High-throughput assay exploiting disorder-to-order conformational switches: application to the proteasomal Rpn10:E6AP complex

Conformational switching is pervasively driven by protein interactions, particularly for intrinsically disordered binding partners. We developed a dually orthogonal fluorescence-based assay to monitor such events, exploiting environmentally sensitive fluorophores. This assay is applied to E3 ligase E6AP, as its AZUL domain induces a disorder-to-order switch in an intrinsically disordered region of the proteasome, the so-named Rpn10 AZUL-binding domain (RAZUL). By testing various fluorophores, we developed an assay appropriate for high-throughput screening of Rpn10:E6AP-disrupting ligands. We found distinct positions in RAZUL for fluorophore labeling with either acrylodan or Atto610, which had disparate spectral responses to E6AP binding. E6AP caused a hypsochromic shift with increased fluorescence of acrylodan-RAZUL while decreasing fluorescence intensity of Atto610-RAZUL. Combining RAZUL labeled with either acrylodan or Atto610 into a common sample achieved robust and orthogonal measurement of the E6AP-induced conformational switch. This approach is generally applicable to disorder-to-order (or vice versa) transitions mediated by molecular interactions.

For GST-tagged Rpn10 RAZUL wildtype and mutants, supernatants were incubated with pre-
Unbound excess dye was removed by buffer exchange into Buffer A four to five times.Labeling of RAZUL mutants with acrylodan was confirmed by LC-MS analysis, Figure S2.
Deconvolution of multiply charged states of RAZUL was processed by MestReNova.Percent labeling was approximated with the Equation 1 with the assumption that unlabeled and labeled RAZUL ionize similarly.Da for Atto610-labeling, and 8500 -10400 Da for DY647P1-labeling.Spectra were also checked for doubly-labeled proteins, but double addition of dye was not observed.

Circular dichroism (CD) spectroscopy
Far-UV range circular dichroism (CD) spectra (240 -190 nm) of 25 µM Rpn10 RAZUL (wildtype or mutant), 25 µM E6AP AZUL, and the mixture of 25 µM Rpn10 RAZUL (wildtype or mutant) and 25 µM E6AP AZUL were recorded on a Jasco J-1500 CD spectrometer using a quartz cuvette with a 1.0 mm path length (Hellma Analytics, Cat.No. 110-1-40) and controlled temperature of 25 ºC.Samples were buffer exchanged into Buffer F (10 mM Na2HPO4, 10 mM NaF, 1 mM TCEP, and 10 µM ZnSO4, pH 6.5).All spectra were collected continuously at a scan speed of 50 nm/min and averaged over accumulation of three spectra.The molar ellipticity () was calculated from the raw millidegrees (m o ) at wavelength lambda using Equation 2: where C is the concentration of the sample in mol L -1 , L is the path length of the cell in cm, and number of amino acids (# AA) for RAZUL is 72 and AZUL is 63.

Isothermal titration calorimetry (ITC)
ITC was performed at 25 °C on a MicroCal iTC200 system (Malvern Panalytical).Rpn10 RAZUL (wildtype protein and unlabeled/labeled mutants) were extensively co-dialyzed with E6AP AZUL in Buffer G (10 mM MOPS, pH 6.5, 50 mM NaCl, 2 mM TCEP, and 10 µM ZnSO4.Eighteen 2.1 µL aliquots of E6AP AZUL were injected at 750 rpm into a calorimeter cell (volume 200.7 µL) that contained 10-fold less concentrated Rpn10 RAZUL.For example, if the titrant E6AP AZUL was 80 µM, Rpn10 RAZUL in the cell was 10-fold less at 8 µM.Reference experiments were performed by replacing protein sample in the cell with Buffer G, and this reference data was subtracted from the experimental data during analysis.The integrated interaction heat values were normalized as a function of the molar ratio of Rpn10 RAZUL to E6AP AZUL, and the data were fit with MicroCal Origin 7.0 software.Binding was assumed to be at one site to yield the binding affinity K a (1/K d ), stoichiometry, and other thermodynamic parameters.
For experiments with E6AP AZUL, E6AP AZUL (or negative control, BSA) was serially diluted within the assay plate.Next, a stock of dye labeled Rpn10 RAZUL was added to all wells.The plate was spun down, and measurements were taken immediately.If experiment was timedependent, samples were protected with foil and stored at 4 ºC prior to measurement.
Measurements by CLARIOstar were taken at room temperature.
For competition experiments with unmodified or phosphorylated Rpn10 322-366 , the peptides were serially diluted within the assay plate, followed by the addition of a stock of E6AP AZUL to all wells.The plate was spun down and stored at 4 ºC for 10 -15 minutes.After equilibration, a stock of dye-labeled Rpn10 RAZUL was added to all wells.The plate was spun down, and measurements were taken immediately.If experiment was time-dependent, samples were protected with foil and stored at 4 ºC prior to measurement.Measurements by CLARIOstar were taken at room temperature.The final DMSO percentage throughout the assay was maintained at 5% (%v/v).
For Z' factor experiments only, Buffer L (10 mM MOPS, pH 6.5, 150 mM NaCl, 5 mM DTT, 10 µM ZnSO4, 1% BSA, 0.1% Tween 20) was used, with BSA acting as a carrier protein to reduce nonspecific interactions.Final DMSO percentage within the assay was at 5% (%v/v).Within the CLARIOstar plate reader, samples were shaken at 300 rpm (double orbital) for 30 seconds prior to measurement.where Signalbound is the average of a triplicate measurement of RAZUL Atto /RAZUL DY with a saturating AZUL concentration.For normalization to compare the free and bound state of various assay conditions, such that the max relative signal is 1.0, Equation 4 (above) was applied.In these conditions, xmax is unbound RAZUL Atto /RAZUL DY .

