Article Text
PDF
Abstract
Background: Fixation is a chemical or physical procedure to prevent the degradation of proteins and tissue morphology. To optimise molecular analysis of archival tissues, it is essential that fixation preserves morphology along with protein epitopes and DNA/RNA integrity.
Methods: A new formalin-free alcoholic-based fixative, FineFIX, was used to fix 15 serous effusions and 38 fine-needle aspirates, and cellular morphology and nucleic acid quality were evaluated.
Results: The cytomorphology of the effusions and fine-needle aspirates obtained with FineFIX fixation was similar to that obtained with formalin-fixed counterparts. Immunocytochemistry showed comparable results with the traditional fixative, but FineFIX preserved higher-molecular-mass DNA and RNA, as demonstrated by successful PCR of large amplification products of >2000 bp.
Conclusions: The formalin-free fixative produced not only satisfactory results for immunocytochemistry on cytological smears and cell blocks, but also excellent preservation of DNA and RNA, which can also be efficiently used for sophisticated molecular techniques.
Statistics from Altmetric.com
Gene and protein expression profiles are increasingly used in pathology to better understand the molecular events leading to disease and to identify new prognostic and therapeutic markers.1–5 Morphological evaluation of bioptic and cytological material remain the “gold standards” of diagnosis, from which any new molecular insight into disease originates. For molecular investigations, pathologists therefore require tissues with satisfactory morphological quality and well-preserved subcellular components.6–8
In pathology, tissue fixation is a critical procedure for the preservation of specimens. Tissues are routinely fixed in formalin, which has the advantages of low cost, rapid diffusion, solution stability and availability, but also has disadvantages such as toxicity, carcinogenicity and nucleic acid fragmentation.7 9 Extraction of good-quality DNA and RNA from formalin-fixed paraffin-embedded tissues (PETs) is compromised because of incomplete removal of protein–nucleic acid cross-links.7–11 Non-cross-linking alcoholic fixatives provide superior results to aldehydes for preservation of nucleic acids, but they properly fix only small tissue fragments with poor morphology, because of cell shrinkage.7 An alcohol-based fixative has recently shown good results not only for tissue morphology, but also for DNA and RNA quality and protein preservation.12
In this paper, we present morphological, immunocytochemical and molecular results for an alcohol-based fixative used on cytological material from effusions, lung, liver, breast and lymph nodes, thyroid and subcutaneous nodules.
MATERIALS AND METHODS
Sample collection
We examined 15 effusions and 38 fine needle aspirates (FNAs) at the Department of Pathology, Padova University.
The effusions arrived at our laboratory unfixed and were centrifuged immediately at 2000 rpm for 20 min. Part of the pellet was smeared on to two slides for cytological examination; one-third of the fresh cells were used immediately for DNA/RNA extraction. The remaining material was fixed in formalin or FineFIX (Milestone, Bergamo, Italy). FineFIX is a mixture of ethanol, distilled water, glycerol, poly(vinyl alcohol) and monomeric carbohydrates. The fixation time varied from 24 h to 1 week.
The FNAs arrived at our laboratory in formalin or FineFIX within 1 h of sampling. Cells were centrifuged at 2000 rpm for 20 min for cell blocks.
Cell blocks
Cells were resuspended in 1% low-melting-point agarose equal to the volume of the pellet, and solidified in Microfuge tubes at 4°C, placed in cassettes, processed and embedded in paraffin.
Immunocytochemistry (ICC)
Sections 5 μm thick from the cell blocks of FineFIX/formalin-fixed FNAs and effusions were deparaffinised before H&E staining and ICC.
Table 1 shows the antibodies and dilutions used. Staining was performed using an indirect immunoperoxidase technique with an automated system (Bond Polymer Refine Detection; Bond-maX; Vision BioSystems, Milton Keynes, UK). Antigen retrieval was performed by heating sections in Bond Epitope Retrieval Solution 1. Antigen was detected by incubation with labelled polymer and diaminobenzidine.
DNA extraction and amplification
The fresh cells were resuspended in extraction buffer (5 M NaCl, 1 M Tris/HCl, pH 8, 0.5 M EDTA, pH 8, 10% SDS) with proteinase K (20 mg/ml). Three 10 μm sections of formalin/FineFIX-fixed PETs from effusions were cut into small pieces, deparaffinised, rehydrated and air-dried at room temperature. Finally, the sample sections were resuspended in extraction buffer containing proteinase K and incubated at 56°C for 60 min.
