003) (for endogenous β-actin), rabbit anti-PA, or M1 rabbit antibodies [kindly provided by Dr. K. Nagata (Kawaguchi et al., 2005; Takizawa et al., 2006)] in skim milk at 4°C overnight. After washing thoroughly in TBS-T, the membranes were incubated with the goat anti-mouse IgG-HRP antibody (Invitrogen/Thermo: 31420) or anti-rabbit IgG-HRP antibody (Invitrogen/Thermo: 31423) diluted at a 1:5,000 in TBS for 30 min at room temperature. The membranes were washed with TBS-T/TBS and visualized with a gel imaging system (DNR Bio-Imaging system) using chemiluminescence substrate (H&C, RPN2232).

The expression and intracellular localization of the proteins in transiently transfected HeLa cells were defined with an immunofluorescence assay. The cells were grown in 12-well plates on glass coverslips for 24 h and transfected with plasmid DNAs using PEI. After 45–48 h of transfection, the cells were fixed in 3% paraformaldehyde for 10 min, rinsed three times with phosphate buffered saline (PBS), and permeabilized with 0.1% NP-40. The cells were blocked in 1% skim milk in PBS for 30 min and treated with primary antibodies for 60 min at 1/250–1/500 dilution in the milk. Filamentous actin proteins in transiently transfected cells were labeled with Alexa 568-Phalloidin (Invitrogen, Thermo Fisher Scientific # A12380) at 1/200 dilution. After rinsing three times with PBS, the cells were blocked a second time in the milk for 20 min and treated with anti-mouse Alexa 488 (Abcam: ab150117) and/or anti-rabbit Alexa 568 secondary antibodies (Abcam: ab175471) at 1/300 dilution for 60 min. For visualization of the cell nuclei, DAPI (4′, 6-diamidino-2-phenylindole) was added to the antibody solutions at a concentration of 0.5 μg/mL. The cells on the glass coverslip were rinsed three times with PBS and placed upside-down on 3–5 μL of antifade mounting media (1 mg/mL p-phenylenediamine in 90% glycerol and 10% 20 mM Tris, pH 8.8) and sealed with nail polish. The labeled proteins were analyzed with a conventional fluorescent microscope equipped with a digital camera (Olympus, DP72) or a laser confocal microscope (Carl Zeiss LSM 700).

HEK293 cells were seeded in 6 cm Petri dishes (1 × 10⁶/Petri) and incubated for 20–24 h under the standard culture conditions. Then, the cells were co-transfected with pCHA-ACTB (3 µg) and pCAGGS-GST.cPA (2 µg) or pCAGGS-GST.ACTB (3 µg) and pCAGGS-cPA (2 µg) using PEI. After 48 h of transfection, the cells were rinsed with PBS and treated with 1 mM dithiobis (succinimidyl propionate (DSP) for cross-linking at 4°C for 60 min. The cells were then collected in 500 µL buffer A (150 mM NaCl, 50 mM Tris, pH 7.5, 1 mM EDTA, 0.1% NP-40, 1 mM PMSF) using a cell scraper and lysed with 10 cycles of freezing/thawing (liquid nitrogen/water bath at 37°C). After passing through a 26-gauge needle attached to a 1 mL syringe (10 times), the lysates were centrifuged at 15,000 rpm for 5 min at 4°C. The supernatants were used for protein precipitation with glutathione-Sepharose beads. For this, 200 μL of cell lysate, 300 μL of wash-binding buffer (150 mM NaCl, 50 mM Tris, 1 mM DTT, pH 8), and 50 μL of 50% slurry of glutathione-Sepharose beads were mixed and rotated with a vertical tube mixer for 4 h at 4°C. The glutathione-Sepharose beads were precipitated by centrifugation at 3,000 rpm for 2 min, and washed three times in 500 μL wash-binding buffer by rotating with a vertical tube mixer for 5 min. After the last wash, the precipitates were suspended in 50 μL of SDS-PAGE loading buffer, and the precipitated proteins were analyzed with SDS-PAGE/Western blotting.

