Raw and processed data and analysis scripts to reproduce the results presented in "Live-cell micromanipulation of a genomic locus reveals interphase chromatin mechanics", Keizer et al. These Zenodo repositories contain the raw data, the processed data, the code to reproduce the integrality of the results presented in the preprint/paper (bioRxiv link to come) : "Live-cell micromanipulation of a genomic locus reveals interphase chromatin mechanics", Veer I. P. Keizer1,2,3,#, Simon Grosse-Holz4, Maxime Woringer1,2, Laura Zambon1,2,3, Koceila Aizel2, Maud Bongaerts2, Lorena Kolar-Znika1,2, Vittore F. Scolari1,2, Sebastian Hoffmann3, Edward J. Banigan4, Leonid A. Mirny4, Maxime Dahan2,§, Daniele Fachinetti3,* & Antoine Coulon1,2,*,¶ 1. Institut Curie, PSL Research University, Sorbonne Université, CNRS UMR3664, Laboratoire Dynamique du Noyau, 75005 Paris, France 2. Institut Curie, PSL Research University, Sorbonne Université, CNRS UMR168, Laboratoire Physico Chimie Curie, 75005 Paris, France 3. Institut Curie, PSL Research University, Sorbonne Université, CNRS UMR144, Laboratoire Biologie Cellulaire et Cancer, 75005 Paris, France 4. Department of Physics and Institute for Medical Engineering and Science, Massachusetts Institute of Technology, Cambridge, 02139 MA, USA # Current address: National Cancer Institute, NIH, Bethesda, MD, USA § Deceased * Correspondence: daniele.fachinetti@curie.fr, antoine.coulon@curie.fr ¶ Lead contact. Content of the Zenodo repositories /!\ The dataset had to be split into multiple Zenodo archives for technical reasons. This dataset is dataset 9/10, please refer to the other Zenodo repositories for the rest of the data. List of the datasets Raw data 1/6 (zenodo 4626942). Raw microscopy data acquired on 2019-04-10 – 30’ Pull-Release. Embargo: YES (avail. to reviewers) ; publication Raw data 2/6 (zenodo 4627034). Raw microscopy data acquired on 2019-04-15 – 30’ Pull-Release. Embargo: YES (avail. to reviewers) ; publication Raw data 3/6 (zenodo 4626909). Raw microscopy data acquired on 2019-12-17 – 30’ Pull-Release. Embargo: YES (avail. to reviewers) ; publication Raw data 4/6 (zenodo 4626914). Raw microscopy data acquired on 2019-12-23 – 30’ Pull-Release. Embargo: YES (avail. to reviewers) ; publication Raw data 5/6 (zenodo 4627010). Raw microscopy data acquired on 2020-02-21 – 30’ Pull-Release. Embargo: YES (avail. to reviewers) ; publication Raw data 6/6 (zenodo 4626981). Raw microscopy data acquired on 2020-03-04 – 100s Pull-Release. Embargo: YES (avail. to reviewers) ; publication Concatenated data 1/1 (zenodo 4674507). Concatenated raw data ; 8 TIFF files (one per position) with a separate timestamps file. Embargo: YES (avail. to reviewers) ; publication Processed data 1/1 (zenodo 4674438). Embargo: No Steps performed on the 9 TIFF movies (details below): drift correction, reference/alignment, rotation with respect to force direction locus tracking (TXT file) Calibration 1/2 (zenodo 4627062). Raw/processed data and code to estimate the force field from microscopy experiments. Embargo: No Calibration 2/2 (zenodo 4674531). Raw/processed data and code to estimate the fluorescence of individual magnetic nanoparticles. Embargo: No List of the softwares. Software 1/4 (links zenodo tbd/ github:CoulonLab/Keizer-et-al). Code to reproduce the results and figures presented in the paper. Embargo: YES (avail. to reviewers) ; publication Software 2/4 (links zenodo:4672595, github:CoulonLab/MagSim). Code to simulate force maps. Embargo: No Software 3/4 (links zenodo:4674399, github:SGrosse-Holz/rouselib). Code to simulate Rouse polymers and infer forces. Embargo: No Software 4/4 (links: zenodo:4674417, github:CoulonLab/chromag-pipeline). Code to concatenate raw files from MicroManager acquisitions. Embargo: No Data re-use policy As standard practice in the field (https://www.4dnucleome.org/policies.html), researchers using this public, but as yet unpublished data must contact the specific data producer (antoine.coulon@curie.fr) to discuss possible coordinated publication. Unpublished data are those that have never been described and referenced by a peer-reviewed publication. Overview of the raw data repositories Refer to the article for full details on cell line generation and cell culture conditions. Cells were imaged using a custom widefield epifluorescence microscope (Nikon, Eclipse Ti2) with a Spectra-X light source (Lumencor), a 100x/1.49NA oil objective (Nikon, CFI Apochromat), an iXon Life 888 EM-CCD camera (Andor), a HLD117NN stage (Prior), a Nano-ZL100 stage piezo (Mad City Labs) and a microscope enclosure chamber (Okolab) to maintain temperature and humidity. The microscope was controlled by the MicroManager software [doi:10.1002/0471142727.mb1420s92]. Unless stated otherwise, z-stacks of 21 to 24 planes were taken (∆z=0.21 or 0.3 µm) with 100ms exposures and Spectra-X powers at 4% or 5% for the GFP, mCherry and SiR-DNA channels. Illumination powers were chosen to be minimal and blue/UV light was excluded, as to avoid light-induced