Article Text


Cervical cytology biobanking: quality of DNA from archival cervical Pap-stained smears
  1. G A V Boulet1,
  2. C A J Horvath1,
  3. S Berghmans2,
  4. L M Moeneclaey1,
  5. I S M Duys1,
  6. M Arbyn3,
  7. C E Depuydt2,
  8. A J Vereecken2,
  9. S Sahebali1,
  10. J J Bogers1,2
  1. 1
    Applied Molecular Biology Research (AMBIOR), Laboratory of Cell Biology & Histology, University of Antwerp, Antwerp, Belgium
  2. 2
    Laboratory for Clinical Pathology (Labo Lokeren, campus RIATOL), Antwerp, Belgium
  3. 3
    Unit of Cancer Epidemiology, Scientific Institute of Public Health, Brussels, Belgium
  1. Gaëlle Boulet, Applied Molecular Biology Research group (AMBIOR), Laboratory of Cell Biology and Histology, University of Antwerp, Groenenborgerlaan 171, BE-2020 Antwerp, Belgium; gaelle.boulet{at}


Aims: Cervical cytology biobanking is a feasible concept in cervical pathology and could be an indispensable tool for fundamental and applied molecular biological research. PCR is a powerful molecular technique that can be performed on a variety of cervical sample types including Pap-stained cervical smears. However, since the quality of DNA from such specimens is inferior to that from fresh tissue, the correct processing methods are required. This study evaluates three commercial isolation methods and one digestion procedure for their ability to obtain DNA suitable for PCR from fixed and stained Pap smears.

Methods: The High Pure PCR Template Preparation kit, the NucliSENS easyMAG system, the QIAamp DNA Mini Kit and crude proteinase K digestion were used to obtain DNA for subsequent PCR applications. Amplification of β-globin was performed to verify the presence and integrity of target DNA. The influence of PCR inhibitors and extent of DNA fragmentation were analysed.

Results: All commercial DNA isolation techniques provided DNA suitable for PCR amplification, and DNA isolated from 10-year-old archival smears yielded amplicons up to 400 base pairs. Conversely, crude proteinase K digestion limited the amplicon size to 300 bp and did not consistently yield amplifiable digests, as these were contaminated with PCR-inhibiting factors and debris.

Conclusion: The study indicates that commercial DNA isolation techniques are suitable for PCR amplification of DNA isolated from archival smears, yielding amplicons up to 400 base pairs. Proteinase K digestion is not suitable to obtain amplifiable DNA from fixed and stained Pap-stained smears.

Statistics from

Organised biobanking is a feasible concept in cervical pathology, because slides and paraffin blocks from the cervix are routinely stored for an average of 10 years for medicolegal reasons and quality control.1 2 PCR is a powerful molecular biological technique3 that can be applied to a variety of cervical sample types, including stored Pap-stained cervical smears.4 However, since the quality of DNA from such specimens is inferior to that from fresh material, proper processing methods are required.46 Recovery of good-quality DNA would allow ancillary testing for human papillomavirus (HPV) and retrospective studies based on archival material. These, in turn, may provide new insights into the molecular pathways of carcinogenesis and the optimal use of HPV tests in screening and vaccination strategies.

This study evaluates methods to retrieve DNA from archival conventional smears, and assesses DNA quality in terms of integrity and applicability for PCR purposes.


Study population

Forty Pap-stained smears, dating from 1997, were retrieved from the Laboratory for Clinical Pathology (Labo Lokeren, campus RIATOL, Antwerp, Belgium). The number of cells on each smear was determined by microscopic evaluation of seven ×400 fields distributed over the slide, followed by multiplication of the mean number of cells per field with the total number of fields per slide. Subsequently, the smears were classified as high, medium or low cellular. The study was performed in accordance with local ethics committee advice.

Processing of archival Pap-stained smears

Each smear was rinsed with 100% ethanol and immersed in xylene for 5 days until the coverslips could be easily removed. Cells were scraped from one half of the glass with sterile razor blades and suspended in 500 µl 100% ethanol. They were pelleted by centrifugation and washed twice with 100% ethanol. Pellets were air dried at 37°C for 30 min.

DNA extraction

Three commercial DNA isolation procedures and one in-house digestion method were tested to obtain DNA suitable for subsequent PCR applications. The ability to provide DNA was tested on cells from one-half of the smears. Ten smears were used for each technique. For each set of 10 smears, three showed low, three showed medium and four showed high cellularity. DNA extracts were stored at −20°C until PCR reactions were performed.