As with the GST pull-down assay, the HEK293 cells were grown in 6 cm Petri dishes and co-transfected with plasmids coding viral cPA and HA-tagged β-actin (hACTB) proteins. After 48 h of transfection, the cells were cross-linked with DSP and harvested in 500 µL Buffer A as described above. 200 μL of cell lysate and 300 μL of wash-binding buffer (150 mM NaCl, 50 mM Tris, pH 8, 1 mM EDTA) were mixed in a 1.5 mL tube. After adding 4 μL of mouse monoclonal anti-HA antibody, the samples were mixed with a vertical mixer at 4°C for 2 h. Then, a 50 μL 50% slurry of protein A-Sepharose was added and mixed at 4°C overnight. After incubation, the protein A-Sepharose was washed three times with wash-binding buffer as described above and suspended in 50 µL of SDS-PAGE loading buffer. Precipitated proteins were analyzed with SDS-PAGE/Western blotting.

The HEK293 cells were seeded in 12-well plates (2 × 10⁵ cells/well) and transfected with 2 µg of plasmid DNA coding hACTB or deleted hACTB proteins. After 36 h of transfection, the medium was removed, and the cells were washed twice with OPTI-MEM (Gibco, United States). The cells were inoculated with A/WSN/33-H1N1 viruses diluted in 1% bovine serum albumin fraction V (BSA) at a 1–2 moi for 30 min. Then, the inoculum was removed, and the cells were incubated in a maintenance medium (DMEM containing 0.2% BSA fraction V and 4 μg/mL TPCK-trypsin) for 8 h. After incubation, the cells were lysed in a 200 µL SDS-PAGE loading buffer. Endogenous β-actin, viral proteins (PA and M1), and hACTB proteins encoded from the plasmids were analyzed with SDS-PAGE/Western blotting using mouse anti-actin, mouse anti-HA, rabbit anti-PA antibodies, and M1 rabbit antiserum.

The effects of over-expression of hACTB variants on the influenza A virus RdRp enzyme were investigated with the influenza A virus mini-replicon system as described above (Turan et al., 2004; Pham et al., 2018). The HEK293 cells were seeded in 24-well plates (5 × 10⁴ cells/well) and co-transfected with a set of mini-replicon plasmids (pHH21-vNS-luc, pCAGGS-PB1, pCAGGS-PB2, pCAGGS-PA, and pCAGGS-NP) and 200 ng of a plasmid encoding the β-actin proteins. The plasmid DNA coding SEAP enzyme under the regulation of the Pol-II polymerase promoter was used as an internal control. After 48 h of transfection, the cells were lysed and luciferase activities in the lysates were defined with a commercial kit (Promega #E1483, Madison, United States) according to the manufacturer’s instructions.

In silico methods were used to predict possible interaction points between human β-actin and the influenza A virus PA protein. For this, homologous protein models were created. Based on the possible protein models, the interaction sites were predicted and analyzed by the protein-protein docking and the molecular visualization software. Three-dimensional structural models of the deleted β-actin and PA proteins were predicted by using the I-TASSER algorithm ¹ (Zhang, 2008). The best-predicted models were determined by using the C-score [−5 to +2], TM-score (>0.5), and root-mean-square deviation (RMSD). The putative complexes of β-actin and PA proteins were created using the ClusPro protein docking algorithm ² (Kozakov et al., 2017). In these complexes, the models with the highest number of cluster members and the lowest free energy were selected, and possible interaction sites and amino acids were determined with the PyMOL molecular visualization software.

