Characterization of Effect of Repeated Freeze and Thaw Cycles on Stability of Genomic DNA Using Pulsed Field Gel Electrophoresis

Abstract

In this study, we systematically investigated the effects of repeated cycles of freeze/thaw on stability of genomic Deoxyribonucleic acid (DNA) samples as evaluated by changes in DNA size, concentration, and freeze/thaw protocols. The DNA was isolated using standard extraction procedures including phenol/chloroform and commercially available Gentra Puregene and Qiagen QIAmp kits. Changes in DNA were monitored over 18 cycles of freezing and thawing utilizing several freeze/thaw protocols. DNA samples from multiple subjects prepared from whole blood samples were examined by pulsed field gel electrophoresis (PFGE), and shown to have different average molecular sizes and size distribution patterns depending on the extraction method. Results of freeze/thaw experiments, analyzed by PFGE, showed progressive DNA degradation of the samples, with DNA sizes larger than 100 kb most sensitive to freeze/thaw degradation. Increasing the DNA concentration of stored samples from 10 μg/mL to 100 μg/mL had a somewhat protective effect on DNA stability. Variations in freeze/thaw protocols did not have a significant impact on DNA stability during repeated freeze/thaw cycles. At freeze/thaw cycle 18, average molecular size and size distribution of all DNA samples tested approached 25 kb, regardless of their initial average size and size distributions. This study provides insight on DNA degradation during freeze/thaw cycles and offers guidance to storage and handling of DNA samples.

Introduction

As advances in genomics and genetics are made, scientists are gaining a greater understanding of the roles that genetic determinants play in diseases, including cancer, heart disease, mental illness, and infectious diseases. Scientific advances in the field have been strongly dependent on the developments of innovative genetic technologies such as polymerase chain reaction, sequencing, and whole genome scan, to compare Deoxyribonucleic acid (DNA) from patients with disease-free controls. DNA samples have been one of the most important sources for understanding the fundamental basis of human life and disease, and have been instrumental in the development of a novel group of therapeutics, transgenes for gene therapy, and oligonucleotides for antisense and antigene applications, modeled on the endogenous structure of DNA. Collection, handling, and storage of DNA samples are, thus, of critical importance in gaining reliable data in disease association studies and development of disease treatments.^1–3

The collection, handling, and storage of DNA samples often involves freezing and thawing of the samples, although newer technologies hold the promise of dry storage.^4–5 The freeze/thaw cycle is usually defined as the phase transition of the aqueous solution from liquid state to solid state, and vice versa. The freeze/thaw effect on biological systems has long been recognized. Many studies have been conducted to look into how the freeze/thaw effect affects the stability and viability of cells (tissues), bacteria, and virus, as well as the stability of biomolecules such as proteins and nucleic acids.^2,6–8 Grecz' studies on the stability of DNA in Escherichia coli, or cells that were frozen and thawed, revealed a reduction in DNA sizes.^9,10 Lyscov's study, from measurements of intrinsic viscosity and sedimentation constants of calf-thymus DNA and phage-T2 DNA, revealed that a decrease of molecular weights of the DNA samples was caused by freeze/thaw.¹¹ Lindley assessed DNA damage of human cumulus cells caused by freeze/thaw. They found that DNA was damaged using a modified comet assay.¹² Sanlidag and Krajden studied the freeze/thaw effect on the stability of viral DNA and RNA, because they were primarily concerned that the freeze/thaw cycles might affect the accuracy of viral copy number determination. They concluded that the copy numbers of viral DNA of infected cultured cells did not change remarkably after 18 cycles of freeze/thaw.^13–15

The majority of the studies conducted to date have involved the actual source materials used to obtain DNA or RNA, such as tissues, cells, and whole blood. Effect of freeze/thaw cycles on extracted genomic DNA stability during storage has not been systematically studied, especially using pulsed field gel electrophoresis (PFGE). As we advance toward defining “best practices” for biospecimen storage and handling, it is important to systematically study the effect of freeze/thaw cycles on the stability of genomic DNA in order to offer guidance on effective storage of genomic DNA samples. In this article, we submit the results of a series of experiments designed to investigate freeze/thaw cycle effects on the stability of genomic DNA using PFGE.

