Effects of preparation and fixation on three quantitative fluorescent cytochemical procedures

Summary A series of experiments was undertaken in which cells dissociated from the abdominal lymph nodes of mice were lightly centrifuged into slides and fixed either wet or after drying in 70% ethanol, 1% glutaraldehyde, 1% formaldehyde, or neutral formalin. Three fluorescent cytochemical methods were evaluated: staining of DNA with mithramycin; fluorochroming of basic groups of proteins with brilliant sulfaflavine (BSF); and staining of sulfhydryl and disulfide groups with N-(7-dimethylamino-4-methylcoumarinyl)maleimide (DACM). In the case of mithramycin, the best results were obtained after fixation in 70% ethanol without drying. Staining of dried preparations fixed in 1% glutaraldehyde also yielded reasonably consistent results, although the fluorescence was lower, and the variability higher, than in the group fixed without drying in 70% ethanol. The use of fixatives containing formaldehyde resulted in fluorescence values of only about onethird those of the other two groups, and the variability of the data was higher. In material stained with BSF, satisfactory results were obtained in preparations fixed without drying in neutral formalin containing mersalyl acid. Other fixatives could be used, but the resulting coefficients of variation were higher than those of formalin-fixed material. Sulfhydryl to disulfide ratios approaching those expected from biochemical evidence were obtained in DACM-stained material only after fixation without drying in neutral formalin containing mersalyl acid. Inverted sulfhydryl-disulfide ratios were observed in material fixed without drying in 70% ethanol; and in dried metarial fixed in 1% formaldehyde, neutral formalin, or 1% glutaraldehyde.     

The influence of fixation and tissue preparation on the immunohistochemical demonstration of fibronectin in human tissue

The influence of fixation and tissue preparation on the immunohistochemical localization of human fibronectin in gastrointestinal tract tissue has been examined using indirect immunoperoxidase technique. The most optimal staining result with strong intensity and well defined localization was obtained on frozen sections of unfixed material. Nearly identical results with improved morphology were obtained when staining paraffin sections of tissue fixed in 96% ethanol, 96% + 1% acetic acid and absolute acetone. All other fixatives tested, 10% neutral buffered formalin. Lillie's AAF, Bouin's fixative, Clarke's fixative, 4% formaldehyde, 4% formaldehyde + 0.5% cetylpyridiniumchloride (F-CPC), 4% formaldehyde +0.1% glutaraldehyde gave unsatisfactory results. However, proteolytic digestion with pepsin of paraffin sections prior to staining of buffered formalin and F-CPCfixed material gave results comparable with those obtained on unfixed frozen sections are regards definition of the staining whereas staining intensity was decreased in some degree. No improvement was observed when using proteolytic digestion of tissue fixed in other fixatives.     

The influence of fixation and tissue preparation on the immunohistochemical demonstration of fibronectin in human tissue

Summary The influence of fixation and tissue preparation on the immunohistochemical localization of human fibronectin in gastrointestinal tract tissue has been examined using indirect immunoperoxidase technique. The most optimal staining result with strong intensity and well defined localization was obtained on frozen sections of unfixed material. Nearly identical results with improved morphology were obtained when staining paraffin sections of tissue fixed in 96% ethanol, 96%+1% acetic acid and absolute acetone. All other fixatives tested, 10% neutral buffered formalin, Lillie's AAF, Bouin's fixative, Clarke's fixative, 4% formaldehyde, 4% formaldehyde+ 0.5% cetylpyridiniumchloride (F-CPC), 4% formaldehyde +0.1% glutaraldehyde gave unsatisfactory results. However, proteolytic digestion with pepsin of paraffin sections prior to staining of buffered formalin and F-CPCfixed material gave results comparable with those obtained on unfixed frozen sections as regards definition of the staining whereas staining intensity was decreased in some degree. No improvement was observed when using proteolytic digestion of tissue fixed in other fixatives.     

