GATA3 is a zinc finger transcription factor essential for the proper development of various tissues and organs, especially in the hematopoietic system. GATA3 has been recently shown to be involved in the development of hair follicles and in the determination of skin cell lineage in mice [14
]. In human skin, GATA3 is expressed in suprabasal layers of the epidermis [19
] and is also regulated during the differentiation of cultured keratinocytes: GATA3 accumulates during keratinocyte differentiation induced by addition of calcium [19
]. GATA3 is therefore strongly expressed in differentiated cultured keratinocytes, which were the cell models used in our previous transcription study [5
]. GATA3 plays a key role in the transition between proliferation and differentiation in human epidermal cells through the p63/IKKα/SMAD pathway [19
], but its function in the control of the cellular response to genotoxic stress has never been explored.
To monitor the function of GATA3 in the low-dose IR response, we set up a cell model of primary keratinocytes with stable knock-down of GATA3. We decided to study cultured keratinocytes in a differentiated state mimicking the suprabasal layer of human epidermis, since these cells are the first to be exposed in therapeutic or accidental irradiation. GATA3 was knocked down by a lentiviral-mediated shRNA expression. As previously reported by others [21
], lentiviral vectors appear to be the most suitable for long-term transgene expression, especially in quiescent cells such as stem cells or confluent keratinocytes, whereas chemical methods or electroporation are completely inefficient. We were able to infect more than 70% of our cultured cells and obtained a marked decrease in GATA3 protein level, which became almost undetectable on western-blot, in infected cells. What are the biological consequences of this knock-down? In a normal context, keratinocytes infected by the shGATA3 vector exhibit a higher rate of proliferation (J. Lamartine. In preparation). This effect was clearly visible in our colony-forming assay, where shGATA3 cells were found to be more clonogenic than shSCR cells, whatever the radiation dose. It can therefore be postulated that this long-term effect on proliferation conceals shGATA3 cell hyper-sensitivity to certain doses of radiation. Indeed, the XTT assay, performed 72 h after exposure, revealed specific shGATA3 cell radiosensitivity to the 1 cGy dose, with a 16% decrease in cell viability compared to shSCR cells. After the 2 Gy dose, both cell lines exhibited the same viability. These results indicate that silencing GATA3 modifies the cellular response of irradiated cells, specifically to the low 1 cGy dose, during the first hours following irradiation. Specific radiosensitivity to low IR doses has been previously described for human keratinocytes [22
]. Low-dose hyper-radiosensitivity (HRS) is an effect in which cells die from excessive sensitivity to low doses (< 0.5 Gy) while becoming resistant to higher doses. The suggested mechanism for HRS is related to the absence at low doses of the inducible DNA repair mechanism observed at higher doses [24
]. Our data indicate that GATA3 knock-down caused increased cell death after 1 cGy IR compared to shSCR cells, highlighting the important role of this protein in the 1 cGy response during the first 72 hours following irradiation. To more closely delineate the relative role of low-dose HRS in this increased cell death, additional studies using a larger range of dose between 1 cGy and 0.5 Gy will be useful.
To further test the role of GATA3, we compared the transcriptional response of shGATA3 and shSCR cells after 1 cGy IR exposure: a modified gene response profile was observed, with a burst of IR-responding genes in shGATA3 cells at 48 h. This time-point of 48 h seems to be a key moment in the transcriptional response. Our previous microarray studies found that the response to 1 cGy, contrary to the classical bimodal response after higher doses, was characterized by almost complete absence of transcriptional changes at early time points, followed by a large modulation at 48 h post-irradiation [5
]. It can be postulated than GATA3 is involved in the control of this 48 h gene response.
Many of the 266 genes responding at 48 h in shGATA3 cells were involved in fundamental mechanisms known to participate in the cellular response to genotoxic stress. Such is the case of EGR1, the gene showing the strongest induction ratio (> 4-fold). EGR1 is a transcription factor strongly activated by a broad spectrum of radiation, and promoting apoptosis and growth arrest through its targets, especially some members of the p53 families and via activation of the EGFR/ERK1/2 pathway [25
]. The DUSP1 gene, encoding a dual-specific threonine and tyrosine phosphatase, is also strongly induced in shGATA3 cells. DUSP1 is controlled by p53 during the cellular response to genotoxic stress and is a potent inhibitor of MAPK activity through dephosphorylation of MAPK [26
]. The functional annotation of IR-responding genes in shGATA3 cells (Table ) revealed several biological processes known to participate in the cellular response to stress: this is the case with TGFβ signalling, which concerns 7 of the 174 genes induced at 48 h in shGATA3 cells (TGIF, TSC22D1, TGFBR1, INHBE, SMAD4, ID1, E2F5). TGFβ1 is involved in the initiation of keratinocyte differentiation, by blocking their proliferation. This cytokine is also involved in wound healing, by inducing cell migration and keratinocyte matrix secretion [27
]. Moreover, TGFβ1 is induced in skin within hours following acute irradiation [28
] and plays a complex role in regulating the canonical cellular DNA damage response through ATM activation [29
]. A possible link between TGFβ1 expression and regulation by GATA proteins has been proposed in hematopoietic cell systems, but the involvement of GATA3 in the regulation of the TGFβ pathways has never been described in keratinocytes. Our functional annotation of IR-regulated genes in shGATA3 cells also revealed enrichment in genes involved in chaperone activity and protein folding. Molecular chaperones prevent protein aggregation and keep proteins in a state suitable for either refolding or degradation after a proteolytic stress such as ionizing radiation [30
]. The present study showed that GATA3 knock-down led to excessive sensitivity to low IR doses. It is possible that, in these cells, the fraction of proteins in an unfolded state is increased after irradiation, inducing the chaperone response.
The central question raised by our study is the link between GATA3 and the genes that are deregulated in shGATA3 cells. To answer this question, we looked for GATA consensus sites in the promoter region of genes that are induced or repressed after irradiation in shGATA3 cells, as compared with a random sample of non-responding genes. We did not observe any enrichment in the fraction of genes that are up-regulated: these genes are probably not direct targets of GATA3. Their up-regulation appears to be secondary to knock-down, or a specific response of cells sensitized to low-dose radiation. On the other hand, the frequency of GATA3 binding sites increased in the fraction of down-regulated genes (Table ). Some of these genes, such as PPIL2 and GRCA, which we have shown to be bound in vivo by GATA3 (figure ), might be direct targets of GATA3. Nevertheless, further exploration of the GATA3 targets in the low-dose IR response will be necessary to clarify this point. The recent development of methods allowing genome-wide identification of transcription factor targets, such as ChIP chip [31
] or ChIP seq [32
], will offer the possibility of achieving this goal.