BIOPRESERVATION AND BIOBANKING

Volume 13, Number 5, 2015

ª Mary Ann Liebert, Inc.
DOI: 10.1089/bio.2015.0029

The Impact of Different Preservation Conditions

and Freezing-Thawing Cycles on Quality of RNA,
DNA, and Proteins in Cancer Tissue

Yaogeng Wang,1,2 Hong Zheng,1 Jie Chen,1 Xiaorong Zhong,1

Yu Wang,1 Zhu Wang,1 and Yanping Wang1,2

Background: High-quality cancer tissues are essential for future research, especially molecular research. For the
sake of better quality of tissues, some storage methods are chosen according to lab conditions. But the impact of
different storing conditions on the quality of RNA, DNA (especially the degree of DNA methylation), and protein
of tissues that have undergone a thawing process, is not clear.
Methods: We analyzed the influence of different storage conditions including in RNALater solution, normal
saline, Opti-mum Cutting Temperature compound (OCT), and snap frozen with no protective reagent (as
control) in paired tissue samples on the quality of RNA (RNA Integrity Number value and mRNA expression),
DNA quality (DNA amplification and DNA methylation degree of gene RASSF1a), and protein quality.
Further, we analyzed the RNA quality of tissues that underwent three freeze–thaw cycles.
Results: The RNALater-treated group retained good RNA quality as expected on three repeated freeze-thaw
cycles (RIN > 8), but the snap-frozen tissues showed relatively poor results after one freeze-thaw cycle
(RIN < 7) and three times repeated freeze-thaw cycles (RIN < 6). RNA from saline- and OCT-treated groups
also yielded good results when we repeated freezing and thawing one time (RIN > 7) and two times (RIN > 6).
The impact of different storing conditions on DNA amplification is small. However, DNA methylation and
protein quality are different with different storing conditions. OCT seems to be more secure and stable com-
pared with other two experimental groups, and show a similar trend with control group.
Conclusions: In consideration of budget and efficiency, we suggest OCT as the best storing method that not
only preserves RNA quality during the freezing-thawing process well, but also ensures more secure and stable
DNA and protein.

Introduction Nevertheless, obtaining and storing these tissues is not as

straightforward as many people think, especially the proper

A s a link between basic medical research and

clinical research, the tissue biobank plays an important
role in original research of life sciences and is considered to
be a precious resource in translational medicine research.1
Researchers hope to obtain biological information to iden-
tify the biological characteristics of cancer or other diseases
more efficiently and conveniently. Clinicians wish to obtain
the most timely research results and clinical information to
formulate therapeutic strategies. Many of the hopes for
achieving the clinical benefit of basic medicine hinge on the
ability to develop an efficient tissue biobank from clinic to
laboratory.2,3 Therefore, the availability and quality assur-
ance of cancer tissues are critical in making progress in
developing a biobank, especially in terms of improving ef-
ficiency, utilization, and reducing the cost.4
1
Laboratory of Molecular Diagnosis of Cancer, West China Hospital, Sichuan University, Chengdu, China.
2
State Key Laboratory of Biotherapy and Cancer Center, West China Hospital, Sichuan University, and Collaborative Innovation Center
for Biotherapy, Chengdu, China.
and efficient storage of tissues samples. Furthermore, the
system of collecting and storing tissue samples has not yet
been standardized, and many systems have been used, de-
pending on the conditions at the institution or laboratory.5
Of late, the method of collecting and storing fresh frozen
without reagent in liquid nitrogen or freezer (at -80C) has
been widely used to ensure the quality of tissue samples.1,6,7
Tissue samples must be stored properly to ensure stable and
high-integrity samples for long-term preservation. However,
severe conditions such as the thawing process inevitably
compromise the quality of these valuable tissues. In such
events, researchers should be prepared to analyze the damage
and evaluate the potential usefulness of the tissue samples
rather than blindly discarding or keeping them.8 Considering
that the samples are very precious and that storage space and

