Phosphorylation of Arp2 is not essential for Arp2/3 complex activity in fission yeast

Abstract

LeClaire et al presented evidence that phosphorylation of three sites on the Arp2 subunit activates the Arp2/3 complex to nucleate actin filaments. We mutated the homologous residues of Arp2 (Y198, T233, and T234) in the fission yeast genome to amino acids that preclude or mimic phosphorylation. Arp2/3 complex is essential for the viability of fission yeast, yet strains unable to phosphorylate these sites grew normally. Y198F/T233A/T234A Arp2 was only nonfunctional if GFP-tagged, as observed by LeClaire et al in Drosophila cells. Replacing both T233 and T234 with aspartic acid was lethal, suggesting that phosphorylation might be inhibitory. Nevertheless, blocking phosphorylation at these sites had the same effect as mimicking it: slowing assembly of endocytic actin patches. Mass spectrometry revealed phosphorylation at a fourth conserved Arp2 residue, Y218, but both blocking and mimicking phosphorylation of Y218 only slowed actin patch assembly slightly. Therefore, phosphorylation of Y198, T233, T234, and Y218 is not required for the activity of fission yeast Arp2/3 complex.

Figure S1.. Proposed phosphorylation sites on Arp2.

(A) Ribbon diagram of B. taurus Arp2 residues Y202, T237, and T238 (green, with oxygens in red) with neighboring subunits Arp3 and ARPC4. The orange mesh surrounding these proposed phosphorylation sites is the electron density map (contour level 2.5 σ), which reveals no density corresponding to a phosphate group at any of the oxygens (Robinson et al, 2001; pdb: 1k8k). Yellow: Arp3 residue R409 and ARPC4 residues R105/R016 that are proposed to interact with Y202, T237, and T238. (B) Sequence alignment of S. pombe Arp2 residues 187–207 and 222–245 with homologous sequences in nine other eukaryotes. Y198, T233, and T234 and homologous residues are highlighted. (C, D) Mass spectra from fragmentation of (C) Y198 and (D) Y218 phosphopeptides identified using liquid chromatography–tandem mass spectrometry of Arp2 from the Arp2/3 complex purified in the presence of phosphatase inhibitors. Red peaks were identified by the Mascot algorithm as associated with a fragment of the phosphopeptide; selected fragments are labeled on both the spectrum and the peptide sequence. y(n) peaks represent n-residue C-terminal fragments; a/b(n) peaks represent n-residue N-terminal fragments. * indicates loss of water and ⁺ represents positive charge.

Figure 1.. S. pombe strains with mutations of Arp2 residues Y198, T233, and T234.

(A) Viability of haploid strains with mutations blocking or mimicking phosphorylation of residues Y198, T233, or T234 of Arp2, measured by growth of tetrads on YE5S plates at 25°C. Dark red: mutations blocking phosphorylation; light green: mutations mimicking it. * indicates a high-temperature growth defect; x indicates that the strain was not viable. (B–F) Sum projection of confocal fluorescence images (six z-sections with 0.6 μm spacing). (B, C) Haploid S. pombe expressing Fim1-GFP with (B) wild-type Arp2 or (C) Y198F/T233A/T234A mutant Arp2. Images taken from the first frame of actin patch time-lapse movies; both have identical contrast settings. (D–F) Diploid S. pombe strains with one copy of untagged wild-type Arp2 and one copy of (D) wild-type Arp2-GFP, (E) Y198F/T233A/T234A Arp2-GFP, or (F) lethal mutant T233D/T234D Arp2-GFP. All three images have identical contrast settings. (G) Measurements of Arp2-GFP molecules per cell in diploid strains (mean ± SD, n = 47–82). Source data are available for this figure.

Figure 2.. Time course of actin patch assembly and disassembly by strains with Arp2 mutations either blocking or mimicking phosphorylation at Y198, T233, and T234.

(A) Time series of fluorescence micrographs at 1 s intervals of individual actin patches in cells expressing Fim1-GFP, reconstructed from sum projections of six Z-sections. (B–G) Mean numbers of Fim1-GFP molecules over time in 56–115 aligned actin patch tracks from haploid S. pombe strains with (B) wild-type Arp2, (C–D) Arp2 with mutations blocking phosphorylation, and (E–G) Arp2 with phosphomimetic mutations. Shaded regions indicate standard deviations. Dashed lines represent mean numbers of molecules per patch from the wild-type strain in panel A. (H) Rates of patch assembly for 3–4 replicates (27–115 patches per replicate) of time-lapse movies of S. pombe with wild-type Arp2 and Arp2 with mutations blocking or mimicking phosphorylation. Points indicate assembly rates of each replicate; horizontal bars denote the mean assembly rates.