Staining of interphase nuclei and mitotic figures in cultured cells with alcian blue 8GX

Cultured endothelial cells derived from bovine calf pulmonary artery were subjected to a variety of fixatives and stained with 1% Alcian blue 8GX at pH 2.59 to 3.26. Within this range of pH, interphase nuclei and especially mitotic figures were (a) strongly stained in cells fixed with 10% formalin (phosphate buffered or unbuffered) or 2.5% buffered glutaraldehyde, (b) weakly stained or unstained in cells fixed in formaldehyde containing divalent cations, and (c) unstained in cells fixed in acetic acid-containing fluids. However, optimal nuclear staining with Alcian blue under the conditions of this study was judged to be achieved after fixation with neutral phosphate buffered 10% formalin. Endothelial cell cytoplasm exhibited a similar fixative-dependent staining. At pH 2.59 the cytoplasm of interphase cells fixed in formaldehyde (containing no divalent cations) or glutaraldehyde remained unstained; however, at higher pH cytoplasmic staining did occur and it increased as pH increased. In contrast, when these latter fixatives were employed the cytoplasm of mitotic cells stained at all pH levels tested. In cultured endothelial cells after appropriate fixation, 1% Alcian blue 8GX (pH 2.59) was found to possess the ability to stain nuclei with a selectivity and intensity that compared favorably to those of the Feulgen reaction of Heidenhain iron hematoxylin but without the latters' length and complexity. Therefore, this procedure may provide a rapid, simple, and selective method for visualizing interphase nuclei or mitotic figures, or both in the majority of cultured cells.     

Staining of interphase nuclei and mitotic figures in cultured cells with Alcian blue 8GX

Summary Cultured endothelial cells derived from bovine calf pulmonary artery were subjected to a variety of fixatives and stained with 1% Alcian blue 8GX at pH 2.59 to 3.26. Within this range of pH, interphase nuclei and especially mitotic figures were (a) strongly stained in cells fixed with 10% formalin (phosphate buffered or unbuffered) or 2.5% buffered glutaraldehyde, (b) weakly stained or unstained in cells fixed in formaldehyde containing divalent cations, and (c) unstained in cells fixed in acetic acid-containing fluids. However, optimal nuclear staining with Alcian blue under the conditions of this study was judged to be achieved after fixation with neutral phosphate buffered 10% formalin. Endothelial cell cytoplasm exhibited a similar fixative-dependent staining. At pH 2.59 the cytoplasm of interphase cells fixed in formaldehyde (containing no divalent cations) or glutaraldehyde remained unstained; however, at higher pH cytoplasmic staining did occur and it increased as pH increased. In contrast, when these latter fixatives were employed the cytoplasm of mitotic cells stained at all pH levels tested. In cultured endothelial cells after appropriate fixation, 1% Alcian blue 8GX (pH 2.59) was found to possess the ability to stain nuclei with a selectivity and intensity that compared favorably to those of the Feulgen reaction or Heidenhain iron hematoxylin but without the latters’ length and complexity. Therefore, this procedure may provide a rapid, simple, and selective method for visualizing interphase nuclei or mitotic figures, or both in the majority of cultured cells.     

Effects of fixation and processing on the immunohistochemical visualization of type-I, -III and -IV collagen in paraffin-embedded liver tissue

Summary We compared the effects of different fixatives and enzymatic-digestion procedures on the immunohistochemical visualization of type-I,-III and-IV collagen in paraffin-embedded normal human liver sections. None of the fixatives tested allowed the staining of these antigens without prior enzymatic digestion. The best results i.e. strong staining intensity and well-defined localization, were obtained when liver tissue was fixed in Bouin's fluid or in other solutions containing picric acid. Several other fixatives, including Carnoy's fluid, Lillie's AAF, 10% neutral formalin and 96% ethanol, gave unsatisfactory results. Pepsin was ineffective for unmasking type-I and-III collagen antigens, and was only partially effective for visualizing the type-IV collagen antigen. The best results were obtained when material fixed in Bouin's fluid was embedded in paraftin and digested with trypsin. Using this procedure, the results were comparable to those obtained in unfixed frozen sections with respect to the staining intensity, specificity and non-specific staining.     

A Comparison of Techniques Useful for Preparing Nematodes for Scanning Electron Microscopy

Second-stage juveniles of Meloidogyne incognita were prepared by several different techniques for scanning electron microscopy (SEM). Sequential fixation in the cold (4-8 C) was superior to rapid fixation at room temperature, glutaraldehyde and glutaraldehyde-formalin were better fixatives than formalin alone, and critical point drying with carbon dioxide or Freon gave similar results that were only slightly better than air drying with Freon. Freeze drying sequentially fixed nematodes from 100% ethanol in liquid propane produced the best preserved specimens with the fewest artifacts. Specimens of various free-living and plant-parasitic nematodes were prepared for SEM by freeze drying. This technique was adequate for most genera but unsatisfactory for a few. Although each genus may require a different procedure for optimum preservation of detail, sequential fixation with glutaraldehyde and freeze drying are comparable and often superior to commonly used techniques for preparing nematodes for SEM.     