335
336
cost need to be controlled, it is necessary to improve the uti-
lization rate of samples and know how much value is left in the
samples. Some of the tissues that undergo repeated freeze-thaw
cycles have potential value, but the effect on tissue quality of
the number of freeze-thaw cycles at room temperature with
different storage conditions is not clear. One goal of our study
was to evaluate the tissues that underwent freeze-thaw cycles
using four commonly used methods in the laboratory, so that
the utilization rate of samples can be improved.
The quality of tissue samples is vital to a tissue biobank. In
cancer research, high-quality tissues are essential for future
research, especially molecular research.2,9–11 The quality of
tissues normally reflects the quality of biological macromole-
cules, especially the quality of RNA, DNA, and protein. With
regard to RNA, a number of studies have evaluated the influ-
ence of different collection methods and storage condi-
tions,2,3,12–18 for example, RNALater Solution, which has
already been described, can be used to store tissues and cells in
a way that the RNA quality and quantity are preserved.
A few studies have addressed the impact of tissue han-
dling on the quality of DNA.19–21 Epigenetic modification
caused by DNA methylation plays an important role in
generating the cancerous phenotype.22–25 However, the in-
fluence of different collection methods and storage condi-
tions on the detection or analysis of DNA methylation has
hardly been addressed in the literature. Likewise, the eval-
uation of the quality of human tissue protein is scarce. Only
a few studies have focused on the impact of RNALater on
protein quality and were restricted to animal tissues. The
results illustrated that different storage methods contribute
to different results in different animal tissue types.26,27
In the current study, we collected and subdivided breast
tissue samples and stored split aliquots in pairs using four
methods including RNALater solutions, normal saline, Opti-
mum Cutting Temperature compound (OCT), and snap
freezing without any reagent (as control). We analyzed the
influence of these four storage conditions in paired tissue
samples on the quality of RNA, DNA, and proteins.
Materials and Methods
Sample collection and storage
Breast cancer tissue samples were collected from 10 pa-
tients (numbered from A to J) who underwent surgery for
breast cancer at West China Hospital of Sichuan University,
Chengdu, Sichuan Province, from May 2013 to Jan 2014.
Table 1. Characteristics of Clinical Pathologic and Physiology of Patients
Patient ID Patient Number Histology at diagnosis
SICK0008147 A IDC,L
SICK0008186 B IDC,R
SICK0008209 C IDC,R
SICK0008431 D IDC,R
SICK0008566 E IDC,R
SICK0008107 F IDC,L
SICK0008594 G IDC,R
SICK0008725 H IDC,R
SICK0008743 I IDC,L
SICK0008742 J IDC,L
IDC, Invasive ductal carcinoma; L, Left breast; N, Negative; P, Positive; R, Right breast.
WANG ET AL.
Table 1 shows the characteristics of the clinical pathology
and physiology of these 10 patients. The patients were not
treated with chemotherapy or radiotherapy before their
surgery. The ethics committee of West China Hospital of
Sichuan University gave approval for the banking and use
of the procured specimens, and informed consent for the use
of tissue samples was obtained from all patients.
At the time of resection, the specimens were obtained by
the attending surgeon and transported at 4C immediately to
the laboratory. Tissue samples were cut into split aliquots of
0.2 · 0.2 · 0.2 cm3 at 4C; the whole process was completed
within an hour. Tissues that exceeded the time limit were
discarded by rule, and one of the split aliquots was used for
histological identification with paraffin embedding.
The split aliquots were treated using four methods: Tissues
with RNALater Solution were stored overnight at 4C and
transferred to liquid nitrogen for 1 day, and then all tissues
were transferred to -80C. Tissues with normal saline, Opti-
mum Cutting Temperature compound (OCT), and without
any reagent were stored in liquid nitrogen for 1 day and
transferred to -80C. The collected tissue included 88 sam-
ples for RNA quality analysis from 6 patients, which were
numbered A to F, and 60 samples for DNA and protein quality
analysis from 5 patients, which we numbered F to J; all the
tissues grouped in pairs.
There were four groups among the 88 samples for RNA
quality analysis, including samples that did not undergo a
thawing process, samples that underwent a thawing process
only once, samples that underwent a thawing process twice,
and samples that underwent a thawing process three times.
Each group contained 22 tissue samples and the thawing pro-
cess was conducted for 1 h at room temperature. Table 2 shows
the experimental design and groups. In this study, we chose
tissues without protective reagents as controls in the quality
and statistical analysis, as this method is considered as a
standard method in recent research studies.
Histology
One of the split aliquots was used for hematoxylin and
eosin (H&E) staining. Histological diagnosis of all samples
was made and confirmed by experienced pathologists. The
inspection of tissue samples was conducted to ensure that
there were no necrotic tissues, and the tumor cell content
was calculated on the basis of the ratio of the tumor area to
the normal areas. All specimens were evaluated to ensure
that tumor cell content was 70%.
ER status PR status HER2 status Tumor size, cm
P P P 3*2.5*2
P P P 2.5*2*1.5
P P P 3*2.5*2
P P P 2*1.8*1
N P P 3*3*2.5
P P P 5.5*5*2.5
P P N 2.8*2.2*1.6
N N P 3*2.5*2
P N N 3.5*3*2.5
P P P 4.5*4*3
RNA, DNA, AND PROTEIN IN FREEZE-THAW CANCER TISSUE 337

5 tissues from patient F

5 tissues from patient F

R-NO-T0 5 tissues from patient A

R-NO-T1 5 tissues from patient A

R-NO-T2 5 tissues from patient A

R-NO-T3 5 tissues from patient A

NO, no protective reagent; NS, normal saline; OCT, Opti-mum Cutting Temperature compound; RL, RNALater solution; T0, no freezing and thawing process; T1, freezing and thawing once;
RNA extraction

Number and Patients

Tissues were put into 1.5 mL Eppendorf tubes and sub-
merged in about 35 mL Trizol reagent (Invitrogen Corporation,

to patient E

to patient E

to patient E

to patient E
Carlsbad, CA). Tissues were cut into pieces and Trizol solution

to patient J

to patient J
was added up to 1 mL. In addition, a shaft rotor hand-held
homogenizer (Omni International, Kennesaw, GA, USA) was
used to homogenize tissues. Finally, RNA was isolated ac-
cording to the standard operating procedure. In brief, the
aqueous phase was resolved by addition of chloroform, and
RNA precipitated from the aqueous phase by addition of iso-
propyl alcohol. Pelleted RNA was washed with 70% ethanol,
NO

