Versatile fluorescent probes for actin filaments based on the actin-binding domain of utrophin

Abstract

Actin filaments (F-actin) are protein polymers that undergo rapid assembly and disassembly and control an enormous variety of cellular processes ranging from force production to regulation of signal transduction. Consequently, imaging of F-actin has become an increasingly important goal for biologists seeking to understand how cells and tissues function. However, most of the available means for imaging F-actin in living cells suffer from one or more biological or experimental shortcomings. Here we describe fluorescent F-actin probes based on the calponin homology domain of utrophin (Utr-CH), which binds F-actin without stabilizing it in vitro. We show that these probes faithfully report the distribution of F-actin in living and fixed cells, distinguish between stable and dynamic F-actin, and have no obvious effects on processes that depend critically on the balance of actin assembly and disassembly.

Fig. 1

Comparison of GFP-UtrCH and fluorescent phalloidin in wounded, fixed oocytes. (a) Low magnification images showing that in samples fixed with an aldehyde-based fixative, fluorescent phalloidin (AX-568-PH) and GFP-UtrCH colocalize, and reveal a striking concentration of F-actin around wounds (arrow). Immunolocalization of tubulin reveals modest preservation of microtubules in the interior of the wound (arrowhead). (b) In samples extracted with methanol after aldehyde fixation, phalloidin staining is completely eliminated, while staining with GFP-UtrCH is preserved. Microtubule preservation is superior under these conditions. (c) High magnification view showing colocalization of fluorescent phalloidin and GFP-UtrCH on cables of F-actin extending away from the wound. (d) High magnification view showing colocalization of fluorescent phalloidin and GFP-UtrCH on actin “fingers” that extend into the wound. (e) Images showing that in cells treated with latrunculin to disrupt F-actin, phalloidin and GFP-UtrCH colocalize in large, unorganized patches around wounds (W). Scale bars is 20 μm.

Fig. 2

The UtrCH probes do not perturb the F-actin cytoskeleton. (a) Microinjection of 40 or 160 ng of GFP-UtrCH mRNA into Xenopus oocytes does not increase the amount of pelletable (P) F-actin compared to uninjected (Unj) controls as indicated by Western blotting. (b) Oocytes microinjected with GFP-UtrCH appear to heal normally after wounding. Comparisons of several timepoints from 4D movies showed no physiological differences in overall appearance or structure, healing rate, or myosin-based cortical flow between GFP-UtrCH or AX-568 actin injected oocytes. (c) Kymographs of oocytes injected with GFP-UtrCH, RFP-UtrCH or AX-568 actin reveal similar rates of F-actin movement toward the wound site (W). Actin flow rates were determined by measuring the distance travelled by individual fluorescent points (arrowheads) over time (T). (d) Quantification of actin flow rates in cells injected with 40 or 160 ng GFP-UtrCH mRNA, 40 ng RFP-UtrCH mRNA, or AX-568 actin. There are no significant differences in the measured flow rates (P > 0.05).

Fig. 3

GFP-UtrCH labels F-actin structures in living echinoderm embryos and oocytes. (a) Frames from a 4D sequence of a 16-cell purple urchin embryo expressing GFP-UtrCH from injected mRNA; vegetal view. The four small cells in the center are the micromeres; their sisters, the macromeres, divide in this sequence. Each image is a brightest-point projection of ten 1-μm sections. Probe accumulation (arrowhead) in the equatorial cortex is approximately coincident with the onset of furrowing. (b) Frames from a time-lapse sequence at a single focal plane through the blastula epithelium in a sand dollar embryo. GFP-UtrCH accumulates on the nuclear membrane in inter-phase, brightening just before nuclear envelope breakdown (see cells labeled 1–4; the pair of cells labeled “3” has just divided at the beginning of the sequence, and by 3 min. have accumulated F-actin on the nuclear membrane). (c) Germinal vesicle (GV) breakdown in a sea star oocyte. Actin assembly proceeds in a wave starting at the interior side of the GV. Each timepoint is a projection of eight 1-μm sections. Scale bar is 25 μm.

Fig. 4

Labeling and distinguishing between relatively more dynamic and stable actin structures in Xenopus oocytes. (a) Frames from a 4D movie show that GFP-UtrCH more effectively labels the stable actin that resides on the interior of contractile ring (arrowhead) than AX-568-actin. (b) Comparisons of GFP-UtrCH to total actin distributions by antibody staining in fixed oocytes show few differences at the leading or on the contractile ring but do exhibit differences at the trailing edge. (c) Quantifications of fluorescent intensities at the leading edge (LE), on the contractile ring (CR), and locations away from the contractile structure (BG) showed no differences (P > 0.05, n = 24) between antibody staining and the UtrCH probe. Significant differences in fluorescent intensities were observed at the trailing edge (TE, P < 0.05, n = 24). (d) Quantifications of actin comets from oocytes injected with AX-488 actin and RFP-UtrCH revealed that the ratio of AX-488 actin to RFP-UtrCH is greater at the newly assembled head of the comet than the older tail regions (Asterisks indicate P < 0.05, n = 24). (e) High magnification images showing the growing actin comet head (arrowhead) is labelled more intensely with AX-488 actin than RFP-UtrCH.

Fig. 5

Using photobleaching and photoactivation with the UtrCH probes to monitor the contributions of the stable and dynamic F-actin populations. (a) Frames from a 4D sequence comparing the simultaneous fluorescence recovery of AX-568 actin and fluorescence loss of PaGFP-UtrCH in the same region of the oocyte cortex. (b) Measurements of fluorescence recovery of AX-568 actin parallels the inverse of fluorescence loss of PaGFP-UtrCH. (c) Frames from a 4D sequence contrasting the fluorescence loss of PaGFP-UtrCH in control and jasplakinolide treated oocytes. (d) Quantification of fluorescence loss of PaGFP-UtrCH in control and jasplakinolide treated cells reveals that fluorescence loss is reduced in jasplakinolide treated oocytes. (e) Measurements of PaGFP-UtrCH fluorescence loss in point-activated control and jasplakinolide treated oocytes also exhibited a reduced rate of fluorescence loss in jasplakinolide treated oocytes. (f) Fluorescence decay half-times of point-activated PaGFP-UtrCH in jasplakinolide-treated and untreated controls. Asterisks indicate P < 0.05. (g) Combined use of RFP-UtrCH and PaGFP-UtrCH reveals a clear subdivision in F-actin populations around wounds. The oldest (i.e. most stable; green around the wound) F-actin becomes enriched on the interior of the actin array. This population of F-actin becomes surrounded by more dynamic F-actin (red). Scale bars = 30 μm.