Simultaneous demonstration of bone alkaline and acid phosphatase activities in plastic-embedded sections and differential inhibition of the activities

Summary Bone alkaline (AlP) and acid phosphatase (AcP) activities were simultaneusly demonstrated in tissue sections obtained from mice, rats, and humans. The method involved tissue fixation in ethanol, embedding in glycol methacrylate (GMA), and demonstration of AlP and AcP activities employing a simultaneous coupling azo dye technique using substituted naphthol phosphate as a substrate. AlP activity was demonstrated first followed by AcP activity. Both enzyme activities were demonstrated in tissue sections from bones fixed and/or stored in acetone or 70% ethanol for up to 14 days or stored in GMA for 2 months. AlP activity in tissue sections from bones fixed in 10% formalin, 2% glutaraldehyde, or formal-calcium, however, was markedly inhibited after 3–7 days and was no longer detectable after 14 days of fixation. Moreover, AlP activity was diminished in tissue sections from bones fixed in 70% ethanol or 10% formalin and subsequently demineralized in 10% EDTA (pH7) for 2 days, and the activity was completely abolished in tissue sections from bones subsequently demineralized in 5% formic acid: 20% sodium citrate (1:1, pH 4.2) for 2 days. Methyl methacrylate (MMA) embedding at concentrations above 66% completely inhibited AlP activity. AcP activity, however, was only partially inhibited by formalin, glutaraldehyde, or formal-calcium after 7 or 14 days of fixation or by MMA embedding and was unaffected by the demineralizing agent formic acid-citrate for 2 days. While AcP activity was preserved in bones fixed in formalin and subsequently demineralized in EDTA, the activity was completely abolished when EDTA demineralization was carried out on bones previously fixed in 70% ethanol. These results indicate that bone AlP and AcP activities can be demonstrated simultaneously in the same section using a simple tissue preparation technique and that the activities are retained in tissues fixed and/or stored in acetone, 70% ethanol or GMA, but are differentially inactivated by other fixatives studied, and by EDTA, formic acid-citrate, and MMA embedding.Abbreviations AcP acid phosphatase - AlP alkaline phosphatase - GMA glycol methacrylate - MMA methyl methacrylate - EDTA ethylenediaminetetraacetic acid     

Effects of fixation and processing on the immunohistochemical visualization of type-I, -III and -IV collagen in paraffin-embedded liver tissue

We compared the effects of different fixatives and enzymatic-digestion procedures on the immunohistochemical visualization of type-I, -III and -IV collagen in paraffin-embedded normal human liver sections. None of the fixatives tested allowed the staining of these antigens without prior enzymatic digestion. The best results i.e. strong staining intensity and well-defined localization, were obtained when liver tissue was fixed in Bouin's fluid or in other solutions containing picric acid. Several other fixatives, including Carnoy's fluid, Lillie's AAF, 10% neutral formalin and 96% ethanol, gave unsatisfactory results. Pepsin was ineffective for unmasking type-I and -III collagen antigens, and was only partially effective for visualizing the type-IV collagen antigen. The best results were obtained when material fixed in Bouin's fluid was embedded in paraffin and digested with trypsin. Using this procedure, the results were comparable to those obtained in unfixed frozen sections with respect to the staining intensity, specificity and non-specific staining.     

The effect of different fixatives and length of fixation time on subsequent AgNOR staining for frozen and paraffin-embedded tissue sections

Summary The AgNOR technique has been used extensively in studies investigating the possibility that the numbers and appearances of the intranuclear structures stained are markers of malignancy. The method has the advantage of being applicable to many different types of histological material, including paraffin-embedded tissue. However, it has been suggested that the visualization of AgNORs is dependent on the type and time of fixation employed. This study set out to measure this effect with the following commonly-used fixatives: acetone, absolute ethanol, methanol, Carnoy's fluid, Bouin's fluid, 4% glutaraldehyde, 10% neutral buffered formalin and 10% formol-saline. Both frozen sections and blocks of fresh tonsil were fixed for varying times, the blocks of tissue then being processed routinely. With the frozen sections AgNORs were easier to discern than in sections of paraffin-embedded tissue, and more intranucleolar AgNORs were visible when alcoholic fixatives were used than with aldehyde fixation. The effects of different fixatives on AgNOR appearance in paraffin sections is, however, more complex. Despite the variation caused by different fixatives, AgNORs could be demonstrated adequately with all the fixatives studied. It is concluded that fixation is not a limitation to the study of AgNORs provided that the time and type of fixative is controlled.     