D-NO

P-NO
dried, and suspended in RNAse-free water treated with DEPC.
R-OCT-T0 5 tissues from patient A

R-OCT-T1 5 tissues from patient A

R-OCT-T2 5 tissues from patient A

R-OCT-T3 5 tissues from patient A

5 tissues from patient F

5 tissues from patient F

RNA integrity
Number and Patients

The RNA quality of the tissue samples was assessed by

RNA Integrity Number (RIN). By means of this method,
to patient E

to patient E

to patient E

to patient E

to patient J

to patient J
Table 2. Grouping of Cancer Tissues for RNA, DNA, and Protein

RNA integrity was determined using the entire electropho-

retic trace of the RNA sample (total of 10 features), in-
cluding the presence or absence of degradation products.
RIN index ranges from 1 to 10, with 1 being the most de-
graded and 10 being the most intact RNA. The Agilent 2100
Bioanalyzer and RNA 6000 Nano Labchip kit (Agilent
Biotechnologies, Santa Clara, CA, USA) were utilized to
OCT

D-OCT

evaluate the RNA quality of the tissue samples. The RNA

P-OCT

sample (1 mL) was transferred to the NanoLabchip, together

with 1 mL of RNA 6000 ladder. The analysis was performed
according to the manufacturer’s instructions.
R-NS-T0 6 tissues from patient A

R-NS-T1 6 tissues from patient A

R-NS-T2 6 tissues from patient A

R-NS-T3 6 tissues from patient A

5 tissues from patient F

5 tissues from patient F

Number and Patients

RNA reverse transcription and real-time

quantitative PCR
to patient F

to patient F

to patient F

to patient F

to patient J

to patient J

cDNA was transcribed from 6 mL of total RNA, adding

1 mL oligo (dT) primers (Takara, Japan) and 5 mL RNAse-
free water with DEPC. The mixture was heated to 70C for
5 min and then immediately chilled on ice. To yield a total
volume of 20 mL, 4 mL of 5 · M-MLV Buffer (Promega
Corp., Madison, WI, USA), 2 mL of 10 mM dNTP Mix (Ta-
kara), 1 mL of RNAse Inhibiter (BioBRK, Chengdu, China),
NS

D-NS

P-NS

and 1 mL of M-MLV reverse transcriptase (200 IU, Promega)

were added. The reaction mixture was incubated at 42C for
T2, freezing and thawing twice; T3, freezing and thawing thrice.

1 h and then 95C for 10 min and stored at -20C.

R-RL-T0 6 tissues from patient A

R-RL-T1 6 tissues from patient A

R-RL-T2 6 tissues from patient A

R-RL-T3 6 tissues from patient A

5 tissues from patient F

5 tissues from patient F

Number and Patients

We assessed the expression of five genes in all RNA groups

(Table 3). The PCR reactions were carried out in a final
volume of 10 mL. Briefly, reactions consisted of 1 mL of
to patient F

to patient F

to patient F

to patient F

to patient J

to patient J

cDNA, 3 mL of nuclease-free water, 5 mL of SsoFast Eva-

Green Supermix (Bio-Rad, Hercules, CA) and 0.5 mL of the
primers described in Table 3. The reaction mixture was ini-
tiated with 10 min at 95C, followed by 40 cycles of 94C for
30 s, 47C for 30 s, and 72C for 45 s, and ended with 72C for
7 min. All reactions were performed on the Chromo4 Real-
Time PCR Detector (Bio-Rad) under conditions re-
RL

D-RL

commended by the manufacturer. Expression levels were

P-RL

calculated from the mean Ct values of triplicates for each of

the samples of 5 or 6 independent RNA preparations.
Methods

DNA extraction and real-time quantitative PCR

DNA was extracted using a TIANamp Genomic DNA Kit
Contents

Protein

(Tiangen, Shanghai, China) according to manufacturer’s

RNA

DNA

instructions. DNA was dissolved in the eluent buffer and

stored at -20C.
338 WANG ET AL.

Table 3. Primer Sequences and Concentrations Used for Real-time Quantitative PCR of RNA
Primer Forward Reverse Length
GAPDH(10nM)* CCCATGTTCGTCATGGGTGT TGGTCATGAGTCCTTCCACGATA 174
PCNA (10nM){ GCGCTAGTATTTGAAGCACCAA CGATCTTGGGAGCCAAGTAGTA 469
c-MYC (10nM){ GCCACGTCTCCACACATCAG TGGTGCATTTTCGGTTGTTG 132
MKI67(10nM){ CCACACTGTGTCGTCGTTTG CCGTGCGCTTATCCATTCA 123
ACTB(10nM) GGGACCTGACTGACTACCTC CTTAATGTCACGCACGATTT 92
*Impact of Thawing on RNA Integrity and Gene Expression Analysis in Fresh Frozen Tissue. Diagn Mol Pathol, 2009;18:44–52.
{
Primer sequences were obtained from public RTPrimerDB database (http://medgen.UGent.be/rtprimerdb/).
ACTB, nonmuscle cytoskeletal actins beta; c-MYC, myelocytomatosis cellular oncogene; GAPDH, glyceraldehyde-3-phosphate
dehydrogenase; MKI67, marker of proliferation Ki-67; PCNA, proliferating cell nuclear antigen.

