Article Text

Download PDFPDF
Nucleic acid sequence-based amplification in formalin-fixed and paraffin-embedded breast-cancer tissues
  1. Ulrike Riehle1,
  2. Andreas Mader2,
  3. Thomas Brandstetter3,
  4. Jürgen Rühe3,
  5. Axel zur Hausen2,
  6. Elmar Stickeler1
  1. 1Department of Gynecology and Obstetrics, University Hospital Freiburg, Freiburg, Germany
  2. 2Institute of Pathology, University Hospital Freiburg, Freiburg, Germany
  3. 3Laboratory for Chemistry and Physics of Interfaces, Department of Microsystems Engineering (IMTEK), University Freiburg, Freiburg, Germany
  1. Correspondence to Professor Axel zur Hausen, Breisacherstrasse 155 a, D-79106 Freiburg, Germany; axel.zurhausen{at}uniklinik-freiburg.de

Abstract

Aim To evaluate the nucleic acid sequence-based amplification (NASBA) technique to amplify mRNA isolated from formalin-fixed and paraffin-embedded (FFPE) breast-cancer tissues.

Methods RNA was extracted from archived, 10-year-old FFPE tissues, and selected genes, namely ribosomal protein S18 (RPS18), epidermal growth factor receptor 2 (HER2), estrogen receptor alpha (ERα), Y box binding protein (YBX-1), matrix metallopeptidase 11 (MMP11), caspase 8 (CASP8) and superoxide dismutase 2 (SOD2), were amplified by NASBA.

Results Despite strong degradation of the template, RNA amplification of all tested genes resulted in strong hybridisation signals. Sensitivity tests showed that the RPS18 NASBA assay was more sensitive than real-time RT-PCR used as a reference method. The sensitivity of the HER2, ERα, MMP11, YBX1, CASP8 and SOD2 NASBA assay was comparable with RT-PCR targeted to the respective genes.

Conclusions The results obtained indicate that NASBA is suitable to amplify with high specificity and sensitivity, even strongly degraded RNA isolated from FFPE tissues, and therefore can complement already-existing amplification techniques such as RT-PCR for analysis of such tissues.

  • NASBA/FFPE tissues/breast cancer
  • fixation

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Introduction

Formalin-fixed and paraffin-embedded (FFPE) tissues represent one of the largest tissue sources, for which well-documented clinical follow-up is available, and therefore large-scale retrospective studies are possible.1 However, depending on the history of the sample before fixation and storage time and conditions, the RNA in FFPE tissues is usually more or less degraded. As a result, the efficacy of RNA extraction and amplification is strongly reduced.2 3 It has been demonstrated that, despite significant degradation, RNA in such samples can be amplified by reverse transcription and PCR amplification (RT-PCR), and also that for such fixed and stored tissue samples, meaningful quantitative gene expression data can be obtained.4

Nucleic acid sequence-based amplification (NASBA)5 is a simple, alternative method for nucleic acid amplification. In medical biology NASBA has been used frequently and proven to be a sensitive method for the detection of various viral5 6 and bacterial7–9 pathogens. This isothermal ‘one-pot’ reaction is highly specific for mRNA targets and results in single-stranded RNA amplicons, which can be immediately detected by sequence-specific probes.10 Therefore, it was recently reported that NASBA products can be easily used as analytes for microarray hybridisation.11 12 New application platforms for the NASBA assay arose in the last few years, such as the development of microfluidic diagnostic chips, that use a chip-based NASBA rather than a PCR for nucleic acid amplification in order to bypass thermal cycling, genomic DNA elimination and reverse transcription steps, and to speed up postamplification analysis.13–16

We have recently developed a platform that combines RNA amplification and labelling by NASBA and microarray analysis in a one-step process.17