artefacts in chromatin motion [doi:10.1016/j.cub.2006.03.059]. In the 30 min pull-release (30’-PR) experimental scheme, individual z-stacks were taken in all 3 fluorescence channels prior to the time-lapse experiment. Then, the external magnet was place onto the microarray holder and a ∆t=2min time-lapse acquisition was started for 60 min over 6 to 8 stage positions (i.e. micropillar tips), with only the GFP and SiR-DNA channels. The external magnet was removed 30 min into the time-lapse acquisition. Whenever the external magnet is not on the holder, a counterweight was placed onto the holder to minimize mechanical drifts. However, on occasions where placement or removal of the external magnet led to a loss of focus, the acquisition was stopped, adjusted and resumed (note that all analyses rely the actual timestamp of each image to accommodate for the resulting variable frame intervals – see ‘Image analysis’ section). In the 100 sec pull-release (100”-PR) experimental scheme, a single z-plane was acquired every 5 section in the GFP and SiR-DNA channels for 3000 sec (600 frames) with the same imaging setting as the 30’-PR scheme. After starting the acquisition, the external magnet was placed and removed every 100 sec (20 frames) 10 times, adjusting the focus while acquiring when necessary. After the 10th magnet removal the acquisition was left to finish for the remaining 1000 sec. Each Zenodo dataset represents one day of acquisition, including the data that was not retained (QC) for further downstream analysis. Each dataset contains: The raw MicroManager folder architecture (one folder contains multiple positions on the coverslip). On occasions where placement or removal of the external magnet led to a loss of focus, the acquisition was stopped and restarted, creating a new MicroManager folder each time. For instance: The various positions were imaged before injection, in a folder with the _preInjection ; this folder contains one TIFF file per position These positions were imaged again after injection and before the magnet was added, leading to a folder with the suffix _beforeAttraction They were imaged again with the magnet added ; folder _attraction1 They were finally imaged after the magnet was removed ; folder _release1 If the focus had to be adjusted during the attraction of release period, this leads to additional folders (e.g. _attraction2, _release2) A text file named labjournal.txt, that contains extra information about specific details of the acquisition Note: the MicroManager metadata in the TIFF file are fully populated Overview of the concatenated datasets In this Zenodo repository, each position (acquired in different folders), is concatenated into a single TIFF movie using code available in the Github repository s4. The folder contains: One TIFF file per selected position One .xls file per selected position, with extra timestamp information Timestamps.xls data format This file contains one line per frame, ordered chronologically, from frame 1 to n. path (Relative path): Reference to the original (raw MicroManager) file start_time (Timestamp): Timestamp saved by MicroManager when the acquisition was started (the «acquire » button was pressed). time_in_file (seconds): Number of seconds between start_time and the acquisition of the current timepoint start_time_s (seconds): Variable start_time converted to a number of seconds time (seconds): Sum of start_time and time_in_file timestamp (Timestamp): Variable time, back-converted to a timestamp timeOn (Timestamp): Time(s) when the magnet was added. This timestamp is provided in the datasets.cfg file in the github repository chromag-pipeline timeOff (Timestamp): Time(s) when the magnet was removed. This timestamp is provided in the datasets.cfg file in the github repository chromag-pipeline forceActivated (Boolean): If the magnet is present during the current frame (calculated from timeOn and timeOff) seconds_since_first_magnet_ON (seconds): Number of (relative) seconds since the magnet was added for the first time. Frame (Integer) Frame number (1-indexed) Positions (Integer): The position number Overview of the processed datasets For each one of the selected concatenated movies, band-passed versions of the GFP and SiR-DNA channels were generated using Fiji (σ1=0.8 px, σ2=10 px) and treated alongside the raw channels in all subsequent steps. Correction of mechanical drifts was performed using the ‘Correct 3D drift’ Fiji plugin using the micropillar and/or debris on the coverglass as cues. The single-MNP 3D force map (see article for detail; code and maps available at https://github.com/CoulonLab/MagSim) was aligned with the micropillar, merged with the microscopy movie and multiplied by the number of MNPs deduced from the fluorescence intensity of the locus: the amount of MNP is calculated on a cell-by-cell basis from the first z-stack of the time-lapse acquisition, and is multiplied by the 3D force vector from the single-MNP force map on a frame-by-frame basis (since the locus and the cell move over time relative to the micropillar). The resulting movie was rotated to align the force vertically, and the Fx and Fy channels of the force map were adjusted accordingly. The motion and deformation of the cell nucleus was corrected solely based on the SiR-DNA channel, first using the ‘Correct 3D drift’ Fiji plugin and then refined manually using visual landmarks in the area of interest in the nucleus, visualized both on top and side views. 