Ti e microscope

Variably timed photoconversion experiments (Fig. 2) were performed on a Nikon Ti-E microscope fitted with an Andor W1 Dual Camera (Andor CMOS ZYLA), dual spinning disk, Coherent Lasers (OBIS 405, 488 and 561 nm) and the Andor Mosaic 3 micro-mirror system, controlled by Andor iQ software, and Nikon elements for image acquisition. PulsePal software was used to time lock the stimulus with the mosaic sequence start. A custom 5×5, 250 µm² square grid was drawn to illuminate different areas (squares) at different times relative to the electrical stimulation. Photoconversion was induced using a train of 5 action potentials (50 Hz) paired with a 100 ms violet light pulse (405 nm, 50 mW cm⁻²) at different delays. The protocol was repeated 5 times at 30 s intervals.

Images were acquired at 100× magnifications using Nikon Ti-E microscope equipped with a Neo sCMOS camera (Andor) and Nikon NIS-Elements software. For the cell length measurements, phase contrast technique was used, and images have 2560×2160 resolution and 16 bit grayscale. The fluorescent images for msfGFP and YFP cells were obtained with illumination by an OBIS 488 nm laser from Coherent and the 59022 filter cube from Chroma. For each cell, the fluorescence signal was integrated and normalized by the projected area of the cell after background subtraction. Illumination across the field of view was homogeneous with less than 5% variations. For each experimental condition, we acquired 150–200 images containing at least 10,000 cells. We developed and used custom high-throughput image analysis software optimized for our experiments using Python and OpenCV library. The length of filamentous cells for SJ_XTL229 (ftsZ CRISPRi) in Fig. 6c were measured manually using a contour of each cell because of their abnormal lengths, curved morphologies, and frequent intersections with each other. For [arabinose] = 0.025% and above, at least 300 filamentous cells were measured for each growth condition.

After 6, 12, 25 or 50 h of light induction or the respective dark controls, images for GFP and RFP were acquired on an inverted Nikon Ti-E microscope with a ×4 1.4-numerical aperture objective and Andor iXON Ultra DU-888 camera; Z-stacks had 1.5 µm spacing over a 40 µm range. GFP had 300 ms of exposure; Sola 50% on 6–12 h, 12% on 25–50 h. RFP had 100 ms of exposure; Sola 20%. All these images were taken with live cells in black 96-well plates. Z-stacks were used for maximum intensity projection within ImageJ, and the projection were used for signal quantification with a macro running the ImageJ Subtract Background plugin with a rolling ball radius of 50, and then the Measure function for signal intensity. This quantification assumes that all wells contain on average the same number of cells, which were seeded in the beginning of the experiment. For some of the wells, we noticed a pipetting artifact on a side, producing an area devoid of cells. We manually selected a ROI that excluded this area for all wells, and we applied this before running the signal measurement macro. This experiment was performed blindly: I.L. transfected the cells and performed the light stimulation, C.C.J. performed the imaging without knowing or seeing the sample labels, then I.L. ran the quantification and then reassigned the original labels to the well names.

SPD-5 condensates were formed by diluting 10 µM SPD-5 (1:10 mixture of SPD-5 and SPD-5::TagRFP) in condensate buffer (25 mM Hepes, pH 7.4, and 150 mM KCl) containing PEG 3350 (Sigma-Aldrich) and fresh 0.5 mM DTT. Before use, the SPD-5 stock solution was centrifuged for 5 min at 80,000 rpm to remove residual aggregates. 5 min after formation, SPD-5 condensates were placed in glass-bottom 96-well dishes (Corning, 4850, high content imaging dish) precleaned with 2% Hellmanex and washed in water. For each sample, half was placed in the well undisturbed (control), and the other half was diluted 10-fold, pipetted five times, and placed in a well (induced disassembly). 96-well plates were imaged on an inverted Nikon Ti-E microscope using a 60× 1.4-NA Plan Apochromat oil objective, a Zyla cMOS camera (Andor), and MicroManager control software. For each image, SPD-5 condensates were identified through applying a threshold then using the particle analyzer function in Fiji. When analyzing condensate formation, we report the sum of the integrated intensities of each condensate per image (total condensate mass). Survival percentage plotted in Fig. 5 assumes a 10-fold loss in total condensate mass due to dilution.

Plasmid ptfLC3 was a gift from Tamotsu Yoshimori (Obtained from Addgene, plasmid # 21074)^{66 (link)}. Transfections of cell monolayers were done with the Lipofectamine Plus reagent (Invitrogen), according to the manufacturer’s instructions. Transfected cells were incubated at 37 °C for 18 to 24 h, unless otherwise stated. For immunofluorescence studies, the cells were permeabilized by incubation with 0.1% Triton X-100/PBS for 10 min at room temperature followed by incubation with 5% BSA/PBS for 10 min. The primary antibodies indicated for any particular experiment were added to cell monolayers in 5% FBS/PBS and incubated for 2 h at rt. After washing the monolayers three times with 5% FBS/PBS, cells were incubated for 1 h with secondary antibodies. Images were alternatively obtained with an Olympus DP-71 digital camera mounted on an Olympus BX51 fluorescence microscope, or with an Olympus IX70 equipped with a TillPhotonics camera. Confocal images were acquired either with an Olympus FV1000 confocal microscope or with an Andor Dragonfly spinning disk confocal system mounted on a Nikon TiE microscope equipped with a Zyla 4.2 PLUS camera (Andor). Colocalization was quantified as Pearson’s coefficient using JACoP, ImageJ plugin^{67 (link)}.

M2 macrophages were stained with CellTrace Far-Red Cell Proliferation Kit (ThermoFisher Scientific) and cocultured 1:2 with tumor cells (DLD-1) in NS media. After 24 hours, cells were infected with parental or oHSV-1 expressing PD-L1 BiTEs at MOI 1. After 3 hours of infection, T cells stained with CellTrace Violet Cell Proliferation Kit (ThermoFisher Scientific) were added to the coculture. Uninfected cells were used as a negative control. Also, 2 µM NucView 530 Caspase-3 Substrate (Biotium) was added to each well to visualize apoptosis. Brightfield and fluorescence images were captured on a Nikon Ti-E Microscope fitted with Andor Zyla 4.2 sCMOS camera (10× optical objective) at intervals of 15 min for 4 days. Time-lapse movies (14 frames/s) were generated using Fiji software. Caspase-3 substrate signal was detected in Cy3 channel of fluorescence microscope (excitation/emission wavelength of 528/563 nm) and was undisturbed by viral EGFP cassette (FITC channel, excitation wavelength 488 nm). Green colour for Caspase was selected to display better contrast during image/movie processing using Fiji software (Figure 6D, Supplementary Figure 15 and Supplementary Movies 1-4).