For clinical applications of this multiplex on-chip NASBA assay, it is crucial to elucidate how efficient NASBA is in case RNA is potentially degraded, and chemicals such as paraffin are present in the analysis process. For this purpose, RNA was extracted from archived FFPE breast-cancer tissues, and a set of genes chosen that reflect both routine and potential biomarkers for diagnosis of breast cancer. Specifically they are genes encoding the epidermal growth factor receptor 2 (HER2), the estrogen receptor alpha (ERα), Y box binding protein (YBX-1),18 the matrix metallopeptidase 11 (MMP11),19 caspase 8 (CASP8)20 and superoxide dismutase 2 (SOD2).21 Furthermore, ribosomal protein S18 (RPS18) was chosen as a reference gene. We report on the results of experiments on NASBA amplification of these genes in RNA isolated from 10-year-old and accordingly more or less degraded FFPE samples.

Materials and methods

Tumour samples

FFPE breast-cancer tissues were collected from the archives of the Institute of Pathology at the University Hospital Freiburg, Germany. Breast-cancer tissues included in this study were derived from patients operated on in 2000 at the Department of Obstetrics and Gynecology, University Hospital Freiburg, Germany. All investigations were carried out according to the national ethical guidelines and approved by the local ethical committee (No 275/2006).

RNA extraction

RNA from FFPE tissues was isolated using the High Pure RNA Paraffin Kit (Roche Applied Science, Mannheim, Germany). For reference experiments, universal human RNA (total RNA) was purchased from the BioChain Institute (Hayward, California).

RNA quantity and intactness were assessed by the Agilent Bioanalyzer 2100 and the RIN (RNA integrity number) software algorithm22 (Agilent Technologies, Waldbronn, Germany).

NASBA assay

Primer and probes

The NASBA oligonucleotide primer and probe sequences used to target RPS18, HER2, ERa, YBX1, MMP11, CASP8 and SOD2 are shown in table 1. Biotin (BIO)-labelled oligonucleotide probes corresponding to an internal region defined by NASBA oligonucleotide primer pairs were synthesised to hybridise with the antisense NASBA RNA products.

Table 1

Sequences of NASBA oligonucleotide primer, RT-PCR oligonucleotide primer and BIO-labelled oligonucleotide probe.

Standard RNA

A standard curve with a known input of RNA transcribed from a plasmid was used to assess the sensitivity of the HER2 NASBA reaction. The plasmid was generated by cloning a specific HER2 PCR product into a pJET1.2/blunt Cloning Vector (CloneJET PCR Cloning Kit Fermentas, St Leon-Rot, Germany) and verified by sequencing. HER2-transcripts were generated from this plasmid using T7 RNA polymerase (TranscriptAid High Yield Transcription Kit, Fermentas, St Leon-Rot, Germany).

NASBA reactions

Different quantities of RNA (universal RNA-human, RNA extracted from paraffin blocks or HER2 transcripts) were amplified at 41°C for 60 min in 20 μl reaction volumes containing 2 pmol of each oligonucleotide primer, 15% (v/v) DMSO (Dimethylsulfoxid), 40 mmol/l of Tris-HCl (pH 8.5), 12 mmol/l of MgCl2, 70 mmol/l of KCl, 5 mmol/l of DTT (Dithiothreitol), 1 mmol/l of each dNTP, 2 mmol/l of rATP, rUTP and rCTP, 1.5 mmol/l of rGTP and 0.5 mmol/l of ITP. After an initial incubation at 65°C, for 2 min, the enzyme mixture, which included 6.4 U of AMV-RT, 0.08 U of RNase H and 32 U of T7 RNA polymerase, was added. NASBA reagents were obtained from Life Science (St Petersburg, Florida). The NASBA products were analysed by Northern/dot blotting and visualised using a corresponding BIO-labelled oligonucleotide probe as well as an anti-BIO-horseradish peroxidase (HRP) conjugate (1:2000, DAKO, Clostrup, Denmark).

Signal intensity was determined by 2D densitometry using the AIDA image analyser software Version 3.21.001 (Raytest GmbH, Straubenhardt, Germany).

