Report

Imaging Protein Kinase Cα Activation in Cells

See allHide authors and affiliations

Science  26 Mar 1999:
Vol. 283, Issue 5410, pp. 2085-2089
DOI: 10.1126/science.283.5410.2085

Abstract

Spatially resolved fluorescence resonance energy transfer (FRET) measured by fluorescence lifetime imaging microscopy (FLIM), provides a method for tracing the catalytic activity of fluorescently tagged proteins inside live cell cultures and enables determination of the functional state of proteins in fixed cells and tissues. Here, a dynamic marker of protein kinase Cα (PKCα) activation is identified and exploited. Activation of PKCα is detected through the binding of fluorescently tagged phosphorylation site–specific antibodies; the consequent FRET is measured through the donor fluorophore on PKCα by FLIM. This approach enabled the imaging of PKCα activation in live and fixed cultured cells and was also applied to pathological samples.

For many proteins, there is a need to integrate spatial data with information on catalytic function. This is a particular concern in signal transduction processes, where the networking of multiple inputs affects the output, leading to grossly different cellular consequences. Methods applicable to the functional analysis of particular (signaling) proteins in situ would provide a significant advance in reaching a molecular description of cellular behavior.

The classical and novel protein kinase isotypes (cPKC and nPKC, respectively) undergo conformational changes in response to their second messenger, diacylglycerol (DAG) (1–3). This PKC activator is restricted to membrane compartments, and the stable membrane/DAG-associated complexes formed by PKC have traditionally provided a useful means of monitoring PKC isotype activation (4). However, PKC isotypes can associate with membranes before cell stimulation, making membrane association an insufficient criterion for determining PKC activity. We identified a marker of PKCα activation and used FLIM to describe the activation state of PKCα in situ. The utility of this approach is exemplified in analyses of fixed and live cells and additionally of archived human tissues.

PKCα is phosphorylated in vivo on threonine-250 (5). A site-specific antiserum to phosphorylated Thr250 [T(P)250] reacted with PKCα expressed in COS-7 cells only after treatment of cells with tetradecanoyl phorbol acetate (TPA), and this response was blocked by the PKC inhibitor bisindolylmaleimide I (Fig. 1A). Optimal phosphorylation was induced after 30 min with 400 nM TPA (Fig. 1B). The protein we observed to be T(P)250 immunoreactive is PKCα, as judged by immunoprecipitation and immunoblot (6).

Figure 1

Identification and detection of an autophosphorylation site on PKCα. (A) COS-7 cells were transfected with a cDNA expression plasmid for PKCα. After 48 hours, the cells were treated with TPA for 0 (untreated) or 30 min either in the presence (+) or absence (−) of the PKC inhibitor bisindolylmaleimide I (BIM, 10 μM). The upper panel indicates immunoreactivity with the T(P)250 antiserum (20). The lower panel shows immunoreactivity with the MC5 antibody to PKCα protein. (B) Transfected COS-7 cells expressing PKCα were treated with TPA (400 nM) for the times indicated. Extracts were prepared and subjected to protein immunoblotting. The upper panel shows the immunoreactivity for the T(P)250 antibody, and the lower panel shows there was no change in PKCα during this treatment. (C) COS-7 cell–expressed, histidine-tagged PKCα was purified by nickel-agarose chromatography (7). Autophosphorylation of purified PKCα was carried out in the presence of TPA (2 μM), phosphatidylserine (1.25 mg/ml) in 1% (v/v) Triton X-100, 50 mM Hepes (pH 7.5), 12.5 mM MgCl2, and 100 μM ATP. Reactions were initiated with ATP and terminated with Laemmli sample buffer at the times indicated. Phosphorylation of the Thr250 site was monitored with the T(P)250 antiserum (upper panel) and the amount of PKCα protein with MC5 (lower panel). (D) Quiescent Swiss 3T3 cells (21) were treated with TPA for the times indicated. Extracts were prepared and blotted for immunoreaction with the T(P)250 antibody or for PKCα protein. The dot in the upper panel indicates an immunoreactive band that is variably observed and not competed by the phosphopeptide antigen (6). (E) Quiescent 3T3 cells were treated with PDBu (500 nM) for 20 min, and then washed with medium at 4°C and incubated at 37°C in the absence of PDBu for the times indicated. Cell extracts were blotted with the T(P)250 antiserum, and immunoreaction was quantified by scanning densitometry. The graph illustrates the lag period observed before an exponential decay in T(P)250 immunoreaction.