Several host proteins that interact with influenza virus PA were identified using the Y2H screening assay carried out with the HEK293 Matchmaker cDNA library and the pGBD-cPA bait plasmid that expressed the GAL4-BD-cPA bait protein. The first step selection of candidate yeast colonies was carried out based on the histidine marker gene activated by protein-protein interaction which prevented the yeast growth in the medium lacking histidine in case of no interaction. The sequences of candidate cDNAs isolated from the yeast colonies grown on synthetic dextrose (SD) agar (lacking tryptophan, leucine, histidine and adenine), were identified with BLAST analysis. Among the identified genes, HCLS1 associated protein X-1 (HAX1), chromosome 14 open reading frame 166 (C14orf166), and minichromosome maintenance complex component 7 (MCM7) have been previously reported to be associated with influenza A viruses. In addition to these genes, the yeast cells carrying the cDNAs of human β-actin, mitochondrial ribosomal protein L27 (MRPL27), coiled-coil domain-containing protein 90B (CCD90B), small ubiquitin-like modifier 2 (SUMO2), fatty acid synthase (FASN), and ADP-ribosylation factor 4 (ARF4) were found to form colonies on SD media and to have high beta-galactosidase activity. In this study, considering the possible role of the above on the intracellular trafficking of influenza A virus vRNPs and/or virus transcription, we focused on the β-actin among them. Growing of the yeast cells expressing GAL4-BD-cPA bait and GAL4-AD-cDNA (β-actin) protein in SD medium without histidine showed that the viral PA and β-actin protein interact with each other in yeasts cells (Figure 1). The S. cerevisiae strain PJ69-4A carries the β-galactosidase reporter gene as a second marker that shows protein-protein interaction in the cells. The yeast cells expressing GAL4-BD-cPA bait protein and GAL4-AD-cDNA (β-actin) protein had a high β-galactosidase activity compared to the control cells (transformed with pGBD-cPA bait plasmid alone), supporting the interaction of viral PA and β-actin protein.

The relationship between the influenza A virus PA and β-actin was evaluated in HEK293 cells, the natural host cells for influenza A viruses with co-precipitation assays. In the first step, a GST pull-down assay was carried out, as mentioned above. The hACTB and cPA proteins, transiently expressed in the HEK293 cell, were co-precipitated with GST.cPA and GST.β-actin, respectively. The results indicated a direct interaction between β-actin and viral PA protein in mammalian cells (Figure 2A). This interaction of cPA and β-actin protein was also confirmed with a co-immunoprecipitation assay. Lysates were prepared from transfected HEK293 cells with pCHA-ACTB and pCAGGS-cPA or pCAGGS-nPA (Pham et al., 2018). Immunoprecipitation assays with anti-HA antibody against hACTB revealed that viral cPA was associated with β-actin in cells (Figure 2B). In contrast, the amino-terminal half of the PA (nPA) was not detected in the immunoprecipitates under the conditions employed here, indicating that the amino terminal half of the PA was not functional in this interaction (Figure 2B).

The possible domains of the proteins involved in interaction of PA and β-actin were evaluated by immunoprecipitation assays by using hACTB variants, certain exon regions of which were removed by gene manipulations, and native PA, nPA and cPA proteins. The human β-actin gene is located on the short arm of chromosome 7 and codes a protein with ∼42 kD. This gene consists of six exons separated by introns. Considering the exon-intron junction points of the gene, four β-actin gene variants with deletions of certain exons were generated by inverse PCR from β-actin complete cDNA cloned into pCHA plasmid (pCHA-ACTB) (Figure 3A). The expressions of the HA-tagged β-actin (hACTB) and deleted β-actin proteins (hACTBΔ2, hACTBΔ2.3, hACTBΔ5.6, and hACTBΔ6) were detected with Western blotting in transiently transfected HEK293 cells (Figure 3B).

To analyze the interaction between viral PA and β-actin with the co-immunoprecipitation assay in detail, HEK293 cells were co-transfected with plasmid coding viral PA proteins (PA, nPA, or cPA) and β-actin variants. HA-tagged β-actin proteins co-expressed with viral PA proteins were precipitated with protein A-Sepharose by using a monoclonal anti-HA antibody, and both proteins in the precipitates were analyzed with Western blotting (Figure 4). It was observed that full-size PA protein co-precipitated with C-terminal deleted β-actins proteins (hACTBΔ5.6 and hACTBΔ6), but not with N-terminal deleted (hACTBΔ2.3) protein (Figure 4A). Furthermore, it was shown that the cPA protein co-precipitated in higher amounts with hACTBΔ5.6 and hACTBΔ6 proteins (Figure 4B). In contrast, the amino-terminal half of the PA (nPA) protein did not co-precipitated with the β-actin proteins (Figure 4C). This result showed that the amino-terminal domains of the β-actin protein bind to the carboxy-terminal moiety of the viral PA protein.