Materials and Methods

DNA sample extractions

Six vials (Becton Dickinson; ACD vacutainer tube) of whole blood samples were collected from each of 10 anonymized donors through the NCI-Frederick normal research donor program. These samples were stored at −70°C until DNA extractions were performed. Three extraction methods were chosen for DNA extraction: Qiagen QIAamp Blood Maxi Kit (Qiagen). Gentra Puregene Blood Kit (Qiagen); and phenol/chloroform method. For the commercial kits, extractions were performed following the manufacturer's instructions. Two vials from each donor were extracted with each of the three extraction methods. Concentrations of the extracted DNA samples were determined using an optical density at 260 nm (SpectraMax Plus; Molecular Devices) and PicoGreen assay (PicoGreen reagents from Molecular Probes; TECAN GENios microplate reader). For assessment of quality of these extracted DNA samples, we determined the ratios of optical density at 260 nm over at 280 nm. The integrity of the DNA samples was visualized on 1% agarose gel stained with ethidium bromide under UV light. Finally, TaqMan assay for single nucleotide polymorphism (SNP) at VDR (vitamin D receptor) gene TaqI site¹⁶ was performed to assess the suitability for use in downstream assays as a further measure of the quality of the extracted DNA samples. The DNA samples had been stored at 4°C for four months before the freeze/thaw experiments.

Freeze/thaw experiments

A total of nine samples from three donors were used for the freeze/thaw experiments. Samples 1, 4, and 7 were from donor I; samples 2, 5, and 8 were from donor II; and samples 3, 6, and 9 were from donor III. The first sample from each donor (samples 1, 2, and 3) was extracted with Qiagen kit; the second sample (samples 4, 5, and 6) was extracted by phenol chloroform protocols; and the last sample (samples 7, 8, and 9) was extracted using Puregene kit. The concentrations of the DNA samples were adjusted to 100 μg/mL with Tris-EDTA (TE) buffer (pH 8.0), which contains 10 mM trishydroxymethylaminomethane–hydrogen chloride/1 mM ethylenediaminetetraacetic acid purchased from Quality Biologicals. Each of the nine samples was aliquoted into 84 0.5 mL screw cap tubes (3 sets of 28 tubes) with each tube containing 20 μL. Of these, the first set was diluted with TE to 10 μg/mL, the second set was diluted with TE to 50 μg/mL, and the final set of 28 was kept at 100 μg/mL. Thus, a total of 756 DNA aliquots were created from the nine DNA samples for the freeze/thaw experiments. All the DNA samples were in TE buffer (pH 8.0).

Four freeze/thaw protocols were designed for the experiments as shown in Figure 1. Both protocols A and B cycled through −70°C, but protocol A had a gradual change in temperature. Protocol C only went to −20°C, while protocol D did not go through the freeze/thaw stage. Thus, protocol D served as a control for comparison for the experiments. Each group of 28 DNA aliquots, from the same sample, with the same concentration, was further subdivided into four sets. In each set, there were seven samples (seven vials) for each specified freeze/thaw protocol. The first vial of each set was directly stored at 4°C. The number of freeze/thaw of the first control sample vials was assigned as zero, as it would not go through the freeze/thaw cycles. At the end of every three cycles of freeze/thaw, a vial from each set was removed and stored at 4°C until analysis. The last vial from each set was removed after 18 cycles of freeze/thaw. The freeze/thaw experiments were completed in 18 days. In order to minimize any differences in pipetting error rate due to differences in viscosity of DNA at different concentrations, all samples were diluted to 10 μg/mL for analysis.

FIG. 1.

Freeze/thaw experiments diagram.