To reverse cross-linking, sections were deparaffinised by adding 100 μl 0.5% Tween-20 and heating to 90°C for 10 min on a thermal cycler. They were then cooled to 56°C and digested in extraction buffer with proteinase K.
DNA was extracted by the phenol/chloroform protocol. Samples were amplified with specific primers for the human beta-catenin 1 gene (CTNNB1; gene ID, 1499), and PCR was performed in a final volume of 25 μl using the following conditions: denaturation at 95°C for 5 min, 35 cycles of 95°C for 30 s, annealing temperature for 30 s, and 72°C for 30 s; elongation step at 72°C for 5 min. Table 2 gives the PCR primers and amplicon lengths.
RNA extraction, reverse transcription (RT) and RT-PCR
Three 10 μm formalin/FineFIX-fixed paraffin-embedded sections from effusions were cut into small pieces. RNA was extracted using the PureLink FFPE Total RNA Isolation kit (Invitrogen, Carlsbad, California, USA). DNase I digestion was performed to eliminate genomic DNA contamination.
RNA integrity was controlled using the RNA 6000 Pico assay kit with Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, California, USA).
RT was performed using 100–500 ng total RNA, Moloney murine leukaemia virus reverse transcriptase and 250 μM random primers (Invitrogen). Each cDNA sample was amplified using specific primers for the human beta-actin gene (ACTB; gene ID, 60). PCR was performed in a final volume of 25 μl using standard conditions. Table 3 gives the PCR primers and amplicon lengths.
High-resolution melting analysis (HRMA)
HRMA is a screening method for mutations and deletions based on PCR amplification and melting curve analysis. DNA from the 11 lung cancer aspirates was amplified using primers for the human epidermal growth factor receptor gene (EGFR; gene ID, 1956). The sequences for amplification of the DEL region in exon 19 were: Fw Ex19, 5′-CGTCTTCCTTCTCTCTCTGTC-3′; and Rev Ex19, 5′-GACATGAGAAAAGGTGGGC-3′. The sequences for amplification of the region of the L858R mutation in exon 21 were: Fw Ex21, 5′-GCATGAACTACTTGGAGGAC-3′; and Rev Ex21, 5′-GGCTGACCTAAAGCCAC-3′. DNA was amplified by real-time PCR in the presence of a proprietary saturating DNA dye contained in the LightCycler 480 High Resolution Melting Master. A melting curve was produced using high data acquisition rates, and data were analysed with the LightCycler 480 Gene Scanning Software Module.
RESULTS
Effects of FineFIX and formalin on cell morphology and ICC
Cell morphology and ICC were analysed on cell blocks obtained from 15 effusion samples and 38 FNAs after fixation and paraffin-embedding using either formalin or FineFIX as fixative (⇓tables 4 and 5).
FineFIX clearly preserved cell integrity, giving morphological information similar to that obtained with formalin: no differences were identified in cell architecture, cytoplasmic and nuclear morphology, or tinctorial reactions (fig 1A,B). The only exception was moderate erythrocyte swelling in FineFIX samples fixed for more than 48 h, but this did not obscure the cellular details.
Immunocytochemistry (ICC) staining. Sections 5 μm thick from formalin-fixed (A, C, E) and FineFIX-fixed (B, D, F) cell blocks (mesothelioma, pleural effusion) (40×). H&E staining (A and B) and ICC with anti-(human cytokeratin 7) (C and D) and anti-(human vimentin) (E and F). The insert in (F) is a particular of the section under 63× magnification.
Immunoreactivity was similar in FineFIX-fixed and formalin-fixed samples, except for a higher sensitivity for vimentin antibody in FineFIX-fixed samples. No cell shrinkage was observed, and nuclear details were satisfactory (fig 1C–F). In FineFIX-fixed samples, nuclear antigenicity was also investigated using MIB-1, Wilms tumour and vimentin, with satisfactory results.
Quality of DNA extracted from FineFIX-fixed and formalin-fixed samples
DNA was extracted from the cellular component of fresh effusions and from FineFIX/formalin-fixed cell block counterparts. The DNA extracted from formalin-fixed samples was clearly degraded, as demonstrated by the smear observed on the agarose gel (fig 2), and failed to amplify large PCR products (table 6).
