Decoding the molecular interplay of endogenous CD20 and therapeutic antibodies with fast volumetric nanoscopy
Data files
Nov 04, 2024 version files 26.97 GB
-
Fig._1.zip
48.71 MB
-
Fig._2.zip
1.20 GB
-
Fig._3.zip
9.33 GB
-
Fig._4.zip
237.63 MB
-
Fig._5.zip
1.30 GB
-
Fig._S1.zip
7.61 KB
-
Fig._S10.zip
437.32 MB
-
Fig._S11.zip
183.19 MB
-
Fig._S12.zip
2.21 GB
-
Fig._S13.zip
4.21 GB
-
Fig._S14.zip
805.87 MB
-
Fig._S15.zip
835.11 MB
-
Fig._S16.zip
430.90 MB
-
Fig._S17.zip
1.18 GB
-
Fig._S18.zip
374.64 MB
-
Fig._S19_S20.zip
170.66 MB
-
Fig._S2.zip
8.53 KB
-
Fig._S21.zip
720.03 MB
-
Fig._S22.zip
667.99 MB
-
Fig._S24.zip
753.38 MB
-
Fig._S25.zip
588.06 MB
-
Fig._S26.zip
485.27 MB
-
Fig._S3.zip
2.84 KB
-
Fig._S4.zip
7.27 MB
-
Fig._S5.zip
73.64 MB
-
Fig._S7.zip
546.54 MB
-
Fig._S8.zip
140.24 MB
-
Fig._S9.zip
34.61 MB
-
README.md
10.37 KB
Abstract
Elucidating the interaction between membrane proteins and antibodies requires fast whole-cell imaging at high spatiotemporal resolution. Lattice light-sheet (LLS) microscopy offers fast volumetric imaging but suffers from limited spatial resolution. DNA-PAINT achieves molecular resolution but is practically restricted to two-dimensional imaging due to long acquisition times. Here, we introduce two-dye imager (TDI) probes, manifesting negligible background and amplified fluorescence signal upon transient binding, enabling ~15-fold faster imaging. Using a combination of TDI-DNA-PAINT and LLS microscopy on B cells, we reveal the oligomeric states and interaction of endogenous CD20 with type I and II therapeutic monoclonal antibodies (mAbs) rituximab (RTX), ofatumumab (OFA), and obinutuzumab (OBZ), respectively, unperturbed by surface effects. Our results demonstrate that CD20 is abundantly expressed on microvilli that are concatenated by mAb binding accompanied by a concentration-dependent B cell polarization and stabilization of microvilli protrusions. These findings, we believe, will aid rational design of improved immunotherapies targeting tumor-associated antigens.
README
The provided .zip folders contain the data and plot variable values, arranged figure-wise.
A. Ghosh*†, M. Meub†, D. A. Helmerich, J. Weingart, P. Eiring, T. Nerreter, K. M. Kortüm, S. Doose, and M. Sauer*. Decoding the molecular interplay of endogenous CD20 and therapeutic antibodies with fast volumetric nanoscopy
Comments and requests should be addressed to Markus Sauer and Arindam Ghosh: m.sauer@uni-wuerzburg.de; arindam.ghosh@uni-wuerzburg.de. All material is free of use, but we would appreciate being told, and this dataset and the paper cited where applicable.
Fig.1B-D contains loc. files of 2D-TDI-DNA-PAINT imaging of cellular microtubules that are generated in the SMAP analysis platform (Jonas Ries lab) and the variables/columns follow exactly the SMAP guidelines. In particular, the provided .mat files contain localization attributes such as coordinates x, y, and z, the number of photons per localization, background values, localization precisions, frames in which localization was found, etc. Any user can open these files in SMAP (free to install and has a MATLAB-free version). 1E-G are plot variables with self-explanatory naming as .mat files. For example, x-axis variables are stored in array 'x' and the same for y-axis variables as 'y'. 1H-I are loc. files .txt from 2-color TDI-DNA-PAINT imaging.
Fig. 2 contains a localization .csv file from whole cell LLS-TDI-DNA-PAINT imaging of cellular microtubule networks using TDI imager R4. The columns in the .csv files containing parameters such as coordinates x, y, and z, number of photons per localization, background values, and localization precisions, are intuitive to any non-specialist. Also, a .mat file containing calculated loc. precision values are provided.
