For many cell or tissue based experiments, microscopic imaging is likely to be the end point, and in this series of articles we want to illustrate how some joined-up thinking regarding tissue or cell collection, fixation and processing, subsequent staining, microscopic imaging and analysis will produce better results.
Make imaging easier
For example, intense well-stained fluorescent cells or tissues will make imaging much easier whether you are using widefield, confocal or super resolution microscopy. Good staining will be achieved by using appropriate detection systems and suitably fixed cells or tissues.
The aim of fixation
The aim of fixation is to preserve cells or tissues in as near a life like condition as possible, prevent autolysis and putrefaction, and protect the tissue from subsequent processing. Fixatives have different actions e.g. crosslinking, precipitative, coagulative etc. They also have different penetration rates, usually measured in depth penetrated per mm per hour. I normally recommend fixing in 20x the volume of fixative to tissue and fix for consistent periods of time- for example, I would normally fix samples 0.5 cm thick in at least one dimension for 24 hours in Neutral Buffered Formalin, or for 6 hours in Bouins fluid. I would then transfer samples into 70% ethanol (the first stage in many routine paraffin processing protocols) prior to processing.
Temperature is also a consideration- low temperature will reduce autolysis in tissues, but lead to slower penetration rate, so choose what is important to you. I normally fix at room temperature with exceptions for tissue to be processed to resin or for some sensitive antigens where low temperature fixation is recommended.
Choice depends on the end
The choice of fixative is determined by the end point- for example, if an extremely high degree of morphological preservation is required for light or electron microscopy on resin sections, then the likelihood is that a gluteraldehyde based fixative (perhaps in combination with paraformaldehyde) would be the method of choice. As neither of these are effective at preserving lipid (which would be lost in the subsequent processing to resin or plastic) the tissue would need to be post-fixed in osmium tetroxide.
A word or two about glutaraldehyde
As glutaraldehyde penetrates tissue slowly, tissue pieces would need to be much smaller or fixation times much longer. However, like many things in life, there is a trade-off for obtaining this high degree of morphological preservation- generally it is much more challenging to perform empirical staining techniques (such as Haematoxylin and Eosin, PAS etc.) or molecular tissue probing techniques using antibodies (immunohistochemistry etc.) or nucleic acid probes(in-situ hybridisation) etc.
Brrr- the cold world of the cryostat!
The other end of the scale is to use frozen (or cryostat) sections with no or minimal fixation. For example, a frozen section can be produced air dried onto a slide and directly stained. Therefore, it’s not had the chemical exposure of fixatives or the alcohols, solvents and the heat which is associated with paraffin processing. For certain techniques this has huge advantages e.g. the demonstration of lipids, immunohistochemistry and immunofluorescence for certain antigens, and for brain tissue where it can give excellent results. However, the generally poorer morphology obtained for other tissues overall makes this less mainstream.
Like our politicians, most of us try to appeal to the middle ground! For the time being, the middle ground in fixation is formalin fixed paraffin embedded tissue. It is used extensively in diagnostic pathology- pathologists are familiar with the chromatin pattern of tissues fixed in this way and it provides a good degree of morphological preservation and chemical preservation such that proteins are well preserved, although often masked (which can be undone) and mRNA and DNA can also be hybridised.
Different disciplines, different preferences
Antibody manufacturers often give conditions for use on FFPE tissues, probably because they have optimised and validated these reagents on this easily available type of sample. But like political parties, different disciplines have different preferences! For example, for studying macrophages in one of our research centres, Methacarn is the fixative of choice.
One solution to fix them all? Not quite…
However, as the title says “think before you fix” it doesn’t mean 4% Neutral Buffered Formaldehyde is the best fixative for every application, it could in fact be detrimental to obtaining the best results in an antibody based experiment, or indeed morphological preservation. For example, in the early 90’s I used to study spermatogenesis and therefore worked on the testis. The testis is a fairly fluid filled organ with a heterogeneous population of cells contained within two main compartments, the tubules and interstitium.
Tissue dictates fixation
In the rat there are 14 stages of spermatogenesis so that in any cross section of a testis the seminiferous tubule could be characterised as one of 14 stages. These stages are distinguished by subtle differences in the morphological appearance of different cell types. In the testis, I wanted to look at mRNA expression using in-situ hybridisation- most literature recommended 4% paraformaldehyde for in-situ hybridisation, so what fixative did I choose? The answer is of course, we chose, well………several! At least for the initial evaluation- the reason was that we could see from the published literature that paraformaldehyde would not give the degree of morphological preservation which was needed to recognise the subtle differences in the cellular population of the testis.
Decades old, but still giving results
We might get an mRNA signal, but if we could not identify precisely the cell type and stage of spermatogenesis, then the information obtained would be of little value. So we chose several and, in the end, selected Bouins fluid which although it gave a slightly less sensitive signal by ISH, it gave the morphological preservation I needed to identify at which stage of spermatogenesis the gene was expressed. Perhaps I was lucky because these 20 year old samples can still be used for ISH today and it has proven in my hands to be an extremely good fixative for immunohistochemistry.
Suck it and see!
Would I recommend you fix everything in Bouins? Of course not! It was the ideal fixative for our tissue and our desired end point, but it does not mean it is ideal for every application. Finding your ideal fixative might involve trying different ways of fixing your samples, preparing them in different ways and applying different staining strategies.
For most applications, sticking to the middle ground, certainly initially, is a good strategy for fixation but if it is either not giving you the end point that you need, or is just plain not working, then keep an open mind and explore different fixation strategies to obtain the best results. Do this near the start of a project where you have the flexibility to change direction.
And finally, once you have made your fixation choices (e.g. fixative, time and temperature) then keep it consistent and, who knows, you could still be using these samples in 20 years’ time!