Dynamic changes in protein interaction between AKAP95 and Cx43 during cell cycle progression of A549 cells

Abstract

Here we show that A-kinase anchoring protein 95 (AKAP95) and connexin 43 (Cx43) dynamically interact during cell cycle progression of lung cancer A549 cells. Interaction between AKAP95 and Cx43 at different cell cycle phases was examined by tandem mass spectrometry(MS/MS), confocal immunofluorescence microscopy, Western blot, and co-immunoprecipitation(Co-IP). Over the course of a complete cell cycle, interaction between AKAP95 and Cx43 occurred in two stages: binding stage from late G1 to metaphase, and separating stage from anaphase to late G1. The binding stage was further subdivided into complex binding to DNA in interphase and complex separating from DNA in metaphase. In late G1, Cx43 translocated to the nucleus via AKAP95; in anaphase, Cx43 separated from AKAP95 and aggregated between two daughter nuclei. In telophase, Cx43 aggregated at the membrane of the cleavage furrow. After mitosis, Cx43 was absent from the furrow membrane and was located in the cytoplasm. Binding between AKAP95 and Cx43 was reduced by N-(2-[P-Bromocinnamylamino]-ethyl)-5-isoquinolinesulfonmide (H89) treatment and enhanced by Forskolin. dynamic interaction between AKAP95 and Cx43 varies with cell cycle progression to regulate multiple biological processes.

Figure 1. Proteins interacting with Cx43 in A549 cells detected by MS/MS analysis.

Figure 2. Nuclear morphology of different cell cycle phases.

Figure 3

Co-localization of Cx43, AKAP95 and DNA in A549 cells (A) Co-localization of Cx43 and AKAP95 in the nucleus. (Aa) Blue fluorescence represents DAPI-labeled nuclei; (Ab) green fluorescence represents FITC-labeled AKAP95; (Ac) red fluorescence represents TXRD-labeled Cx43; (Ad) merged image of Aa and Ab; (Ae) merged image of Aa and Ac; (Af) merged image of Ab and Ac; and (Ag) merged image of Aa, Ab, and Ac; (B) Rows 1–5 are the enlarged images of cells 1–5 in Af. (magnification 200 × for A; 200 × 4 for B). (C) Proximity ligation assay and immunofluorescence detected by Laser scanning confocal microscope (200 × ). Labeling with DAPI is shown in (Ca), and lamin B1 is shown in (Cb). In (Cc), the combination of AKAP95 and Cx43 was labeled with TXRD. Results showed interaction between AKAP95 and Cx43 (red). Both proteins were localized in the nucleus. Panel (Cd) is a merged image of (Ca) and (Cb). Panel (Ce) is a merged image of (Cb) and (Cc). Panel (Cf) is a merged image of (Ca) and (Cc). In panel (Cg) (merged image of Ca, Cb and Cc), cells highlighted as I and II point to the combination of AKAP95 and Cx43 crossing the nuclear membrane. Cells II are enlarged in (Ch) (200 × 4). At the lower right corner (200 × 4 × 4), the junction between complex and membrane was demonstrated by yellow fluorescence(arrow), suggesting movement into the nucleus via the nuclear membrane. (D) is the negative control to C. (Di) is DAPI, (Dj) is Lamin B1, (Dk) is Cx43 primary antibody replace by PBS as a negative control in PLA assays; red fluorescence was not detected. Panel (Dl) is a merged image of (Di), (Dj) and (Dk). (E) We detected the quantitative changes of AKAP95/Cx43 complexes in G1, S, G2, M phases using Olympus software in PLA assays. The number of the detected cells was 37 in G1 phase, 30 in S phase, 13 in G2 phase and 10 in M phase. In addition, there were statistically significant differences of the amount of AKAP95/Cx43 complexes between G1 and S phase, G1 and G2 phase respectively.

Figure 4. Co-localization of AKAP95 and Cx43 in the nuclear membrane.

Using nuclear-cytosol extraction kits according to the manufacturer’s instructions, cytoplasmic and nuclear fractions were separated by using two of protocols: 1. Adding CEB-B solution to produce the Nucleus1 and Cytoplasm1 fractions containing nuclear membrane proteins; 2. Without adding CEB-B solution to produce the Nucleus2 and Cytoplasm2 fractions without nuclear membrane proteins. Samples (300 μg) of each group were subjected to Co-IP using an anti-Cx43 antibody (3 μg). Immunoprecipitated lysates (100 μg) were then analyzed by Western blotting using anti-AKAP95 and anti-Cx43 antibodies. The experiment was repeated three times.

Figure 5. Expression, localization, and co-localization of AKAP95 and Cx43 in A549 cells during different cell cycle phases.