Fig. 3B contains localization .mat files from 2D TDI-DNA-PAINT experiments on CD20 labeled with Rituximab (RTX), 2H7, Ofatumumab (OFA), and Obinutuzumab (OBZ). These localization files are generated in SMAP analysis platform (Jonas Ries lab) and the variables/columns follow exactly as the SMAP guidelines. In particular, the provided .mat files contain localization attributes such as coordinates x, y, and z, number of photons per localization, background values, localization precisions, frame in which localization was found, etc. Any user can open these files in SMAP (free to install and has a MATLAB-free version). Any user can open these files in SMAP (free to install and has a MATLAB-free version). Fig. 3C-F contains localization files .csv from whole cell LLS-TDI-DNA-PAINT imaging of CD20 labeled with RTX, 2H7, OFA and OBZ. The columns in the .csv files are intuitive to any non-specialist containing coordinates x, y, and z, number of photons per localization, background values, and localization precisions, etc. Fig. 3G has the .mat file with localization density values calculated for each antibody sample.
Fig. 4 contains deconvolved and deskewed lattice light-sheet (LLS).czi files of the images from two-color live-cell imaging of CD20-RTX (5 μg/mL) and actin, CD20-2H7 (5 μg/mL) and actin, fixed cell two color images of CD20-RTX and CD45, and CD20-2H7 and CD45. All files were generated using Zeiss LLS microscope's Zen Blue software. Can be opened in Fiji/ImageJ.
Fig. 5A-E contains deconvolved and deskewed lattice light-sheet (LLS).czi files of the images of two-color live-cell imaging of CD20-OFA (5 and 10 μg/mL) and actin, CD20-OBZ (5, 10, and 20 μg/mL) and actin generated using Zeiss LLS microscope's Zen Blue software. Can be opened in Fiji/ImageJ. Fig. 5F contains self-explanatory plot variables (protrusion lengths) as a .mat file.
Fig. S1 contains absorbance spectra values of Ox2-P3 and Ox2-R4 TDI imagers as .mat files.
Fig. S2 contains fluorescence spectra values of TDI imagers labeled with four different dyes as .mat files.
Fig. S3 contains fluorescence spectra of Ox2-R4 and FCS analyzed data as .mat files.
Fig. S4 contains plot variables as .mat files for panels A,B,C,D,F,G and intensity grey values for panel E as a .csv file.
Fig. S5 contains localization file .txt from two-color 2D TDI-DNA-PAINT imaging of microtubules and clathrin-coated pits. The files can be opened through ImageJ plugin Thunderstorm and columns include parameters such as coordinates x, y, and z, number of photons per localization, background values, and localization precisions, etc.
Fig. S7 contains localization file .csv from whole cell LLS-TDI-DNA-PAINT imaging of cellular microtubule networks using TDI imager P3. The columns in the .csv files are intuitive to a non-specialist in the field.
Fig. S8 contains localization .mat files from 2D TDI-DNA-PAINT experiments on CD20-RTX. These loc. files are generated in SMAP analysis platform (Jonas Ries lab) and the variables/columns follow exactly as the SMAP guidelines. In particular, the provided .mat files contain localization attributes such as coordinates x, y, and z, number of photons per localization, background values, localization precisions, frame in which localization was found, etc. Any user can open these files in SMAP (free to install and has a MATLAB-free version). The folder also contains .txt files containing number of localizations plotted in S8C.
Fig. S9 contains localization .mat files from 2D TDI-DNA-PAINT experiments on CD20-2H7. These loc. files are generated in SMAP analysis platform (Jonas Ries lab) and the variables/columns follow exactly as the SMAP guidelines. In particular, the provided .mat files contain localization attributes such as coordinates x, y, and z, number of photons per localization, background values, localization precisions, frame in which localization was found, etc. Any user can open these files in SMAP (free to install and has a MATLAB-free version). The folder also contains .txt files containing number of localizations plotted in S9C.
