Fig 1.
Domain structure of rat hnRNP U.
(A) Domain diagram of rat hnRNP U used in this study. SAP (amino acids 8–42), NTP-binding (amino acids 478–485), and RG (amino acids 676–767) domains are indicated as colored boxes on full-length rat hnRNP U (amino acids 1–798). The RG domain is newly defined in this report whose amino acid sequence is shown in the box below. It contains a cluster of 16 RG dipeptides interrupted by the central glycine-rich spacer (GRS). The sequence is folded like a hairpin to highlight the symmetrical distribution of RGs. (B) Sequence alignment of the RG domain which is highly conserved in higher eukaryotes. RGG-box is originally described as a conserved motif in RNA binding proteins [10]. Glycine-rich spacer (GRS) is marked by a black line.
Fig 2.
DNA-binding activities of domain deletion mutants.
(A) Illustration of wild type (WT) and deletion mutants. The orange and brown boxes at the N-terminus represent His-Myc and GST tags, respectively. (B) Gel images from competitive magnesium-agarose EMSA are shown with band density plots. Concentrations for WT, ΔSAP, ΔRG, NC, and NCΔRG were varied. Error bars indicate mean ± standard deviation (SD) from three independent experiments. Half-maximal concentrations (C50) for WT, ΔSAP, and NC are indicated by arrowheads.
Fig 3.
Comparison of DNA-binding activities of SAP and RG domains by on-bead DNA binding assay.
(A) On-bead binding assay of SAP domain (amino acids 1–44). Fixed volume (2 μl) of glutathione-Sepharose beads coupled with increasing amounts of GST-SAP protein were incubated with 100 ng (37.3 fmol) of pBS/FTZ. Bead-bound DNA was analyzed by agarose gel electrophoresis. The quantified band densities relative to those of input are plotted against the amounts of GST-SAP on beads. Error bars indicate mean ± SD from at three independent experiments. In drawing the regression curve, a nonlinear curve fitting program (Graphpad prism) was applied to mean values. (B) On-bead binding assay of RG domain (amino acids 676–767). The same procedure as above was employed. Error bars indicate as described above. (C) On-bead binding assay of RG domain with or without excess amounts of E. coli DNA against 50 ng of pBS/FTZ (10, 33, and 100-folds). GST-RG (56 pmol) on 7.5 μl beads was used for these experiments. The GC content of pBS and E. coli DNA fragments used as competitor was 50.4% and 50.8%, respectively. Error bars indicate mean ± SD from three independent experiments.
Fig 4.
Effects of RNA on S/MAR-selective binding of hnRNP U.
Aggregation assays using 20 ng of pBS/FTZ incubated with varying dose of recombinant hnRNP U proteins in the presence or absence of RNA. After the binding reaction, reaction mixture was centrifuged to separate into pellet (aggregate) and supernatant (soluble) fractions. Gel images and densitometric plots are shown in pairs. (A) Increasing concentrations of hnRNP U (WT) alone (left panels). The same experiment was done at 30 nM (right panels) of WT with or without excess amounts of total RNA (indicated by ‘fold’ against pBS/FTZ DNA). Bars plotted are average of 3 experiments with standard deviation. The decrease in aggregated DNA at 1x fold RNA was statistically significant by Student’s t test (marked by asterisk, p<0.01). (B) Increasing concentrations of N659 alone (left panels). The same experiment was done at 200 nM of N659 with or without excess amounts of total RNA (right panels). Error bars indicate mean ± SD from three independent experiments. No statistical significance in the reduction of aggregated DNA was detected at 1x RNA (p<0.05).
Fig 5.
Effects of synthetic homoribopolymers on S/MAR-selective binding of hnRNP U.
Aggregation assays using 20 ng of pBS/FTZ incubated with 20 nM of WT protein in the presence and absence of synthetic homoribopolymers that are indicated on top of the Figure. Added RNA amounts shown on bottom are expressed by ‘fold’ against pBS/FTZ DNA. Error bars indicate mean ± SD from three independent experiments.
Fig 6.
Effects of total substitution of arginines to lysines in the RG domain (RK mutant) on S/MAR-binding activity of hnRNP U.
Aggregation assays with WT (A) and RK mutant (B) using 25 ng of pBS/FTZ at different NaCl concentrations (50, 150 and 200 mM). Gel images from aggregation assays are shown with band density plots below. Error bars indicate mean ± SD from three independent experiments.
Fig 7.
Effects of netropsin on the binding of FTZ to RG domain and possible synergism between SAP and RG domains.
(A) Effects of netropsin on the selective binding of FTZ. For on-bead binding assays, GST-fusion proteins immobilized on glutathione-Sepharose beads (7.5 μl) was incubated with 100 ng of pBS/FTZ (37.3 fmol) and the bead-bound DNA fragments were analyzed. pBS/FTZ was pre-incubated with or without netropsin at the molar ratio indicated in the Figure (drug vs base pair in DNA). GST-RG (15 pmol) was used for the assays. Error bars indicate mean ± SD from three independent experiments. (B) On-bead binding assay of GST-SAP-RG. The same procedure as Fig 3A and 3B was employed. Data points for GST-SAP and GST-RG are transcribed from Fig 3A and 3B, respectively, and overlayed on the same Figure. Error bars indicate mean ± SD from three independent experiments. In drawing the regression curve, a nonlinear curve fitting program (Graphpad prism) was applied to mean values. (C) Schematic representation of the mechanism for synergistic binding.
Fig 8.
Schematic representation of the binding mechanism between hnRNP U and S/MAR DNA.
To aid understanding, the binding process is expressed stepwise, which may not be the case in actual environment. Step1: RG domain recognizes and binds to A/T-tracts of S/MAR DNA. Step2: SAP domain interacts with RG domain and secure the binding. Step3: Ambient RNA binds to RG domain. Step4: RG domain is released from DNA and the interaction between SAP domain is disrupted. Step5: SAP domain with limited affinity is dissociated from DNA and hnRNP U entire molecule is released.