Host-cell factor 1 (HCF-1) is an unusual transcriptional regulator that undergoes a process of proteolytic maturation to generate N-(HCF-1 N ) and C-(HCF-1 C ) terminal subunits noncovalently associated via self-association sequence elements. Here, we present the crystal structure of the self-association sequence 1 (SAS1) including the adjacent C-terminal HCF-1 nuclear localization signal (NLS). SAS1 elements from each of the HCF-1 N and HCF-1 C subunits form an interdigitated fibronectin type 3 (Fn3) tandem repeat structure. We show that the C-terminal NLS recruited by the interdigitated SAS1 structure is required for effective formation of a transcriptional regulatory complex: the herpes simplex virus VP16-induced complex. Thus, HCF-1 N -HCF-1 C association via an integrated Fn3 structure permits an NLS to facilitate formation of a transcriptional regulatory complex.crystallography | chromatin | nuclear localization sequence H ost cell factor-1 (HCF-1; also known as HCFC1) is a metazoan transcriptional regulator. HCF-1 was initially identified as a human coactivator for immediate-early gene expression of herpes simplex virus (HSV) by forming a complex (VP16-induced complex, VIC) with the HSV protein VP16 and the cellular transcriptional regulator Oct-1 (1). HCF-1 regulates cell-cycle progression by mediating associations between DNA-binding transcription factors and chromatin modifying complexes (2-12).Many HCF-1 proteins, including all known vertebrate HCF-1s, undergo a process of proteolytic maturation. Thus, human HCF-1 is synthesized as a 2035-aa precursor, which is cleaved and modified by O-GlcNAC transferase (13) to generate HCF-1 N and HCF-1 C subunits possessing different roles in, for example, cell-cycle regulation (14). After cleavage, the two resulting HCF-1 N and HCF-1 C subunits remain noncovalently associated via two "selfassociation sequences" called SAS1 and SAS2: SAS1 is the primary association element and its sequence is conserved in HCF-1, whereas SAS2 is secondary, less well conserved, and not considered in this study (15).SAS1 is composed of a short 43-aa HCF-1 N segment called SAS1N and a 197-aa HCF-1 C segment called SAS1C (Fig. 1A). Wilson et al. (15) suggested that SAS1C has two tandem fibronectin type 3 (Fn3) repeat structures forming a binding site for the 43-aa SAS1N to promote HCF-1 N -HCF-1 C association.To understand the mechanism of HCF-1 self-association, we have determined the crystal structure of SAS1 with the neighboring nuclear localization signal (called here SAS1-NLS). Contrary to expectation, SAS1N and SAS1C together-not SAS1C aloneform an integrated tandem Fn3 structure. Furthermore, we show that the self-association tethers the basic NLS to the HCF-1 N Kelch domain to facilitate VP16-induced complex formation.
ResultsCrystal Structure of the HCF-1 SAS1 Self-Association Domain. The SAS1-NLS crystal structure (Fig. 1B) was determined to 2.7 Å resolution using the Se single-wavelength anomalous dispersion (SAD) method (Table S1 and Fig. S1). Fig. S2 details schematically...