Protocol for quantitative shape analysis of deformities in early larval European seabass Dicentrarchus labrax

This study established an optimized protocol for quantifying the shape variation of newly hatched larvae of European seabass Dicentrarchus labrax, focusing on the effect of fixatives and mounting on body shape from hatching until 14 days post hatching, while also minimizing the error introduced by handling. This assessment was performed based on both biometric and geometric shape data, with the latter relying on outline based elliptic Fourier analysis. The fixation and mounting effect on the total length and shape of newly hatched larvae of D. labrax was tested for four fixation treatments: (1) 8% formalin, (2) 70% ethanol, (3) 8% formalin for 48 h and then to 70% ethanol and (4) 3% phosphate‐buffered glutaraldehyde. The analyses showed that no significant size and shape effect was observed on anaesthetized specimens 5 months post‐fixation in glutaraldehyde, and that the glycerol mounting process of specimens fixed in glutaraldehyde provided the best results for further ontogenetic qualitative and quantitative analysis. The protocol proved to be sufficiently sensitive to even quantify and visualize subtle pre‐fixation shape differences originating from a different egg batch.     

Fixation and Staining of Planaria for Histological Study

Fixation and staining of planaria can affect the interpretation of histopathological changes following their exposure to various agents. We assessed several fixation protocols with various stains in planaria to determine an optimal combination. Planaria were fixed in each of the following: 10% neutral buffered formalin, 2.5%, glutaraldehyde, Bouin's, Zenker's, 70% ethanol, and relaxant. In addition, planaria were fixed in relaxant and postfixed in each of the fixatives above. Paraffin embedded sections from each fixation protocol were stained with hematoxylin and eosin (H & E), toluidine blue, periodic acid-Schiff (PAS), or phosphotungstic acld-hematoxylin (PTAH). Relaxant fixed planaria were also stained with Steiner's, Holmes, trichrome, Giemsa, Grocott's methenamine silver (GMS) and antibodies for intermediate filaments (cytokeratin, vimentin and desmin). Relaxant and Zenker's gave the best fixation with minimal artifacts. Formalin, glutaraldehyde, and ethanol were unacceptable because they caused contortions of the body, crenation, and a darkly pigmented epidermis. Gastroderm could be differentiated from stroma best when stained with H & E, toluidine blue and PTAH. Other organ systems differentially stained included the epidermis, marginal adhesion gland, nervous tissue, and muscle. PAS, Steiner's, Holmes, trichrome and the intermediate filament stains were not useful for planaria staining. The most morphological information was obtained with relaxant fixative and a combination of sections stained with H & E and PTAH.     

Methods of cytologic smear preparation and fixation. Effect on the immunoreactivity of commonly used anticytokeratin antibody AE1/AE3

OBJECTIVE: To evaluate the effect of fixation and methods of cytologic smear preparation on the immunoreactivity of commonly used anticytokeratin antibody AE1/AE3. STUDY DESIGN: Scrape cytology smears and formalin-fixed, paraffin-embedded tissue sections (FPTS) of 20 unfixed, fresh specimens submitted for intraoperative consultation were studied by the immunoperoxidase method. In addition to the morphologic examination, the smears and FPTS were evaluated for intensity and proportion scores. For each specimen, two scrape cytology smears were wet fixed in 95% ethanol, and 12 smears were air dried without fixation. Air-dried smears were either postfixed after rehydration in saline or fixed directly without rehydration by one of the three fixatives: alcoholic formalin, 95% ethanol with 5% acetic acid or 95% ethanol. RESULTS: Both intensity and proportion scores were higher with rehydrated, air-dried smears as compared to those without rehydration and were comparable to those with wet-fixed smears and FPTS. In the rehydrated group, the optimum results were achieved when the smears were postfixed with alcoholic formalin. CONCLUSION: The method of preparation and fixation had variable effects on the immunoreactivity of anticytokeratin antibody AE1/AE3. The optimum results were achieved with saline-rehydrated, air-dried smears post-fixed in alcoholic formalin. To evaluate the role of inter-sample variation, further, larger studies are recommended on this and other antibodies before applying them to different types of cytologic smears.     

