to BioTechniques free email alert service to receive content updates.
Correlative light microscopy for high-content screening
Benjamin Flottmann*1, Manuel Gunkel*1, Tautvydas Lisauskas*1, Mike Heilemann1,2, Vytaute Starkuviene1, Jürgen Reymann1, and Holger Erfle1
Full Text (PDF)

Confocal microscopy

Selected cells were imaged on a Leica TCS SP5 with a 63× objective lens (Leica HCX PL APO, NA 1.40; Leica, Wetzlar, Germany), a scan speed of 400 Hz at 512 × 512 pixels and with a pixel size of 69 nm, resulting in a field of view of 49 μm × 49 μm (pinhole diameter of 96 μm). For each cell, 47 image planes with an axial distance of 0.2 μm were acquired. Selected cells were initially positioned in the center of the z-stack by an autofocus routine before acquisition. Image acquisition took 196 s at each position, subsequent positions could be acquired every 212 s including positioning of the stage and autofocus routines. In total, confocal multicolor 3-D stacks were imaged at 796 single positions within 46.9 h.

Super-resolution imaging

Super-resolution imaging of selected cells was performed with a custom-built microscope using experimental protocols as described earlier, following direct stochastic optical reconstruction microscopy (dSTORM) (11). The beam of a multi-line argon-krypton laser (Innova70C, Coherent, Santa Clara, CA) is coupled into an acousto-optical tunable filter (27896; 8-channel) (AAOptics, Orsay, France) for wavelength selection and then combined with a UV laser beam (405 nm diode, Cube 405–100C, 100 mW, Coherent). The combined beams are focused onto the back-focal plane of an oil-immersion total internal reflection fluorescence objective (PlanApo N 60×, NA 1.45, Olympus) mounted onto an inverted microscope (IX71; Olympus). Emission light is separated from excitation light by appropriate dichroic mirrors and filters (AHF, Tübingen, Germany) and focused on an electron multiplying charged coupled device (EMCCD) camera (Ixon, Andor, Ireland). The samples were mounted on the stage, and the sample chamber was filled with ~200 μL switching buffer (100 mM β-mercaptoethylamine; MEA in PBS, pH 8). The exposure time was 50 ms with an electron multiplier gain of 200. Imaging intensities were 35 mW at 647 nm excitation wavelength and 65 mW at 532 nm, respectively. To adjust the photoswitching rates of Alexa Fluor 647, illumination with UV light was gradually increased from zero intensity to a maximum intensity of 0.1 mW. The two channels were recorded sequentially (first Alexa Fluor 647, followed by Alexa Fluor 532), chromatic shifts were corrected by using multi-color fluorescent beads (Tetraspeck, Invitrogen). dSTORM images with were typically reconstructed from 7000 frames using rapidSTORM (15).

Relocation of structures of interest

Relocation was done via coordinate transformation. In order to obtain a coordinate system on the sample that is independent of the coordinate system for the specific microscope system, we fixed three reference structures on the sample. These small pieces of titan coated glass, which each had a circular hole in the titan coating of 400 μm in diameter, were imaged first in transmission mode. Their central positions (x and y) were determined by applying an automatic Otsu threshold and calculating the centroid of the resulting area. The z position was taken from the position of the z-stage of the microscope after an autofocus routine. Based on these reference positions, an orthonormal basis of the sample was created, with the first reference position in the top left corner as origin. Widefield images of the sample were acquired, and the position of the lateral microscope stage and the axial z-drive was again recorded for each image. Individual cells of interest were selected for further imaging. The absolute position of these cells (position of the stage + position within the image) was transferred to the microscope stage independent coordinate system of the sample defined by the reference points. To transfer these coordinates to a second system, the referencing procedure was repeated and a coordinate transformation was performed.

Cell selection

Cellular positions for further confocal imaging were selected within the widefield images in two ways: 196 cells were selected manually by an expert marking the cellular positions within the image on a screen, and 600 cells were identified automatically based on nucleus segmentation within the images of the DAPI channel. The confocal images were then investigated for candidates for dSTORM acquisition and the appropriate cells were chosen manually.

Correlative workflow

Correlative image acquisition was performed as listed here:

  1. Image reference markers on widefield setup

  2. Image sample automatically on widefield setup

  1    2    3    4