Interference reflection microscopy

Abstract

Interference reflection microscopy (IRM) is an optical technique used to study cell adhesion or cell mobility on a glass coverslip. The interference of reflected light waves generates images with high contrast and definition. IRM can be used to examine almost any cell that will rest upon a glass surface, although it is most useful in examining sites of close contact between a cell and substratum. This unit presents methods for obtaining IRM images of cells with particular emphasis on IRM imaging with a laser scanning confocal microscope (LSCM), as most LSCM are already capable of recording these images without any modification of the instrument. Techniques are presented for imaging fixed and live cells, as well as simultaneous multi-channel capture of fluorescence and reflection images.

Figure 1

Suitable light paths for IRM imaging (A)Light path in a Laser Scanning Confocal Microscope. The laser, usually 453nm or 633nm, is reflected by the main beamsplitter into the microscope, through the objective where it reflects from the sample. The galvanometric mirrors in the scanner deflect the laser, scanning across the entire field to produce a full image. The reflected beams pass back through the neutral dichroic, 80/20 in this diagram, and are sent to the detector. The emission filter is chosen to allow the incident light to reach the detector. The pinhole can be either completely open or closed down to one Airy Unit.(B) Light path in an epi-illumination microscope. White light from a high pressure mercury or xenon lamp passes through a line isolating filter and a polarizer to produce monochromatic, linearly polarized incident light. The main beamsplitter reflects this light into the microscope, through the objective and then through a quarter-wave retardation plate. The emerging light is now circularly polarized. The reflected light is circularly polarized but with the opposite handedness from the incoming beam. After passing back through the quarter-wave plate, the reflected light is now linearly polarized with an orientation 90° with respect to the incident light. The reflected light passes through the neutral dichroic, 80/20 in this diagram, and through the analyzer, a second polarizing filter oriented 90° relative to the first polarizer, to reach the detector. Stray light from internal reflections in the microscope will be linearly polarized with the original orientation and will not pass through the analyzer to the detector.

Figure 2

Suitable configurations for obtaining IRM images with a Laser Scanning Confocal Microscope. (A) Detection of an IRM image alone. Any laser line can be used but usually a red or far-red line is used. This example shows a configuration using a 633 nm laser. The main beamsplitter is a neutral 80/20 or 70/30 dichroic depending on which is available in the microscope. This will direct the incoming laser to the sample but will still allow the reflected light to reach the detector. The reflected light could be sent to either of the two detectors in this diagram, as long as the emission filter will allow the light to pass. Here, the LP600nm filter will pass the 633 nm reflected light, but almost any LP filter would be suitable. Therefore, any detector in the system can usually be used fro capture an IRM image. (B) Simultaneous detection of fluorescent and IRM images. In this case, the configuration settings will be based primarily on those needed to detect the emitted fluorescence and the IRM image will be captured with whatever detector is free and has a suitable emission filter. This example show simultaneous imaging of green fluorescence (505–530nm) and reflected light. The laser line, 488 nm, and the main beamsplitter, NFT 488, are chosen to fit the characteristics of the fluorophore. The IRM image will be formed with the reflected light that passes through the main dichroic because it is not a perfect optical device. If the main dichroic is so efficient that no incident light is allowed through, it will be necessary to use a neutral dichroic for the IRM image. The emission side beamsplitters are chosen to send the emitted fluorescence to one detector and the incident wavelength to another. In this diagram, all light longer than 505 nm is gathered for the fluorescence image and all shorter wavelengths including 488 nm are used for the IRM image. If the fluorescence is very weak, this beamsplitter can even be set at 470 nm, to collect as much light as possible for the fluorescent image. The remaining reflected light will be sufficient to produce an IRM image. While the emission filter for the fluorescence image should match the emitted fluorescence, almost any long pass filter will allow the reflected light to reach the detector.

Figure 3

Diagrams representing the reflections generated by a cell on a coverslip. (A) Cell separated from the coverslip by a thin layer of thickness d. The first reflection, r₁, is generated at the glass/medium interface with incident angle θ₁, followed by refraction angle θ₂. The refracted beam continues through the medium and generates another reflection, r₂, at the medium/cell interface. R₂ will undergo a phase reversal because it occurs at a transition from lower to higher refractive index. The light that continues through the cell can be used to form a transmitted light image if an appropriate detector is available. (B) An additional reflection from the top of the cell may also contribute to the IRM image. When higher order interference fringes contribute to the image, the top of the cell will also contribute a third reflection, r_3.

