Real-time imaging of myosin II regulatory light-chain phosphorylation using a new protein biosensor

Abstract

Phosphorylation of the RMLC (regulatory myosin light chain) regulates the activity of myosin II, which is critically involved in the motility of both muscle and non-muscle cells. There are both Ca2+-dependent and -independent pathways for RMLC phosphorylation in smooth-muscle cells, and the latter pathway is often involved in an abnormal contractility in pathological states such as asthma and hypertension. Therefore pharmacological interventions of RMLC phosphorylation may have a therapeutic value. In the present study, we developed a new genetically encoded biosensor, termed CRCit (ECFP-RMLC-Citrine, where ECFP is enhanced cyan fluorescent protein), that detects RMLC phosphorylation using fluorescence resonance energy transfer between two variants of the green fluorescent protein fused to both the N- and C-termini of RMLC. When expressed in primary cultured vascular smooth-muscle cells, CRCit detected the Ca2+-dependent RMLC phosphorylation with a high spatiotemporal resolution. Furthermore, we could specifically assay the agonist-induced Ca2+-independent phosphorylation of RMLC when Ca2+ signalling in cells expressing CRCit was suppressed. Thus CRCit may also be used for the high throughput screening of compounds that inhibit abnormal smooth-muscle contraction.

Figure 1. Construction and properties of CRCit

(A) Schematic structure of CRCit with the linker sequences shown above. The two phosphorylation sites (Thr¹⁸ and Ser¹⁹) on the RMLC domain of CRCit are indicated. MLC, myosin light chain. (B) Western blots of whole-cell lysate from wild-type smooth-muscle cells or cells expressing 75 kDa CRCit probed with the anti-GFP antibody. (C) Fluorescence image of CRCit expressed in cultured vascular smooth-muscle cells. (D) Upper panel: time courses of fluorescence intensities at 535 nm (green line) and 480 nm (blue line) of a single CRCit-expressing vascular smooth-muscle cell on application of 80 mM K⁺ solution. Application of 80 mM K⁺ solution is indicated by a shaded box. The sampling interval was 1 s. Lower panel: time course of change in FRET response (F₅₃₅/F₄₈₀). (E) F₅₃₅/F₄₈₀ fluorescence intensity ratio images before and after the application of 80 mM K⁺ solution in the same cell as in (D). Each panel shows the average of five consecutive images around the indicated time.

Figure 2. Effects of the MLCK inhibitors and withdrawal of external Ca²⁺ on the high-K⁺-induced change in FRET efficiency of CRCit

(A) Time course of FRET response of CRCit during repeated applications of high-K⁺ solution. The high-K⁺-induced response was observed reversibly and repetitively more than five times. Ratios of F₅₃₅/F₄₈₀ were normalized by the value before the high-K⁺ application. The sampling interval was 5 s. (B) The application of high-K⁺ solution in the absence of external Ca²⁺ resulted in no significant increase in FRET signal. (C, D) Inhibition of high-K⁺-induced FRET response by 10 μM ML-7 or by 10 μM wortmannin (WM). (E) Composite results of experiments shown in (A–D). White bars indicate the FRET response to the first stimulation, and black bars indicate those to the second stimulation (for B) or the third stimulation (for A, C and D). Means±S.E.M., n=14, 6, 5 and 7 for control, Ca²⁺-free, ML-7 and WM respectively. **P<0.01 (t test).

Figure 3. High-K⁺-induced change in FRET signal of CRCit with amino acid substitution at the phosphorylation site

(A) Time course of FRET efficiency of wild-type (WT) and mutant CRCit when [Ca²⁺]_i was increased by application of 80 mM K⁺ or 1 μM ionomycin. The sampling interval was 5 s. (B) Compiled results of effects of mutations on the Ca²⁺-dependent change in the FRET efficiency. Means±S.E.M., n=11, 8, 15 and 15 for wild-type (WT), T18A, S19A and T18A/S19A respectively. **P<0.01 (t test).

Figure 4. Rho-kinase-dependent change in FRET signal of CRCit

(A) Time course of FRET efficiency of wild-type CRCit after application of U46619 in the absence (black continuous line) and presence (broken line) of Y-27632, a Rho-kinase inhibitor. Time course of FRET efficiency of mutant CRCit (T18A/S19A) is also shown (grey line). The intracellular Ca²⁺ store was depleted by treatment with thapsigargin before the application of the thromboxane A₂ analogue in the absence of extracellular Ca²⁺. The sampling interval was 5 s. (B) Composite results of effects of Y-27632 and T18A/S19A mutations on the U46619-mediated change in FRET efficiency of CRCit. Means±S.E.M., n=22 and 11 for wild-type (WT) CRCit in the absence and presence of Y-27632 respectively and n=24 for T18A/S19A. **P<0.01 (t test).