To investigate how different storage conditions may induce
alterations in DNA amplification, the amplification of three
different genes was analyzed in all DNA groups using quan-
titative real-time PCR (Table 4). The PCR reactions were
carried out using a final volume of 10 mL. Briefly, reactions
consisted of 1 mL of DNA, 3 mL of nuclease-free water, 5 mL of
SsoFast EvaGreen Supermix (Bio-Rad), and 0.5 mL of the
primers described in Table 4. The reaction mixture was initi-
ated with 10 min at 95C, followed by 40 cycles of 94C for
30 s, 47C for 30 s, and 72C for 45 s, and ended with 72C for
7 min. All reactions were performed on the Chromo4 Real-
Time PCR Detector (Bio-Rad) under conditions recommended
by the manufacturer. DNA amplification levels were calcu-
lated from the mean Ct values of triplicates for each of the
samples of three independent genes.
DNA methylation analysis
RASSF1 is one of the tumor suppressor genes localized at
chromosome 3p21.3. Seven different transcript variants of this
gene encoding distinct isoforms have been reported (RASS-
F1A-G), and RASSF1a is the main isoform.28 The CpG Island
in RASSF1a is located at the 5¢-end regulatory regions, ex-
tending from the promoter to the first exon region.29 In addi-
tion, this CpG Island, crossing the promoter and the first exon
regions of RASSF1a, is hypermethylated in primary breast
tumors and in many other types of tumors.30–33
In the current study, we assessed the methylation level of 8
CpGs in the first exon of RASSF1A (from CpG6 to CpG14)
using the Sequenced method by detecting pyrophosphate
release in all DNA groups, including tissues with RNALater
(D-RL), normal saline (D-NS), OCT (D-OCT), and no pro-
tective reagent (D-NO), which are all paired tissues. Analysis
of DNA methylation patterns by pyrosequencing is a modi-
fication of the combined bisulfite restriction analysis, which
combines a simple reaction program with repeated and ac-
curate measurement of the degree of methylation at several
specific CpGs.34,35
In brief, the DNA solution was modified by bisulfite using
PCR with a total volume of 140 mL and the modified DNA
after PCR was purified after introducing EpiTect Bisulfite
Handbook-2009 (Qiagen, Dusseldorf, Germany). Methyla-
tion-specific PCR was achieved with a total volume of 50 mL,
including 10 mL of 5X Buffer CG (KAPA, London, UK), 1 mL
of 10 mM dNTP Mix (Takara), 2 mL of modified and purified
DNA template, 34.8 mL of nuclease-free water, 0.2 mL of Taq
DNA Polymerase (KAPA), and 2 mL of the primers described
in Table 5. PCR products were addressed with streptavidin-
coated Sepharose beads (GE Healthcare, Little Chalfont,
UK), and the degree of each methylation at each CpG position
in a selected sequence was determined by the ratio of T and C
using Pyromark Q96 MDx (Qiagen).
Protein extraction
Tissues were put into 1.5 mL EP tubes and submerged in
40 mL RIPA Lysis Buffer (Beyotime, Shanghai, China). The
tissues were then cut into pieces and 53 mL RIPA Lysis Buffer
and 7 mL PMSF were added up to 100 mL. Afterwards,
granular tissues were broken up ultrasonically three times
using 225 W (working 5 s and pausing 10 s for one cycle).
Then, the tubes were put at 4C for 50 min and centrifuged at
4C for 10 min, 13,800 g. Supernatant was transferred to a
new EP tube and stored at -20C.
Western blot analysis
Western blot analysis was performed using standard op-
erating procedures. Equal amounts of protein were separated
by electrophoresis on 8% SDS-PAGE and transferred elec-
trophoretically onto PVDF membrane. After blocking, the
membranes were incubated overnight with anti-GAPDH
(1:1000 dilution) (Zen Bioscience, Chengdu, China), and
anti-VEGFC (1:400 dilution) (Boster, Wuhan, China) at 4C.
Intensity values were normalized relative to control values.
The blots were scanned and exposed using Universal Hood II

Table 4. Primer Sequences and Concentrations Used for Real-time Quantitative PCR of DNA
Primer Forward Reverse Length
GAPDH (10 nM)* CCCATGTTCGTCATGGGTGT TGGTCATGAGTCCTTCCACGATA 174
P16 (10 nM) CTTGTGTGGGGGTCTGCTT TGGTTACTGCCTCTGGTGC 152
RASSF1a (10 nM) GCCAACTCCTCACACACC TGCCCACATTCACACAGA 179
*Impact of thawing on RNA integrity and gene expression analysis in fresh frozen tissue. Diagn Mol Pathol, 2009;18:44–52.
GAPDH, Glyceraldehyde-3-phosphate dehydrogenase; P16, Cyclin-dependent kinase inhibitor 2A; RASSF1a, Ras association domain-
containing protein 1.
RNA, DNA, AND PROTEIN IN FREEZE-THAW CANCER TISSUE 339