Figure 4

Gray level intensities observed in a zero-order IRM image. A cell adhering to the coverslip will generate black areas on an IRM image in areas of close contact where the gap between the cell and coverslip is less than 15 nm (black arrow). If the gap is between 15 and 100 nm, the IRM image will show shades of gray (double headed arrow). The brightest areas on the IRM image will come from areas where the cell is about 100 nm from the coverslip (gray double headed arrow). If the cytoplasm is less than 1 m thick (feathered arrow), reflections from the top of the cell will affect the intensity of the IRM image, so it will not be possible to approximate the distance between the cell and the coverslip unless the thickness of the layer of cytoplasm is known.

Figure 5

IRM images of fixed NIH 3T3 cells. Images were obtained with a Zeiss LSCM 510 using a 63X oil objective NA 1.4. The configuration shown in Figure 2A was used for A and B. A similar configuration using a 543nm laser line was used for the last part of C. (A) Comparison of DIC and IRM images of the same cell. Left panel: low contrast DIC image showing intracellular organelles, details of the cell shape but containing little information about the adherent surface. Bar=10 μm. Middle panel: High contrast IRM image taken with an open pinhole. Right panel: IRM image taken with the pinhole set a 1 Airy Unit (1 AU). This is similar to the other IRM image, but the smaller pinhole gives a sharper image. (B) Effects of focal plane in images taken with narrow or open pinhole. Top panels: Images of the same field taken at successively higher focal planes with the pinhole set to 1 AU. The higher order interference fringes are evident with little change in focus. Bottom panels: A similar series of images taken at successively higher focal planes with the pinhole completely open. These images are primarily formed from the zero order interference fringe. Bar=5 μm. (C) The zero order interference image is similar when taken with light of different wavelengths. Top panels: The zero order (left side) and higher order (right side) images taken with a 633 nm laser. Bottom panels: The zero order (left side) and higher order (right side) images taken with a 543 nm laser. Bar=20 μm. The dark areas of close contact indicated with black arrowheads outlined in white remain fairly constant through this set of images. While there is some variation in the remaining gray level intensities in the two zero-order images, the overall appearance of the images is very similar. In contrast, the higher order interference are shifted substantially by changes in incident wavelength.

Figure 6

Application of IRM to migrating Dicytostelium discoideum cells. Dicytostelium discoideum amoeba were developed according to standard protocols to produce mobile cells. The images were taken using the same settings as Figure 5 with the addition of a transmitted light detector to the track. DIC optics installed on the microscope were used for the transmitted light image. Instead of single images, z stacks were collected at each time point and different focal planes were chosen to show the DIC or IRM images. Top panels: Time series of DIC image from the top of the z stack. The cells appear to glide over the coverslip. Bottom panels: IRM images from the bottom of the z stack in the same time series. Bar=10 μm These images clearly show that the amoeba move by extending a pseudopod that is not in contact with the coverslip and then forming a new contact site some distance from the cell body.

Figure 7

Simultaneous capture of fluorescence and interference reflection images of transfected Jurkat T cells expressing ZAP70-YFP. The configuration shown in Figure 2B was used to obtain these images. Top panels: Time series of YFP fluorescent images. The ZAP70-YFP clusters as the T cell spreads on the stimulatory coverslip. Bottom images: IRM images captured in the same track clearly show the T cell spreading across the coverslip. Bar=5 μm. A comparison of the two images shows that ZAP70 clusters at the contact sites.

Figure 8

Quantification of an IRM image. Left panel: IRM image of fixed mouse T cells obtained using the same conditions as the cells in Figure 5. The T cells were allowed to spread on a stimulatory coverslip and were then fixed in 2.5% paraformaldehyde. Middle panel: An auto-thresholding algorithm was used to outline the T cells based on the gray level contrast and object shape. Some touching cells were improperly outlined together and were then separated by hand, for examples # 23 and #26. Right panels: The same field of cells visualized with an LUT where the gray level of each cell is proportional to its area. Lighter colors indicate cells with larger areas.