Quality of genomic DNA. DNA from fresh (lane 1), FineFIX-fixed (lane 2) and formalin-fixed (lane 3) samples. Lane 4, 1 kb Plus DNA Ladder (Invitrogen; bands ranging in size from 100 to 12 000 bp). High-molecular-mass DNA was extracted from FineFIX-fixed and fresh samples, whereas the DNA obtained from the formalin-fixed sample was completely fragmented.
For formalin-fixed samples, cross-link reversion obtained DNA of <300 bp length.
High-quality genomic DNA was obtained from FineFIX-fixed samples, comparable to that from fresh samples (fig 2). In addition, the integrity of DNA from FineFIX-fixed samples permitted amplification of a 2361 bp amplicon for CTNNB1, exactly the same as from the fresh counterpart (table 6, fig 3), allowing more sophisticated analyses such as HRMA. This technique can detect missense mutations, deletions and insertions in isolated tumour DNA.13 14
Genomic DNA and PCR. PCR amplification of DNA extracted from fresh (lane 1), FineFIX-fixed (lane 2) and formalin-fixed (lane 3) samples. Lane 4, negative control. Lane 5, DNA molecular mass marker XIII (Roche Diagnostics Spa, Milano, Italy; bands ranging in size from 50 to 2462 bp). CTNNB1 amplicons of: (A) 199 bp; (B) 941 bp; (C) 1900 bp; (D) 2361 bp.
In HRMA, mutation/deletion detection is only possible if the amplification curve reaches the plateau between 20 and 40 amplification cycles. In the analysis of exon 19 of EGFR, we obtained two comparable amplification curves using DNA from fresh and FineFIX-fixed material at the 23rd and 25th cycles, respectively. The plateau was reached with fresh and FineFIX-fixed samples, but not with formalin-fixed samples. Amplification of DNA extracted from formalin-fixed samples was delayed beyond the 40th cycle, which is inadequate for detection of deletions/mutations (fig 4). Similar results were obtained for the analysis of exon 21 of EGFR. Mutation in exon 21 was detected only in FineFIX-fixed samples, and was not possible in formalin-fixed specimens because of DNA degradation (fig 4B,C).
High-resolution melting analysis. (A) Amplification curves of DNA extracted from fresh, FineFIX-fixed (FF) and formalin-fixed (Form) samples in the analysis of EGFR exon 19. The curves for the fresh and FineFIX-fixed samples are comparable: PCR amplification occurred at the 23rd and 25th cycles, respectively. In comparison, amplification of DNA extracted from formalin-fixed samples was delayed, occurring after the 40th cycle. (B) A sample fixed in FineFIX and formalin shows different amplification curves. The plateau was reached only with the FineFIX-fixed sample. (C) The FineFIX-fixed sample was analysed with 10 different FineFIX-fixed specimens. The plot shows two different melting profiles, corresponding to the wild-type exon 21 in the 10 samples and a positive sample corresponding to the mutation L858R in exon 21 (arrow).
FineFIX and RNA quality
The quality of RNA extracted from the fresh and FineFIX samples was comparable, whereas RNA extracted from formalin-fixed cells was significantly degraded, as indicated by the absence of 28S and 18S ribosomal RNA bands (fig 5). To estimate the lengths of preserved transcripts in samples prepared with different fixatives and embedded in paraffin, we amplified various sizes of cytoplasmic beta-actin mRNA fragments by RT-PCR from total RNA. In six out of 15 (40%) formalin-fixed samples, all successfully amplified amplicons were a maximum of 287 bp; no amplification products were detected for longer fragments (table 7, fig 6).
RNA quality. Electropherogram profiles of RNA extracted from fresh sample (A) and FineFIX-fixed (B) and formalin-fixed (C) cell blocks of a pleural effusion. Peaks of 18S and 28S ribosomal RNA are evident in the FineFIX-fixed sample, whereas they are absent from the formalin-fixed sample. Total RNA extracted from fresh samples had an RNA integrity number of 9.6, that from FineFIX-fixed samples was 6.3, and that from formalin-fixed samples was 2.1. FU, fluorescence units.
cDNA and PCR. PCR amplification of cDNA obtained after retro-transcription of total RNA extracted from a fresh (lane 1), FineFIX-fixed (lane 2) and formalin-fixed (lane 3) sample. Lane 4, negative control. Lane 5, molecular mass marker; 1 kb Plus DNA Ladder (Invitrogen; bands ranging in size from 100 to 12000 bp). beta-actin amplicons of: (A) 126 bp; (B) 287 bp; (C) 617 bp; (D) 1199 bp.
