Loss of CENP-F results in distinct microtubule-related defects without chromosomal abnormalities

Abstract

Microtubule (MT)-binding centromere protein F (CENP-F) was previously shown to play a role exclusively in chromosome segregation during cellular division. Many cell models of CENP-F depletion show a lag in the cell cycle and aneuploidy. Here, using our novel genetic deletion model, we show that CENP-F also regulates a broader range of cellular functions outside of cell division. We characterized CENP-F(+/+) and CENP-F(-/-) mouse embryonic fibroblasts (MEFs) and found drastic differences in multiple cellular functions during interphase, including cell migration, focal adhesion dynamics, and primary cilia formation. We discovered that CENP-F(-/-) MEFs have severely diminished MT dynamics, which underlies the phenotypes we describe. These data, combined with recent biochemical research demonstrating the strong binding of CENP-F to the MT network, support the conclusion that CENP-F is a powerful regulator of MT dynamics during interphase and affects heterogeneous cell functions.

FIGURE 1:

CENP-F^−/− MEFs lack directionally persistent migration. Migration patterns of (A–D) CENP-F^+/+ and (E–H) CENP-F^−/− acquired by DIC microscopy over a period of 1 h. False-colored overlays of cells. Orange, time zero; red, 20 min; purple, 40 min; green, 60 min. CENP-F^−/− MEFs fail to establish directionally persistent migration patterns as quantified in I. A value of 1.0 corresponds to movement in a straight line, and 0 equals random movement. CENP-F^+/+ = 0.81 and CENP-F^−/− = 0.24 (n = 20; *p < 0.001; error bars represent SEM).

FIGURE 2:

FAs of CENP-F^−/− MEFs are larger and denser and disassemble more slowly than CENP-F^+/+ MEFs. Immunofluorescence using α-vinculin antibodies visualizing FAs in CENP-F^+/+ and CENP-F^−/− MEFs. Quantification of (A) mean size (p = 0.001), (B) total fluorescence intensity (p < 5.1E-05), and (C) FA fluorescence intensity/area (p < 5.6E-05; n > 1500). Immunofluorescence using vinculin antibodies in MEFs plated on fibronectin crossbows. (D, G) Fluorescent fibronectin crossbows. (E, H) Vinculin staining in representative CENP-F^+/+ and CENP-F^−/− cells. When fluorescence of many cells is averaged, CENP-F^−/− cells show greater FA and cytoplasmic vinculin fluorescence (D, I). n = 78 CENP-F^+/+, n = 69 CENP-F^−/−. (J) Fluorescence recovery curve from CENP-F^+/+ and CENP-F^−/− MEFs. Error bars represent SEM. (K) Quantification of immobile fraction from fluorescence recovery curve (p = 0.05; n = 10).

FIGURE 3:

Decrease in cilia formation in CENP-F^−/− MEFs. (A, B) Fluorescence micrographs of MEFs after serum starvation to induce ciliogenesis. Cilia are labeled with antibodies raised against acetylated tubulin (green) and γ-tubulin (red). Nuclei are stained with DAPI. Images represent z-projections. Ciliated cells are labeled with an asterisk. (A′, B′) Increased magnification of ciliated MEFs. Cilia are the same length in both cells but occur less frequently in CENP-F^−/− MEFs. (C) Quantification of cilia length in MEFs (p = 0.91; n = 25; error bars represent SEM). (D) Quantification of portion of ciliated cells (p < 5.0E-05; n = 30; error bars represent SEM).

FIGURE 4:

CENP-F is required for establishment of MT array asymmetry, and loss of CENP-F leads to an abundance of posttranslational modifications and resistance to depolymerization by NOC. Immunostaining of MTs with DM1A antibody (green) and DAPI (blue) in CENP-F^+/+ (A) and CENP-F^−/− (C) mouse embryonic fibroblasts. Scale bar, 15 μm. Schematic for fluorescence intensity quantification is shown on the right in A and C. Representative fluorescence intensity plots (30, each) of CENP-F^+/+ (B) and CENP-F^−/− (D) for each sample. Immunostaining of MTs (green) in wild-type MEFs (E–E′′) shows minimal amounts of acetylated (E′) and Glu (E′′) tubulin in relation to the typical staining of the complete MT network (E). In contrast, the MT network of CENP-F^−/− MEFs (F) contains an abundance of posttranslational modifications, as indicated by immense acetylated (F′) and Glu (F′′) tubulin staining. Scale bar, 10 μm. (G) Western blot demonstrating the abundance of acetylated tubulin and tubulin in CENP-F^+/+ and CENP-F^−/− MEFs. Immunofluorescence analysis of the MT network (green) in CENP-F^+/+ (H–H′′) and CENP-F^−/− (I–I′′) MEFs treated for 2 h with no NOC (H, I), 1 μg/ml NOC (H′, I′), or 2.5 μg/ml NOC (H′′, I′′). Numerous short MTs persist in CENP-F^−/- MEFs (I′, I′′), as compared with full depolymerization of MTs in CENP-F^+/+ MEFs (H′, H′′). n = 25. Scale bar, 10 μm. Time-lapse imaging of CENP-F^+/+ (J–J′′) and CENP-F^−/- (L–L′′) MEFs expressing 3xEGFP-EMTB (black) to highlight the MT network. Images showing dynamic pixel differences were created by projecting pixel differences (purple) between sequential movie frames. CENP-F^−/− MTs are very stable and move as an entire unit, as opposed to dynamic instability, exclusively at the tips of CENP-F^+/+ MTs. Life history plots of MT dynamic instability can be observed in CENP-F^+/+ MEFs (K) with periods of growth and shrinkage at the MT tips. In contrast, CENP-F^−/- MEFs (M) experience very few periods of growth and catastrophe with long periods of pause.