Molecular Cancer (Mar 2012)

A STAT3-inhibitory hairpin decoy oligodeoxynucleotide discriminates between STAT1 and STAT3 and induces death in a human colon carcinoma cell line

  • Souissi Inès,
  • Ladam Patrick,
  • Cognet Jean AH,
  • Le Coquil Stéphanie,
  • Varin-Blank Nadine,
  • Baran-Marszak Fanny,
  • Metelev Valeri,
  • Fagard Remi

DOI
https://doi.org/10.1186/1476-4598-11-12
Journal volume & issue
Vol. 11, no. 1
p. 12

Abstract

Read online

Abstract Background The Signal Transducer and Activator of Transcription 3 (STAT3) is activated in tumor cells, and STAT3-inhibitors are able to induce the death of those cells. Decoy oligodeoxynucleotides (dODNs), which bind to the DNA Binding Domain (DBD) of STAT3, are efficient inhibitors. However, they also inhibit STAT1, whose activity is essential not only to resistance to pathogens, but also to cell growth inhibition and programmed cell death processes. The aim of this study was to design STAT3-specific dODNs which do not affect STAT1-mediated processes. Results New dODNs with a hairpin (hpdODNs) were designed. Modifications were introduced, based on the comparison of STAT3- and STAT1-DBD interactions with DNA using 3D structural analyses. The designed hpdODNs were tested for their ability to inhibit STAT3 but not STAT1 by determining: i) cell death in the active STAT3-dependent SW480 colon carcinoma cell line, ii) absence of inhibition of interferon (IFN) γ-dependent cell death, iii) expression of STAT1 targets, and iv) nuclear location of STAT3 and STAT1. One hpdODN was found to efficiently induce the death of SW480 cells without interfering with IFNγ-activated STAT1. This hpdODN was found in a complex with STAT3 but not with STAT1 using an original in-cell pull-down assay; this hpdODN also did not inhibit IFNγ-induced STAT1 phosphorylation, nor did it inhibit the expression of the STAT1-target IRF1. Furthermore, it prevented the nuclear transfer of STAT3 but not that of IFNγ-activated STAT1. Conclusions Comparative analyses at the atomic level revealed slight differences in STAT3 and STAT1 DBDs' interaction with their DNA target. These were sufficient to design a new discriminating hpdODN that inhibits STAT3 and not STAT1, thereby inducing tumor cell death without interfering with STAT1-dependent processes. Preferential interaction with STAT3 depends on oligodeoxynucleotide sequence modifications but might also result from DNA shape changes, known to modulate protein/DNA interactions. The finding of a STAT3-specific hpdODN establishes the first rational basis for designing STAT3 DBD-specific inhibitors.

Keywords