FineFIX-fixed samples produced different results. Table 7 shows the results of cDNA amplification from fresh, formalin-fixed and FineFIX-fixed cytological samples. cDNA amplification efficiency for FineFIX-fixed samples was comparable to that for the fresh samples for fragments no longer than 300 bp, whereas a decreased efficiency for FineFIX-fixed samples was noticed in the amplification of PCR fragments longer than 617 bp. In particular, we obtained an RT-PCR product of 1199 bp in only three FineFIX-fixed effusions out of 15 (20%) .
DISCUSSION
Archived tissues are a source of DNA and RNA for molecular biological studies in cancer research and for screening for genetic-based diseases, requiring optimal preservation of proteins and nucleic acids. Up until now, the commonly used technique of fixation with formalin has prevented extraction of intact nucleic acids, and the molecular information conserved in pathology archives has not been able to be exploited. Furthermore, formalin is toxic, it is an allergen and carcinogen, and its disposal is becoming increasingly costly for laboratories.15–18 Formaldehyde has been evaluated by the International Agency for Research on Cancer as “carcinogenic to humans” (group 1) on the basis of induction of nasopharyngeal cancer, although there is uncertainty (“strong but not sufficient evidence”) about a causal association with myeloid leukaemia and a limited association with nasopharyngeal carcinoma.19
Take-home messages
FineFIX is an ethanol-based fixative.
It can be successfully used for cytological smears, providing good definition of cytoplasmic and nuclear detail, and on cell blocks for immunocytochemistry.
FineFIX efficiently protects nucleic acids from degradation, and DNA and RNA can be extracted and amplified up to 2300 bp.
FineFIX is non-toxic and will possibly replace formalin in pathology laboratories.
It is not surprising therefore that, in recent years, many fixatives have been introduced as formalin replacements. Alcohol-based fixatives exert their effect by protein precipitation, and are superior in preserving a number of cellular antigens.20 21 Advantages also include elimination of carcinogenic vapours, greater staining avidity, no enzyme predigestion in ICC, and simple and rapid disposal. Disadvantages include slightly increased viscosity, variability of tissue staining and nuclear shrinkage, artifactual pigment deposition in bloody specimens, and increased flammability.22
A new ethanol-based fixative, FineFIX, has recently been developed, which overcomes these problems and allows improved molecular analysis. FineFIX fixation has been described for the study of PETs by Stanta et al,12 with good tissue histomorphology, low denaturation of proteins and nucleic acid preservation. In histopathology, FineFIX is best used with microwave-processing apparatus for better fixation of the deeper layers of the specimen.12 23 We considered the use of microwaves unnecessary in this cytopathology study because of the monolayer nature of the cytological preparation. In this study, FineFIX was validated for DNA integrity in fixed and paraffin-embedded cytological material by showing high yields of extracted DNA and amplification of fragments >2300 bp in all the FineFIX-fixed samples. The quality of the DNA from FineFIX-fixed material is excellent, allowing application of fine and sophisticated methodologies, such as HRMA, with better results than with formalin.
FineFIX is a non-cross-linking fixative that preserves RNA by precipitation and inactivation of endogenous RNAses. We obtained high-quality RNA from FineFIX-fixed samples, in contrast with that obtained from formalin-fixed and paraffin-embedded samples. To test the quality of RNA, we evaluated RT-PCR amplification from extracted RNA samples. From the FineFIX-treated specimens, we were able to amplify products that, in some cases, were longer than 1000 bases.
FineFIX is an efficient fixative not only in protecting nucleic acids from degradation, but also in preserving cell morphology. ICC experiments showed excellent preservation of cell morphology, maintaining antigen properties and giving morphological information similar to that from formalin-fixed tissues.
In conclusion, FineFIX fixation has been shown to preserve cytomorphology, immunoreactivity, DNA and RNA of effusions and aspirates with a quality comparable to that of fresh material, without the need to modify routine laboratory procedures, and is safe and non-toxic. With the use of this satisfactory replacement for formalin, the pathology archives should yield even more important information.
Acknowledgments
We thank Dr Serena Bonin, Professor Giorgio Stanta and Professor Rodolfo Costa for critical reading of the manuscript.
REFERENCES
Footnotes
Competing interests: None.