Evaluation of tissue preparation methods and paired immunofluorescence staining for immunocytochemistry of lymphomas

Eight cross-linking fixatives were tested for preservation of extracellular or intracellular IgG, IgA, IgM, IgD, kappa and lambda light chains, J chain and secretory component. Most of the selected fixatives have been used in recent immunohistochemical studies of lymphoproliferative processes and comprised routine formalin, glutaraldehyde(1%)-formalin, Baker's formalin-calcium, formalin-sublimate, acetic acid(2%)-formalin-saline, Bouin's fluid, Susa fixative, and carbodiimide. The results obtained in artificial test substrates with defined amounts of IgG or IgA and in biological substrates (colon mucosa, tonsils, and different types of lymphomas) were compared by immunofluorescence with the antigenic preservation afforded by fixation in cold 96% ethanol (with or without inclusion of a pre-fixation 48 h washing period). An antigen concentration at least an eight-fold higher was necessary for detection with most other fixatives. Bouin's and Susa fixatives were peculiar in that they required antigen concentration 150 times higher for detection of IgG but only 3-8 times higher for IgA. Light chains were relatively well preserved by all fixatives except glutaraldehyde. For all cross-linking fixatives, the extent of antigenic masking depended on the concentration of environmental proteins, and the efficiency of unmasking with pronase or trypsin, therefore, varied with the location in the tissue. The J chain was particularly vulnerable to degradation during proteolytic treatment. The extensive masking of extracellular immunoglobulin in formalin-fixed tissue afforded a relatively good signal-to-noise ratio for immunoglobulin-producing cells when kappa and lambda chains were traced. Thus, differentiation between polyclonal and monoclonal B-cell processes on the basis of cytoplasmic labelling was often better in undigested sections. However, the light-chain type of membrane immunoglobulin could usually not be determined in directly fixed tissue. Ethanol fixation preceded by washing in saline afforded such determination and also preserved certain T-cell and HLA-DR antigens as well as diffuse alpha-naphthylbutyrate esterase. Reactive and malignant macrophages could further be traced by their cytoplasmic expression of L1 antigen, both in formalin- and ethanol-fixed material.     

Volumetric Instillation of Fixatives and Inert Substances Into Mouse Lungs

Mouse lungs were instilled with various fixatives to establish the optimum volume necessary to fix the lung without distortion and to compare the efficacy of the fixatives. Fixation with either Stieve's or Bouin's fluid was found preferable to 2.5% and 5% glutaraldehyde, 4% neutral buffered formalin, and to a mixture of formalin and Stieve's fixative. In addition, a comparison was made between diluted Ames O.C.T. Compound and 4% aq. gelatin as supportive substances for unfixed lungs in preparation of cryostat sections and for histochemistry. A 1:2 dilution of O.C.T. was found to be superior to 4% gelatin in preparative, cutting and adhesive properties. The optimal instilled volume for mouse lungs was found to be 0.1 ml for every 7 grams of body weight, introduced at a rate of 0.1 ml per 10 seconds.     

Effects of different fixatives on detection of nucleic acids from paraffin-embedded tissues by in situ hybridization using oligonucleotide probes

Detection of nucleic acids from paraffin-embedded material by in situ hybridization with oligonucleotide probes is increasingly being used. To determine the effect of fixation on the preservation of DNA and mRNA, we studied 18 lymphoid tissues fixed in B5, formalin, OmniFix, ethanol, and Bouin's fixatives and embedded in paraffin by in situ hybridization, using biotinylated oligonucleotide poly d(T) probes and immunoglobulin light chain probes. Detection of DNA using the poly d(T) probe was most consistent and most intense in tissue fixed in formalin, followed by OmniFix and ethanol, with B5 and Bouin's fixatives yielding unsatisfactory results. Detection of mRNA, using the light chain probes, was most consistent and most intense with tissue fixed in formalin and Bouin's solution, followed by B5 fixative, with OmniFix and ethanol fixatives yielding unsatisfactory results. The results of mRNA detection using the poly d(T) probe were found not to correlate with mRNA content as determined by the light chain probes for several fixatives, possibly owing to selective degradation of portions of the mRNA molecule.     