Pulsed field gel electrophoresis

The experiments of PFGE were performed approximately one month after completion of the freeze/thaw experiments. We loaded 250 ng of each sample on a 1% agarose gel. Several DNA size markers were chosen for the experiments: 5 kb DNA marker from Bio-Rad; lambda DNA mix 19 from Fermentas; lambda DNA HindIII digest and intact lambda DNA from New England Biolab; and ProMega-Markers lambda ladders from Promega. Electrophoresis conditions of 16 h at 180 V with a switch time of 4 to 20 s were used for the initial runs. The electrophoresis conditions were optimized for different samples based on the initial PFGE results. Using Bio-Rad CHEF MAPPER, PFGE was performed on Qiagen kit extracted DNA samples; for example, sample 2 at 180 V with a switch time of 4 to 15 s for 16 h; size markers were 5 kb DNA marker from Bio-Rad. Phenol extracted DNA samples were electrophoresed at 60 V for 16 h and 180 V for 8 h, with a switch time of 7 to 60 s; size markers were lambda DNA HindIII digest and intact lambda DNA from New England Biolab, and ProMega—lambda ladders from ProMega. Gentra Puregene kit DNA samples were electrophoresed at 180 V for 16 h with a switch time of 4 to 20 s; 5 kb DNA marker was from Bio-Rad. The gel images were taken using a KODAK EDA 290 system, and the DNA bands were analyzed using KODAK 1D image analysis software. Relative fluorescence intensity on Y axis in gel profiles were expressed in terms of arbitrary densitometric units indicating the intensity of fluorescence emitted by DNA-bound ethidium bromide after the background had been subtracted. Relative fluorescence intensities labeled on all profile drawings within the same figure were on the same scale. As representatives, only the gel profiles of samples 2, 5, and 8, which were from the same donor, were shown.

Statistical analysis

Data from linear ranges of high DNA size side of pulsed field gel electrophoresis profiles among freeze/thaw cycles, protocols, and DNA concentrations were compared using paired student's t-test (Fig. 3).

Results

Before the freeze/thaw experiments were performed, the extracted DNA samples were evaluated. The samples, on 1% agarose gels, appeared generally intact, and genotyping tests on VDR gene TaqI SNP showed that all samples could be amplified and genotyped. The DNA samples were selected for subsequent freeze/thaw experiments based on DNA, which was recovered at concentrations >100 μg/mL using all three extraction platforms. The original DNA recovered using each extraction method was subjected to PFGE on agarose gels to compare size and size distribution. DNA samples from a representative donor extracted using the three extraction methods prior to being subjected to freeze/thaw transition are shown in Figure 2A. The figure shows that DNA extracted with phenol/chloroform has the largest size and a relative uniform size distribution pattern over the range of 12 kb to over 300 kb. Gentra Puregene generated the next highest DNA fragments in size, averaging approximately 60 kb, with a size distribution ranging from 12 to 100 kb, while Qiagen extracted DNA had the lowest average size, approximately 30 kb with a size distribution of about 8–80 kb. Although the ranges of size distribution of DNA samples extracted with Puregene and Qiagen blood kits were similar, there was a clear difference in the shapes of the size distribution curves, with the Puregene extracted DNA curve shifting toward a higher average size. After 18 repeated freeze/thaw cycles, the DNA size and size distribution shifted toward the smaller size direction, as shown in Figure 2B. The average length of DNA after up to 18 freeze/thaw cycles approached 25 kb, regardless of the method of extraction or starting DNA size.

FIG. 2.

Comparison of size distribution of genomic Deoxyribonucleic acid (DNA) samples at freeze/thaw cycles 0 and 18. (A) at freeze/thaw cycle 0 and (B) at freeze/thaw cycle 18. Conditions: protocol B, samples 2, 5 and 8 with concentration of 10 μg/mL, 1% agarose gel. 180 V, 16 h, switch time 4–20 s.

FIG. 3.

Gel image of pulsed field gel electrophoresis (PFGE) of DNA sample 8 with concentration of 10 μg/mL with a different number of freeze/thaw cycles and under different freeze/thaw protocols. (A) DNA at protocol A; (B) DNA at protocol B; (C) DNA at protocol C; and (D) DNA at protocol D. Seven samples in each group represent the number of the freeze/thaw cycle at 0, 3, 6, 9, 12, 15, and 18. Lanes at both sides are 5 kb markers. 1% agarose gel, 180 V, 16 h, switch time: 4–20 s.

The effect of the number of freeze/thaw cycles and freeze/thaw protocols on DNA sample stability was evaluated using PFGE. A representative gel image is shown in Figure 3, which displays a pulsed field gel image of a Puregene DNA sample after multiple freeze/thaw cycles using the four protocols, outlined in Figure 1. All seven samples exposed to protocol D only went through repeated cycles of a temperature change from 4°C to room temperature, and back to 4°C. They had essentially no change in DNA size over the time course of the experiment, indicating that routine procedures associated with sample handling did not cause the obvious degradation of DNA in size. In contrast, the other three DNA sets that underwent repeated cycles of freeze/thaw protocols A, B, and C showed signs of progressive degradation. Although only minor size distribution changes were observed between adjacent lanes representing three cycles of freeze/thaw, the changes between adjacent lanes were significant (P<0.05) for Puregene and phenol/chloroform extracted samples at 10 and 50 μg/mL. For Puregene and phenol/chloroform extracted samples at 100 μg/mL and Qiagen extracted samples except for protocol C at 100 μg/mL, the changes in every 6 cycles became significant (P<0.05), but not after 12 cycles.