Fig. S10 contains localization .mat files from 2D TDI-DNA-PAINT experiments on CD20-OFA. These loc. files are generated in SMAP analysis platform (Jonas Ries lab) and the variables/columns follow exactly as the SMAP guidelines. In particular, the provided .mat files contain localization attributes such as coordinates x, y, and z, number of photons per localization, background values, localization precisions, frame in which localization was found, etc. Any user can open these files in SMAP (free to install and has a MATLAB-free version).
Fig. S11 contains localization .mat files from 2D TDI-DNA-PAINT experiments on CD20-OBZ. These loc. files are generated in SMAP analysis platform (Jonas Ries lab) and the variables/columns follow exactly as the SMAP guidelines. In particular, the provided .mat files contain localization attributes such as coordinates x, y, and z, number of photons per localization, background values, localization precisions, frame in which localization was found, etc. Any user can open these files in SMAP (free to install and has a MATLAB-free version).
Fig. S12 contains localization files .csv and reconstructed .tiff files from whole cell LLS-TDI-DNA-PAINT imaging of CD20 labeled with RTX and 2H7. The columns in the .csv files are intuitive to any non-specialist. MATLAB files provided contain loc. precision values plotted in S12E.
Fig. S13 contains localization files .csv and reconstructed .tiff files from whole cell LLS-TDI-DNA-PAINT imaging of CD20 labeled with OFA and OBZ. The columns in the .csv files are intuitive to a non-specialist in the field.
Fig. S14 contains deconvolved and deskewed lattice light-sheet (LLS).czi files of the images of two-color live-cell imaging of CD20-RTX (5 μg/mL) and actin generated by Zeiss LLS microscope's Zen Blue software. Can be opened in Fiji/ImageJ.
Fig. S15 contains deconvolved and deskewed lattice light-sheet (LLS).czi files of the images of two-color live-cell imaging of CD20-RTX (10 μg/mL) and actin generated by Zeiss LLS microscope's Zen Blue software. Can be opened in Fiji/ImageJ.
Fig. S16 contains deconvolved and deskewed lattice light-sheet (LLS).czi files of the images of two-color live-cell imaging of CD20-2H7 (5 μg/mL) and actin generated by Zeiss LLS microscope's Zen Blue software. Can be opened in Fiji/ImageJ.
Fig. S17 contains deconvolved and deskewed lattice light-sheet (LLS).czi files of the images of two-color live-cell imaging of CD20-2H7 (10 μg/mL) and actin generated by Zeiss LLS microscope's Zen Blue software. Can be opened in Fiji/ImageJ.
Fig. S18 contains deconvolved and deskewed lattice light-sheet (LLS).czi files of the images of two-color live-cell imaging of CD20-2H7 (20 μg/mL) and actin generated by Zeiss LLS microscope's Zen Blue software. Can be opened in Fiji/ImageJ.
Fig. S19_20 contains deconvolved and deskewed lattice light-sheet (LLS).czi files (large and small field of views) of the images of two-color fixed cell imaging of CD20-RTX (5 μg/mL) and CD45 generated by Zeiss LLS microscope's Zen Blue software. Can be opened in Fiji/ImageJ.
Fig. S21 contains deconvolved and deskewed lattice light-sheet (LLS).czi files of the images of two-color live-cell imaging of CD20-OFA (5 μg/mL) and actin generated by Zeiss LLS microscope's Zen Blue software. Can be opened in Fiji/ImageJ.
Fig. S22 contains deconvolved and deskewed lattice light-sheet (LLS).czi files of the images of two-color live-cell imaging of CD20-OFA (10 μg/mL) and actin generated by Zeiss LLS microscope's Zen Blue software. Can be opened in Fiji/ImageJ.
Fig. S24 contains deconvolved and deskewed lattice light-sheet (LLS).czi files of the images of two-color live-cell imaging of CD20-OBZ (5 μg/mL) and actin generated by Zeiss LLS microscope's Zen Blue software. Can be opened in Fiji/ImageJ.
Fig. S25 contains deconvolved and deskewed lattice light-sheet (LLS).czi files of the images of two-color live-cell imaging of CD20-OBZ (10 μg/mL) and actin generated by Zeiss LLS microscope's Zen Blue software. Can be opened in Fiji/ImageJ.