Table 5. Primer Sequences and Concentrations Used for Detection of the Degree of Methylation of DNA
Primer Primer sequences Target sequence
RASSF1a- Forward GTTTAGTTTGGATTTTGG GCCCAGTCTGGATCCTGGGGGAGGCGCTGAAGTCG
GGGAG GGGCCCGCCCTGTGGCCCCGCCCGGCCCGCGCTTG
CTAGCGCCCAAAGCCAGCGAAGCACGGGCCCAAC
CGGGCCATGTCGGGGGAGCCTGAGCTCATTGAG
CTGCGGG (144bp amplified sequence contains 8 CpGs)
RASSF1a- Reverse* CCCRCAACTCAATAAACTC
AAACTC
RASSF1a-Sequencing AAGTYGGGGTTYGTTTTGTG GTTTC/TGTTC/TGGTTC/TGC/TGTTTGTTAGC/TGTT
TAAAGTTAGC/TGAAGTAC/TGGGTTTAATC/TGG
*biotin labeled in the 5¢ end.
C/TG, CpG island; RASSF1a, Ras association domain-containing protein 1.

(Bio-Rad), and the band intensity was transformed into In-
tegrated Optical Density (IOD) and analyzed using the IPP
software.
Statistical analysis
Statistical analysis was performed using SPSS Statistics
20.0 (IBM, Chicago, IL, USA). The differences of RNA in-
tegrity, DNA Ct values, DNA methylation, and protein quality
among groups were validated by the paired-sample t-test. P-
Value <0.05 was considered to be statistically significant. The
Spearman correlation (r) analysis was performed to evaluate
the correlation of RIN and RNA Ct values with p < 0.01 and
95% confidence interval (two-tailed test).
Results
Histology
One of the split aliquots of each patient’s tissue, was
pathologically diagnosed as breast cancer tissue, and the
tumor cell content was >70% (Fig. 1).
RNA integrity with four methods
A total of 88 breast tissue specimens with four different
storage conditions that underwent different thawing process
times were analyzed for RNA integrity. Figure 2 shows the
line chart. As a standard for RNA integrity, the RIN was
calculated for all specimens. In general, the cut-off RIN
values for samples used in gene expression arrays appeared
between 6 and 7,36,37 and one study used a RIN value of 6.5 as
a cut-off for RNA quality.38 In this study, a RIN value of 7
was chosen as the cut-off. RIN values >7 seemed to work well
in most conventional experiments. RIN values <7 required
extra validation steps before we could perform experiments.
RIN values >5 but <7 might not work for a microarray ex-
periment, but might work for an RT-PCR experiment. RIN
values <5 might not work for most of our experiments.2,39
Figure 3 shows some representative data graphs.
In our study, tissues using four storage methods with no
freeze-thaw cycles showed good results as expected. All
RINs were >9 and there was no significant difference be-
tween the group with no protective reagent (R-NO-T0) and

FIG. 1. Paraffin-embedded
section of cancer tissues sam-
ples. (A) Normal breast tis-
sues. (B) Sample shows 80%
of sample is necrotic. (C) Less
than 20% of sample is tumor.
(D) More than 70% of sample
is tumor (X10).
340 WANG ET AL.

FIG. 2. The RIN of all tissue-samples and RIN of 7 is chosen as the value cutoff. A–F, patient number; NO, no protective
reagent; NS, normal saline; OCT, Optimum Cutting Temperature compound; RL, RNALater solution; T0, no freezing and
thawing process; T1, freezing and thawing once; T2, freezing and thawing twice; T3, freezing and thawing three times.

the other three groups (with RNALater, normal saline, and
OCT). But when the tissue samples underwent one freeze-
thaw cycle, differences were visible: tissues with RNALater
(R-RL-T1) yielded good results, showing a mean RIN of
9.4 – 0.1; tissues with normal saline (R-NS-T1) and OCT
(R-OCT-T1) yielded normal results, showing a mean RIN of
7.3 – 0.3 and 7.7 – 0.3, which were above the cut-off of 7.
However, the Mean RIN of tissue samples without any
protective reagent (R-NO-T1) was only 6.7 – 0.2, which was
below the cut-off value 7, but >5.
The difference was more obvious when tissue samples
underwent two freeze-thaw cycles: The Mean RIN of tissue-
samples with RNALater (R-RL-T2) was 9.1 – 0.1, which
was >9; tissue samples with normal saline (R-NS-T2) and
OCT(R-OCT-T2) yielded relatively poor results, showing a
mean RIN of 6.6 – 0.2 and 6.1 – 0.3, which were below the

