McEwan et al.

Immunocytochemistry and in situ Hybridization - is the order important?

Neil R. McEwan Ph.D.*

Department of Biological Sciences, University of Stirling, Stirling, FK9 4LA Scotland
*Current address: Land Resources Department, Scottish Agricultural College, Aberdeen AB21 9SJ, Scotland

Correspondence should be addressed to: Neil McEwan Ph.D.

Keywords: in situ hybridization, immunocytochemistry, filamentous actin.


The importance of performing either immunocytochemistry or in situ hybridization first in experimental work was investigated. It was found that it was essential to perform immunocytochemistry first, otherwise the antigenicity of the proteins may be lost following incubation in the hybridization solution.


Immunocytochemistry and in situ Hybridization (ISH) are two techniques used routinely for the determination of either cell type distribution, or sub-cellular distribution, of proteins and nucleic acids respectively. In general, most researchers will use either one or the other, but there are fewer reports of both techniques being used on a single sample. To date, there is no documentation to demonstrate which of the two techniques is the one which should be performed first. This paper presents evidence that it is essential that the immunocytochemistry be performed first.


Growth of cells
Oligodendrocytes were grown on glass coverslips as previously described [1]. Cells were fixed in 4% (w/v) paraformaldehyde in PBS for 20 minutes at 37'C. Unreacted aldehydes were quenched with filter-sterilized 1M glycine, PBS (pH 7.4) for 2 x 10 minutes and washed briefly in PBS. Cells could be stored for several days at 4'C in filter-sterilized PBS, 0.05% (w/v) sodium azide (pH 7.4) before being used for experimental work.

Cells were washed several times in autoclaved PBS to remove all traces of sodium azide, followed by a 5 minute incubation in 0.1% Triton X-100, PBS to permeablise the cells. After permeablisation, cells were washed several times in PBS to remove all traces of detergent. All solutions for immunocytochemistry were treated with DEPC, where possible. With all other solutions (e.g. those containing Tris) being autoclaved instead. Rhodamine Phalloidin (RP) was purchased from Sigma Biochemicals. Non-specific binding sites for phalloidin were blocked using Blocking Buffer [0.2% (w/v) gelatin, PBS] for 45 minutes. Filamentous actin was detected by incubating with RP by diluting it in Blocking Buffer (1:200 dilution), and incubating it in the dark for 1 hour. Unincorportated RP was removed by washing for 3 x 15 minutes in Blocking Buffer at room temperature, followed by a single overnight wash in Blocking Buffer at 4'C. All washes were performed in the dark.

in situ Hybridization
Oligonucleotide probes were raidoactively labeled using T4 polynucleotide kinase (BRL), subject to the manufacturer's specifications. It was found that 32P labeled probes provided sufficient resolution for use of light microscopy. in situ Hybridization (ISH) was performed as previously described [2, 3], using the oligonucleotide ATA GCA CAG CTT CTC TTT [2].
Briefly, this entailed a prehybridization for a minimum of 1 hour at 37'C in a humidified chamber, using 50 µl of buffer per 10mm2 coverslip. The buffer comprised 0.1% [w/v] dextran sulphate, 5 x SSPE (1 x SSPE = 180mM NaCl, 10mM sodium phosphate (pH 7.7), 50% [v/v] formamide, 8 µg/ml E. coli tRNA. After the prehybridization incubation 0.8 pmoles of 32P- labeled probe was added to the buffer and incubated for 2 hours at 37'C. Cells were washed 10 times for 5 minutes at room temperature with PBS, 0.9% [w/v] NaCl to remove all non-hybridized probe, followed by a single wash at 37'C for 5 minutes to remove any non-specifically hybridized probe. Finally, cells were washed for a further 2 x 5 minutes at room temperature.
Coverslips were attached to microscope slides with the cell side upwards. Slides were then dipped in Ilford K5 emulsion (diluted 1:2 with distilled water), allowed to expose for 14-16 hours in a dehumidified box and developed in Kodak developer. Silver grains developing were fixed with 30% [w/v] sodium thiosulphate and mounted in 0.1 x PBS, 90% [v/v] glycerol.

Order of incubations
Both orders of incubation were attempted; i.e. either ISH followed by immunocytochemistry, or vice versa. As an additional precaution, after performing one type of incubation cells were again fixed in 4% [w/v] paraformaldehyde (as described above) before performing the second type of reaction. Where immunocytochemistry was performed first, all ISH reactions were performed in the dark to minimise bleaching of the fluorescent signal.


Figure 1 shows an oligodendrocyte which has been viewed using Nomarski optics to show the typical structure of the cell type. It can clearly be seen that there are a considerable number of major processes radiating from the central body of the cell and that the major processes in turn subdivide to give smaller minor processes.

Figure 2 shows an oligodendrocyte whose actin mRNA distribution has been assessed by ISH. The central area of the cell is quite dark, and the areas around the cell body have the major amount of signal present. This is in keeping with the observations previously published [2]. This type of pattern resulted independently of the order of ISH versus immunocytochemistry, showing that the actin messages tend to be transported to the peripheral areas of the cell.

Figure 3 shows an oligodendrocyte whose filamentous actin distribution has been assayed using Rhodamine-Phalloidin prior to performing in situ hybridization. The pattern of the protein distribution is similar to that described for the actin messages in Figure 2, and is in keeping with published observations for filamentous actin in oligodendrocytes [1, 2].

Figure 4 shows an oligodendrocyte whose filamentous actin distribution has been assessed after previously performing ISH. Clearly there is a considerable difference in the patterns detected in Figures 3 and 4. In Figure 4 there is no longer the same clearly defined structure of cell type as can be seen in any of the three previous figures.


It is not clear what causes this alteration in the conformation detected, although it may be surmised that it has something to do with the alteration in the structure of the filamentous actin within the cell. It is not well established how radically the incubation in ISH buffer is likely to affect other proteins within the cell, as few proteins have been described which show such a pronounced distribution in as highly defined a cell shape as the oligodendrocyte. For example, if a cell is less highly structured, such as a fibroblast, it is quite possible that changes such as the ones described in Figures 3 and 4, may be more difficult to detect. Nevertheless, it is clear from this work that the order of performing either ISH or immunocytochemistry is particularly important in this cell type for this particular protein, and that it is possible that similar results might apply to other systems, but that it might be difficult to detect. Thus, it appears that the optimal sequence for performing ISH and immunocytochemistry on a single piece of tissue is immunocytochemistry followed by in situ hybridization. In addition, it seems that adding an extra fixation step between immunocytochemistry and ISH is advantageous, as there appeared to be a slight loss, although nowhere near as pronounced as that seen in Figure 4, in the filamentous actin distribution when the extra fixation step is omitted.


This work was funded by a Science and Engineering Research Council studentship.


  1. Wilson R. and Brophy P.J. (1989) Role for the oligodendrocyte cytoskeleton in myelination. J. Neurochem. Res. 22 439-448

  2. McEwan N.R. (1996) 2'3'-CNPase and Actin Distribution in Oligodendrocytes, Relative to their mRNA. Biochem. & Mol. Biol. Internat. 40 975-979

  3. McEwan N.R. (1997) Comparison of the Subcellular Distribution of the Messages Encoding PLP and DM-20. Biochemal Genetics 35 51-58

© 1997 Epress Inc.