Fig. S26 contains deconvolved and deskewed lattice light-sheet (LLS).czi files of the images of two-color live-cell imaging of CD20-OBZ (20 μg/mL) and actin generated by Zeiss LLS microscope's Zen Blue software. Can be opened in Fiji/ImageJ.
Methods
Ensemble absorbance and fluorescence spectroscopy. UV/visible absorbance spectra and steady-state fluorescence emission spectra of fluorescently labeled imager probes were recorded using a Jasco V650 absorbance spectrometer and Jasco FP-6500 spectrofluorometer respectively. For absorbance spectra, 500 nM Ox2 and Ox1 imagers were used (fig. S1). In case of fluorescence spectra, 1 nM imager strands and respective docking strands were used in 104 molar excess (10 µM). Fluorescence emission spectra of P3 labeled with 2×ATTO655, 2×Cy5, and 2×TMR, 2×ATTO520 R4 (fig. S2), and 2×Ox2-R4 imager strands (fig. S3A). Samples were measured in a 3 mm path-length cuvette (Hellma, 105.251-QS) in PBS (Sigma-Aldrich, #D8537-500ML) containing 500 mM NaCl adjusted at pH 7.4.
Fluorescence correlation spectroscopy (FCS). FCS measurements on Ox2-P3 probes (Fig. 1C and fig. S3B-C) were carried out using a commercial time-resolved confocal fluorescence microscope setup (MicroTime200, PicoQuant) as described previously (42). Imager strands were used at a concentration of 1 nM dissolved in PBS (Sigma-Aldrich, #D8537-500ML) containing 500 mM NaCl adjusted at pH 7.4, while the complementary docking strand was 104 times higher i.e., 10 µM.
Single color TDI-DNA-PAINT imaging. TDI-DNA-PAINT imaging of microtubules in COS7 cells (Fig. 1D), DNA origami structures (fig. S4D-G), and CD20 in Raji cells (Fig. 3B, fig. S8-S11) were performed utilizing an inverted wide-field fluorescence microscope (IX-71; Olympus). For excitation of Ox2 probes, a 640 nm diode laser (Cube 640-100C, Coherent) in combination with a clean-up filter (laser clean-up filter 640/10, Chroma) was used. The laser beam was focused onto the back focal plane of the oil-immersion objective (100×, NA 1.51; Olympus). Emission light was separated from the illumination light using a dichroic mirror (HC 560/659; Semrock) and spectrally filtered by a band-pass filter (FF01-679/41-25, Semrock). Images were recorded with an electron-multiplying CCD camera chip (iXon DU-897; Andor). Pixel size for data analysis was measured to be 104 nm. For classical DNA-PAINT and TDI-DNA-PAINT experiments on cellular microtubule networks, 5 nM Ox2-P3 / Ox2-R4 probe and 100 pM Ox1-P3 probe was added to the respective glass-bottomed chamber wells containing COS7 cells fixed and immunolabeled with corresponding MetTet-P3’ docking strand (Fig. 1D). For measurements on Raji cells on CD20 labeled via RTX or 2H7, 5 nM Ox2-R4 probe was used. Blinking movies (Supplementary movies S1-S12, S18-S21) were recorded for representative 128 × 128-pixel areas (for microtubule in COS7 cells) and 256 × 256-pixel areas or similar (for CD20 with RTX or 2H7 in Raji cells) with TIRF excitation with an exposure time of 100 ms (frame rate 9.96 Hz) and irradiation intensity of 0.8 kW/cm2. Images were recorded for 18k frames (30 min) for each dataset (Table S2). We recorded images of DNA-origami samples using a different setup but with same optical parameters and components as for the previous one. We used this second setup as it is a more robust “drift-free” system in our laboratory and can be beneficial for origami samples. We added 25 nM of Ox2-R4 probes to the well containing DNA-origami nanostructures labeled with MetTet -7×R4’ docking strand incubated in the imaging buffer for origami. Images were recorded for 128 × 128-pixel areas (Movie S3, Table S2) with exposure times of 30 ms for 90k frames at 1.1 kW/cm2. Pixel size for data analysis was measured to be 128 nm.