Figure S3 .
Figure S3.Detergent reduces aggregation in a plate reader assay without significant effect on fluorescent properties.(A) RAZUL S358C Acr binding curves in 384-well plates with E6AP AZUL where buffer includes (left, same as Figure 2D) or lacks (right) 0.1% Tween 20.A constant concentration of RAZUL S358C Acr [25 nM (red), 50 nM (green), 100 nM (blue), or 500 nM (purple)] was tested with increasing concentrations of AZUL in 10 mM MOPS, 50 mM NaCl, 5 mM DTT, and 10 µM ZnSO4.Results were plotted by GraphPad Prism and fit to a Hill curve for specific binding.(B -C) Fluorescence spectroscopy of 500 nM RAZUL S358C Acr in 10 mM MOPS, pH
washed glutathione sepharose beads (Cytiva 17-0756-05) for 1 hour at room temperature.Beads were washed with Buffer A three times.Rpn10 RAZUL was cleaved from GST on-bead overnight at 4 ºC with PreScission Protease (Cytiva 270884301) in Buffer B (50 mM Tris-HCl, pH 7.0, 150 mM NaCl, 1 mM EDTA, 1 mM DTT).For His-tagged E6AP AZUL, the supernatant was incubated with pre-washed Ni-NTA beads (Qiagen 30210) for 1 hour at room temperature and washed three times with Buffer A. E6AP AZUL was released from the His-tag on-bead overnight at 4 ºC with thrombin protease (EMD Millipore 605195) in Buffer C [1X Phosphate Buffered Saline (PBS) or 10 mM Na2HPO4, 1.8 mM KH2PO4, 2.7 mM KCl, 137 mM NaCl, pH 7.4].Eluted cleaved proteins were purified by size exclusion chromatography on an FPLC ÄKTA pure (GE Healthcare) using a HiLoad 16/600 Superdex 75 in Buffer A. Purity of protein was assessed by SDS-PAGE, and samples were dialyzed in Buffer A at 4 ºC prior to use.If buffer exchange was required, samples were buffer exchanged with columns (Thermo Fisher Cat.No. 89882) or dialyzed prior to experimental application.
As a control, Rpn10 RAZUL wildtype was buffer exchanged into Buffer D and treated with acrylodan under the same conditions.Excess dye was removed, and sample was analyzed by LC-MS to ensure no reactivity, FigureS2C.Atto610-and DY647P1-labeling of cysteine-substituted Rpn10 RAZULCysteine-substituted Rpn10 RAZUL mutants were buffer exchanged into Buffer E (50 mM HEPES, pH 8.0, 150 mM NaCl) and prepared at 10.52 µM.Atto610 maleimide (Atto-tec Cat.No. AD 610-45) and DY647P1 maleimide (Dyomics Cat.No. 647P1-03) were prepared in DMSO at 4000 µM.At a 1 mL scale, 950 mL of 10.52 µM protein and 50 µL of 4000 µM dye were combined.Overall, this reaction was performed at a 1:20 protein:dye ratio with 10 µM protein and 200 µM dye.Samples were rotated at 4 ºC overnight (16 -20 hours) and protected from light.Unbound excess dye was removed by buffer exchange into Buffer A four to five times.Labeling of RAZUL mutants with Atto610 or DY647P1 was confirmed by LC-MS analysis, Figure S4 -S5, respectively.Percent labeling was approximated with the Equation 1 with the assumption that unlabeled and labeled RAZUL ionize similarly.As a control, Rpn10 RAZUL wildtype was buffer exchanged into Buffer E and treated with Atto610/DY647P1 under the same conditions.Excess dye was removed, and sample was analyzed by LC-MS to ensure no reactivity, FigureS4C and S5C.LC-MS collection and analysis of labeled mutants10 µM of labeled protein (or unlabeled protein for control samples) was prepared in 50/50 water/acetonitrile for LC-MS collection and analysis.Mass spectrometry data were acquired on an Agilent 6130 Quadrupole LC-MS System, (Agilent Technologies, Inc.) equipped with electrospray source, operated in positive-ion mode.Separation was performed on a 300SB-C3 Poroshell column (2.1 mm x 75 mm; particle size 5 µm).Approximately 25 pmol of protein was injected and were eluted at a flow rate of 1 mL/min with a 5 -100% increase of mobile phase B over 5 minutes and holding with mobile phase B for 1 minute.Mobile phase A contained 5% acetic acid in water, and mobile phase B was acetonitrile.Mass spectra were analyzed and deconvoluted with MestReNova software.Full protein peaks were analyzed and deconvoluted with tolerance set to 50 -150 ppm, abundance threshold at 1%, charged state range from 4 -11, and mass-to-charge range from 600 -2300 Da.For each labeled sample, deconvoluted mass range varied: 8000 -9500 Da for acrylodan-labeling, 8500 -10200