FIG. 3. The representative RIN data of tissue-samples with various methods underwent different times of freeze-thaw
cycles. The y-axis represents the fluorescence intensity, while the x-axis shows the time. RIN index ranges from 1 to 10,
with 1 being the most degraded and 10 being the most intact RNA. NO, no protective reagent; NS, normal saline; OCT,
Optimum Cutting Temperature compound; RL, RNALater solution; T0, no freezing and thawing process; T1, freezing and
thawing once; T2, freezing and thawing twice; T3, freezing and thawing three times.
RNA, DNA, AND PROTEIN IN FREEZE-THAW CANCER TISSUE 341

cut-off value of 7, but >6. However, tissue samples without
any protective reagent (R-NO-T2) showed the poorest re-
sults, with a mean RIN of 5.4 – 0.4.
We then processed tissue-samples that underwent three
freeze-thaw cycles and the RIN value was detected. The
mean RIN of tissue samples with RNALater (R-RL-T3) was
8.3 – 0.3, which was significantly reduced when compared
with group R-RL-T2 but was also a good result when
compared the with other three groups. Tissue samples with
normal saline (R-NS-T3) and OCT (R-OCT-T3) yielded
poor results, with a mean RIN of 5.1 – 0.6 and 5.5 – 0.3.
Tissue samples without any protective reagent (R-NO-T3)
obtained a worse result, with a mean RIN of 4.0 – 0.6.
Figure 4 shows the analysis of RIN values for each storage
group, classified by different storage conditions (RL vs. NS
vs. OCT vs. NO) and freeze-thawed times (T0 vs. T1 vs. T2
vs. T3).
Ct values analysis and correlation between
Ct values and RNA integrity
The recovery of specific genes at the RNA level was
analyzed by qRT-PCR. We studied three specific genes
described in the literature (Table 3) that are involved in
human breast cancer and two endogenous control genes.
Supplementary Figure 1 (Supplementary material is avail-
able online at: www.liebertpub.com/bio) shows the Ct value
of triplicates for each of the tissue samples and the mean CT
of tissues in five genes classified by storage conditions and
the number of freeze-thaw cycles. Results show that Ct

FIG. 4. (A) Comparative analysis between each age group that is classified by different storing methods. Significant
differences ( p < 0.05) were found between each two adjacent groups of freeze-thaw times except the group R-RL; (B)
Comparative analysis between each storage group that is classified by different freeze-thaw times. There are significant
differences ( p < 0.05) between R-NO and the other three groups in T1, T2, and T3. NO, no protective reagent; NS, normal
saline; OCT, Optimum Cutting Temperature compound; RL, RNALater solution; T0, no freezing and thawing process; T1,
freezing and thawing once; T2, freezing and thawing twice; T3, freezing and thawing three times.
342
FIG. 5. The correlation between RIN Integrity and Ct values in five genes. The Spearman correlation (r) was respectively
-0.817, -0.762, -0.824, -0.781, and -0.807 with p < 0.01 and 95% confidence interval (two-tailed test).
values had an opposite trend compared to the RIN (the
higher RIN, the lower Ct values) when groups were classi-
fied by methods, but statistical differences by method were
not found when there were statistical differences in RIN
analysis (Ct statistical analysis not shown).
When groups were classified by thawing times, we found
that Ct values were relatively similar where RIN was >5 but
<8. Moreover, the Ct value of the group of RNAlater and
OCT were lower. To further evaluate the correlation between
RIN Integrity and the quality of reverse transcription, we
made a correlation analysis between RIN Integrity and Ct
values. The correlation between RIN Integrity and Ct values
in five genes was calculated using Spearman’s correlation (r)
analysis with p < 0.01 and 95% confidence interval (two-
tailed test) (Fig. 5). All five genes showed a significant
correlation ( p < 0.01), and five graphs show the correlation
between RIN Integrity and Ct values in GAPDH, PCNA,
MYC, MKI67, and ACTB, respectively, and Spearman cor-
relation (r) was -0.817, -0.762, -0.824, -0.781, and -0.807,
respectively.
DNA amplification and DNA methylation analysis
At the DNA level, we studied the amplification of three
specific genes described in the literature (Table 4) that are
involved in human breast cancer and the one endogenous
control gene. Supplementary Figure 2 shows the results of
mean Ct values in all three genes, which showed no sig-
nificant differences between the groups ( p > 0.05).
The methylation status of CpG islands of the RASSF1a
was assessed in a series of 20 frozen tissues with different
storage conditions by the pyrosequencing technique, and the
data graph of patient H was chosen as representative (Fig.
6A). We chose the methylation level of group D-NO as the
standard in consideration of other studies concerning DNA
methylation.40–42
WANG ET AL.
Table 6 illustrates the situation of the methylation
level and average methylation of five tissue samples. The
results showed that in all five tissue samples, group D-OCT
showed a similar trend as group D-NO, and the methyla-
tion of these two groups were always in a relatively stable
level compared with other groups. However, the methyla-
tion levels of group D-RL and D-NS varied in five tissue
samples. For example, the methylation level of group D-
RL was the lowest in the second and third tissue samples,
but the highest in the first and fourth tissue samples, and
the second highest only after group D-NS in the fifth tis-
sue samples. Figure 6B illustrates the results of statistical
analysis, which showed significant differences ( p < 0.05)
among the other three groups compared with group D-NO,
except the group D-OCT and D-NO in the first and third
tissue samples.
Protein quality
We evaluated the protein quality by intensity values in
protein VEGF-C and the internal reference protein GAPDH
and transformed band intensity values into IOD (Fig. 7A). Our
results showed that the group with no protective reagent (P-
NO) yielded the highest IOD values, that the mean IOD value
was 966.7 – 151.5 in internal reference protein GAPDH and
320.9 – 73.2 in protein VEGF-C. The group with RNALater
(P-RL) yielded a poor result showing intensity values which
were the lowest in two proteins, the mean IOD value of
GAPDH and VEGF-C was only 46.1 – 21.9 and 12.0 – 2.6,
respectively. The group with normal saline (P-NS) yielded a
result that the mean IOD value was 263.1 – 49.3 and
94.5 – 28.9 in protein GAPDH and VEGF-C, respectively,
while the mean IOD value of the group with OCT (P-OCT) in
GAPDH and VEGF-C was 554.9 – 156.1 and 315.3 – 48.4,
respectively. Figure 7B shows the IOD value and the statistical
analysis of each group.
RNA, DNA, AND PROTEIN IN FREEZE-THAW CANCER TISSUE 343