Intracellular localization of transiently expressed β-actin variants and/or viral PA proteins in HeLa cells were determined with an immunofluorescence assay by using mouse monoclonal anti-HA and rabbit polyclonal anti-PA antibodies, respectively. The endogenous filamentous actin proteins were labelled with Alexa 568-Phalloidin. It was determined that the hACTB protein has a radial localization in the cell cytoplasm, similar to endogenous filamentous actin, mostly in regions close to the cell membrane. hACTBΔ2 and hACTBΔ2.3, amino-terminal deleted variants, localize as aggregates homogeneously distributed in the cell cytoplasm, incompatible with the hACTB. Carboxy-terminal deleted hACTBΔ5.6 and hACTBΔ6 showed the cytoplasm localization similar to the hACTB, close to the cell membrane (Figure 5A). Similar localization patterns of HA-tagged β-actin variants are also observed in the HeLa cells whose the filamentous actin proteins are labeled with Alexa 568-phalloidin (Figure 5B).

The influenza A virus PA protein encoded by the plasmid DNA shows nucleo-cytoplasmic localization in the HeLa cells. The carboxy-terminal domain deleted β-actins (ACTBΔ5.6 and ACTBΔ6), having a similar localization pattern with hACTB when expressed alone, also show nucleo-cytoplasmic localization with the viral PA protein in the cells (Figure 5C). This co-localization patterns supported the interaction of amino terminal domains of β-actin with viral PA protein.

During viral infection, there is a complex and tight interaction between the host cell and viral proteins which may affect viral replication. To detect the possible effects of the β-actin protein on the viral replication, the HEK293 cells over-expressing the HA-tagged ACTB variants were infected with influenza A WSN viruses, and 8 h after infection, the viral proteins and both endogenous β-actin (ACTB) and transiently expressed β-actin proteins (hACTBs) were analyzed with SDS-PAGE/Western blotting (Figure 6A). The results showed that over-expression of the hACTB had a low negative effect on the viral protein level. In contrast, deleted hACTB proteins had no significant effect on the viral replication at the early stage of infection. The effects of β-actin variants on virus RNA polymerase were also investigated in transiently transfected HEK293 cells by using the influenza A virus mini-genome. The luciferase enzyme activities, which are dependent on the viral polymerase enzyme in the cell lysates, were determined. The results showed that the over-expression of deleted hACTB proteins created no significant change in reported luciferase activities. In contrast, the HA-tagged native β-actin (hACTB) protein resulted in an average 65% decrease of reporter luciferase activity in the cells compared to the control (Figure 6B).

The co-precipitation of β-actin protein with viral PA protein in GST pull down and immunoprecipitation assays is an important indicator that these two proteins interact with each other in mammalian cells as well. Although truncated proteins are not native proteins, the co-precipitation of C-terminal deleted β-actin variants with the viral PA protein suggests that the N-terminal region of the protein may be important in this interaction. The fact that many factors interfere with protein interactions in the cell makes it extremely difficult to reveal binding patterns. Here, possible interaction patterns were predicted by in silico algorithms, ignoring the intracellular factors and infection conditions. In these interaction patterns, the models with the highest values of cluster number and the lowest free energy were selected, and the possible interaction sites and amino acid residues were determined with the PyMOL molecular visualization software. Considering the co-precipitation properties of the deleted proteins and docking assessment results, the first and most probable interaction patterns for the protein complexes are given in Figure 7. In these models, the amino acid residues in N-terminal halves of β-actin, hACTBΔ5.6 and hACTBΔ6 proteins seem to be prominent. These patterns agree with the results of the co-immunoprecipitation results.