The plasma membrane of polarized epithelial cells is divided into two clearly separate parts; an apical membrane domain facing the organ lumen or, in cell culture, the medium, and the basolateral domain, which is connected with adjacent cells or the extracellular matrix. The two membrane domains are separated by tight junctions and have their own specific protein and lipid content. A classic model system for studying the intracellular trafficking of these components into their particular target membrane is provided by MDCK cells (Madin-Darby-Canine-Kidney, which are used in this article). The study of intracellular sorting processes is of fundamental importance for understanding the cell’s protein and lipid trafficking system. Wrong targeting of these components can cause organ defects such as those found in various diseases of the lungs, the intestinal tract and the kidney. Prominent examples of this are cystic fibrosis or congenital sucrase-isomaltase deficiency.
The subject of our team’s research at the Institute of Cell Biology and Cell Pathology of the Philipps University Marburg is the elucidation of protein sorting to the apical cytoplasmic membrane of MDCK cells. The application of TIRF microscopy (in short: TIRFM) proved extremely useful for this work. TIRFM employ the physical phenomenon of total reflection to generate a so-called evanescent field. This field only extends over a penetration depth of a few hundred nanometers and excites fluorophores in its range near to the coverglass (Figure 1).
With the Leica AM TIRF MC mentioned in this article and based on the Leica DMI6000 B fluorescence microscope, the penetration depth of the excitation light into the cell can be continuously adjusted from 70 to 300 nm. In practice, this can be used to excite fluorescence-tagged proteins in the immediate vicinity of the cytoplasmic membrane, in our case of polarized epithelial cells.
Fig. 1: Experimental setup for apical TIRFM: MDCK cells are sown on a porous membrane (PET filter) and incubated for several days. To view the apical part of the cells, the PET filter is removed from its plastic mount and placed upside down in a Bachofer chamber in close contact with the coverglass that is fitted there. To make sure the cells stay close to the coverglass, a 4 g weight is put on top of the PET filter. Due to this arrangement, the apical part of the MDCK cells that is lined with microvilli lies within the evanescent field produced at the surface of the coverglass by the total reflection of the laser beam. Only fluorescent particles within this field are excited and therefore visible.
MDCK cells are sown on a porous membrane (PET filter) and incubated for several days. To view the apical part of the cells, the PET filter is removed from its plastic mount and placed upside down in a Bachofer chamber in close contact with the coverglass that is fitted there. To make sure the cells stay close to the coverglass, a 4 g weight is put on top of the PET filter. Due to this arrangement, the apical part of the MDCK cells that is lined with microvilli lies within the evanescent field produced at the surface of the coverglass by the total reflection of the laser beam. Only fluorescent particles within this field are excited and therefore visible.
Observation of the apical membrane domains of MDCK cells is a particular technical challenge, because, unlike TIRF excitation of fluorophores in the basolateral membrane domain, for which cells can be directly cultivated on a coverglass, cells for inspecting the apical side first have to form a polarized epithelium in a petri dish. This is done by growing them on porous PET filters to obtain an epithelial monolayer. The filters with the polarized cell layer are then cut out of their mount and put upside down onto the coverglass of a Bachofer chamber. Physical proximity between the apical membrane of the cells and the evanescent field at the coverglass is achieved by putting an additional weight on the filter (Figure 1).
Figure 2a shows a comparison of conventional epifluorescence images and the image of various fluorophores on the apical membrane domain of MDCK cells made by TIRFM. Here, a predominantly apical membrane protein, the neurotrophin receptor (p75), was studied in comparison with its apical sorting receptor, galectin-3. The two proteins were labelled with fluorescence proteins, p75 with eGFP (p75-GFP) and galectin-3 with DsRed (Gal3-DsRed) for imaging in the fluorescence microscope. Whereas conventional epifluorescence images tend to produce a diffuse light over the whole cell body in both channels, the TIRF images reveal extremely distinct and clearly structured objects. Most of these are finger-like protuberances of the apical membrane, so-called microvilli, decorated with p75 or, in some cases, with galectin-3.
Looking at the basolateral side of the cells (Figure 2b), the TIRF images again show clearer structures despite the smaller amounts of protein in this membrane domain. Cell borders are far more clearly demarcated and have a more amorphous line compared with the epifluorescence image. The image also shows the limited excitation range of the TIRFM in a narrow field around the cytoplasmic membrane. This is shown by intracellular Gal3-DsRed-positive objects, which appear in epifluorescence mode, but are not detected by the evanescent field of the TIRFM.
Fig. 2a–b: Comparison of epifluorescence microscopy and TIRFM,
a: For apical TIRFM, MDCKp75-GFP/Gal3-DsRed cells were cultivated on PET filters and viewed using the Leica AM TIRF MC based on the Leica DMI6000 B fluorescence microscope. Whereas the epifluorescence images (left) only show a diffuse light, the TIRF images of the same cells (right) are capable of resolving tiny structures (arrow tips). These are often tubular and exhibit the fluorescence of both markers used.
b: For the basolateral TIRFM, the same cell line was directly cultivated on the coverglass of a Bachofer chamber and also viewed in the TIRF microscope. Again, TIRF images delivered better resolution. Also, red signals far below the cell surface are not excited in the TIRFM (arrow tips). Scale bar: 10 µm
To find out whether the long structures on the apical side of the cells illustrated in Figure 2a are microvilli or intracellular structures, MDCKp75-GFP or MDCKGal3-DsRed cell lines were subjected to immunofluorescence staining without permeabilizing the cells first. Antibodies therefore had no access to the cell interior, but were only able to stain structures on the cell surface. Here, it was evident that some objects were intracellular, probably endosomal compartments located immediately below the apical membrane of MDCK cells. These objects can only be visualized with fluorescing fusion proteins, but not with immunofluorescence. As a control, the cells were permeabilized with the detergent saponin before antibody treatment. In this case the immunofluorescence images were identical with those of the GFP or DsRed signal (Figure 3c).
Fig. 3a–c: The TIRFM images show structures both on and just below the cytoplasmic membrane.
a: MDCKp75-GFP cells were cultivated on PET filters and subjected, without previous permeabilization, to immunofluorescence stain-ing using an anti-p75 antibody. Thus only p75 structures on the cell surface were stained, not inside the cell (left). A parallel look at the p75-GFP signal shows that the TIRFM picks up structures on and inside the cytoplasmic membrane as well as in a small area below it. Here, tubular elements appear (arrow tips) that vary from the antibody signal on the surface.
b: A similar picture is obtained with MDCKGal3-DsRed cells. Here too, intracellular Gal3-DsRed signals appear that were not picked up by the antibody on the surface (arrow tips).
c: In the control, MDCKp75-GFP cells were permeabilized with a detergent before immunofluorescence staining. Here, the two signals are identical. Scale bar: 10 µm
TIRFM is a useful tool for viewing cellular processes in the plasma membrane or its immediate environment. It can be used for observing either fixed or living tissue, even in several fluorescence channels if necessary. These preconditions give reason to hope that apical membrane proximal processes or structures that could not be clearly detected with techniques available so far will soon be visualized with the TIRFM technique.