Reverse transcription and real-time PCR

For comparison of NASBA with a commonly used nucleic acid amplification method (RT-PCR), serial dilutions of RNA extracted from paraffin blocks (100 ng to 100 fg) were reversed using the QuantiTect Reverse Transcription Kit from Qiagen (Hilden, Germany). Subsequently the cDNA was amplified by real-time PCR utilising the QuantiFast SYBR Green PCR Kit (Qiagen, Hilden, Germany) and specific oligonucleotide primer pairs (1 μmol/l) for RPS18, HER2, ERα, MMP11, YBX1, CASP8 and SOD2 respectively (table 1). Fluorescence was collected by the mastercycler ep real-plex (Eppendorf, Germany) and product specificity confirmed by melting curve analysis, sequencing (GATC Biotech AG, Konstanz, Germany) and gel electrophoresis.

Real-time PCR efficiencies of RPS18, HER2, ERα, MMP11, YBX1, CASP8 and SOD2 were determined using a standard curve of Ct versus the log of cDNA input (0.032–50 ng cDNA). The slope of the standard curve was used to calculate the PCR efficiency (E) according to the equation: E=10[−1/slope].23 All real-time RT-PCR assays showed high real-time PCR efficiency rates (RPS18, 2.09; HER2, 1.97; ERα, 1.94; MMP11, 1.93; YBX1, 1.88; CASP8, 2.17; SOD2, 2.11) with high linearity (r2>0.95).

Statistical analysis

The non-parametric Spearman correlation was used to analyse the association of HER2 expression between HER2 assessment by NASBA and real-time RT-PCR in the 10 tumours. Data were analysed using standard statistical software (SPSS version 9). A p value of <0.05 was considered significant.

Results

Evaluation of RNA integrity

The Agilent Bioanalyzer 2100 was used to assess RNA integrity. All of the tested RNA samples extracted from FFPE tissues were strongly fragmented, and RIN values of 2–2.5 were determined.22 In contrast, universal human RNA, which was evaluated as a reference, had an RIN value of 7.4.

NASBA in FFPE samples

To test the feasibility of NASBA in FFPE samples, RNA was extracted from three different 10-year-old FFPE tumour blocks. Each RNA extraction was repeated, and seven different mRNA transcripts (RPS18, HER2, ERα, YBX1, MMP11, CASP8 and SOD2) amplified using target-specific NASBA oligonucleotide primer pairs (table 1). The NASBA products were separated in a non-denaturing agarose gel, transferred onto a positively charged nylon membrane and hybridised with specific BIO-labelled oligonucleotide probes. Figure 1 shows the amplification of the set of seven genes in the tested tumour samples. All transcripts could be amplified by NASBA. HER2 and SOD2 NASBA resulted in strong hybridisation signals, while the ERa was especially in the tumour tissues from patients 2 and 3 only weakly amplified. In the respective pathology report, the tested tumours were all classified by immunohistochemistry as HER2-positive (score 3) and ERα-negative explaining the difference in signal intensity. The NASBA hybridisation signal intensities were evaluated using the AIDA (2D densitometry) software. Statistical analysis revealed a significant correlation between the NASBA amplification signal of the respective genes in independent RNA preparations of the same tumour samples (Tumour 1.1 and tumour 1.2: rs=0.964, p=0.000; tumour 2.1 and tumour 2.2: rs=0.929, p=0.003; tumour 3.1 and tumour 3.2: rs=0.999, p=0.000.)