Fluorescence intensity measurements to
monitor fluorescently labeled RAZUL were performed with an excitation of 390 (± 15) nm and emission of 475(± 20) or 500 (± 20) nm for acrylodanlabeled RAZUL (RAZUL Acr ) , excitation of 580 (± 15) nm and emission of 632 (± 20) nm for Atto610-labeled RAZUL (RAZUL Atto ), and an excitation of 617 (± 15) nm and emission of 667 (± 20 20) nm for DY647P1-labeled RAZUL (RAZUL DY ).Gain was set by adjusting towards the max signal of the experiment, which was fluorophore-dependent: bound RAZUL Acr or unbound RAZUL Atto /RAZUL DY .Relative fluorescent units (RFUs) were calculated by Equation 3 Acr :x (in RFU) = Signalmeasured -Signalbackground (3 Acr )where Signalbackground is the average of a triplicate measurement of RAZUL Acr in buffer only.For normalization to compare the free and bound state of various assay conditions such that the max relative signal is 1.0, the following Equation 4 was applied to the RFU calculated in Equation3.Relative Signal = x / xmax (4)where x is the normalized signal from Equation 3 Acr of a particular condition and xmax is the maximum signal within the titration or experiment.xmax is fully bound acrylodan-labeled RAZUL.Data was then plotted on GraphPad Prism 9.For fitting, curves were analyzed with non-linear regression parameters, either the Specific binding with Hill slope or Dose-Response with variable slope (four parameters) with asymmetric confidence intervals.RFU of RAZUL Atto or RAZUL DY was normalized using Equation 3 Atto/DY :x (in RFU) = Signalbound -Signalmeasured (3 Atto/DY )

2 : 4 :where
′ = 1 − .(0 * &0 ( ) | 3 * 4 3 ( | (5) 21 where µp and µn are the average signal of the positive (RAZUL 322-366 ) and negative (DMSO, vehicle) control, respectively, and σp and σn are the standard deviations of the positive and negative control samples, respectively.The robust Z' factor (Z'M) was calculated by Equation 6 as described previously 3, Mp and Mn are the median signal of the positive (RAZUL 322-366 ) and negative (DMSO, vehicle) control, respectively, and MADp and MADn are the median absolute deviations (MAD) of the positive and negative control samples, respectively.

Table of Contents
Supplementary Figures & Captions .
, fluorescence was recorded in Buffer A. For Figure S3B and S3C,