Technical Comments

Response to Comment on "PDK1 Nucleates T Cell Receptor-Induced Signaling Complex for NF-κB Activation"

See allHide authors and affiliations

Science  07 Apr 2006:
Vol. 312, Issue 5770, pp. 55
DOI: 10.1126/science.1115362

Abstract

In their comment, Gruber et al. report constitutive phosphorylation of protein kinase C θ at threonine residue 538. However, they fail to note that, consistent with our results, a number of other groups have previously reported inducible phosphorylation of Thr-538 in T cells. Although the physiological relevance of this discrepancy is unclear, substantial differences in experimental conditions and reagents may account for the conflicting observations.

Gruber et al. (1) report that they observe constitutive phosphorylation of protein kinase C θ (PKCθ) at threonine residue 538 (Thr-538). However, our report (2) is not the first to demonstrate inducible phosphorylation of PKCθ on Thr-538 in T cells (35). Gruber et al. have also reported constitutive phosphorylation in previous publications (6, 7). Clearly, investigators are divided on whether Thr-538 phosphorylation is inducible or constitutive, with the caveat that, to date, reports of constitutive phosphorylation have been based mostly on overexpression experiments. Therefore, although the comment by Gruber et al. finally showing an example in which endogenous PKCθ appears to be constitutively phosphorylated at Thr-538 in T cells is interesting, the usefulness of this data as it applies to our findings is limited because their experiments are not technically comparable to our own.

Gruber et al. (1) describe an experiment in which they transiently transfect Jurkat T cells with PKCθ-expressing constructs and examine the status of phosphorylation of the expressed protein [figure 2 in (1)]. However, we fail to see how such an experiment might work. It is well known that transient transfection of Jurkat cells is an inefficient process and that only a fraction of cells can be transfected. Therefore, it seems that the signal from the endogenous PKCθ should overwhelm that of transfected PKCθ unless, on a per cell basis, PKCθ was grossly overexpressed. Likewise, if transfection is to be used, it is necessary to assess activation in transfected cells rather than observing total extracellular signal–regulated kinase 1/2 (ERK1/2) phosphorylation as was done in their experiment. It is also important to point out that in the experiment shown in figure 1 in (1), where Gruber et al. claim to observe basal phosphorylation from primary CD3+ mouse cells, their experimental protocol differs substantially from ours. T cell receptor (TCR) stimulation in our experiments was performed using soluble cross-linked antibodies rather than plate- or bead-bound antibody, a difference that may well affect TCR signaling through PKCθ (8). Our analysis in primary T cells was performed using CD4+ T cells, whereas Gruber et al. used whole CD3+ mouse splenocytes obtained by negative selection. In addition, it is unclear whether they ensured that the isolated T cells were in a quiescent state by controlling the quality and quantity of serum in the medium. We argue that abruptly altering culture conditions from 10% to 0.5% serum is not an appropriate method of addressing this issue.

Although the experiments provided by Gruber et al. are difficult to compare directly with ours, we are perplexed by their observations regarding Thr-538 phosphorylation in unstimulated cells. In the experiments that we published (2), PKCθ was immunoprecipitated before immunoblotting, raising the somewhat unlikely possibility that there might have been selective loss of phosphate from the PKCθ from unstimulated cells during the time required for immunoprecipitation. We have therefore repeated the analysis with straightforward immunoblotting of extracts from unstimulated and stimulated Jurkat cells. As shown in Fig. 1, extracts from unstimulated cells do not show phosphorylation of PKCθ on Thr-538, whereas stimulation leads to a striking increase in phosphorylation. Furthermore, we have consistently obtained similar results using pT538 antibodies from two different vendors and in multiple Jurkat cell stocks. We are currently in the process of generating mouse knock-ins of phosphorylationsite mutants of PKCθ—experiments that we believe will unequivocally establish the importance of phosphorylation in regulation of PKCθ.

Fig. 1.

Jurkat cells were stimulated as described previously and after centrifugation were lysed directly with radioimmunoprecipitation assay (RIPA) buffer (containing protease and phosphatase inhibitors) for 15 min on ice. After addition of SDS-loading buffer, equal amounts of the samples were fractionated on SDS–polyacrylamide gel electrophoresis, followed by transfer to polyvinylidene fluoride (PVDF) membranes. Immunoblotting of the membrane was carried out using standard protocols (BD Pharmingen).

So, what might be the basis of the difference between our results and those reported by Gruber et al.? One factor to which we pay particularly close attention is the basal state of activation of cells. It has been our experience that Jurkat cells in particular can be easily stimulated by different batches of serum. We therefore characterize their state of activation by following the degree of basal nuclear factor κB (NF-κB) activity, as measured by electrophoretic mobility shift assay or luciferase reporter assays. Demonstrating inducible ERK phosphorylation does not necessarily ensure that Jurkat cells are not abnormally activated with regard to PKCθ-dependent signaling pathways. It might be helpful for the authors to examine the state of activation in their unstimulated cells (e.g., NF-κB activity, PI-3 kinase activity) before stimulating the cells with antibodies to CD3 and CD28. Nevertheless, despite the difficulties in comparing our data to that generated using disparate experimental approaches, there is ample evidence, both here and elsewhere in the literature, to suggest that the sequence and importance of individual PKCθ phosphorylation events does require further clarification. Therefore, in our own research, we continue to investigate this issue by generating knock-in mouse strains, as well as through accurate repetition of the experiments of Gruber et al. and others to determine both the origin and the importance of the discrepancies discussed herein.

References and Notes

View Abstract

Navigate This Article