In Sect. 6.1, we described many examples of collaborative action between STAT1 and other transcription factors needed for the expression of certain genes in response to IFN-y. Similarly, it is likely that IFN-Y-activated STAT3 or STAT5 cooperates with other transcription factors. However, it is possible that the activation of additional transcription factors leads to the induction of genes without the help of any STAT. It remains to be discovered whether the IFN-Y-dependent, STAT1-independent pathways such as PI3K-dependent monocyte adhesion and induction of concentrative nucleoside transporters
(CNT) in macrophages are truly independent of any STAT (Navarro et al. 2003; Soler et al. 2003). Importantly, a novel type of GAS element with an IRF/ETS binding site has been described that functions independently of STAT1 (Con-tursi et al. 2000). In a murine macrophage-like cell line, PU.1 and IRF-8 bind to the GAS element present in the promoter of irf8 (Contursi et al. 2000; Kanno et al. 2005). DNA sequence motif comparison revealed that GAS elements can be divided into two subtypes. One is the classical AAA/TTT palindrome, which contains no IRF/ETS binding motif; examples from this group are PML-GAS and GBP-GAS. The other includes a novel IRF/ETS composite element (Kanno et al. 2005), and examples are GAS elements in the promoters of irf8, irf1, CCL2, TLR3, and cathepsinEpreprotein (Kanno et al. 2005). Notably, all of these genes are important for immune cell functions, particularly in macrophages and dendritic cells.
Other IRF family members can also mediate IFN-y-stimulated gene induction in a STAT-independent manner. IFN-y stimulates the expression of the polymeric immunoglobulin receptor (PIGR), which is expressed constitutively on the basolateral surfaces of secretory epithelial cells, where it directs polymeric IgA and pentameric IgM to exocrine secretions (Piskurich et al. 1997). There are three ISRE elements in the promoter of PIGR; the two upstream elements are bound to constitutive transcription factors and the third binds to IRF-1 upon IFN-y stimulation of a human epithelial cell line (Piskurich et al. 1997). Furthermore, constitutive expression of the exonuclease ISG20 depends on the constitutive transcription factors SP1 and USF-1. In contrast, IFN-y-stimulated expression of ISG20 depends on the binding of IRF-1 to the ISRE element in the promoter, which contains no functional GAS elements (Gongora et al. 2000). These two IFN-y-activated promoters are driven by the binding of activated IRF-1 to ISRE elements independently of any STAT.
It is becoming clear also that not all ISRE elements are equivalent. Sequence alignment of ISREs has revealed three subtypes thus far (Meraro et al. 2002). The classical ISRE can recruit only IRF dimers, in addition to ISGF3. In contrast, some ISRE subtypes harbor an ETS/IRF binding site named EIRE (ETS/ IRF response element; Meraro et al. 2002), different in composition from the ETS/IRF binding element, called EICE, described earlier by Brass et al. (1996). Of note, IRF-8 expression is high in myeloid and B cells, whereas IRF-4 is highly expressed in T and B cells. EICE binds to the immune cell-restricted factor PU.1, which forms a complex with either IRF-4 or IRF-8, and is therefore present in genes whose expression is restricted to immune cells (Kanno et al.
2005). IRF-4 and IRF-8 do not bind effectively to the ISRE element alone, but can do so only when in a complex with PU.1. In contrast, other members of the IRF family (IRF-1, 2, 3, and 7) bind to ISRE elements directly (Honda et al.
2006). Both EICE and EIRE have only one ETS binding site, but EIRE possesses two IRF binding sites, in contrast to EICE which has only one (Meraro et al. 2002). Because of this difference, promoters that contain EIRE can also become activated after binding to IRF dimers, whereas both types of promoters can become activated by IRF4 or IRF-8/PU.1 heterocomplexes (Meraro et al. 2002). In contrast, IRF heterocomplexes consisting of IRF-8 with either IRF-1 or IRF-2 can bind only to classical ISREs and EIREs and not to EICEs and have been suggested to function as repressors (Bovolenta et al. 1994; Sharf et al. 1995). Interestingly, classical ISRE-containing genes in macrophages, and possibly in other immune cells, can be repressed by IFN-y through protein-protein interaction between IRF-8 and another ETS family member, TEL, resulting in recruitment of the histone deacetylase HDAC3 (Kuwata et al. 2002). Examples of human genes with EIREs are ISG15, 6-16, 9-27, IP10, ISG54, and CCYBB (encoding GP91(PHOX)). However, other ways to activate EIRE-containing genes by IFN-y have also been described. For instance, the human IP10 and ISG54 genes possess only an EIRE and no GAS in their promoters and both can bind to STAT1 and IRF-9, probably as a STAT1 homodimer/IRF-9 complex (Majumber et al. 1998; Bluyssen et al. 1995). In addition, the mouse ip10 promoter can bind to STAT1 and IRF-1 in hepatocytes, but it is not clear what kind of complex is formed (Jaruga et al. 2004). Importantly, the ability of EIRE motifs to recruit not only IRFs but also PU.1/IRF heterocomplexes predicts that some of the genes harboring such elements might be regulated differentially by IFN-y in immune cells, which constitutively express IRF-4, IRF-8, and PU.1, as has been shown for the ISG15 gene in macrophages (Meraro et al. 2002) and the CYBB gene in myelomonocytic cells, the latter gene showing cooperation with PU.1/ IRF-1/IRF-8/CBP (Eklund et al. 1995).