Dual color TDI-DNA PAINT imaging. All two-color TDI-DNA-PAINT measurements were performed on an inverted wide-field fluorescence microscope (IX-71; Olympus) equipped with an oil-immersion objective (60×, NA 1.49; Olympus), a nose-piece stage (IX2-NPS, Olympus), a 514 nm and 639 nm lasers (Genesis MX STM-Series, Coherent) and a suitable dichroic mirror (R442/514/635, Chroma). The fluorescence emission of ATTO Oxa14 and ATTO520 coupled imager strands were acquired simultaneously and projected on two electron-multiplying charge-coupled device (EMCCD) cameras (iXon Ultra 897, Andor) by means of a dichroic mirror (630 DCXR, Chroma) and two bandpass filters (582/75 and 679/41 BrightLine series, Semrock). The excitation intensity was set to ~ 1 kW/cm2 for both light sources. Each color was measured for 20 min at 10 Hz (100 ms exposure time) in the total internal reflection fluorescence (TIRF) illumination mode if not stated otherwise. Pixel size for data analysis was measured to 128 nm (red camera) and 132 nm (green camera). To ensure the alignment of both detection channels, a registration sample exhibiting identical patterns across the spectral ranges of both acquisition channels was obtained. For this purpose, fluorescent microsphere samples (TetraSpeck, Invitrogen, #T14792) with diameters of 0.2 µm were employed. The microspheres were positioned on the sample holder and it was verified that both cameras were physically aligned with a precision of one pixel. The excitation intensity was adjusted so that the fluorescence intensity was approximately equal across both channels. Data from both channels was recorded simultaneously, and the acquisition was repeated in various areas to achieve a high density of alignment markers in the reconstructed image of all microsphere recordings. Afterwards, the two-color samples were imaged and the corresponding imaging buffer, containing 3.5 nM of dye coupled imager strands respectively, were applied to the sample prior measurements.
Data analysis for 2D TDI-DNA-PAINT. TDI-DNA-PAINT data for microtubules and DNA-origami were analyzed using the MATLAB-based Super-resolution Microscopy Analysis Platform (SMAP) (43). Using SMAP, data analysis was done by following the steps: loading of camera images, filtering, background estimation, finding of candidate molecules, fitting, and saving of the results. First, we loaded the raw blinking movies into the SMAP GUI, which was followed by manually setting up the camera parameters. Next, under the Localize\Peak Finder tab, we set the parameters for the initial estimation of single molecule positions, such as filtering cutoff with intensity values. To ensure selecting the brightest localizations, we select a higher cutoff value. After previewing an exemplary frame, we chose the Gaussian 2D fitting tab PSF free for localizing the full image set. The analysis returns coordinates such as x, y and z, number of photons per localization, background values, localization precisions, frame in which localization was found etc. Grouping (merging) of localizations persistent over several frames is performed after fitting. We can manually set the following grouping parameters dX (maximum distance two localizations can be apart) and dT (maximum number of dark frames between localizations). We grouped the localizations by setting dT = 1, and limiting the localization precision range from 0-15 nm, which filters out localizations with worse precision. On applying these grouping parameters, SMAP calculates the maximum distance from the localization precision of the two candidate localizations. That means that localizations are only grouped if it is likely they stem from the same event. We also quantified the background signal from different concentrations of Ox2-P3/R4 probes from 100 pM to 10 nM, and 5 nM appeared to be the most optimal concentration considering it had very similar background as compared to 100 pM Ox1-P3 (fig. S3A). We also tested if the attachment of two fluorophores to the imager P3 introduced any change to binding time (on-time). Figure S3C depict the on-time comparison between Ox1-P3, Ox2-P3, and Ox2-R4. Figures S3B and S3D-S3G presents photon yields, localization precision, background signal, and detected origami structures from TDI-DNA-PAINT experiments with Ox2-R4. The reported distances from origami experiments were from N = 33 values. For CD20 TDI-DNA-PAINT datasets, analysis was performed following the same workflow as described and reconstructed images from 10 min acquisition is illustrated in Fig. 3B and figs. S8-S11.
