The multifunctional protein YB-1 potentiates PARP1 activity and decreases the efficiency of PARP1 inhibitors

Abstract

Y-box-binding protein 1 (YB-1) is a multifunctional cellular factor overexpressed in tumors resistant to chemotherapy. An intrinsically disordered structure together with a high positive charge peculiar to YB-1 allows this protein to function in almost all cellular events related to nucleic acids including RNA, DNA and poly(ADP-ribose) (PAR). In the present study we show that YB-1 acts as a potent poly(ADP-ribose) polymerase 1 (PARP1) cofactor that can reduce the efficiency of PARP1 inhibitors. Similarly to that of histones or polyamines, stimulatory effect of YB-1 on the activity of PARP1 was significantly higher than the activator potential of Mg²⁺ and was independent of the presence of EDTA. The C-terminal domain of YB-1 proved to be indispensable for PARP1 stimulation. We also found that functional interactions of YB-1 and PARP1 can be mediated and regulated by poly(ADP-ribose).

Keywords: PARP1 inhibitors; Y-box binding protein 1 (YB-1); olaparib; poly(ADP-ribose) (PAR); poly(ADP-ribose) polymerase 1 (PARP1).

Figure 1. YB-1 and PARP1 are able to form a heteromeric complex with damaged DNA

(A) The reaction mixtures contained 1× RB, 0 (red curve) or 200 nM (blue curve) PARP1, 100 nM FAM-labelled DNA Nick and 0–1600 nM YB-1. The formation of YB-1-PARP1-DNA complexes was followed by the fluorescence spectroscopy technique. (B) Reaction mixtures contained 1× RB, 0 or 200 nM PARP1, 100 nM radioactively labelled DNA Nick and 0–400 nM YB-1. The formation of YB-1-PARP1-DNA complexes was analyzed by gel-shift and autoradiography as described in Materials and Methods. Lanes 1–7: YB-1 binding to DNA in the absence of PARP1; lanes 8–14: in the presence of 200 nM PARP1. The concentrations of YB-1 in the mixtures are presented at the bottom of the panel. Positions of DNA and its complexes were visualized by phosphorimaging; the data acquired were analyzed by the Quantity One analysis software, providing the Transform and Crop Plot tools to optimize the image display. The experiment was performed 2 times.

Figure 2. YB-1 is a preferable target of poly(ADP-ribosyl)ation

(A) The “lag-period” for DNA release at high YB-1 concentration. The curves presented illustrate fluorescence anisotropy change of FAM-labelled DNA Nick measured by kinetic scanning. The reaction mixtures contained 1× RB, 200 nM PARP1, 100 nM Nick and 0–3200 nM YB-1 (the increase in YB-1 concentration is shown by the increased intensity of the color of the curve). Poly(ADP-ribosyl)ation was started by the addition of NAD+ at 25 s to a final concentration of 500 μM. All the measurements were carried out in duplicates for each reaction mixture. (B) The presence of PARP1 in the complex is necessary for DNA release after NAD+ addition (control). The curves presented illustrate the change of fluorescence anisotropy of FAM-labelled DNA Nick measured by kinetic scanning. The reaction mixtures contained 1× RB, 100 nM Nick and 0–3200 nM YB-1 (the increase in YB-1 concentration is shown by the increased intensity of the color of the curve). NAD+ was added at 25 s to a final concentration equal to 500 μM. All the measurements were carried out in duplicates for each reaction mixture. (C) PARP1 autopoly(ADP-ribosyl)ation reaction performed with the use of radioactively labeled NAD+^*. Lanes (1–4): without YB-1; (5–8): in the presence of 800 nM YB-1; (9–12): in the presence of 3200 nM YB-1. PARP1^*, YB-1^* designate poly(ADP-ribosyl)ated PARP1 and YB-1, respectively. Reaction times and YB-1 concentrations are shown at the bottom of the panel. Positions of protein bands were visualized by phosphorimaging; the data acquired were analyzed by the Quantity One analysis software, providing the Transform and Crop Plot tools to optimize the image display. The experiment was performed 3 times. (D) Gel-mobility shift assay analysis of YB-1 interaction with radioactively labelled DNA Nick (40 nM) during the time of poly(ADP-ribosyl)ation. C1: control for Nick. Lane 1: Nick bound by 400 nM YB-1 and 100 nM PARP1 before NAD+ addition. Lanes 2–7: appearance of free Nick during YB-1 and PARP1 repulsion from the protein-DNA complex upon poly(ADP-ribosyl)ation. The reaction time is shown at the bottom of the panel. The positions of DNA and DNA-protein complexes were visualized by phosphorimaging; the data acquired were analyzed by the Quantity One analysis software, providing the Transform and Crop Plot tools to optimize the image display. The experiment was performed at least 3 times.

Figure 3. Real-time assay for poly(ADP-ribosyl)ation

Reaction mixtures were prepared in Corning black 384-well polystyrene assay plates and irradiated with polarized light. Fluorescence anisotropy was defined as the ratio of the polarized component to the total intensity: A = (I₁ – I₂) / (I₁ + 2I₂), where I₁ and I₂ are the intensities of the light emitted by a fluorophore along different axes of polarization. The anisotropy level was used to estimate the size of the complex containing fluorescent DNA. During irradiation, excitation of the fluorophore can occur only if the electric field of the light is oriented in a particular axis towards the molecule. The anisotropy value (A) is maximum when the rotation of the fluorophore is confined by proteins bound to fluorescently labelled DNA and I₁ >> I₂ (that correspond to the case of unmodified PARP1, purple curve). The minimum A level is observed when the fluorophore has high mobility and I₁ ~ I₂ (that in our conditions corresponds to the control sample containing only DNA (blue curve) or free DNA after repulsion of poly(ADP-ribosyl)ated PARP1 (green curve)). The method allows one to detect protein binding to / dissociation from fluorescent DNA in the real-time.