A strong increase in IRF-8 in response to IFN-y was observed originally in macrophages (Politis et al. 1994), probably because activated PU.1 and IRF-8 binding to the IRF/ETS composite GAS element present in the irf8 promoter (Contursi et al. 2000; Kanno et al. 2005). Because of the increase in IRF-8 expression, a model was suggested in which IRF-8 and PU.1 play a role in amplifying the expression of genes containing the IRF/ETS composite GAS element by generating a second wave of transcription (Kanno et al. 2005). In addition, it is likely also that the expression of EICE- or EIRE-containing genes is amplified in a macrophage-specific manner through a second wave of transcription. Although it was believed that IRF-8 is only expressed in certain immune cells, it has been shown recently that IFN-y-stimulated primary colon carcinoma cells also express IRF-8, probably because activated STAT1 binds to the composite GAS element in the IRF8 promoter (Liu et al. 2003). Similarly, an increase in IRF-1 levels after stimulation by IFN-y in nonimmune cells is a result of the binding of STAT1 to the IRF/ETS composite GAS element in the IRF1 promoter. Therefore, since nonimmune cells do not express PU.1, an increase in IRF-8 may have a quite different effect on classical ISRE or EIRE-containing genes, because complexes formed between IRF-8 and IRF-1 or IRF-2 might function as repressors (Bovolenta et al. 1994; Sharf et al. 1995). However, it cannot be excluded that a concomitant increase in IRF-8 and IRF-1 in immune cells might also lead to the repression of certain genes. However, when IRF-8 levels are not too highly induced in non-immune cells, an increase in IRF-1 homodimers will certainly lead to increased expression of genes containing an ISRE or EIRE element. Moreover, activated STATs may collaborate with IRF-1 to induce the expression of genes that harbor both a classical GAS and an ISRE element. Summarizing these results, we propose that the availability of STATs, IRF, or ETS family members, constitutively present or induced, will determine how a certain cell type responds to IFN-y, and that the balance among these transcription factors will determine which subtype of ISRE- or GAS-containing genes will be turned on in the first and second phases of transcriptional activation.
CIITA and IRF-9 are examples of transcription factors whose expression is induced by IFN-y and which therefore play an important role in the second wave of the IFN-y response. A novel IFN-responsive cis-acting enhancer element, y-IFN-activated transcriptional element (GATE), distinct from GAS and ISRE, but partly homologous to ISRE, has been identified in the promoter of the irf9 gene (Weihua et al. 1997). Two transcription factors bind to GATE, GBF1, and GBF2, after they are both synthesized de novo in response to IFN-y. GBF2 was identified subsequently as C-EBP-P. Its induction by IFN-y is JAK1-and STAT1-dependent and is probably mediated by the binding of STAT1 to a putative GAS element in the promoter of the cebpb gene (Roy et al. 2000). IFN-y activates C-EBP-P expression by activating the MEKK1/MEK1 or MEK2/ERK1 or ERK2 cascade, resulting in the phosphorylation of C-EBP-P at threonine 294 in the consensus ERK phosphorylation site of the regulatory domain RD2 (Hu et al. 2001). Furthermore, MLK activation by IFN-y leads to the dephosphory-lation of C-EBP-P at serine residue 64 in the transcription activation domain, permitting recruitment of transcriptional co-activators such as p300 (Roy et al. 2005). To date, it is not known whether the MLK-driven dephosphorylation is caused by enhanced phosphatase activity or by inactivation of the kinase that normally phosphorylates serine 64 constitutively (Kalvakolanu et al. 2005). GBF-1 is a novel transcription factor and the gene is located at human chromosome 9q34.13 (Hu et al. 2002). It is not clear which signals from the activated IFNGR induce GBF-1 expression. GBF-1 possesses glutaredoxin-like, PTP-like, RNA Pol II-like, and ribonucleoside diphosphate reductase-like domains, but its precise mechanism of action is not yet known (Hu et al. 2002). The fact that GBF-1 does not bind to monomeric GATE, but does bind to multimeric GATE in a DNA screen, shows that it has very weak DNA binding activity by itself.
However, similarly to IRF-8, which does not bind to DNA alone, GBF-1 possesses strong transactivating activity (Hu et al. 2002). Recent data indicate that C-EBP-P interacts with GBF1 after phosphorylation of C-EBP-P at threonine 294 (Meng et al. 2005). Just as IFN-P induces IRF-9, the induction of IRF-9 by IFN-y seems to be independent of activated STAT1, but dependent on unphos-phorylated STAT1 through an as-yet-unidentified mechanism (Rani et al. 2005). Unphosphorylated STAT1 and STAT3, upregulated in response to IFN-Y or GP130-linked cytokines, respectively, induce sets of genes distinct from those that respond to phosphorylated STAT1 and STAT3 (Chatterjee-Kishore et al. 2000; Yang et al. 2005).
In summary, we describe several IFN-Y-stimulated signaling pathways that involve the activation of different IRF and ETS family members and that function either independently of any STAT or collaborate with STAT1 for the induction of certain genes. These signaling pathways play a very important role in the response of immune cells to IFN-y. It is not yet known which transcription factors collaborate with activated STAT3 or STAT5 in specific cell types in response to IFN-y. We expect that many more IFN-Y-activated transcription factors will be discovered that either collaborate with STATs or induce gene expression totally independently of any STAT.
Was this article helpful?
Parenting is a challenging task. As a single parent, how can you juggle work, parenting, and possibly college studies single handedly and still manage to be an ideal parent for your child? Read the 65-page eBook Single Parenting Becoming The Best Parent For Your Child to find out how. Loaded with tips, it can inspire, empower, and instruct you to successfully face the challenges of parenthood.