IFI16 Is an Essential Mediator of Growth Inhibition, but not Differentiation, Induced by the Leukemia Inhibitory Factor/JAK/STAT Pathway in Medullary Thyroid Carcinoma Cells*

of in the human medullary thyroid carcinoma (MTC) cell line, TT, induces growth arrest and differentiation via two parallel, yet independent pathways. One of these pathways is intracellular and the other is a cell-extrinsic, autocrine/paracrine pathway mediated by the leukemia inhibitory factor (LIF)/JAK/STAT pathway. Here, we show that IFI16 is a necessary and sufficient downstream effector for LIF effects in MTC cells, specifically required for the LIF/JAK/STAT pathway-induced growth inhibition in these cells. IFI16 was induced by Raf or LIF. Dominant negative STAT3 could block the induction, indicating that Raf can induce IFI16 only via the cell-extrinsic pathway. Knock-down of IFI16 using siRNA abrogated LIF-induced changes in cellular levels of E2F1, cyclin D1, and p21 WAF / CIP1 , and cell cycle arrest. In addition, adenovirus-mediated overexpression of IFI16 was sufficient to induce growth arrest. In contrast to its essential role for LIF-mediated growth arrest, IFI16 was not required for differentiation induced by LIF. Knock-down of IFI16 could not block changes in differentiation markers of the MTC cells, including calcitonin, RET, and cell morphology. Our study identifies IFI16 as an essential growth-specific effector of the cell-extrinsic growth inhibitory pathway of Ras/Raf signaling in MTC cells. downstream components of the LIF/JAK/STAT-mediated growth inhibition and differentiation signaling to characterize the different responses of MTC and SCLC cells to LIF/JAK/STAT activation. This study was performed to dissect the mechanism of the Ras/Raf-induced cell-extrinsic signaling pathways in MTC cells. Using a cDNA microarray, we have identified IFI16, an interferon-inducible transcriptional repressor, as a downstream target of the LIF/JAK/STAT pathway. We show that IFI16 is essential and sufficient for the LIF/JAK/STAT-induced growth arrest in MTC cells, but is not required for differentiation, which is also induced by the LIF/JAK/STAT pathway along with growth arrest. Thus, IFI16 appears to be specific for cell cycle regulation mediated by the cell-extrinsic LIF/JAK/STAT pathway, and separates the regulation of growth arrest from that of differentiation. We also show that, unlike in MTC cells, IFI16 cannot be induced by LIF in the SCLC cell line DMS53. However, IFI16 can induce growth arrest in DMS53 cells, suggesting that IFI16 may be a missing component for the LIF-induced growth inhibitory signaling in some SCLC cells. Our study defines IFI16 as an essential growth specific effector of the LIF/JAK/STAT-mediated cell-extrinsic pathway of Ras/Raf. IFI16 LIF/JAK/STAT


Ras/Raf
Autocrine/Paracrine PathwayPreviously, we showed that Ras/Raf-induced growth arrest and differentiation in the human MTC cell line, TT, is mediated via two parallel, yet independent pathways (18).
One is intracellular and the other is the LIF/JAK/STAT-mediated autocrine/paracrine pathway. These two pathways induce growth arrest accompanied by a differentiation program characterized by changes in morphology, expression of the neuroendocrine markers calcitonin and calcitonin-gene-related peptide (CGRP), and downregulation of RET, the proto-oncogene responsible for MTC development in the MEN 2 syndromes (17,27).
To further elucidate the mechanism of the growth inhibition mediated by Ras/Raf-induced cell-intrinsic or extrinsic signaling, we sought to identify downstream effectors of these parallel pathways. For this, we examined changes in gene expression induced by these signaling pathways using microarray analysis techniques. Information on genes regulated by the cell-extrinsic growth inhibitory pathway was collected by examining total RNA extracted from TT cells treated with LIF, in the absence or presence of a dominant negative STAT3, and subsequently selecting genes induced in a STAT3-dependent manner. We have shown that STAT3 is essential for the LIFmediated growth inhibitory response in TT cells (18). We then examined gene expression upon Raf activation, which represents both intracellular and the autocrine/paracrine pathways. This result was then compared with the LIF-regulated gene expression profile. In this way, we could identify genes specifically regulated either by cell-intrinsic signaling or by the LIF/JAK/STATmediated extrinsic pathway.
We were particularly interested in the genes regulated by LIF in a STAT3-dependent manner, since they represent potential downstream targets specific for the cell-extrinsic pathway.
Among these STAT3-induced genes in TT cells, IFI16, which has been shown to inhibit growth in other cell types (28), was a fascinating candidate to mediate LIF-induced growth arrest. We further analyzed the LIF/JAK/STAT-dependent induction of IFI16 by RT-PCR and Western blotting. Indeed, increase in mRNA level of IFI16 could be detected upon Raf activation or LIF treatment, while the induction was blocked in cells infected with a dominant-negative STAT3 adenovirus ( Fig. 1A and B). A similar pattern of STAT3-dependent induction was also observed at the protein level (Fig. 1C). Since STAT3 is activated and utilized only by the cell-extrinsic pathway, these data indicate that Raf induces IFI16 via the LIF/JAK/STAT-mediated cell extrinsic pathway, but not via intracellularly mediated signaling pathways. contrary to its sufficiency for cell growth arrest, IFI16 did not induce RET downregulation, calcitonin/CGRP, and morphological changes in TT cells (Fig. 5, A to C).

IFI16 Is Necessary and Sufficient for Growth Inhibition, but Is Not Required for
Taken together, these data show that IFI16 is necessary and sufficient to mediate growth arrest induced by the LIF/JAK/STAT pathway, but it is not required for differentiation that is also induced by the LIF/JAK/STAT pathway in TT cells.  (Fig. 6). Nevertheless, IFI16 expression could be induced by the known inducers of IFI16, interferon-α or γ (Fig. 6), indicating that the endogenous IFI16 gene is competent to be expressed in DMS53 cells. Data (mean ± standard error) are from a representative experiment performed in triplicate.

IFI16 Is Not Induced in the SCLC
Experiments were repeated at least 3 times with similar results.