Three commercial DNA extraction kits were used according to the manufacturer’s protocol.

The High Pure PCR Template Preparation (HPPTP) kit (Roche Applied Science, Mannheim, Germany) is based on the specific binding of nucleic acids to the surface of glass fibres in the presence of chaotropic salt guanidine HCl.

The NucliSENS easyMAG system (BioMérieux, Marcy l’Etoile, France) is based upon the nucleic acid binding qualities of magnetic silica in combination with guanidine thiocyanate (GTC).

The QIAamp DNA Mini Kit (Qiagen, Hilden, Germany) is based on the adsorption of DNA onto the QIAamp silica-gel membrane of QIAamp spin columns.

In addition, crude proteinase K digestion, which is commonly used for DNA extraction in liquid-based cytology, was assessed.7 8 In brief, the cell pellet was resuspended in 50 µl digestion solution (10 mM Tris, 1 mM EDTA, 200 µg/ml proteinase K) and incubated for 3 h at 56°C. The digestion was followed by a 10 min incubation at 92°C to inactivate proteinase K.

To evaluate the presence and influence of inhibitors on PCR performances, this digestion procedure was repeated for the ten remaining halves of the smears, with an increased volume of digestion solution (200 µl), resulting in a fourfold dilution of the extracts.


Absorbance (A) at wavelengths 260 nm and 280 nm was measured. DNA concentration was determined based on A260, while the ratio A260/A280 indicated DNA purity.

PCR amplification

PCR amplification of β-globin was performed to verify the presence and integrity of target DNA (table 1).

Table 1 β-Globin primer and probe sequences in the conventional and qPCR assays

Conventional PCR was performed in a 20 µl volume containing 10 µl RedTaq (Sigma-Aldrich, St Louis, Missouri, USA), 7 µl water, 1 µl (10 µM) forward β-globin primer (GH20), 1 µl (10 µM) reverse β-globin primer (PC04) and 1 µl DNA template. The thermocycling profile consisted of 35 three-step cycles: denaturation at 95°C for 1 min, annealing at 55°C for 1 min, and polymerisation at 72°C for 5 min. PCR products were identified by gel electrophoresis. The intensity of bands at 268 base pairs (bp), corresponding with the β-globin amplicon, was measured according to a molecular mass marker, designed for size determination and quantification (SmartLadder; Eurogentec, Liège, Belgium) (Chemigenius2; Syngene, Frederick, MD, USA).

Quantitative real-time PCR (qPCR) was performed using the LightCycler 480 system (Roche Diagnostics, Basel, Switzerland) with primers and probes for the detection of an 80 bp fragment of the β-globin gene (table 1). All qPCRs were performed in a 5 µl volume containing LightCycler 480 Probes Mastermix (Roche Diagnostics), 200 nM of each β-globin primer and probe, and 1 µl DNA template. The 45 two-step cycles consisted of 15 s at 95°C and 50 s at 60°C. Standard curves for β-globin were generated by amplification of dilution series of the different DNA samples to evaluate qPCR efficiency.

To further asses the degree of DNA fragmentation, a conventional multiplex control gene PCR for amplicons of 100, 200, 300, 400 and 600 bp was used, as previously described.10

Several precautions were taken to prevent false-positive PCR results. Different steps such as DNA extraction, sample preparation, amplification and post-PCR were performed in separated rooms. Distilled water was included as negative PCR control, while HeLa, a HPV-positive cell line containing 10–50 copies of HPV18 per cell, served as a positive control sample.


Data were analysed using SPSS 14.0 for Windows 2000 (SPSS Inc., Chicago, Illinois, USA). Mann–Whitney and Kruskal–Wallis tests were performed to compare different DNA extraction methods and to assess the influence of cellularity on DNA recovery and PCR performances. A p value ⩽0.05 was considered statistically significant.


Absorbance measurements were not significantly different between the HPPTP, easyMAG, and QIAamp method, but the proteinase K method resulted in significantly higher A260 values (p<0.001) (table 2).

Table 2 Spectrophotometrical and PCR results for DNA obtained with different extraction methods

Conventional PCR amplification of a 268 bp β-globin fragment depended on the extraction method (fig 1). The HPPTP and QIAamp methods produced amplicons for all smears. The easyMAG technique yielded a β-globin amplicon in nine out of 10 smears and proteinase K digestion in seven cases (fig 1). β-globin amplicons were quantified. No difference in amount of amplified DNA was found between the commercial isolation methods. Only the QIAamp isolation and proteinase K digestion resulted in significantly different amounts of amplified DNA (p = 0.043) (table 2).

