To the Editor:

Reidl et al. have described the use of a small polypeptide, Lifeact, derived from yeast, to visualize filamentous actin (f-actin)1. This reagent, when fused to GFP, allows the study of actin remodeling in living cells without consequences to actin polymerization. The authors concluded that “Lifeact can be used as a universal marker for actin imaging...”1.

We fused the Lifeact sequence to either monomeric (m)Cerulean blue2 or mRFP3 and expressed these proteins in mouse striatal neuron–derived STHdh cells to study the formation of nuclear cofilin-bound actin rods that form during cell stress (Supplementary Methods online).

We found that Lifeact-CFP could be used to visualize f-actin under optimal growth at 37 °C, with Lifeact signal very similar to that of actin-mRFP (Fig. 1a–c). Upon cell stress induction (Supplementary Methods), we observed the formation of typical cofilin-bound actin rods in the nuclei of cells with endogenous amounts of cofilin by analyzing immunofluorescence of an antibody to cofilin (Fig. 1d). However, we did not detect actin in these structures with Lifeact-CFP (Fig. 1e,f), but Lifeact highlighted nuclear actin puncta (Fig. 1b,e). Using CFP-cofilin to localize the position of stress-induced rods, we detected the presence of actin in the nuclear or cytoplasmic rods either by immunofluorescence analysis (Fig. 1g) together with CFP-cofilin (Fig. 1h,i) or by actin-mRFP expression (Fig. 1j) together with CFP-cofilin (Fig. 1k,l). To test whether the fused fluorophore protein may be affecting Lifeact activity, we switched from using CFP to mRFP. Whereas Lifeact-mRFP could be used to visualize actin similar to Lifeact-CFP (Fig. 1m), we found that Lifeact could not be used to visualize nuclear rods that contain f-actin (Fig. 1m–o). Heat shock did not appear to affect the ability of Lifeact to recognize f-actin in cytoplasmic stress fibers (Fig. 1e,m).

Figure 1: Visualization of actin and cofilin in nuclear rods after stress.
figure 1

(a–o) Fluorescence images of STHdh cells transfected with the indicated fusion proteins or of endogenous proteins in STHdh cells detected with the indicated antibodies. Cells were grown at 37 °C without heat shock (a–c) or after 30 min heat shock at 42 °C (d–o). Merged red-cyan signal is pseudocolored white. Scale bar, 10 μm.

Full size image

Cofilin responds to cell stress by arresting actin dynamics, thus freeing cellular ATP4. Cofilin-bound actin rods persist after cell stress in Alzheimer's disease–affected neurons and can be induced by a variety of cell stresses5. Others have shown that cofilin-actin rods cannot be stained with fluorescent dye–labeled phalloidin, as cofilin binding to f-actin stabilizes the twisted state of f-actin in the rods, which prevents phalloidin binding6. The most parsimonious explanation of our results is that f-actin conformational changes upon cofilin binding prevent Lifeact from binding to twisted f-actin. To date, the only way we can accurately visualize actin in cofilin-actin rods in live cells is with the use of an Aequoria sp. fluorescent protein (AFP) fused to actin.

These data provide an important caveat for those using Lifeact and studying actin dynamics during cell stress and indicate that Lifeact may not be a universal tool for studying actin dynamics in live cells. These data also suggest that Lifeact may be a useful tool to distinguish certain forms of f-actin in live cells.

Note: Supplementary information is available on the Nature Methods website.