Schematics of the cryo-SMLM microscope.

Top: wide-field fluorescence microscope. The excitation lasers shown in cyan output from a multi-mode (MM) optical fiber (150×150 µm, 0.39 NA, M103L05) are collimated using an aspheric lens (L10) and focused at the back focal plane of an objective using the focusing lens (L250). The fluorescence signal shown in brown is back-reflected through the objective, transmitted through the quad-band dichroic mirror (DM), and collected using the tube lens (TL). The fluorescence intermediate image created at the focal plane of the tube lens is projected on an sCMOS camera using a 4f telescope system comprised of the L200 and L100 lenses. The irises in the excitation and detection paths, respectively, adjust the size of the illumination and the detection field of view. The far-red fluorescence signal from dark-red fiducial markers, shown in red, is reflected to the focus-lock detection arm using a short-pass dichroic (SP-DM, cut-off at 697 nm). This signal is aberrated via a weak cylindrical lens (CL) and relayed to an auxiliary camera using another 4f telescope system made of lenses L150 and L75. Bottom: cryo-stage. The air objective is mounted in an upright configuration inside a cage system with a copper adapter, where heating foils and temperature sensors control its temperature. The cage, clamped on the upper optical table, is made of heat-insulating side walls and a conducting plate at the bottom, where the cryo-compatible translation stages and the sample are mounted. A liquid nitrogen (LN2) micro-dosing pump cools the stages and sample by injecting LN2 inside the thermal insulator, where the cage is immersed. The laboratory jack adjusts the height of the thermal insulator. The aluminium cryo-shield reduces the humidity around the stages and the sample. In addition, it enhances heat transfer by directing the cold nitrogen gas that enters from openings in the bottom base plate. The sample holder (cartridge) is transferred from the right side using the cryo-transfer shuttle and sits on a heat-conducting device containing a heater and a temperature sensor for controlling the sample temperature. L: lens (focal length specified in mm), M: mirror, RA: rectangular aperture, BPF: band-pass filter, EXC-F: excitation filter, FW: filter wheel, TC: temperature control.

Thermal and mechanical stability of the cryo-SMLM open stat platform.

a) From bottom to top: lateral and axial translation stages, the metallic plate containing the heater and temperature sensor, adapter for holding the sample cartridge, and cartridge with two slots to accommodate autogrid-mounted cryo-EM grids. b) Stage temperature cooling and thermal stability over more than six hours. The inset shows the average temperature values of every 50 binned data points over 2 hours. c) Vitrified E. coli cells on a cryo-EM grid together with tetra-spectral beads shown in red and cyan circles. d) Upper and lower panels showing correspondingly the average lateral drift and lateral drift-corrected positions (blue and red) measured for the beads encircled in c. Axial position values in yellow are the same in both panels, indicating the focus-lock precision. The inset in the lower panel plots the averaged axial position values of every 50 binned data points over 1 hour. e) Histograms and standard deviations of drift-corrected lateral positions and axial positions for the single bead highlighted in a red circle in c.

Reduced humidity and ice contamination in the cryo-SMLM open stat platform.

a) Relative humidity measured inside the plexibox. b) Cryo-EM micrograph from a random hole in grid #1 after loading, incubation for 5 minutes, and unloading from the cryo-SMLM microscope. Two additional random holes from two different grids (#2 and #3), are shown after being imaged independently in the cryo-SMLM microscope for 3.5 hours.

SR-cryo-CLEM results.

a) First frame of the fluorescence imaging sequence showing several E. coli cells on an Ultrafoil grid. b) Increase in the number of detected localizations following photoactivation (at the indicated time point). c) Diffraction-limited rendering of FtsZ–rsEGFP2 (gray) and the corresponding SMLM reconstruction (red-hot color scale) from the ROI indicated by the yellow dashed square in (a). d) Cryo-fluorescence image of another batch of cells, with grid holes highlighted by magenta dashed circles. The inset shows a zoomed-in view of a single cell. e) Low-magnification cryo-EM image and inset of the same regions as in (d), with detected grid holes indicated by green dashed circles for correlation with the fluorescence image. f) Central slice from the cryo-ET volume corresponding to the white square in (e) (grayscale), correlated with the cryo-SMLM image of FtsZ–rsEGFP2 (red-hot color scale). The white triangle indicates the cell constriction site.