Column 1 (A1-I1): DAPI-labeled nuclei with morphological changes during a complete cell cycle from G1 to telophase; column 2 (A2-I2): expression and localization of TRITC-labeled AKAP95 in each cell cycle phase; column 3 (A3-I3): expression and localization of FITC-labeled Cx43 in each cell cycle phase; column 4 (A4-I4): merged images of columns 1 and 2; column 5: merged images of columns 2 and 3; and column 6: merged images of columns 1, 2, and 3 (200 × 3). The experiment was repeated four times.

Figure 6. Interaction between Cx43 and AKAP95 in A549 and L-O2 cells at different cell cycle phases.

A549 and L-O2 cells were treated with L-mimosine (G1) (75 and 80 μg/mL), aphidicolin (S) (0.8 and 1 μg/mL), nocodazole (G2) (0.6 and 1 μg/mL), and colchicine (M) (0.5 and 0.5 μg/mL) for 24 h. A portion of the cells were subjected to total protein extraction and analysis. (A) Changes in AKAP95 and Cx43 expression determined by Western blot; and (B) statistical analysis of AKAP95 and Cx43 expression levels derived from images acquired by the BioRad Chemi DOC XRS + Imaging System. The gray value of protein bands was scanned using Image Lab5.0, and the ratio of band intensity in the treatment group versus the control group was determined using one-way analysis of variance in SPSS 21.0. * indicates statistically significant difference between groups at p < 0.05. Remaining cells were subjected to cell fractionation. (C) Co-immunoprecipitation assay using anti-Cx43 antibody and Western blot using anti-AKAP95 antibody, when cytoplasm and nuclear protein were separated, a small amount nuclear membarnce protein was ramained in the cytoplasm protein. The experiment was repeated four times.

Figure 7. Binding between AKAP95 and Cx43 is affected by PKA activity.

(A) Western blot examination of Cx43 and AKAP95 expression; GAPDH used as the internal reference. Each lane contains 30 μg of total proteins. The experiment was repeated three times. (B) Statistical data of AKAP95 and Cx43 expression levels in cells after H89 and Forskolin treatments (20 μM each). AKAP95 and Cx43 expression levels decreased in the H89 treatment (p < 0.05) and increased in the Forskolin treatment compared to the control group (p < 0.05). Images were acquired using the BioRad Chemi DOC XRS + Imaging System, and the gray value of protein bands was scanned using Image Lab5.0. The ratio of measurements of the treatment group to the control group was calculated by one-way analysis of variance in SPSS 21.0. (C) Western blot analysis of binding between AKAP95 and Cx43 in cells after H89 and Forskolin treatments. 500 μg of extracted total proteins was co-immunoprecipitated (Co-IP) with 3 μg of anti-Cx43 antibody. The Co-IP products were resuspended in an equal volume of loading buffer, and the supernatant was subjected to western blot with anti-Cx43 and anti-AKAP95 antibodies, and the grayscale value of the bands of the H89 and forskolin groups were compared with those of the control bands and the ratios were subjected to statistical analysis. (D) Western blot analysis of binding between AKAP95 and Cx43 after alkaline phosphatase (ALP) treatment. 500 μg of extracted total proteins was diluted to 0.2 μg/μL with RIPA lysis buffer and split into two groups (500 μL each). The ALP treatment was performed following the manufacturer’s instructions (see ALP treatment in M&M). The protein diluents were subjected to Co-IP with 1 μg of anti-Cx43 antibody, and the products were resuspended in an equal volume of loading buffer, and then were subjected to western blot analyzed with anti-Cx43 and anti-AKAP95 antibodies, the grayscale value of the bands of ALP treatment groups were compared with those of the control bands and the ratios were subjected to statistical analysis. Results showed that the binding capacity of Cx43 and AKAP95 was improved after ALP treatment. The experiment was repeated three times.

Figure 8. Binding and localization of AKAP95 and Cx43 in A549 cells at different cell cycle phases after H89 and Forskolin treatments.

(200 × 3) 1. In anaphase, binding between AKAP95 and Cx43 decreased; only a small portion of these proteins were bound to one other in the H89 and Forskolin treatment groups compared to the control group (A); in telophase, there was no binding between AKAP95 and Cx43 (B). 2. In anaphase, aggregation of Cx43 between two daughter nuclei was weakened by H89 treatment and enhanced by Forskolin treatment (A). 3. After Forskolin treatment, Cx43 was clearly distributed on the cell membrane from the center of the cleavage furrow along the two daughter cells (B). This Cx43 localization might be enhanced by Forskolin treatment. The experiment was repeated four times.