Purification of His-tagged PKCα from unstimulated COS-7 cells (7), followed by an in vitro autophosphorylation reaction, led to a time-dependent increase in immunoreaction with the T(P)250 antiserum (Fig. 1C). These results demonstrate that the Thr250 site is an autophosphorylation site in vitro and in vivo. The equivalent site in PKCβ is also an autophosphorylation site; it is also conserved in PKCγ but not in other PKC isotypes (6).

The response of endogenous PKCα was determined in the murine Swiss 3T3 fibroblast line. Exposure of quiescent Swiss 3T3 cells to TPA induced rapid phosphorylation of PKCα (Fig. 1D). This was time-dependent, with optimum phosphorylation induced within 5 min. The dynamics of Thr250dephosphorylation were monitored upon the removal of the relatively hydrophilic agonist, phorbol dibutyrate (PDBu). After a lag period of variable length (10 to 25 min), which probably reflects the reequilibration of cellular PDBu pools, T(P)250 immunoreaction decreased with a half-life of ∼5 min (Fig. 1E). Thus, the dynamics of PKC activation can be followed through the immunoreaction of T(P)250.

To monitor the activation of endogenous PKCα, we conjugated the fluorophore Cy3 to the MC5 anti-PKCα antibody (MC5-Cy3) and also conjugated the fluorophore Cy5 to purified T(P)250 immunoglobulin G (IgG) [T(P)250-IgG-Cy5]. In unstimulated, fixed 293T cells, heterogeneous staining of the protein was observed with only weak immunoreaction to T(P)250-IgG-Cy5 (Fig. 2, upper panels). Upon stimulation with TPA, T(P)250-IgG-Cy5 immunoreactivity was increased and localized with MC5 immunoreactivity, consistent with activation of PKCα. Colocalization gives only an indication of association, because the determination of proximity is limited by the resolution of the optical system. FRET between Cy3 and Cy5, as measured by FLIM (8–16), provides a molecular proximity assay with nanometer-scale resolution for the quantitative determination of the activation and intracellular location of PKCα (17). Lifetimes of the Cy3 donor fluorophore (conjugated to MC5) were monitored in cells stained with MC5-Cy3 alone or both MC5-Cy3 and T(P)250-IgG-Cy5. The Cy3 lifetimes (Fig. 2, middle panels) were reduced after stimulation of cells with TPA, when staining was done with both antibodies; for MC5-Cy3 alone, there was no change in lifetimes. The largest reduction in lifetimes (from 1.0 to 0.49 ns) was associated with regions of the plasma membrane and vesicular structures in the cytoplasm. These PKCα-containing vesicular structures overlapped in part with the distribution of ectopically expressed PKCα described for NIH 3T3 cells (18). The reduced lifetimes observed in the unstimulated cells were also localized to punctate cytoplasmic structures.