FIG. 6. (A) Analysis of DNA methylation by pyrosequencing. Representative results of Patient-H showing analysis of 8
CpGs in the first exon of RASSF1a (from CpG6 to CpG14). The y-axis represents the signal intensity of luminescence in
relative luminescence units (RLU) emitted following nucleotide base incorporation into the sequence, while the x-axis
shows the nucleotide dispensation order, indicated by the letters below each graph: E, enzyme mix; S, substrate; A, G, C, T,
nucleotides. The sequence in the upper left of the boxes is the region subjected to pyrosequencing analysis, where the nine
letters Y indicate the respective cytosine in the 9 CpG sites (from CpG6 to CpG14) whose methylation status is shown here.
Values in light blue boxes are the percentages of methylation of each CpG. Percentages of methylation represent the ratio
between signal intensities of C and T in each C of a CpG site. Dispensations corresponding to the potentially methylated
cytosine (C or T after bisulfite treatment) are highlighted in gray. (B) The statistical difference analysis of average
methylation level in 8 CpGs islands among five tissue samples. Significant differences ( p < 0.05) are found between group
D-NO and other three groups except the tissue samples from patient F and H. In the tissue sample F and H, there are no
significant differences ( p > 0.05) between group D-NO and D-OCT. NO, no protective reagent; NS, normal saline; OCT,
Opti-mum Cutting Temperature compound; RL, RNALater solution.

Table 6. Methylation Level and Average Methylation of Five Tissue Samples

8 CpGs in CpG island A of RASSF1a
Tissues CpG 6 CpG 7 CpG 8 CpG 9 CpG 10 CpG 11 CpG 12 CpG 13 Average methylation (%)
D-RL-F 6 4 9 4 4 6 4 4 5.13
D-NS-F 3 4 7 2 3 6 0 4 4.14
D-OCT-F 4 3 2 1 2 3 3 4 2.75
D-NO-F 4 2 4 1 2 4 3 3 2.88
D-RL-G 29 28 20 26 23 22 29 30 25.88
D-NS-G 47 49 46 46 42 42 44 45 45.13
D-OCT-G 37 37 38 31 31 32 33 35 34.25
D-NO-G 42 44 40 41 39 40 45 42 41.63
D-RL-H 25 26 24 24 21 22 26 25 24.13
D-NS-H 63 69 67 64 58 57 66 y 63.43
D-OCT-H 43 45 42 40 38 39 44 43 41.75
D-NO-H 44 45 41 39 35 36 44 40 40.50
D-RL-I 57 58 56 52 48 52 58 56 54.63
D-NS-I 38 37 41 32 32 33 36 37 35.75
D-OCT-I 33 32 38 28 27 28 31 34 31.38
D-NO-I 31 23 28 25 25 24 29 29 26.75
D-RL-J 50 55 51 51 46 47 52 50 50.25
D-NS-J 53 56 46 50 47 48 54 57 51.38
D-OCT-J 44 46 44 40 38 38 44 y 42.00
D-NO-J 40 34 40 33 32 32 39 38 36.00
344 WANG ET AL.

FIG. 7. (A) Western Blot analysis examined the expression of GAPDH protein and VEGFC protein, and the protein
quality was assessed by intensity values in reference protein GAPDH and protein VEGFC. (B) The mean IOD values in
protein GAPDH and VEGFC of tissues with different storing methods. Significant differences ( p < 0.05) are found between
group P-NO and any other three groups in GAPDH, while in protein VEGFC, there are significant differences ( p < 0.05)
between group P-NO and any other group except for the group P-OCT. NO, no protective reagent; NS, normal saline; OCT,
Opti-mum Cutting Temperature compound; RL, RNALater solution.