Data analysis for two-color TDI-DNA-PAINT. For the reconstruction of super-resolved two-color images, rapidSTORM3.3 was utilized. To align the two cameras employed, images of the recorded microsphere samples were reconstructed. A registration matrix was generated using the ImageJ plugin bUnwarpJ. The corresponding microsphere images were loaded, and 10 to 20 landmarks were identified using different microspheres in both images. Subsequently, the transformation matrix was converted into a raw transformation matrix. For the reconstruction of super-resolved microtubules and clathrin samples, the same settings used in the microsphere reconstruction were applied. The final reconstructed images for each channel were aligned using the ImageJ plugin bUnwarpJ along with the previously created transformation matrix. Finally, the aligned images were overlaid using ImageJ (v. 2.3.0/1.54j).
Whole-cell lattice light sheet (LLS) TDI-DNA-PAINT imaging. The workflow mostly recapitulates the description from M. K. Iwanski, E. A. Katrukha, L. Kapitein, Methods Mol. Biol. 2694, 151-174 (2024).
Live-cell two-color LLS imaging of actin and CD20. Dual-color LLS imaging of Raji cells was carried out using the Zeiss lattice light-sheet 7. Live-cell incubation with 6% CO2 and 37 ˚C was maintained using the inbuilt stage top Incubator (ibidi, #12722). Actin was stained with the live-cell F-actin dye SPY555-FastAct™ (Spirochrome AG, #SC205) as described previously. RTX 5 μg/mL (Fig. 4A and fig. S14) or 10 μg/mL (fig. S15)), 2H7 5 μg/mL (Fig. 4C and fig. S16) or 10 μg/mL (fig. S17) or 20 μg/mL (fig. S18)), OFA 5 μg/mL (Fig. 5A and fig. S21) or 10 μg/mL (Fig. 5B and fig. S22-S23), OBZ 5 μg/mL (Fig. 5C and fig. S24) or 10 μg/mL (Fig. 5D and fig. S25) or 20 μg/mL (Fig. 5E and fig. S26-S27), and were added to the cells in the respective sample chamber right before the start of experiment. Fluorophores were excited with “Sinc3 30 × 1000” light sheets for wavelengths (λexc) 561 nm (SPY555-FastAct™) and 640 nm (RTX-AF647/2H7-AF647/OBZ-ATTO643/OFA-ATTO643) via 13.3×/ 0.4 N.A. excitation objective lens. After focusing on the sample, imaging parameters such as correction for the X and Y tilt of the coverslip, “Focus Sheet” and “Focus Waist” were adjusted. The FOV in focus was scanned at 0.2 μm y stage position displacement in skewed dimension for a total range of 100 μm. Experiment was started right after adding RTX/2H7 to the chamber containing Raji cells and continued for 40 min with a sample scan repeating at 1 min interval. Emission light was collected through 44.83× /1.0 N.A. detection objective lens, followed by splitting to two cameras (ORCA-Fusion canal sCMOS (Hamamtsu Photonics), final pixel size 145 nm) with the aid of a 640 long-pass filter. Additionally, emission band-pass filters (576-617 and 656-750 nm) were also used before the respective cameras. Image analysis was performed using ZEN 3.9 blue software using the analysis method “Lattice light sheet”: with this protocol, acquired datasets were first deconvolved with a software in-built ‘Constrained Iterative’ deconvolution algorithm (6 iterations), followed by deskewing the image datasets, and finally rotated to coverslip coordinates. Dual color images for RTX or 2H7 or OBZ or OFA (magenta) and actin (yellow) from these measurements are shown in figs. S14-S18 and S21-S25 with the same intensity gray value scaling for the respective channel across images for a final comparison of fluorescence signals between RTX and 2H7 and actin in different conditions tested. Movies S26-S35 depict Raji cell dynamics for 30 min at different RTX,2H7, OBZ, and OFA concentrations.
Two-color LLS imaging of CD45 and CD20. Raji cells were labeled with 5 µg/mL of either anti-CD20 RTX-AF647 or 2H7-AF647 and anti-CD45 HI30-CF568 and fixed after RTX/2H7 incubation following the protocol described earlier. Two-color LLS imaging was performed using the commercial Zeiss lattice light-sheet 7 setup with the same settings as described in the previous section. Dual color Raji cell volumes (CD20 (magenta) and CD45 (green)) shown in Fig. 4B and 4D and figs. S19-20 demonstrates RTX accumulation in a polarized manner and its localization in membrane protrusions, corroborating our findings from live-cell LLS and LLS-TDI-DNA-PAINT measurements.