Figure 1

Feasibility of nucleic acid sequence-based amplification (NASBA) in formalin-fixed and paraffin-embedded (FFPE) tissues. From three different FFPE blocks (tumour 1–3), RNA was extracted twice (extraction 1–2) and amplified using oligonucleotide primer pairs to detect ribosomal protein S18 (RPS18) (149 nt), epidermal growth factor receptor 2 (HER2) (144 nt), estrogen receptor alpha (ERα) (148 nt), matrix metallopeptidase 11 (MMP11) (141 nt), Y box binding protein (YBX-1) (151 nt), caspase 8 (CASP8) (140 nt) as well as superoxide dismutase 2 (SOD2) (155 nt). The NASBA reactions were transferred onto a nylon membrane (RPS18 diluted 1:20) and detected utilising biotin-labelled oligonucleotide probes. In each tumour tissue, the seven transcripts were amplified and could be detected via northern hybridisation.

In a test set of 10 tumour samples from 10 different patients, HER2 NASBA was performed and compared with HER2 transcript levels that had been measured by real-time RT PCR (figure 2). Therefore, total RNA was extracted from each tumour and 100 ng used for amplification of both HER2 and RPS18, respectively, by means of NASBA.

Figure 2

HER2 nucleic acid sequence-based amplification (NASBA) in formalin-fixed and paraffin-embedded (FFPE) tissues and comparison with quantitative real-time RT PCR. (A) From 10 archival tumours (T1–T10), RNA was extracted and epidermal growth factor receptor 2 (HER2) as well as ribosomal protein S18 (RPS18) as reference gene amplified by means of NASBA. The HER2 and ribosomal protein S18 (RPS18) NASBA products were northern-blotted and detected using specific biotin-labelled oligonucleotide probes. The RPS18 NASBA products were diluted (1:20) prior to electrophoresis so as not to overload the gel. (B) The experiment was repeated and the blot intensities analysed (AIDA/2D densitometry). The relative HER2 signal intensities expressed as mean±SEM are shown. (C) In each tumour (1–10), HER2 mRNA expression relative to RPS18 was determined in duplicate and shown as mean relative HER2 expression±SEM. The relative HER2 blot intensities were correlated with the relative HER2 mRNA levels measured by RT-PCR. Significant correlation was found between mRNA levels measured by RT-PCR and HER2 blot intensities (rs=0.717, p=0.03).

In each tumour both transcripts (HER2 and RPS18) could be detected by northern hybridisation. In tumour sample 10, HER2 could be amplified by HER2 NASBA but not with real-time RT-PCR. The hybridisation signal intensities of the HER2 NASBA relative to the RPS18 NASBA were evaluated using the AIDA (2D densitometry) software. Statistical analysis revealed a significant correlation between relative HER2 mRNA levels measured by NASBA and quantitative RT-PCR (rs=0.745, p=0.013).

Sensitivity of NASBA in FFPE samples

Serial dilutions of the reference RNA, the in vitro transcribed HER2 RNA as well as the RNA extracted from archival breast cancers were prepared to estimate the sensitivity of both RPS18 and HER2 NASBA (figures 3, 4). For analysis, 1 μl of each NASBA product was spotted onto a positively charged nylon membrane and hybridised with specific BIO-labelled oligonucleotide probes. Using reference RNA, the RPS18 NASBA was capable of producing a detectable signal with a minimum of 1 pg total RNA input. For RNA extracted from FFPE tissues, the minimum amount of RNA that could be reliably amplified by RPS18 NASBA was 100 pg RNA (figure 3). The amplification of the in vitro transcribed HER2 NASBA RNA (approximately 108–102 HER2 copies) resulted in a strong yield of amplified RNA from as few as 100 molecules of HER2 transcript RNA. The HER2 NASBA in paraffin blocks was sensitive enough in both tested tumour samples to detect a minimum of 10 ng total RNA input (figure 4).