Figure 1 Conventional PCR and quantitative real-time PCR (qPCR) analysis of DNA obtained with different DNA extraction methods from archival Pap-stained smears using β-globin primers. Conventional PCR generating a 268 bp amplicon (a). The samples are organised according to cellularity. Lane 1 (+): positive control sample. Lanes 2–4 (samples 1–3): PCR products of archival smears with low cellularity. Lanes 5–7 (samples 4–6): PCR products of archival smears with medium cellularity. Lanes 8–11 (samples 7–10): PCR products of archival smears with high cellularity. Corresponding outcome of qPCR for each sample (ie, qPCR-positive (+) or negative (−)) (b).

qPCR of an 80 bp β-globin fragment detected the presence of amplifiable DNA in a qualitative manner, but did not show a steady increase of fluorescence in proportion to the amount of PCR product. Because there was no correlation between the crossing point and DNA concentration, standard curves and PCR efficiency could not be obtained. The smear, which was conventional PCR-negative after easyMAG isolation, was also negative by qPCR (fig 1). DNA obtained by proteinase K digestion was not consistently amplifiable by qPCR, as only five smears showed positive results. The three smears that did not result in amplification by conventional PCR, did not result in qPCR amplification either (fig 1).

The cellularity of the smears correlated with DNA recovery and PCR outcome. In case of proteinase K digestion, only high cellular smears rendered amplifiable DNA (fig 1).

A fourfold dilution of the proteinase K digests led to increased PCR positivity for both conventional and qPCR analysis (fig 2).

Figure 2 Conventional PCR and quantitative real time PCR (qPCR) analysis of DNA digests obtained with 50 µl (1:1) and 200 µl (1:4) proteinase K digestion solution from both halves of archival Pap-stained smears using β-globin primers. A fourfold dilution of the digest resulted in an increase of PCR positivity for both conventional (samples 3, 5 and 6) and qPCR analysis (sample 6).

Conventional multiplex PCR for five amplicons in a range of 100–600 bp determined the extent of DNA fragmentation. Isolated DNA generated bands up to 400 bp, while the proteinase K digests produced amplicons of maximal 300 bp (fig 3).

Figure 3 Multiplex PCR of DNA obtained with different DNA extraction methods from archival Pap-stained smears for the assessment of DNA fragmentation. The control amplicons have lengths of 600, 400, 300, 200 and 100 base pairs (bp). Lane 1 (+): positive control sample contains high molecular mass DNA. Lanes 2–4: amplicons up to 400 bp. Lane 5: amplicons up to 300 bp.

Take-home messages

  • Cervical cytology biobanks might prove to be an indispensable tool for fundamental and applied molecular biological research.

  • Proteinase K digestion is not suitable to obtain amplifiable DNA from fixed and stained Pap-stained smears.

  • DNA isolated from 10-year-old archival smears can yield amplicons up to 400 base pairs.


Cervical cytology biobanks are an extension of current clinical practice and might prove to be an indispensable tool for fundamental and applied molecular biological research. Molecular biological techniques have opened a new dimension in the retrieval of information hidden in preserved samples, but require the ability to obtain good-quality nucleic acids. Successful PCR analysis, which does not demand high quality DNA, appears to be the ideal method to assess archival material.11 However, studies have indicated that crude DNA digestion from fixed and stained cells can be a limiting factor in the PCR-based assessment of clinical samples.5 12 13

The importance of DNA suitable for molecular analysis is stressed by the emergence of HPV testing in the screening,1417 triage18 19 and follow-up17 20 of patients with cervical intraepithelial neoplasia or cervical cancer. HPV DNA genotyping and the determination of viral load and integration are currently proposed as risk markers in patient management.2132

DNA extracts from Pap smears often hamper proper evaluation due to substantial staining residues and debris. Spectrophotometrical approaches are considered insufficient because of protein crosslinking and the presence of compounds which interfere with light absorption.33 This explains the high absorbance values for the proteinase K digests in which debris remains present. Therefore, the absorbance results in this study are merely indicative and not considered conclusive for DNA quality.

To evaluate DNA recovery methods, amplification of β-globin was used as predictor of overall PCR feasibility. This assumption can be made when the length of the β-globin amplicon equals the length of other potential targets, such as HPV DNA.34 It should be stressed that the mean amount of amplifiable HPV DNA per cell often exceeds the copies of the β-globin gene.