Figure 2

Endogenous PKCα activation detected by immunofluorescence and FLIM. (Upper panels) Stacked confocal micrographs show an increase in Thr250phosphorylation of endogenous PKCα following 20 min of TPA stimulation (+TPA), as detected by a Cy5-conjugated IgG fraction of T(P)250 [T(P)250-Cy5]. There was no apparent difference in the level of PKCα protein expression before and after TPA treatment, as demonstrated by concomitant detection with MC5-Cy3. Double-label immunofluorescence staining was performed as described (22), except for minor modifications (23). Antibodies were conjugated to fluorophores as described (24). (Middle panels) “I” shows the steady-state fluorescence images from the donor fluorophore (Cy3) conjugated to MC5 in fixed 293T cells, which were stained with either MC5-Cy3 alone or with both MC5-Cy3 and T(P)250-Cy5 (acceptor fluorophore) as indicated (25). FRET from the donor (Cy3) to acceptor (Cy5) fluorophore took place on individual PKCα molecules that were phosphorylated after TPA stimulation, resulting in a reduction of the fluorescence emission lifetime of Cy3 as demonstrated by changes in 〈τ〉 (the average of τpand τm). (Lower panels) The 2D histograms quantify the reduction of the donor lifetimes of Cy3 emission for both the phase (τp) and modulation (τm). A detailed description of the FLIM apparatus used for FRET determination can be found elsewhere (16). FRET efficiency color tables (Eff) were calculated from EFF = 1 − τda/〈τd〉 where τda is the lifetime map (or color table) of the donor in the presence of acceptor and 〈τd〉 is the average lifetime of the donor in the absence of acceptor.

To demonstrate the specificity of the FRET signals as measured by FLIM, cells were stained for PKCα or activated PKCα, or for markers of the endoplasmic reticulum (anti–p62-Cy3) or Golgi (anti–galactosyl transferase-Cy3 or anti–p115-Cy3) (19). There was some overlap in localization of PKCα with both compartments; however, there was no FRET between the fluorophores, and thus no reduction in lifetime was observed. The specificity of the PKCα antibody itself (MC5) is essential for the interpretation of the FLIM measurements. After treatment of cells for 18 hours with TPA (1 μM) in order to down-regulate PKCα, immunoreactivity with MC5 was abolished (19).

FLIM can be used to analyze a donor-fluorophore–tagged PKCα in live cells microinjected with an excess of acceptor labeled T(P)250-IgG. In transiently transfected COS-7 cells, green fluorescent protein–PKCα (GFP-PKCα) accumulated in a partly cytoplasmic and partly perinuclear location in untreated cells (Fig. 3A). After treatment of cells with TPA, there was a progressive increase in GFP-PKCα accumulation in vesicular structures within the cytoplasm; this was most evident after 30 min. The two central GFP-PKCα–transfected cells and one untransfected cell (Fig. 3A) were microinjected with T(P)250-IgG-Cy3.5. The donor fluorophore (GFP), when visualized in the unstimulated cells, was broadly dispersed in the cytoplasm and excluded from the nucleus. Analysis of mean fluorescence lifetime images (〈τ〉, the average of τp and τm) revealed that there was a small degree of GFP-PKCα phosphorylation in unstimulated cells, but following stimulation, there was a progressive increase in Thr250 phosphorylation, evidenced by a decrease in GFP emission lifetimes; this was evident only in the two central GFP-PKCα–transfected and T(P)250-IgG-Cy3.5–microinjected cells. After 15 min, the most intense decrease was observed in punctate areas adjacent to the plasma membrane. By 30 min, there was a greater shift in donor emission lifetimes at the plasma membrane with a broader distribution of shorter lifetimes throughout the cytoplasm. The τm and τp two-dimensional (2D) histograms show the shift in lifetimes that occurred upon TPA stimulation. Control experiments on fixed cells analyzed by confocal imaging and protein immunoblotting confirmed a redistribution of GFP-PKCα to a perinuclear compartment, but indicated little or no protein degradation within this time frame.

