Imaging of Estrogen Receptor-α in Rat Pial Arterioles using a Digital Immunofluorescent Microscope

Previously, we showed that after transient global ischemic brain injury the pial artery capacity to change diameter in response to both vasodilators ^{7, 8} and vasoconstrictors ⁹ is markedly depressed in young ovariectomized rats. Chronic estrogen repletion restores vasomotor responses ^7-9. Further, delayed estrogen replacement reverses the beneficial effects of estrogen on pial artery vasodilatory capacity ¹⁰ leading us to question whether long-term estrogen depletion affects ER-α density in the cerebrovasculature. We planned to use immunofluorescent labeling combined with western blotting to evaluate changes in ER-α expression in response to chronic estrogen depletion. Because we had some difficulty in visualization of discrete groups of receptors we modified the method that we demonstrate here. Therefore, our primary goal in this study was to first develop a relatively simple method to optimize the visualization of the receptors in pial arterioles. In agreement with several reports ^6,11,12 we found it necessary to make adjustments for the type of tissue we were probing, the sample thickness, the distribution of ER-α in the vessel, as well as the degree of non-specific binding.

Our method is very effective to visualize the presence of ER-α in pial vessel slices. But as others have previously pointed out the successful use of these methods is highly dependent on the specificity and concentrations of the antibodies, the subcellular location of the proteins of interest, fixation and blocking protocols and tissue handling ¹². Prior to visualization of our vessel preparations we spent time to work out all these variables. We have highlighted how we needed to adjust some of our approach in this paper.

We set out to optimally visualize ER-α in pial arterioles and this presented a few challenges. First, the average size of a rat pial artery is between 35-50 μm making isolation and removal from the brain surface somewhat tricky. We found that perfusion with Evans blue dye prior to sacrifice makes dissection of the arteries easier. Evan’s blue can affect fluorescence, but we feel that the advantage of using this dye outweigh its minor disadvantages. In our protocol the vessels are washed several times. This minimizes the residual dye in the vessels and reduces the possible effects of the dye on fluorescence. Another difficulty we encountered was that the thickness of the pial vessel segments made clear visualization of the receptors difficult. Additionally, we could not obtain particularly clear images probably due to the fact that ER-α was scattered throughout several cell layers. We tried working with longitudinal vessel slices (as others have), but due to the small pial size it was difficult to handle the vessels and prevent twisting of the vessels while mounting on slides even using a microscope. Finally, we successfully modified our technique by using a cryostat and cutting the vessel cross-wise in to 8 μm thick rings. The rings are easier to handle and several can be mounted on one slide.

Once we acquired the vessels we felt it necessary to test the effectiveness of first flash freezing the tissue then fixing the vessels as suggested by Danseshatalb and associates ¹¹. Although we found marginally less background immunofluorescence, we also noted that the receptors stained less intensely making visualization of the receptors more difficult. Consequently, we prefer to first to freeze the entire brain and store until needed (generally between 1-2 months). We then pull off the vessels as described and fix the vessels prior to slicing with the cryostat. as we demonstrate here.

One necessary step is to run controls with both the primary and the secondary antibody separately to determine specificity and background immunofluorescence. Estrogen receptor-α antibodies can be difficult to work with. In fact, we tried 3 different primary antibodies (1 monoclonal and 2 polyclonal) before we were satisfied with the success of our receptor staining. As part of our controls we first verified the specificity of the primary antibody by omission of the addition of the primary antibody. Figure 2D shows the vessel without primary antibody. There is no staining and just a little background signal. Since the primary antibody is polyclonal, there is some diffuse non-specific background staining. Secondly, we added the primary antibody, but not the secondary antibody (Fig. 2E). One can see a nice margin representing the autofluoresence along the endothelial borders of the vessels. This is consistent with the work of Dan and associates ⁵. However, neither the small points of immunofluorescence that represent ER-α , nor the granular appearance reflecting the presence of numerous receptors can be detected without the combination of both the primary and secondary antibodies. We noted a distinct pattern of autofluoresence along the endothelial margins of the vessels. These independent tests for antibody specificity are important as they confirm the specificity of the antibody for ER-α and the secondary antibody binds to the primary.

In conclusion, as demonstrated by others ^{4, 5} brain blood vessels contain numerous ER-α sub-type receptors. In our hands, preparing 8μm cross sectional slices from isolated pial arteries enhances staining and visualization of the receptors. Traditional methods use confocal microscopy to examine the samples. Having the capacity to view serial optical sections from thick specimens is an important benefit of confocal microscopy. However, purchase of a good confocal microscope can be very expensive. By using Extended Depth (EDF) software, we found that could reduce our equipment costs. Using a fluorescent microscope with EDF we obtained clear images and are able to visualize pial arterial ER-α on several levels using Z-stacks. A significant benefit of the EDF software is that it permits image capture in a manner to effectively create a 3-D picture from the images. For laboratories that do not have access to a confocal microscope or the funds to purchase one a digital fluorescent microscope with appropriated software may be a practical and less costly option depending on the goals of the investigator.