The effect of freeze/thaw protocols on DNA sample 5 that was extracted by phenol/chloroform method is shown in Figure 4. In the figure, the DNA sample had an initial DNA size range of 12–300 kb. Significant DNA degradation occurred in samples subjected to all three experimental freeze/thaw protocols A, B, and C (P<0.05). The 300 kb DNA fragments of the samples degraded to a maximum size of approximately 120 kb after 18 cycles of freeze/thaw, while the proportion of small DNA fragments (<80 kb) increased substantially. In addition, distribution of DNA sizes became narrower as the number of freeze/thaw cycles increased. Meanwhile, the average size of the samples degraded approximately from 100 kb to 35 kb. Although samples exposed to freeze/thaw protocol A experienced relatively slow temperature changes compared with those exposed to protocol B, both sets of samples had very similar DNA size change profiles (P>0.05). Comparing DNA size changes in samples exposed to freeze/thaw protocols A and B with samples exposed to protocol C, it appears that degradation was slightly slower in protocol C samples subjected to freezer temperatures of −20°C. As previously mentioned, size reduction of Qiagen extracted samples at 100 μg/mL for protocol C was not significant every 6 freeze/thaw cycles, while it was significant for protocols A and B. However, the difference in protocols A and B with C was not significant (P>0.05).

FIG. 4.

PFGE profiles of DNA sample 5 at different number of freeze/thaw cycles under the four different freeze/thaw protocols. 1% agarose gel, 60 V, 16 h, 180 V, 8 h, switch time 7–60 s. DNA sample 5 was extracted according to phenol chloroform protocol; seven samples in each group representing number of freeze/thaw cycle at 0, 3, 6, 9, 12, 15, and 18. In the graph, the legends represent sample - DNA concentration - freeze/thaw protocol - freeze/thaw cycle number. For example, 5–10-C-6 stands for sample 5 - DNA concentration of 10 μg/mL - freeze/thaw protocol C - 6 cycles of freeze/thaw.

DNA samples prepared with different DNA extraction methods and with different concentrations exhibited different stabilities against repeated freeze/thaw cycles. Figure 5 shows the degradation of the three DNA samples caused by freeze/thaw cycles. Sizes of the DNA samples ranged from small to large, the order being Qiagen kit (sample 2), Puregene kit (sample 8), and phenol-chloroform method (sample 5). The extent of degradation of the DNA samples is shown in the same order. For DNA extracted using the Puregene kit, the freeze/thaw resulted in a shift in the DNA peak from 60 kb to 35 kb after 18 cycles of freeze/thaw, while DNA extracted using Qiagen QIAmp kits, as represented by sample 2, showed the least change with freeze/thaw, with an average size decrease of approximately 5 kb after 18 cycles of freeze/thaw. Although size change in the DNA samples extracted by Qiagen kit was small, it is significant (P<0.05). The primary target of DNA degradation appeared to be associated with the largest size DNA, for example, in sample 5, with accumulation of smaller size DNA fragments in the range of 10–80 kb. After repeated freeze/thaw cycles, the size distribution of DNA extracted using phenol/chloroform or Puregene begins to approach that of DNA extracted using Qiagen, which is only minimally affected by freeze/thaw cycles.

FIG. 5.

PFGE profiles of DNA samples 2, 8, and 5 after different numbers of freeze/thaw cycles through protocol B at DNA concentration of 10 μg/mL, which were obtained using PFGE. 1% agarose gel. Electrophoresis conditions: sample 2–16 h at 180 V with a switch time 4–15 s; sample 8–16 h at 180 V with a switch time 4–20 s; sample 5–16 h at 60 V and 8 h at 180 V with a switch time 7–60 s. In the figure, the legends represent sample - DNA concentration - freeze/thaw protocol - freeze/thaw cycle number. For example, 2-50-B-12 stands for sample 2 - DNA concentration of 50 μg/mL - freeze/thaw protocol B - 12 cycles of freeze/thaw.