Figure 3

Sensitivity of ribosomal protein S18 (RPS18) nucleic acid sequence-based amplification (NASBA) in both reference RNA and RNA extracted from formalin-fixed and paraffin-embedded (FFPE) tissues, respectively. (A) NASBA was carried out in doublet in a decreasing amount (100 ng to 100 fg) of reference RNA (upper image) and pooled RNA extracted from six different FFPE tissues (lower image), dot-blotted and detected via northern hybridisation. The minimum amount of RNA that could be reliably amplified by RPS18 NASBA in reference RNA was 1 pg and in FFPE RNA 100 pg. Note that at 10 pg of FFPE RNA, a very weak signal was visible. (B) Sensitivity of RPS18 real-time RT-PCR in both reference RNA and RNA extracted from FFPE tissues, respectively. After real-time amplification and a melting curve analysis, the PCR reactions were applied to a 2% agarose gel stained with ethidium bromide (0.2 μg/ml). The RPS18 RT-PCR assay amplified RPS18 in 1 ng of reference RNA (upper gel image) and in 1 ng of RNA extracted from FFPE tissues (lower gel image) to a detectable level.

Figure 4

Sensitivity of epidermal growth factor receptor 2 (HER2) nucleic acid sequence-based amplification (NASBA). A 10-fold serial dilution of in vitro transcribed HER2 RNA (100 ng to 100 fg∼108–102 HER2 copies) was used as template for HER2 NASBA. From each NASBA reaction, 1 μl was spotted onto a nylon membrane and detected with specific biotin-labelled oligonucleotide probes. A minimum of 100 fg of HER2 transcript (∼102 HER2 copies) could still be amplified by NASBA and resulted in a strong hybridization signal. (B) Comparison of HER2 NASBA and HER2 RT PCR in formalin-fixed and paraffin-embedded (FFPE) tissues. Decreasing amounts of RNA (100 ng to 100 fg) extracted from two different archival tumours (tumour 1, HER2 score 1; tumour 2, HER2 score 3) were amplified by HER2 NASBA and real-time RT-PCR. The HER2 NASBA assay (dot blot) amplified HER2 in 10 ng of FFPE RNA of both tumours to a reliably detectable level. Note, also, that at 1 ng of FFPE RNA, a very weak signal was visible. In contrast the HER2 real-time RT-PCR assay (gel image) amplified HER2 in tumour 1, in 10 ng and in tumour 2, in 100 pg of input RNA to a detectable level.

Comparison of NASBA and RT-PCR in paraffin blocks

Serial dilutions (100 ng to 100 fg) from both reference RNA and RNA extracted from paraffin blocks were reverse-transcribed and amplified in real-time (SYBR Green I) using a PCR with specific oligonucleotide primer pairs for RPS18, HER2, ERα, MMP11, YBX1, CASP8 and SOD2. The RPS18 RT-PCR was capable of detecting a minimum of 1 ng input RNA in both RNA extracted from FFPE tissues and reference RNA (figure 3B; melting curves are shown as supplementary data). Using the NASBA technique, as little as 100 pg of RNA extracted from FFPE tissues and 1 pg of reference RNA gave a detectable signal after amplification and dot blot hybridisation.

The minimum level of target RNA to be amplified and detected in HER2 RT-PCR was in the tested tumour samples 10 ng and 100 pg respectively (melting curves shown in the supplementary data). The HER2 NASBA in both tumours was capable of detecting 10 ng total RNA reliably (figure 4B).

The MMP11, YBX1, CASP8 and SOD2 NASBA assays amplified the respective transcripts using as little as 1 ng of total FFPE RNA. The ERα NASBA assay was capable of amplifying a minimum of 10 ng of FFPE RNA. Using real-time RT-PCR, the minimum level of total FFPE RNA to be amplified and detected in ERα, MMP11 and YBX1 RT-PCR was 10 ng of total FFPE RNA. In contrast, the detection limit of the CASP8 and SOD2 RT-PCR was 1 ng of total FFPE RNA (supplementary data).