Figure 3

TPA-induced Thr250phosphorylation of GFP-tagged PKCα in live COS-7 cells detected by FLIM. (A) COS-7 cells were transiently transfected with an NH2-terminal GFP-tagged PKCα construct (details to be described elsewhere) and cultured in serum-containing Dulbecco's modified Eagle's medium for 36 hours before transfer to serum- and phenol red–free Optimax medium (Life Technologies). After an overnight culture, cells were washed and resuspended in phosphate-buffered saline, then microinjected with a Cy3.5-conjugated, protein G–purified IgG fraction of T(P)250. The fluorescence images are shown from the donor (GFP) in four live COS-7 cells which were either left as controls (GFP) or microinjected with T(P)250-IgG-Cy3.5 [T(P)250-Cy3.5]. The bottom four cells were transfected, but only the top three cells (including an untransfected cell) were microinjected with T(P)250-IgG-Cy3.5. By 30 min of TPA stimulation, two separate fluorescence lifetime populations belonging to the microinjected and uninjected cells, respectively, were evident from changes in both the phase (τp) and modulation (τm) lifetimes of GFP emission (bottom panels). Images shown are representative of five independent experiments. A detailed description of the FLIM apparatus used for FRET determination can be found elsewhere (16). (B) COS-7 cells were transiently transfected with a GFP-tagged PKCα construct, stimulated with TPA (400 nM), and fixed at various time points as indicated, then either left as controls (GFP-PKC) or stained with T(P)250-IgG-Cy3.5 [GFP-PKC/T(P)250]. The fluorescence images from the donor (I) are shown. The cumulative lifetimes of GFP-PKC alone (green) and that measured in the presence of the acceptor fluorophore [T(P)250-IgG-Cy3.5] (red) are plotted on the same 2D histogram for each time point. For derivation of Eff, see Fig. 2.

Monitoring of donor emission lifetimes in live cells for up to 40 min after TPA treatment showed that there was a persistent reduction of 〈τ〉 in the transfected, microinjected cell population. In contrast, control experiments done in fixed, GFP-PKCα–transfected COS-7 cells revealed a subsequent Thr250dephosphorylation event after approximately 30 min of TPA treatment, with some degree of variation in time course among cells (Fig. 3B). The persistent reduction in 〈τ〉 observed in live cells may result from a masking effect of the phospho-specific antibody which protects the epitope from subsequent dephosphorylation by phosphatases that act on PKC. This also shows that the FLIM experiments on live and fixed cells are complementary.

We investigated the activation state of PKCα in fixed tissue samples, using FLIM to provide the required specificity of detection. In a series of formalin-fixed paraffin-embedded breast tumors (n = 23), parallel samples were stained either for PKCα (MC5-Cy3) or for both PKCα and Thr250phosphorylation [T(P)250-IgG-Cy5]. The results obtained fall into two categories: those patients where there was no detectable change in fluorescence lifetime (12 of 23 patients) and those where there was a significant decrease in 〈τ〉 (11 of 23) (Fig. 4). Quantification of the τp and τm changes is provided by the 2D histogram. The results demonstrate that PKCα is activated in situ in a significant number of human breast tumors. Notably, whereas some tumors displayed an up-regulation of PKCα, PKC content per se did not correlate with activation.

Figure 4

Demonstration of a variable degree of Thr250 phosphorylation of PKCα in paraffin-embedded, fixed tissue sections of human breast carcinomas. Paraffin-embedded breast cancer sections were dewaxed in xylene, pressure cooked for antigen-retrieval (26), then stained with either MC5-Cy3 alone or with both MC5-Cy3 and T(P)250-IgG-Cy5. The leftmost panels show the donor (upper and middle) and acceptor (lower) fluorescence patterns. The corresponding lifetime images for the upper two sections are shown at middle and right. The 2D histogram (bottom right) shows the donor lifetimes (τp and τm) from these images. For derivation of Eff, see Fig. 2.

The results demonstrate that FRET measured with FLIM is able to detect catalytic functions in vivo as shown by the use of an autophosphorylation site in PKCα to monitor its activation. Our study reveals the dynamics of this phosphorylation process in live cells through donor lifetime properties, revealing the tonic and stimulated juxta-membrane vesicular location of activated PKCα. The studies in fixed cells and tissues complement the analysis and demonstrate the use of these techniques for investigation of human disease.

  • * These authors contributed equally to this work.

  • Present address: Synthelabo Recherche (LERS), 31 Avenue Paul, Valliant-Counturier, 92220 Bagneux, France.

  • To whom correspondence should be addressed. E-mail: (P.I.H.B.) bastiaen{at}icrf.icnet.uk, and (P.J.P.) parkerp{at}icrf.icnet.uk

REFERENCES AND NOTES

View Abstract

Stay Connected to Science

Navigate This Article