We found that the genomic DNA samples degraded as the number of freeze/thaw cycles increased and was directly related to the size of the isolated DNA, with the largest DNA fragments most susceptible to degradation. As the number of freeze/thaw cycles increased, the amount of DNA greater than 100 kb decreased, while the amount of DNA that was less than 100 kb increased for DNA extracted using phenol/chloroform method (Fig. 5). For every three freeze/thaw cycles, sizes of DNA sample 5 decreased approximately by 20 kb, while roughly 5 kb or less was lowered for samples 8 and no significant change for sample 2.

Increasing the concentration of stored DNA resulted in a slight but significant increase in the stability of DNA (P<0.05). After 18 freeze/thaw cycles, the peak of sample 5 shifted from 25 kb to 35 kb, when the concentration of DNA increased from 10 μg/mL to 100 μg/mL (Fig. 6). Although PFGE detected degradation of Qiagen DNA samples at 10 and 50 μg/mL (P<0.05) during repeated freeze/thaw cycles, degradation was not significant in Qiagen DNA stored at the concentration of 100 μg/mL.

FIG. 6.

PFGE profiles of DNA samples 5 after different numbers of freeze/thaw cycles through protocol B at concentrations of 10 μg/mL, 50 μg/mL, and 100 μg/mL from top to bottom. Electrophoresis conditions: 16 h at 60 V and 8 h at 180 V with a switch time 7–60 s.

Discussion

In this study, we chose three extraction methods for DNA sample preparation. The DNA samples were prepared using Qiagen QIAamp extraction kit, Puregene extraction kit, and phenol/chloroform method, respectively, providing DNA of different sizes and size distribution. These outcomes may be the result of the principles of the three extraction methods. Phenol/chloroform extraction is a liquid phase based extraction method that is a more gentle procedure; the Puregene kit is a salting-out procedure, involving solid salt precipitation; and the Qiagen QIAamp is a liquid-solid phase extraction method, in which large DNA molecules tend to bind on the silica surfaces and do not elute as efficiently as smaller molecules, or are eluted by the shearing force of centrifugation. These three methods are representative of commonly used extraction procedures and provide DNA with different average sizes and different size distributions. Using these DNA samples to study storage and effects of repeated freeze/thaw cycles is of both practical and theoretical significance.

From the experiments, we found that the genomic DNA samples degraded as the number of freeze/thaw cycles increased and was directly related to the size of the isolated DNA with a large-sized DNA most susceptible to degradation. The average length for DNA approaches 25 kb after up to 18 freeze/thaw cycles, regardless of the method of extraction or starting DNA size (Fig. 2B). To explain how the freeze/thaw cycles affect the stability of the genomic DNA, it is necessary to consider DNA conformation and the behavior of water molecules during freeze/thaw cycles. DNA is a random coil in solution. When the temperature is reduced to the freezing point of water, water molecules rearrange and form hexagonal ice crystals, which expand to occupy a larger volume than water in the liquid state.^17,18 The formation of ice crystals during freezing and reformation of ice crystals during thawing generates enormous tension forces. This tension force on DNA molecular chains could lead to breakage of DNA chains. When DNA fragments undergo repeated freeze/thaw cycles, the tension forces are likely to be generated on these same DNA fragments until they reach a size where the tension forces are not sufficient to cause DNA chain breaks at a high rate. This would explain the observation that the DNA size distribution moved toward lower molecular size with increased freeze/thaw cycles, but the effect was less prominent on smaller DNA extracted using Qiagen kit compared with the effect on large-sized DNA extracted using phenol/chloroform. With up to three freeze/thaw cycles, DNA degradation of samples was minimal. However, as the number of freeze/thaw cycles increased, the DNA size profile showed progressive degradation of DNA. After 18 freeze/thaw cycles, our study clearly shows that larger molecules degraded more readily. Distribution of DNA sizes becomes narrower as the number of freeze/thaw cycles increased, which also supports the observation that smaller DNA chains in size such as 30–40 kb are more resistant to freeze/thaw than larger DNA chains. In sample 5, degradation in size was most likely a result of degradation of the DNA more than 100 kb in size. It is easy to understand at a molecular level that the larger the size of DNA molecules, the larger the size of the random coil they form. When DNA samples are exposed to a freeze/thaw process, large coils occupy more space and have a higher likelihood of being trapped. Thus, greater tension force causes DNA molecules to break. Theoretical calculations, assuming DNA behaves similar to a rod, suggest that rupturing of molecules occurs at the midsection of the chains.¹⁹ Levinthal and Davison^19,20 found that the tensional force is proportional to DNA sizes.