Microscope layout based on SolidWorks files.

a) angled, b) side, and c) top views of the layout, correspondingly. The cyan, orange, and red dashed lines in the top-view panel indicate the illumination, detection, and focus-lock paths, respectively.

Overview of the graphical user interface of the control software (controls detailed below).

The main camera control.

Numbers shown in red indicate the ROI selection and zoom-in/out options. Numbers on the top right represent the shown ROI size, minimum (MIN), maximum (MAX), and average (AVG) pixel values, and frames per second (FPS). The black camera icon enables acquiring a single frame and the white camera icon enables and disables live imaging mode. ‘save to NDTiff’ enables allocating a folder name for frame saving. The number of saved frames can be set by ‘Num. Frames’. Time-lapse image saving is possible using ‘save each N frames’, where the default value of ‘N=1’ translates to saving all frames sequentially. The center position of the selected ROI, and its size, can be correspondingly set using ‘Next ROI X/Y’ and ‘size’. ‘Auto-contrast full range ignoring’ allows for adjusting the image contrast based on pixel values and signal-to-noise ratio. Besides the pre-defined values of ‘None’, 0.01%, 0.1%, 1%, and 5%, a manual contrast value can be set in percent in the same row. ‘Exposure time (ms)’ can set the camera exposure time. ‘Cooler state’: ‘True’ enables water cooling, while ‘False’ enables internal fan (air) cooling.

Translation stage control.

The coarse movement (step mode, left panel) and fine movement (offset mode, right panel) controls. Step sizes can be adjusted by setting the applied voltages (in volts) to each axis. Frequency (in Hz) determines the speed at which the stepping is performed. One can determine the number of steps per click on the software arrows on the left by setting ‘Num. Steps X/Y’ for the lateral movement and ‘Num. Steps Z’ for the axial movement. Stepping can also be easily controlled using the computer keyboard, via up and down arrows (y-axis), left and right arrows (x-axis), or page up and page down keys (z-axis). Fine stepping can be done by applying offset voltages in the tab named ‘Fine conf.’, where the step size can be determined by setting the offset voltages (DeltaVolts.X,..) in Volts. An offset voltage of 0 V or 150 V represents the lower and upper limits for the stage driver. When either limit is reached, the system stops stepping automatically according to the programmed ‘stop stepping’ function. The ‘Reset’ button switches the stages off and on again, in case the offset voltage reaches 0 or 150 V, accidentally.

Upper and lower panels show coarse and fine z-sweep control tabs, respectively.

Laser switch and intensity control of: 405 nm and 488 nm iBeam Smart Toptica lasers with 200 mW power, 561 nm Coherent Sapphire LPX with 500 mW power, and 640 nm diode laser with a power mapped in a 0-100 % range.

Pulsing mode for the 405 nm laser, specific for modular photo-activation.

The pulse duration and repetition rate can be set using the Time ON and Time OFF options. The shortest ‘ON’ time (pulse duration) can be set to 1 ms.

Emission filter wheel selection tab for five different emission band-path filters.

a) Auxiliary camera and focus-lock settings. All similar buttons to those in Fig. S3 do the same functionalities as described for the main camera, except the cooling option for the auxiliary camera is not provided. In addition, only five different exposure times and two pixel binning modes can be set using the specific camera model used as the auxiliary camera in this work, with a default exposure time set to 100 ms and camera video binning mode set to ‘Mode0’: no Binning, ‘Mode1’: binning 2 pixels. ‘Start Focus Lock’ and ‘Stop Focus Lock’ enable and disable focus lock accordingly. Focus-lock can be conducted in coarse or fine stepping mode. In each mode, the upper and lower bounds of out-of-focus ratio, σx/σy, can be set accordingly. In Kalman variances, one can preset ‘signal’ and ‘noise’ depending on the image signal-to-noise ratio for a more robust fitting. When Focus-lock is activated, the current out-of-focus ratio is shown in the lower panel, where the gray plot shows the actual measured value, and the blue line shows the Kalman-filtered value. The blues dashed lines represent the upper and lower bounds of axial extent, where the focus is locked within. b) Enlarged image from the single bead image with astigmatic PSF at focus.

Navigation tab.

The upper panel shows position 1 (‘ROI_001’), corresponding to the EM grid center at (x = 0, y = 0, z = 0). The lower panel shows position 2 (‘ROI_003’), corresponding to an EM grid square. The displayed x, y, and z values represent the number of steps at a set voltage of 20 V, which correspond to displacements of (−22.4, 4.8, 2.6) μm.