3D tracking of the genomic locus was performed in the resulting movie using a custom-written ImageJ/Python software and the Fx, Fy and Fz components of the force were extracted as the values of the force map at the location of the genomic locus on each frame. The 100”-PR movie was analyzed in a similar manner as the 30’-PR movie, with few differences: All the steps were performed in 2D since the 100”-PR movie is a single-plane acquisition. The SiR-DNA channel was band-passed using σ1=2.5 px (to reduce further imaging noise) and σ2=10 px. To account for the occasional loss of focus of the locus on a few frames due to the placement/removal of the external magnet, we created an extra channel with a Gaussian at the center of mass of the locus (which hence locates precisely the locus whether in focus or not). This was used for both calculating the trajectory and for visualizations. This repository contains one TIFF file per concatenated movie. Overview of the calibration data & simulations Calibration of force maps This Zenodo upload describes the experimental data an associated image-processing code used to to calibrate our simulated force maps (see article for details). Raw data (`raw_data` folder): Images of 6 different pillars obtained as follows. We covered an area of the magnetic microarray with a drop of diluted MNP solution (25 nM) and imaged the spatial distribution of MNPs subject to the magnetic field around the pillars. Using a spinning-disk confocal microscope (Yokogawa CSU-X1 on a Nikon Ti stand with a Plan Apo 100x/1.4 oil objective and an Andor iXon3 camera) to achieve z-sectioning, z-stacks of GFP signal (MNPs) were obtained along with a homogenous control (TMR) in a second channel. Code for image analysis (`scripts` folder): Raw z-stack timeseries images were corrected for intensity offset and non-homogenous illumination/light collection (‘flat-field’ correction), time-averaged and down-sampled to minimize noise, and the GFP channel (MNPs) was normalized by the TMR channel (homogenous control) to correct for shadowing effects due to partial obstruction of the confocal light beam at the vicinity of the pillar. Within an intermediate range of fluorescence intensity, the resulting images are assumed to approximate a Boltzmann distribution of MNPs, so that F=kBT ∇[log(fluo)]. See article for details. For the comparison with the simulated maps, masks were made to exclude areas of low fluorescence signal (prone to noise and inaccurate flat-field correction) and areas of high signal (prone to non-linearity due to steric exclusion and fluorescence quenching). Since even a small amount of photobleaching during the spinning-disk acquisition makes the z component of the experimental force maps inaccurate, only the x and y components were used for comparison with the simulated maps. Processed data (`output` folder): We provide the final output as a single TIFF file with all the channels described above and with all 6 pillars provided as raw data. Fluorescence-to-MNPs conversion factor To estimate the fluorescence of single ferritin MNPs, we used imaging conditions where we can see individual ferritin MNPs. We injected low amounts of MNPs into cells and performed short-exposure (5 ms) max-power imaging. After temporal binning of the images and exclusion of bright and static objects, single MNPs were detected and averaged into a single-MNP image. The integrated fluorescence intensity was computed and converted into a ‘fluorescence-to-MNP conversion factor’ accounting for our default imaging conditions used for time-lapse acquisitions. ; This work received funding from: • the LabEx CELL(N)SCALE (ANR-11-LABX-0038, ANR-10-IDEX-0001-02) (MD, DF, AC) • the Agence Nationale de la Recherche (project CHROMAG, ANR-18-CE12-0023-01) (MD, AC), • the PRESTIGE program of Campus France (PRESTIGE-2018-1-0023) (VK) • the ATIP-Avenir program of CNRS and INERM, the Plan Cancer of the French ministry for research and health (AC, DF), • the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation program (grant agreement No 757956) (AC), • the LabEx DEEP (ANR-11-LABX-0044, ANR-10-IDEX-0001-02) (AC), • the program Fondation ARC (grant agreement PJA 20161204869) (AC) • the Institut Curie (DF, AC) • the Centre National de la Recherche Scientifique (CNRS) (AC, DF) • the European Union's Horizon 2020 research and innovation program under the Marie Skłodowska-Curie grant agreement No 666003 (SH) • the NIH GM114190 grant (LAM), • the MIT-France Seed Fund (LAM, MD), • LAM is a recipient of Chaire Blaise Pascal by Île-de-France Administration (LAM).