The effect of concentration of DNA on DNA degradation was also studied. Increased DNA concentration, to some extent, reduced degradations of DNA fragments caused by repeated freeze/thaw cycles. This can be explained as higher freezing point depression or better vitrification around DNA fragment regions. In addition, DNA fragments are more compact at a higher concentration.²¹

To test the effects of freeze/thaw cycles on DNA stability in storage, we designed four freeze/thaw protocols. They were selected based on differences in freeze/thaw rates and temperatures, and also because storage temperatures in protocols B–D are routinely used in DNA sample storage and analysis. Both samples in protocols A and B were stored at −70°C. However, protocol A was a gradual freeze/thaw cycle, in which the temperature changed relatively slowly, and the DNA samples were frozen and thawed gently. Protocols B and C were direct, more standard freeze/thaw cycles, with a sharp change in temperature, and differentiated by the “storage” temperature to mimic either a standard laboratory freezer or an ultra-low mechanical freezer routinely found in repository facilities. In addition, there is a concern that samples stored at −20°C may not be completely frozen at this temperature.^22,23 In protocol D, the DNA samples were not frozen and thawed. However, they underwent the same sample handling processes, such as mixing and centrifugation, as the other samples. The purpose of protocol D was to exclude effects other than freeze/thaw cycles on DNA degradation over the time period of the study and as such was a control. The effects could be due to sample handling, centrifugation, and possible enzyme digestion, if any. From the PFGE profiles, we found that, in protocols A, B, and C, as the number of freeze/thaw cycles increased, the DNA sizes tended to decrease in all samples, while in protocol D, DNA samples displayed no obvious change in sizes, as these samples were not exposed to a temperature below 0°C. This indicates that DNA samples are stable in general under common sample handling procedures as those in protocol D.

The degradation patterns from the DNA samples exposed to any of the three freeze/thaw protocols A and B were very similar, which indicates that different freeze/thaw speed did not affect DNA degradation in protocols A and B. DNA stored at −20°C (protocol C) appeared degraded at a slightly slower rate than DNA exposed to −70°C (protocols A and B), which suggests that the samples might not be completely frozen at −20°C. However, this slight difference observed in Fig. 4 was not significant.

In summary, freeze/thaw cycles induce progressive degradation of genomic DNA under multiple conditions evaluated in this study. The greater the number of freeze/thaw cycles, the greater the degree of degradation of genomic DNA. The larger the size of DNA, such as DNA samples extracted using phenol-chloroform method, the more susceptible it is to degradation by freeze/thaw cycles. After 18 freeze/thaw cycles, the average size of genomic DNA approached 25–35 kb, regardless of its initial size or the method of extraction. Variations in freeze/thaw protocols had only a minor effect. DNA stored at a concentration of 100 μg/mL was slightly more stable during freeze/thaw phase cycles than DNA stored at 10 μg/mL. Exposure of DNA to multiple freeze/thaw cycles may have an impact on its suitability for use in some downstream applications. The samples from this study were still usable for standard TaqMan genotyping applications even after up to 18 cycles of freeze/thaw. However, applications that are dependent on large size of DNA such as long-distance polymerase chain reaction and multiple displacement amplification methods of whole genome amplification would be expected to be affected by freeze/thaw reduction in template size. Multiple cycles of freeze/thaw could be expected to result in low yield and asymmetric amplification and possibly amplification failure in these applications. In addition, it may affect the quantitation of DNA samples for certain assays. The effect of freeze/thaw on DNA side groups is not clear and will be a focus of future studies.

Footnotes

Acknowledgments

The authors thank Dr. Jimmie B. Vaught for his support and for reviewing the article. The authors also thank Dr. Thomas Beck for kindly allowing them to use the PFGE device and Ms. Leslie Garvey, Mr. Donald Simms, and Ms. Elizabeth Shannon for excellent technical assistance.

This project has been funded in whole or in part with federal funds from the National Cancer Institute, National Institutes of Health, under contract HHSN261200800001E. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does it mention trade names, commercial products, or organizations implying endorsement by the U.S. Government.

