In vitro investigation of the interactions of nuclear factor one proteins with transcriptional regulators PIN1 and LMO4

Nuclear factor one (NFI) family of transcription factors regulates gene expression during embryonic development and in the adult stem cell niches. NFI family comprises four members in vertebrates: NFIA, NFIB, NFIC and NFIX. NFIs bind to the consensus sequence, TTGGC(N5)GCCAA on double-stranded DNA as homodimers or heterodimers. Members of NFI family contain an N-terminal DNA binding and dimerization domain and a C-terminal transcriptional activation and repression domain. The N-terminal domain is highly conserved among members and includes a MH1 motif also found in SMAD proteins of the TGF- signaling pathway. Proline rich C-terminal domain is less conserved and may be involved in differential protein-protein interactions between NFIs and nuclear proteins. Alternative splicing generates multiple isoforms of each member. Expression of NFIA, NFIB and NFIX is required for normal development of central nervous system, and loss of each one leads to severe defects in brain development. The fact that NFIA, NFIB and NFIX knock-out mice display distinct phenotypes suggests that each NFI has a distinct function in the brain. Interactions with different regulatory proteins may account for this specificity. Several proteins have been found to interact with NFI family members, including the transcription factors FOXA1, Sox9, Olig2. However, proteins that bind NFI and regulate NFI function in the developing brain have not been studied well. In order to identify proteins that may interact with NFIB in the developing central nervous system, yeast two-hybrid (Y2H) screen of a human fetal brain cDNA library was previously performed. PIN1 and LMO4 were identified as putative NFIB binding partners. PIN1 belongs to the peptidyl-prolyl isomerase family which catalyzes cis-trans isomerization of peptidyl prolyl bonds. PIN1 recognizes phosphorylated serine or threonine residues in pSer/Thr-Pro peptide sequences on target proteins. Cis-trans isomerization by PIN1 changes conformation of target proteins and affects their function. LMO4 belongs to the LIM-Only (LMO) subfamily of LIM proteins and contains two tandem LIM domains. LMO4 binds a variety of proteins via its LIM domain and acts as an adaptor that links proteins, including transcription factors, to form multiprotein complexes. In the developing brain, PIN1 and LMO4 are expressed in the right location to interact with NFIs: PIN1 is highly expressed within the ventricular zone and the subventricular zone in the cortex. LMO4 is also expressed within both ventricular zone and subventricular zone, the neurogenic regions of the cortex. Moreover, phenotypes of PIN1 and LMO4 null mice show that these proteins are involved in neural development: In the absence of PIN1 development of late-born neurons is impaired in the cerebral cortex. Absence of LMO4 leads to neural tube closure defects in the dorsal brain and defects in cortical neurogenesis affecting both early and late-born neurons. xxiv In this study, we investigated interactions between NFIs and two putative NFIB binding partners, PIN1 and LMO4. To this end, epitope tagged PIN1 or LMO4 and NFIs were co-expressed and binding was assayed via co-immunoprecipitation experiments. We found that each NFI member and PIN1 could be co-immunoprecipitated with each other. GST pull-down assays confirmed the interactions detected in co-immunoprecipitation experiments, bacterially expressed GST-PIN1 pulled down each NFI members. To identify NFI domains required for interaction with PIN1, we created NFIB mutants that contain only the N-terminal DNA binding and dimerization or the C-terminal transactivation domains. GST pull-down experiments with these mutants showed that NFIB C-terminal domain is not involved in PIN1 interactions while the N-terminal domain is required for PIN1 binding. Site-directed mutagenesis of conserved serine residues in pSer/Thr-Pro motifs in the NFIB C-terminal domain did not interfere with NFIB-PIN1 binding, in either co-immunprecipitation or GST pull down assays consistent with the finding that NFI N-terminal domain is necessary for NFI-PIN1 interactions. On the other hand, preliminary data from co-immunoprecipitation and GST pull down experiments indicate that only NFIX interacts with LMO4. We tested the NFIB mutant that lacks the C-terminal domain for LMO4 binding via co-immunoprecipitation assays and found that it did not interact with LMO4. As NFI members differ in their ability to bind LMO4, these data suggest that the C-terminal domain of NFIX is involved in LMO4 interactions. Further site-directed mutagenesis of NFIB and NFIX can be used to pinpoint the exact regions required for PIN1 and LMO4 binding.