Here, we report that Lck protects cells from glucocorticoid-induced apoptosis. In glucocorticoid-sensitive T cells, Lck was downregulated by dexamethasone to inhibit TCR activation and signaling. Because TCR activation antagonizes glucocorticoid-induced apoptosis,9–12
we reasoned that the inhibition of Lck would confer sensitivity to dexamethasone. We found that inhibition of Lck by RNA interference (RNAi) or by the small-molecule inhibitor dasatinib enhanced glucocorticoid-induced apoptosis in lymphoid cells, and particularly in primary CLL cells that were partially resistant to dexamethasone. CLL represents a clinically relevant model of lymphoid malignancy because synthetic glucocorticoids, such as prednisone and dexamethasone, are widely used in combination with other chemotherapeutic agents for treating aggressive or refractory CLL.
Previous studies have shown that glucocorticoids rapidly inhibit Lck by a nongenomic mechanism involving interactions between the ligand-bound GR and TCR signaling complex.22,23
In addition, it has been shown that dexamethasone redistributes Lck out of lipid rafts after T-cell activation, thereby attenuating its activity.19
Although these studies unequivocally show that glucocorticoids inhibit Lck and other Src family kinases by distinct mechanisms, this is the first report providing evidence that Lck transcript and protein levels are downregulated by dexamethasone in a GR-dependent manner.
This finding was initially discovered from microarray analysis of dexamethasone-treated cells. In primary thymocytes, Lck was among a subset of genes that were down-regulated by a signal Log2
ratio of >2.5 (that is, greater than sixfold repression). In addition, we show that Lck expression was downregulated at the protein level in mouse lymphoma lines WEHI7.2 and S49A.2, primary thymocytes, and the human T-ALL cell line CEMC7, which is also sensitive to glucocorticoid-induced apoptosis.42
However, Lck transcript levels were not reported to be differentially expressed in primary ALL cells treated with prednisolone43
or after in vivo
treatment with glucocorticoid-based monotherapy.44
Yet, a recent study by Mansha et al
., found that the Src-like adaptor protein (SLAP), a negative regulator of TCR signaling with significant homology to Lck,45
was upregulated by dexamethasone exclusively in glucocorticoid-sensitive ALL cell lines.46
Thus, SLAP may be upregulated in B- or T-ALL to circumvent lymphocyte activation or Lck activity. Moreover, it is likely that the regulation of Lck in lymphocytic leukemias is heterogeneous. For example, in this report, we observed that Lck expression was not downregulated by dexamethasone in CLL cells, but was modestly elevated.
Of particular interest were other genes that were down-regulated by dexamethasone that are part of the TCR signaling pathway. First, CD3ε
polypeptides were both downregulated in primary thymocytes. Although decreased expression of CD3 may contribute to glucocorticoid-mediated inhibition of TCR signaling, our RNAi experiments clearly show that the downregulation of Lck alone is sufficient to inhibit TCR-induced calcium oscillations. Second, MEK was downregulated by dexamethasone at the transcript level. Although we did not confirm whether glucocorticoids directly affect MEK levels, this result may provide an additional explanation for why dexamethasone and dasatinib have synergistic activity, given that dasatinib effectively inhibits MEK phosphorylation in T cells.33
Finally, we observed that multiple proteins that make up the TCR signaling pathway were downregulated by dexamethasone. In particular, Fyn and ZAP-70 levels were decreased 24 h after glucocorticoid treatment. Adaptor proteins LAT and SLP-76 were also downregulated by dexamethasone, although this effect was far more pronounced in the presence of dasatinib. These observations further support the concept that glucocorticoids strongly inhibit TCR signal transduction by negatively regulating multiple components of the pathway.
Our results suggest that the downregulation of Lck by dexamethasone does not directly mediate glucocorticoid-induced apoptosis in T cells. However, it is likely that the downregulation of Lck by dexamethasone contributes to cell death and apoptosis by blocking lymphocyte receptor signaling. Because it has been previously shown that MEK and ERK are both necessary and sufficient to inhibit glucocorticoid-induced apoptosis in immature T cells,11
we anticipate that Lck inhibition results in the loss of MEK and ERK activation, thereby increasing glucocorticoid sensitivity. The direct influence of MEK and ERK activation on glucocorticoid sensitivity has been documented in other studies,12,47
and supported in part by our result that MEK1/2 phosphorylation is virtually nondetectable in the presence of dexamethasone and dasatinib.
Although dasatinib as a single agent had some cytotoxic effect in CLL cells, the degree of cell killing was variable between patient samples and was relatively modest at a concentration of 100 nM. A recent study showed that the IC50
of dasatinib in 18 CLL samples ranged from 0.78 to 53.3 μ
Although CLL cells are relatively insensitive to dasatinib alone, it is likely that Lck may contribute, in part, to their survival. For example, CLL cells are partially characterized by increased BCR signaling events,35,36
including elevated levels of phosphotyrosine and cytosolic calcium, which are thought to facilitate cell survival.49,50
These events correlate with aberrant expression of Lck29
and constitutive expression of its B-cell homolog Lyn,36
both of which regulate BCR activity. Moreover, our results suggest that other Lck inhibitors (for example, PP2, Y*EEI peptide, and BIBF 1120) also have some degree of cytotoxic activity and all three molecules synergize with dexamethasone. The peptide sequence EGQY*EEIP was specifically designed to bind the SH2 domain of Lck37,38
and derivatives of these peptides have previously been shown to inhibit cell proliferation in 293T cells.51
BIBF 1120 specifically implicates the involvement of Lck because it is ~10-fold more selective for Lck than other Src kinases.41
The most profound cytotoxic responses in CLL cells were observed when dasatinib and dexamethasone were used in combination. Importantly, although glucocorticoids are commonly used to treat lymphoid malignancy, complete responses in CLL lag behind that of ALL and multiple myeloma.1,34,52,53
CLL cells were relatively insensitive to dexamethasone when treated ex vivo
, and their degree of resistance was positively correlated with Lck expression. In addition, CLL cells were resistant to dexamethasone-mediated downregulation of Lck, which was not due to defects in glucocorticoid uptake or aberrations in the GR. Thus, even in the presence of dexamethasone, total and phosphorylated Lck were elevated. Thus, we argue that the high level of Lck in CLL contributes to glucocorticoid resistance in these cells, as Src kinase inhibitors sensitize them to dexamethasone.
Collectively, our data indicate that Lck functions to antagonize glucocorticoid-induced apoptosis. Because inhibition of Lck sensitizes cells to the cytotoxic effects of dexamethasone, small-molecule inhibitors of Lck should be considered for treating glucocorticoid-resistant malignancies.