Discussion

The aim of the present study was to prove the feasibility of NASBA in RNA extracted from archived FFPE breast-cancer tissues. Therefore, NASBA assays were established for several transcripts (RPS18, HER2, ERa, MMP11, YBX1, CASP8 and SOD2) and applied to RNA extracted from 10-year-old FFPE breast-cancer tissues. We show that despite significant RNA degradation, probable binding of RNA to proteins and possible carryover of paraffin which might interfere with the enzymatic NASBA reaction, an input of 100 ng of fragmented RNA was sufficient to generate a strong hybridisation signal for all tested genes in the respective tumour samples. A significant correlation was found between NASBA signal intensities of different RNA preparations of the same tumour, confirming that NASBA with RNA extracted from FFPE samples is working well.

In a set of 10 tumour samples HER2 mRNA was amplified by HER2 NASBA and blot intensities compared with quantitative real-time RT-PCR. Using the Spearman non-parametric correlation, significant concordance was found between relative HER2 mRNA levels measured by NASBA and qRT-PCR underlining the reliability of HER2 NASBA in FFPE tissues.

The sensitivity of the NASBA assay in RNA extracted from FFPE tissues was tested with oligonucleotide primer pairs targeting the RPS18, HER2, ERα, MMP11, YBX1, CASP8 and SOD2 mRNA, respectively. Therefore, serial dilutions of intact reference RNA, HER2 transcripts and RNA extracted from FFPE tissues were prepared. The smallest amount of RNA that could be reliably detected by RPS18 NASBA was 1 pg of reference RNA and 100 pg of total RNA extracted from FFPE tumours, which is even more sensitive than real-time RT-PCR. In samples with smaller amounts of RNA (10 pg FFPE RNA), a very weak signal is observed (figure 3) which is unequal in both reactions and was therefore not considered in the evaluation of the lower detection limit. In the case of HER2 also RNA extracted from FFPE breast-cancer tissues was amplified by NASBA. We show that in both tested tumours, 1 ng of total RNA already gave a detectable signal. However, for reliable results, a minimum of 10 ng of RNA should be used for analyses. In this case, the NASBA process is equal to (sample 1) or less sensitive than (sample 2) HER2 real-time RT-PCR. For ERα, CASP8 and SOD2, we show that the respective NASBA assay is as sensitive as real-time RT-PCR targeted to the same gene. The MMP11 and YBX1 NASBA assay were one order of magnitude more sensitive in the tested samples than MMP11 and YBX1 RT-PCR.

Conclusions

NASBA is a fast, isothermal, one-step and multiplexable nucleic acid amplification method. It has strong potential for bioanalysis, as it can be performed directly on top of a microarray, which in turn requires no additional purification and hybridisation steps.24 So far, NASBA assays have been restricted mostly to the analysis of intact RNA derived from cell lines, blood, urine samples or fresh frozen tissues. Here, we demonstrate that NASBA is also working reliably and sensitively with heavily degraded RNA samples extracted from FFPE tissues. Despite strong degradation of the template, RNA amplification of all tested genes resulted in strong and, if the same tumour is viewed, comparable hybridisation signals. Sensitivity tests showed that the RPS18 NASBA assay was more sensitive than real-time RT-PCR, while the sensitivity of the HER2, ERα, MMP11, YBX1, CASP8 and SOD2 NASBA assay was comparable with a PCR targeted to the respective genes. Our results indicate that NASBA is suitable to amplify even strongly degraded RNA isolated from FFPE tissues and therefore can complement already-existing amplification techniques with high sensitivity such as RT-PCR for analyses of these tissues.

Take-home messages

  • NASBA with RNA extracted from FFPE tissues is feasible and highly sensitive.

  • NASBA can complement already-existing nucleic acid amplification methods in FFPE tissues.

  • NASBA products can be used as analytes for microarray hybridisation, making this assay, also in FFPE tissues, very interesting for a broad range of new applications.

References

Footnotes

  • Funding This work was supported by the German Research Foundation (Sti 153/3-1, Ru 489/15-1, HA 5797/2-2).

  • Competing interests None.

  • Ethics approval Ethics approval was provided by the local ethical committee (No 275/2006).

  • Provenance and peer review Not commissioned; externally peer reviewed.