We have assessed the use of two types of archived specimens for both array CGH and SNP array. Cytogenetic preparations from diagnostic analysis are often archived in laboratory freezers and should be seen as a potential source of archival material for research into the genetics of malignancy and inherited disease. In our laboratory, fixed cytogenetic preparations are often the only stored patient specimen available. Bone marrow specimens received for karyotyping are also a potential source of DNA, but it has been assumed that the DNA is not suitable unless processed or frozen immediately. We have shown that these specimens can be successfully used for array karyotyping.
In 1986 Barker et al. reported that DNA suitable for Southern analysis can be extracted from fixed cytogenetic preparations, using a phenol/chloroform protocol, and they suggested that this DNA might be suitable for other analytical purposes. Our array CGH results confirm the use of DNA extracted from fixed cytogenetic preparations for array CGH [3, 4, 7], and here we show that the quality metrics of the results compare favourably with results from one day old bone marrow. We have shown for the first time that DNA from fixed cytogenetic preparations can also be used for SNP array. DNA prepared from fixed chromosomes was slightly degraded but was not degraded further with longer storage at -80°C (Figure 6).
We describe a modified DNA extraction method for use with cytogenetic suspensions. By processing a known cell number and performing an initial lower volume elution, the amount of specimen used can be kept to a minimum. This is important for limited volume specimens or specimens with low cellularity. In low cellularity specimens this produced an eluant that could be used without further concentration and loss of DNA. Interestingly we have managed to extract a small amount of highly degraded RNA from fixed cytogenetic suspensions which have been used successfully for Real-Time PCR .
Validation of array CGH and SNP array results showed that they were reliable. Direct comparison of the same specimen processed fresh or fixed showed only a small increase in the spread of B allele frequencies in results derived from a fixed specimen when compared with the same cell line processed fresh. A deletion producing a shift in BAF values from 0.5 to 0.4 and 0.6 was clearly visible in both specimens (equivalent to a deletion in 1/3 of diploid cells).
We also achieved reliable SNP array and array CGH results using unprocessed bone marrow specimens refrigerated for 36 days or more before freezing. Agarose gel analysis showed that the DNA in bone marrow stored at 4°C degraded over time (see Figure 6). However, specimens stored for up to 36 or 42 days at 4°C were still suitable for both array CGH and SNP array, respectively.
As it may occasionally take weeks to determine a karyotype, considerable time and resources can be saved if only the specimens of interest are frozen or processed after a karyotype is known. This approach is particularly suitable for bone marrow, which is difficult to re-collect. Also, therapy may have been administered or the genotype of malignant cells may have changed by the time a subsequent collection is contemplated.
Cytogenetic slides are another potential source of DNA for microarray studies. A method has been described for extracting DNA from chromosomes scraped from microscope slides for PCR . However, around 106 nuclei are needed for the extraction protocol we describe, and so a single cytogenetic slide would not produce enough DNA for array karyotyping without amplification. Thus, while DNA could potentially be obtained from slides, use of unspread specimens is simpler, and is less likely to require amplification.
Other methods of DNA analysis may also be possible using DNA extracted from these specimens. DNA extracted from formalin-fixed paraffin embedded tissues (FFPE) can be used for array CGH and SNP array analysis, although sensitivity is much poorer than for fresh tissue [10–12]. Whole genome amplification makes it possible to use a small starting amount of DNA [10, 12]. Improved protocols make sequencing of degraded DNA extracted from FFPE specimens possible [13, 14], suggesting that DNA from cytogenetic preparations will be even more suitable for this and other methods of DNA analysis. Massively parallel sequencing can be carried out on as few as six microdissected and amplified chromosome segments .