Routine Acid Decalcification of Bone Marrow Samples Can Preserve DNA for FISH and CGH Studies in Metastatic Prostate Cancer

Abstract

Production of paraffin-section material from tissue samples that contain bone requires decalcification. Techniques such as acidic decalcification or EDTA chelation are suitable methods. Acid decalcification is generally quicker than EDTA chelation but studies have suggested that it may result in hydrolysis of DNA. Here we show that limited acid decalcification (less than 24 hr) in 5% formic acid can preserve DNA sufficient for fluorescent in situ hybridization (FISH) or comparative genomic hybridization (CGH) and that prolonged 10% formic acid decalcification results in failure of FISH and only limited retrieval of DNA for CGH studies.

Routine acid decalcification (RDO; Apex Engineering Products, Plainfield, IL) for bone marrow trephine and autopsy bone samples has been reported to result in failure to obtain DNA for in situ hybridization (ISH), comparative genomic hybridization (CGH), and flow cytometric studies in trephine and autopsy bone marrow samples from prostate cancer (Alers et al. 1999). Decalcification using 10% EDTA for three autopsy bone marrow specimens resulted in successful FISH, CGH, and flow cytometric studies and gave better preservation of architecture for routine staining and immunohistochemistry. The authors concluded that EDTA was highly preferable to their routinely used acid decalcified bone where studies on DNA were required.

We also routinely use acid decalcification for bone marrow trephine specimens (12–18 hr) but use a 5% formic acid (Becton Dickinson Laboratory Supplies; Mountain View, CA) solution in distilled water rather than a hydrochloric acid-based product (RDO). This method of decalcification has enabled us to success-fully perform FISH on 15 of 15 samples analyzed for androgen receptor (AR) gene copy number in bone metastases from prostate cancer (SpectrumOrange AR FISH probe, SpectrumGreen centromere FISH probe for the X chromosome; Vysis, Downers Grove, IL) (see Figure 1A). AR gene amplification has been implicated in the development of hormone-refractory prostate cancer (HRPC), being found in approximately 30% of locally recurrent hormone-refractory biopsies and rarely in hormone-naïve specimens (Visakorpi et al. 1995). A similar figure has been reported for the prevalence of AR gene amplification in distant metastases from HRPC (Bubendorf et al. 1999), but to date no studies have looked at AR gene amplification in bone metastases. This was the primary aim of our study. Amplification of the AR gene was found in 5/12 (38%) of bone metastases (in the remaining three cases, scoring for AR gene copy number in tumor cells was not performed because of a low number of malignant cells in the trephine biopsies).

The FISH studies on decalcified trephine material required an additional pretreatment, i.e., microwave antigen retrieval technique, before standard tissue digestion with pepsin (M. Farquharson, Department of Pathology, Glasgow Royal Infirmary; personal communication).

Figure 1(A)

Successful FISH study using decalcification with 5% formic acid for a bone marrow trephine biopsy from a patient with hormone-refractory prostate cancer. Background nuclear stained is with DAPI, the androgen receptor gene probe is visible in red, and the X chromosome centromeric probe is in green. The specimen shows amplification of the AR gene, indicated by multiple red AR signals each associated with a single green X chromosome centromeric signal. Paler red signal clusters represent other tumor cells with AR gene amplification slightly out of plane of focus deeper within the 5-μm section. Note gland formation by the tumor within the marrow. Original magnification × 1000. (B) Failed FISH specimen from an autopsy bone marrow sample showing the absence of either green X centromeric probe signal or red AR gene signal after decalcification in 10% formic acid for 10 days. Original magnification × 400. (C) Successful FISH study from a lymph node section from the same patient as in B with hormone-refractory prostate cancer. The tumor cells are disomic for the X chromosome with two copies of the green X centromeric probe and red AR gene, unlike normal male cells that show only one signal of each type. Original magnification × 1000.

Figure 2

Hybridization of extracted tumor DNA to normal metaphase chromosomes for comparative genomic hybridization.

Formic acid decalcification has also enabled us to successfully hybridize extracted tumor DNA from decalcified bone marrow trephines to normal metaphase chromosomes in five cases studied thus far (Figure 2) for CGH. Degenerate oligonucleotide primer (5′-CGACTCGAGNNNNNNATGTGG-3′) polymerase chain reaction (DOP-PCR) products of up to 1500 base pairs in length have been generated from all 15 decalcified trephine samples (Figure 3).

For larger autopsy bone marrow specimens, we use a 10% formic acid solution for decalcification over a 7–10-day period (with replenishment of the solution on one to two occasions during this period). Similarly to Alers et al. (1999), we found that in five autopsy bone marrow specimens studied FISH could not be successfully performed, whereas studies on soft tissue samples from the same autopsy cases were successful (Figures 1B and 1C). Unlike the cases reported by Alers et al., we have been able to produce DNA fragments using DOP-PCR of up to 1500 base pairs suitable for CGH in two of five of autopsy cases studied (Figure 3).

Little published information is available on the effects of formic acid decalcification on DNA degradation. We have been unable to find other reports of successful studies using FISH or CGH on formic acid-decalcified bone marrow trephine biopsies (Medline Search 1966–2001). Sarsfield et al. (2000) have reported that formic acid decalcification of bone marrow trephines degrades DNA for PCR using specific primers. In this comparative study of formic acid and EDTA decalcification in 11 bone marrow trephine specimens, similar quantitative amounts of DNA could be retrieved using both decalcification methods, but formic acid pretreatment resulted in a degraded DNA smear and failure to obtain specific PCR products in the majority of samples tested. Provan et al. (1992) were also able to generate a 294-base pair PCR product in only six of ten after formic acid decalcification of paraffin-embedded bone marrow biopsies. By contrast, EDTA decalcification resulted in successful amplification of specific PCR products up to 643 bp in the study by Sarsfield et al. (2000), and the author's previous experience of EDTA for decalcification of bone marrow trephines alone strongly supports this finding (Wickham et al. 2000).

We conclude that routine acid decalcification with 5% formic acid can satisfactorily preserve DNA for some types of molecular biological studies (FISH and CGH) on prostate cancer bone metastases taken by trephine biopsy. Longer and stronger acid decalcification resulted in unsuccessful FISH in all autopsy specimens, whereas non-calcified material from the same postmortem samples worked satisfactorily. Decalcification using EDTA appears to offer the best chance of successful DNA retrieval from bone tissue and appears to be the method of choice for decalcification in a prospective study of bone marrow trephine and autopsy samples. Five percent formic acid decalcification avoids major acid hydrolysis of DNA, as judged by our results with FISH and CGH studies on decalcified trephine material. Attempts to use techniques such as FISH and CGH on archival material should not automatically be abandoned if retrospectively collected specimens have been decalcified in formic acid.

Figure 3

Agarose gel electrophoresis of DOP-PCR DNA fragments after extraction and formic acid decalcification of bone marrow biopsies. BMT, 5% formic acid-decalcified bone marrow trephine biopsy. PM, 10% formic acid decalcification of an autopsy bone sample. T, tumor; N, normal tissue; L, DNA base pair ladder.