Figure S2.. Effects of mutations that prevent or mimic phosphorylation of Y198, T233, and T234 on the growth of fission yeast and the yield of Arp2/3 complex purification.

(A, B) Semilog plots of optical density over time in liquid cultures of wild-type and Arp2 mutant yeast, at (A) 25°C and (B) 36°C. (C) Growth of 1:10 serial dilutions of wild-type cells and strains with phosphomimetic mutations of Arp2 on YE5S plates at 32°C and 36°C. (D) Anion exchange chromatography of fractions eluted from the GST-VCA affinity column containing the Arp2/3 complex in wild-type cells. No protein from the Y198F/T233A/T234A strain elutes at the position of the wild-type Arp2/3 complex. Source data are available for this figure.

Figure S3.. Time course of Fim1-GFP appearance and disappearance in actin patches.

(A) Numbers of Fim1-GFP molecules over time in 95 individual actin patches in wild-type cells. (B) Mean number of Fim1-GFP molecules per actin patch over time after continuous realignment of patches in panel B. The assembly phase (blue) and disassembly phase (red) are highlighted. (C) Aligned Fim1-GFP actin patch tracks from four separate time-lapse movies of wild-type S. pombe, with assembly phases highlighted. (D) Assembly rates from the four time-lapse movies in panel D, with bar representing the mean. (E, F) Actin patch assembly rates observed in 3–4 movies of wild-type and mutant Fim1-GFP S. pombe strains.

Figure 3.. Phosphorylation at conserved Arp2 residue Y218 is not essential for Arp2/3 complex activity.

(A) Sequence alignment of S. pombe Arp2 residues 208–224 with homologous regions of Arp2 in nine other eukaryotes. Y218 and homologous residues are highlighted. (B–D) Sum projection of confocal fluorescence images (six Z-sections with 0.6 μm spacing) of haploid S. pombe endogenously expressing Fim1-GFP with (B) wild-type Arp2, (C) Arp2 Y218F, or (D) Arp2 Y218E. Images taken from the first frame of actin patch time-lapse movies; all have identical contrast settings. (E–G) Mean numbers of Fim1-GFP molecules over time in 67–76 aligned actin patch tracks from (E) wild-type cells, (F) Y218F Arp2, or (G) Y218E Arp2. Dashed lines represent mean number of molecules per patch from wild-type Arp2 alignment in panel E. (H) Rates of patch assembly for 3–4 replicates (24–76 patches per replicate) of wild-type and Y218 Arp2 mutants. Points indicate assembly rates of each replicate; horizontal bars denote the mean assembly rate.

Figure S4.. Automated calibration curve generation and photobleaching correction for analysis of actin patches in time-lapse movies.

(A) Example segmentations using MAARS software of fluorescence micrographs of fission yeast cells expressing Acp2-GFP or Arp3-GFP. These were two of the seven S. pombe strains used to construct the calibration curve. (B) Calibration curve constructed using seven endogenously expressed GFP-tagged proteins. Numbers of molecules per cell were published (Wu & Pollard, 2005). (C) First frame of a sample time-lapse movie consisting of six optical sections with 0.6 μm spacing, cropped to remove areas of highly uneven illumination. Scale bar: 10 μm. (D) Determination of segmentation threshold using Gaussian fits to a pixel brightness histogram. Circles: histogram of pixel brightness within uncropped area of the first frame. Blue area: Gaussian fit to a section of the pixel brightness histogram preceding the peak. Red area: Two-term Gaussian fit to the entire pixel brightness histogram. Dashed line: Brightness threshold used for segmentation, at which 95% of pixels are thought to be intracellular. (E) Binary segmentation of first 10 frames of the sample time-lapse movie to highlight intracellular regions. (F) Median pixel brightness within the intracellular region over time, fit with a double-exponential function. (G–H) Last frame of the sample time-lapse movie (G) before and (H) after cropping and photobleaching correction. Both images have the same contrast settings used in panel C.