Figure 4. YB-1 stimulates PARP1 activity in the absence of magnesium

The poly(ADP-ribosyl)ation reaction was performed using radioactively labelled NAD+^* as described in Materials and Methods and analyzed using TCA-targets. The radioautographs of TCA-targets optimized by Quantity One Transform Plot tool (A) are presented. The reaction buffer contained 5 mM MgCl₂ or 10 mM EDTA. Reactions were performed without YB-1 (gray columns and the first series of TCA-targets) or in the presence of 400 nM YB-1 (blue columns and the second series of TCA-targets). The reaction time is shown at the bottom of the panel A. The experiment was performed at least three times, the histogram (B) shows the mean values ± SD of three independent experiments.

Figure 5. C-terminal domain of YB-1 is necessary for PARP1 stimulation

Poly(ADP-ribosyl)ation reaction performed using radioactively labelled NAD+^* as described in Materials and Methods and analyzed using TCA-targets. The figure illustrates the radioautographs of TCA-targets (A) and the histogram showing the mean values ± SD of three independent experiments (B). The reaction buffer was without MgCl₂ and additionally supplemented with 10 mM EDTA. Reactions were performed in the presence of YB-1 or its mutants in varying concentrations as shown at the bottom of the panels.

Figure 6. YB-1 stimulates PARP1 activity in the presence of PARP1 inhibitors

Poly(ADP-ribosyl)ation reactions performed using radioactively labelled NAD+^* as described in the presence of 3-aminobenzamide (A), olaparib (B) or EtBr (C) at varying concentrations and analyzed using TCA-targets or by SDS-PAGE. (A and B) The figures present the histograms obtained by analysis of the radioautographs of TCA-targets by Quantity One software. The reactions were performed in the absence of magnesium (10 mM EDTA) without YB-1 (gray columns) or in the presence of 400 nM YB-1 and 10 mM EDTA (blue columns). The relative PARP1 activity is indicated on the left of the histograms, and the concentrations of inhibitors used are shown at the bottom of the panels. Histograms A and B show the mean values ± SD of three independent experiments. (C). The radioautograph of the SDS-PAGE used to analyze the reaction products. Reactions were performed in the absence of magnesium (10 mM EDTA) without YB-1 (lanes 1–7) or in the presence of 400 nM YB-1 and 10 mM EDTA (lanes 8–14). The concentrations of EtBr are shown at the bottom of the panels. The experiment was reproduced at least three times.

Figure 7. YB-1 can “reactivate” PARylated PARP1

(A) PARP1 autopoly(ADP-ribosyl)ation was performed with the use of radioactively labelled NAD+^* as described in section 2.5. The reaction mixtures contained 1x reaction buffer, 10 mM EDTA, 10 nM Nick, 4 μM NAD+^* and 200 nM PARP1. After 20 min of reaction, the mixtures were supplemented with 1× reaction buffer (lanes 1–3) or YB-1 to the final concentration of 400 nM (lanes 4–6) and additionally incubated for 5, 10 or 30 min at 37° C. The incubation time after 1× RB/YB-1 addition is indicated at the bottom of the panel. The experiment was performed twice. (B) The curves presented illustrate the change of fluorescence anisotropy of FAM-labelled DNA Nick measured by kinetic scanning. The reaction mixtures contained 1× RB, 200 nM PARP1, 10 nM Nick and 10 mM EDTA. Poly(ADP-ribosyl)ation was started by the addition of NAD+ to a final concentration of 500 μM. After 35 min of reaction the samples were supplemented with 1× RB (grey curve) or YB-1 to the final concentration of 400 nM (red curve). All the measurements were carried out in duplicates for each reaction mixture.

Figure 8. YB-1 and PARP1 interplay is regulated by poly(ADP-ribose)

(A) Reaction mixtures (10 μl) contained 10 nM Nick, 200 nM PARP1, 0 or 400 nM YB-1, 10 mM EDTA, 4 μM NAD+^* and 0-0.004 A₂₆₀ units of PAR prepared according to Materials and Methods, After incubation for 10 min at 37° C, the reaction mixtures were supplemented with 2.5 μl of Laemmli buffer with subsequent heating for 2 min at 97° C and analyzed by SDS-PAGE. The data acquired were analyzed by the Quantity One analysis software, providing the Transform and Crop Plot tools to optimize the image display. The experiment was performed at least 3 times. (B) Reaction mixtures (10 μl) contained 10 nM Nick, 200 nM PARP1, 0–400 nM YB-1, 10 mM EDTA and 4 μM NAD+^*. After incubation for 10 min at 37° C, the reaction mixtures were supplemented by 2.5 μl of Laemmli buffer with subsequent heating for 2 min at 97° C and analyzed by SDS-PAGE. The data acquired were analyzed by Quantity One analysis software, providing the Transform and Crop Plot tools to optimize the image display. The experiment was performed twice.

Figure 9. Application of PARP1 inhibitors to the treatment of YB-1-overexpressing tumors (scheme)