The ability to provide DNA was tested on cells from one-half of the smears, while the other half was preserved for morphological purposes; this can be important for medicolegal reasons. The DNA recovery procedures were tested in different sets of 10 smears that were standardised in terms of rough categories of cellularity. This could limit the comparability between the methods. However, because the amount of amplified DNA correlated with cellularity, our conclusions can be considered as qualitatively reliable.

The HPPTP and easyMAG systems provided DNA isolates with similar features that were suitable for PCR purposes. The QIAamp kit also recovered appropriate DNA, which gave rise to significantly higher amounts of PCR product than the proteinase K digests. Because the HPPTP and easyMAG kits did not render significantly more amplified DNA, this could indicate superior performance of the QIAamp kit in removal of PCR inhibitors. Proteinase K digests resulted in remarkably less successful conventional PCRs and appeared to be even less appropriate for qPCR purposes. In general, DNA from Pap-stained smears proved to be unsuitable for qPCR analysis, restricting real-time PCR outcome to presence versus absence of the PCR target.

Our findings support those of other studies reporting poor amplification of proteinase K digests from Pap-stained smears.4 5 11 3538 Overall, PCR negativity could be explained by the low cellularity of the smears, the presence of PCR-inhibiting factors or DNA fragmentation. Contrary to the commercial DNA isolation procedures, the crude proteinase K protocol only releases DNA from the cells without binding, washing or isolating it. Staining residues and debris stay present in the DNA solution and can inhibit subsequent assays. Vulnerability of qPCR to debris and staining leftovers could be partially related to the small reaction volume. In conventional PCR, potential inhibitors are four times more diluted. The effect of this dilution was assessed by increasing the proteinase K digestion volume, which resulted in more successful PCR amplifications, because of the reduced concentration of inhibitors (fig 2). These experiments indicate that proteinase K digestion retrieves DNA with sufficient quality, while PCR inhibitors prevent successful amplification. Jacobs et al reported that DNA integrity and quantity showed little improvement after an additional purification step.4 The approach of dilution is not ideal to overcome PCR inhibition as it also reduces the DNA concentration. At low DNA concentrations, qPCR detection even is more susceptible to inhibitors increasing the detection limit.39 This also implicates important limitations to the cellularity of the smears.

Further experiments assessed whether DNA fragmentation would hamper PCRs that generate products longer than the 268 bp β-globin amplicon. Multiplex PCR indicated that integrity of DNA isolated from these 10-year-old archival smears was guaranteed up to 400 bp. For proteinase K digests the amplicon length was restricted to 300 bp (fig 3). The procedure of DNA recovery and coextraction of PCR inhibitors proved to influence the integrity of DNA from fixed and Pap-stained cervical smears and its PCR amplification; this has been described previously for DNA extracted from paraffin-embedded samples.10 Another possible explanation could be that PCR inhibitors in proteinase K digests preferentially hamper the amplification of longer amplicons.

Apart from efficient recovery, different DNA extraction techniques show different merits and pitfalls. The HPPTP and easyMAG systems are automated GTC-based multi-extraction methods that provide sufficient DNA from 10-year-old smears for reliable DNA amplification. The QIAamp kit is also appropriate for similar high-throughput applications. These methods have the advantage of being easy to perform with minor susceptibility to contamination. They open the possibility to perform extensive retrospective studies based on archival cervical samples to investigate the natural history of HPV infection, facilitate longitudinal assessment of the dynamics of HPV infection, large-scale biomarker discovery and validation, and promote the implementation of vaccination strategies. This study emphasises the importance of DNA technology, which is a critical factor in the feasibility of large-scale studies and the reliability of their results.


Thanks to Danny Vindevogel for the language review.


View Abstract


  • Competing interests: None.

  • Funding: GAVB is supported by the Fund for Scientific Research Flanders (FWO-Vlaanderen). CAJH is supported by the Institute for the Promotion of Innovation by Science and Technology in Flanders (IWT-Vlaanderen). JJB is supported by the Fund for Scientific Research Flanders (FWO-Vlaanderen, G.0205.04) and the Belgian Cancer Foundation (Belgische Stichting tegen Kanker). MA received funding from the 6th Framework Programme (European Commission, DG Research, Brussels, Belgium) through the CCPRB Network (University of Lund, Malmö, Sweden), the Institute for the Promotion of Innovation by Science and Technology in Flanders (IWT-Vlaanderen) and Belgian Cancer Foundation (Belgische Stichting tegen Kanker).

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.