Comparison of the Ruthenium hexammine trichloride method to other methods of chemical fixation for preservation of avian physeal cartilage

Summary Several methods of chemical fixation of avian physeal cartilage were compared. The Ruthenium hexammine trichloride method was compared to isotonic glutaraldehyde and neutral buffered formalin for light microscopy and paraffin embedment, and to two osmium-ferrocyanide methods and a combination of 1% glutaraldehyde and 4% formaldehyde for electron microscopy. Only the Ruthenium hexammine trichloride method prevented the loss of matrix proteoglycans and shrinkage of chondrocytes. In undecalcified paraffin-embedded cartilage, preservation of matrix and cellular detail was excellent, but Ruthenium hexammine trichloride interfered with Haematoxylin and Eosin staining. Glutaraldehyde gave more intense eosinophilia than neutral buffered formalin. Ultrastructurally, the Ruthenium hexammine trichloride method was the most consistent and gave the best overall fixation. Matrix elements and cellular and nuclear membranes were well preserved. It did result in vacuolation of the cytoplasm and mitochondria, and it increased granularity of the cytoplasm, chromatin, and rough endoplasmic reticulum. Other fixatives produced minimal vacuolation and finer granularity, but preservation was less consistent, cell/matrix contrast was often excessive, and they caused shrinkage of all chondrocytes. Large dilatations of the rough endoplasmic reticulum that appear to be cytoplasmic inclusions by light microscopy are described for the first time in avian cartilage.     

Scanning electron microscopy of cultured cells of Aedes aegypti

Cultured cells of Aedes aegypti were fixed with glutaraldehyde and prepared for scanning electron microscopy by four procedures: air drying, lyophilization, ethanol dehydration and air drying, and ethanol dehydration and critical point drying. Comparison of the resulting electron micrographs with phase contrast photomicrographs of living cells revealed that although cultured insect cells dried by the critical point method are not completely without artifacts, this method of preservation is superior to other techniques currently used.     

Effect of Fixatives on Silver Impregnation of Plant Cells

The kind of fixative and duration of fixation modify the affinity of plant cell structures, as shown by a 10-15 hr impregnation at 70 C in 2% aqueous AgNO₂, and a 1-2 hr reduction at room temperature by a 1:1 mixture of 10% formalin and 1% hydroquinone. Cytoplasmic staining was enhanced by fixing in salts of heavy metals, in buffered 6.5% glutaraldehyde, and in 0.5% picric acid. Nuclear staining was prominent after mixtures of glutaraldehyde and hydroquinone, after formalin and pyrogallol, and after acetone, propylene glycol or ether. Nucleolar staining was favored by fixing in 10% formalin, in 5% formalin containing 0.5% hydroquinone, in 50% ethanol containing 0.5% pyrogallol, or in ethylene glycol. Chromosome staining was favored by fixation in 50% acetic or propionic acid, in 2% trichloroacetic acid, and in methanol or ethanol. The best morphological preservations were seen after 50% acetic acid, 6.5% glutaraldehyde, or the 5% formalin-0.5% hydroquinone mixture.     

The importance of fixation procedures on DNA template and its suitability for solution-phase polymerase chain reaction and PCR in situ hybridization

Summary Conventional solution-phase polymerase chain reaction (PCR) and in situ PCR/PCR in situ hybridization are powerful tools for retrospective analysis of fixed paraffin wax-embedded material. Amplification failure using these techniques is now encountered in some centres using archival fixed tissues. Such ailures may not only be due to absent target DNA sequences in the tissues, but may be a direct effect of the type of fixative, fixation time and/or fixation temperature used. The type of nucleic acid extraction procedure applied will also influence amplification results. This is particularly true with in situ PCR/PCR in situ hybridization.To examine these effects in solution-phase PCR, -globin gene was amplified in 100 mg pieces of tonsillar tissue fixed in Formal saline, 10% formalin, neutral buffered formaldehyde, Carnoy's, Bouin's, buffered formaldehyde sublimate, Zenker's, Helly's and glutaraldehyde at 0 to 4°C, room temperature and 37°C fixation temperatures and for fixation periods of 6, 24, 48 and 72 hours and 1 week. DNA extraction procedures used were simple boiling and 5 days' proteinase K digestion at 37°C. Amplified product was visible primarily yet variably from tissue fixed in neutral buffered formaldehyde and Carnoy's, whereas fixation in mercuric chloride-based fixatives produced consistently negative results. Room temperature and 37°C fixation temperature appeared most conducive to yielding amplifiable DNA template. Fixation times of 24 and 48 hours in neutral buffered formaldehyde and Carnoy's again favoured amplification.Fixed SiHa cells (containing 1–2 copies of HPV 16) were examined using PCR in situ hybridization for the amplification of HPV 16. Discrete and diffuse amplification signals were obtained. Neutral buffered formaldehyde fixation for 12–24 hours yielded amplifiable material suitable for use with PCR in situ hybridization. Overall amplification success within cellular preparations was 40%, with non-specific background staining also seen. Possible technical problems encountered with PCR in situ hybridization are discussed.