ERK1 nucleocytoplasmic shuttling rate depends on specific N-terminal aminoacids

https://doi.org/10.1016/j.bbrc.2010.06.044Get rights and content

Abstract

Despite ERK1 and ERK2 were considered interchangeable isoforms for a long time, their roles are now emerging as only partially overlapping. We recently reported that the nucleocytoplasmic trafficking of GFP-tagged ERK1 is slower than that of ERK2, this difference being caused by a unique domain of ERK1 located at its N-terminus (ERK1-Nt). In the present report we further investigated this issue by asking which were the specific aminoacids involved in such process. By photobleaching strategy, we demonstrated that ERK1-Nt is a domain capable to slow down the nucleocytoplasmic shuttling rate even of a small cargo protein. ERK1-Nt was then dissected into three regions as follows: 1 (aa 1–9), 2 (aa 10–29) and 3, (aa 30–39) that were deleted or mutated at specific sites. Dynamic imaging assessment of the role played by each region in determining the shuttling rate revealed that: region 1 has no significant role, region 2 and specific aminoacids of region 3 (V31, K33, P36) are critical, but singularly do not totally account for the difference in the shuttling rate between ERK1 and 2. Finally, we demonstrated that the nucleocytoplasmic shuttling rate of a passively diffusing protein (mRED) is inversely related to ERK1-Nt-GFP concentrations inside the cell, thus suggesting that ERK1-Nt-GFP occupies the nuclear pore perhaps because of an important affinity of ERK1-Nt for nucleoporins.

In conclusion, ERK1-Nt is a domain able per se to confer a slower shuttling rate to a cargo protein. Specific regions within this domain were identified as responsible for this biophysical property.

Introduction

The kinases ERK1 and 2 have been considered for a long time under the control of the same regulatory mechanisms and responsible for the same biological outputs. Nevertheless, the presence of the two separate isoforms raises some doubts about their interchangeability. Furthermore, the ERK2 knockout is embryonic lethal while the ERK1 knockout is perfectly viable, exhibiting only minor deficits [1], [2]. Recently, we demonstrated that the different capability of ERK1 and 2 in influencing cellular proliferation is related to their diverse biophysical properties in crossing the nuclear barrier [3]. We showed that ERK1 and 2 are continuously and bidirectionally exchanged across the nuclear envelope and that the rates of their shuttling vary in dependence of their activation state [3]. This mechanism has been demonstrated to be an essential step in the transduction of extracellular signals and it represents a powerful system to control ERK action on nuclear and cytoplasmatic targets. Furthermore, we showed that ERK1 and 2 drastically differ in their capability of crossing the nuclear envelope with ERK1 being three times slower than ERK2. This difference critically affects the capability of ERK1 and 2 of activating downstream effectors, thus providing a sensitive mechanism through which cells modulate their response to extracellular stimuli.

In addition, we showed that the different shuttling rate is caused by the N-terminal region of ERK1 encompassing residues 1 to 39 (ERK1-Nt). Indeed, the deletion of this portion converted the shuttling features of ERK1 into those of ERK2 and, consistently, the fusion of this motif to the N-terminus of ERK2 slowed down ERK2 shuttling rate to a value similar to that found for ERK1 [3].

What remains to be elucidated are the specific molecular determinants underlying the difference in the nucleocytoplasmic shuttling speed between ERK1 and 2. Aim of the present study was to get further insights into this matter by addressing three related issues including: (1) whether ERK1-Nt constitutes per se a domain or masks some important sites on the tertiary structure of the protein; (2) whether specific aminoacids (aa) of ERK1-Nt are critical in slowing down ERK1; (3) whether the increase of ERK1-Nt-GFP concentration slows down the trafficking rate of a reporter protein (mRED) across the nucleus. The use of real time imaging techniques, like Fluorescence Recovery After Photobleaching (FRAP), allowed to positively answer to all the questions outlined above and provided a better and molecular based understanding of the physiological/biophysical differences described for ERK1 and 2.

Section snippets

Plasmid preparation

GFP-ΔN1–39 and GFP-ΔN1–27: GFP-ΔN1–39 and GFP-ΔN1–27 was obtained by digesting pEGFP-C2-rat ERK1 with HindIII/BamHI and cloning the extracted fragment of ERK into pEGFP-C2. The resulting construct was digested with ApaI (ΔN1–39) or SmaI (ΔN1–27) and the plasmid were ligated to let its re-circularization.

ΔN1–39-GFP: ΔN1–39-GFP was obtained digesting the construct carrying ERK1 fused to the N-terminal of GFP (ERK1-GFP) with XhoI/BamHI and cloning the fragment of ERK N-terminus in pEGFP-N3.

The isolated N-terminal of ERK1 decreases per se the nucleocytoplasmic shuttling rate

In the present report, we first investigated if this terminal region is capable of slowing down a generic cargo. In order to test this hypothesis we fused GFP with ERK1-Nt alone and we measured the speed of nucleocytoplasmic exchange by nuclear FRAP experiments.

Fig. 1A shows the aminoacidic sequence of mouse ERK1-Nt that we can split in three regions:

  • 1.

    aa 1–9, an alanine enriched cap in common with the two kinases except for the inversion of P8G9 of ERK1 in G8P9 of ERK2 (region 1);

  • 2.

    aa 10–29, a

Discussion

So far the portion of the ERK1-Nt which ranges from aa 1 to 39 has not been associated to any specific function. Recently, we proved that this region is sufficient and necessary to slow down the nuclear turnover of ERK1 respect to ERK2 [3], but no hints concerning which specific aa involved were uncovered. In this study we investigated this open question.

We demonstrated that the fusion of ERK1-Nt to GFP dramatically slowed down also the trafficking of GFP, so we can exclude that this terminal

Conclusions

We took advantage of live imaging techniques and mutagenesis strategies to demonstrate that:

  • (i)

    ERK1-Nt is a domain per se capable of slowing down ERK1, (ii) there are critical aa in the determination of the distinctive ERK1 shuttling rate, (iii) ERK1-Nt showed an interaction for nucleoporins.

Acknowledgments

We thank Fabio Beltram for his intellectual support. This research was funded by the Italian Ministry of Research (PRIN 20083ZAXYC_003). M. M. is a Dulbecco Telethon Scientist.

References (12)

There are more references available in the full text version of this article.
View full text