Author Disclosure Statement

No competing financial interests exist.

References

Vaught

. Biorepository and biospecimen science: a new focus for CEBP. Cancer Epidemiol Biomarkers Prev, 2006; 15:1572–1573.

Liu

, Shirano

, Fukaki

et al. Complementation of plant mutants with large genomic DNA fragments by a transformation-competent artificial chromosome vector accelerates positional cloning. Proc Natl Acad Sci U S A, 1999; 96:6535–6540.

Anderson

. Human gene therapy. Nature, 1998; 392:25–30.

Kline

, Duewer

, Redman

et al. Polymerase chain reaction amplification of DNA from aged blood stains: quantitative evaluation of the “suitability for purpose” of four filter papers as archival media. Anal Chem, 2002; 74:1863–1869.

Galaal

, Meirovitz

, Hussain

et al. The feasibility of storing ovarian tumor cells on databasing paper: establishing a library of ovarian cancer DNA. Int J Gynecol Cancer, 2007; 17:94–100.

Bellete

, Flori

, Hafid

et al.

Influence of the quantity of nonspecific DNA and repeated freezing and thawing of samples on the quantification of DNA by the Light Cycler^®

J Microbiol Methods, 2003; 55:213–219.

Coil

, Strickler

, Rai

et al. Jaagsiekte sheep retrovirus Env protein stabilizes retrovirus vectors against inactivation by lung surfactant, centrifugation, and freeze-thaw cycling. J Virol, 2001; 75:8864–8867.

Fipps

, Damato

, Brandt

et al. Effects of multiple freeze thaws and various temperatures on the reactivity of human immunodeficiency virus antibody using three detection assays. J Virol Methods, 1988; 20:127–132.

Grecz

, Hammer

, Robnett

et al. Freeze-thaw injury: evidence for double strand breaks in Escherichia coli DNA. Biochem Biophys Res Commun, 1980; 93:1110–1113.

10.

Alur

, Grecz

. Mechanism of injury of Escherichia coli by freezing and thawing. Biochem Biophys Res Commun, 1975; 62:308–312.

11.

Lyscov

, Moshkovsky

. DNA cryolysis. Biochim Biophys Acta, 1969; 190:101–110.

12.

Thomson

, Fleming

, Barone

et al. The effect of repeated freezing and thawing on human sperm DNA fragmentation. Fertil Steril, 2010; 93:1147–1156.

13.

Sanlidag

, Akcali

, Ozbakkaloglu

. Serum hepatitis B DNA: stability in relation to multiple freeze-thaw procedures. J Virol Methods, 2005; 123:49–52.

14.

Krajden

, Minor

, Zhao

et al. Effect of multiple freeze-thaw cycles on hepatitis B virus DNA and hepatitis C virus RNA quantification as measured with branched-DNA technology. J Clin Microbiol, 1999; 37:1683–1686.

15.

Kim

. Effect of fixation time and freeze-thaw cycles on the molecular analysis of viral DNA. J Vet Sci, 2003; 4:203–204.

16.

Peters

, Hayes

, Chatterjee

et al. Circulating vitamin D metabolites, polymorphism in vitamin D receptor, and colorectal adenoma risk. Cancer Epidemiol Biomarkers Prev, 2004; 13:546–552.

17.

Fletcher

. The Chemical Physics of Ice. London: Cambridge University Press, 1970; 73–103ISBN 0521075971.

18.

Petrenko

, Robert

. Physics of Ice. Oxford: Oxford University Press, 1999; 10–18ISBN 0198518943.

19.

Bowman

, Davidson

. Hydrodynamic shear breakage of DNA. Biopolymers, 1972; 11:2601–2624.

20.

Davison

. The effect of hydrodynamic shear on the deoxyribonucleic acid from T(2) and T(4) bacteriophages. Proc Natl Acad Sci U S A, 1959; 45:1560–1568.

21.

Han

, Craighead

. Separation of long DNA molecules in a microfabricated entropic trap array. Science, 2000; 288:1026–1029.

22.

Kuma

, Yan

, Xu

et al. Glass transition in biomolecules and the liquid-liquid critical point of water. Phys Rev Lett, 2006; 97:1–4.

23.

Norberg

, Nilsson

. Glass transition in DNA from molecular dynamics simulations. Proc Natl Acad Sci U S A, 1996; 93:10173–10176.