Discussion RNA. RIN, which contains 10 features of RNA electrophe-

In recent years, biomedical research has played an im-
portant role in fighting cancer and other serious diseases.
Biobanks have become a vital resource for providing well-
stored, high-quality samples and systematic information for
further research. Moreover, biobanks are also a link between
basic medical research and clinical research.1 Tissue sample
quality is the key factor in tissue biobank construction and
development. In cancer research, the availability and quality
assurance of cancer tissues are critical for making progress
in future research, especially molecular research for cancer
researchers and clinicians.9–11
As we know, the quality the tissue samples should be better
maintained to avoid damage during storage. However, severe
conditions or necessary operations compromise the quality of
these precious tissues inevitably; for example, power outages
or thawing process during repeated use. If such events hap-
pen, we need to analyze the damage and evaluate the potential
usefulness of the tissue samples. Considering that the samples
are very precious, we cannot discard them without using
them. On the other hand, in view of storage space and cost,
they should not be maintained if they have no value for future
use. Some repeated freezing and thawing tissues also have
potential value that they maintain good quality and integrity.
It is necessary to improve the utilization rate of samples and
know how much usage value these tissue samples have.
RNA is very easily degraded, and RNA quality has been
regarded as the criterion to measure the quality of tissues,
while RNA integrity effectively represents the quality of
rograms, has proved to be better indicator of quality than the
ratio of 18S rRNA to28S rRNA.17,39,43 It has also been shown
that tissues with RIN >7 work well in most conventional
experiments, including highly demanding gene array assays,
while those with RIN between 5.0 and 6.9 could be used for
RT-PCR experiments, and those with RIN between 1 and 4.9
may not work for most experiments.38,39
There are many methods of storing tissues. But each of
those methods has disadvantages, for example, flash freez-
ing tissues in liquid nitrogen might be expensive and labo-
rious.4 To our knowledge, the standard of storage of tissues
has been in liquid nitrogen or freeze stored at -80C.1,6,7 In
this study, we followed this method of storing in liquid ni-
trogen 1 day for simulating the transportation process and
transferring tissue samples to -80C freezer for long-term
storage using four methods (for experimental need, some
tissues were treated with protective reagents). We detected
the RNA quality of breast-cancer tissues under four different
storage conditions, and independently analyzed the RNA
quality of tissues that underwent different times of freeze-
thaw cycles. In addition, we analyzed the influence of dif-
ferent storage conditions on detection of DNA and protein.
In the current study, RNA quality was assessed by RIN, and
a RIN of ‡7 was chosen as the cut-off value. Our research
showed that the snap-frozen tissue samples without freeze-
thaw process yielded good results as expected. Although the
mean RIN value of tissues with RNALater (group R-RL-T0)
was the highest in four storage methods, there was no sig-
nificant difference ( p < 0.05) between group R-NO-T0 and
RNA, DNA, AND PROTEIN IN FREEZE-THAW CANCER TISSUE
the other three groups. In consideration of all four methods
yielding high RIN values (Mean RIN >9) and cost of storage,
we did not need to use protective reagent deliberately if we
wanted short-term preservation or used them just once.
Tissue samples using the four methods underwent one
cycle of freeze-thaw, which yielded different results. Tissues
of group R-RL-T1 also yielded good results with a mean RIN
of 9.4, tissues of group R-NS-T1and R-OCT-T1 yielded
normal results with mean RINs of 7.3 and 7.7, tissues without
any reagent (group R-NO-T1) yielded a poor result with mean
RIN being 6.7, which is below the cut-off value of 7. Simi-
larly, when tissues underwent the freeze-thaw process twice,
the mean RIN of group R-RL-T2, R-NS-T2, and R-OCT-T2
was 9.1, 6.6, and 6.1, respectively. The group of R-NO-T2
yielded the worst result with the mean RIN of 5.4.
Finally, all tissues underwent freeze-thaw one more time
so that the mean RIN of group R-RL-T3 was 8.3, which was
significantly reduced compared with group R-RL-T1 and R-
RL-T2, but was also a good result above 8, and tissues of
group R-NS-T3, R-OCT-T3, and R-NO-T3 yielded bad re-
sults, with mean RIN of 5.1, 5.5, and 4.0, respectively. The
results suggested that tissues without any protective reagent
would be compromised severely by a freeze-thaw process or
other conditions. We found that, if we could use protective
reagent during storage, such as normal saline or OCT, we
could reduce the tissue damage to some extent and improve
the utilization rate of tissue samples, still retaining RIN >5
even when tissues underwent the freeze-thaw process three
times. On the other hand, the results also reminded us that
tissues might be severely damaged when they underwent
freeze-thaw without any protective reagent. We suggest
caution in using them next experiments.
In addition, we analyzed the mRNA expression level of
specific genes by qRT-PCR and evaluated the correlation
between RIN integrity and the quality of reverse transcription.
The results showed that the difference between Ct values was
not as significant as RIN, especially when RIN was in the
range from 5 to 8. However, the Ct values had an opposite
trend with RIN. The correlation analysis showed that RIN did
have linear correlation with Ct values, and they showed
negative correlation in all five genes. The correlation value (r)
was -0.817, -0.762, -0.824, -0.781, and -0.807 in genes
GAPDH, PCNA, MYC, MKI67, and ACTB, respectively.
The outcomes suggested that higher the RIN values, the lower
are the Ct values. However, further studies are warranted to
identify the exact trend. In addition, once this trend is con-
firmed, we could evaluate the RNA quality with qRT-PCR if
we cannot detect the RNA integrity in some special condi-
tions, for instance, in some laboratories without equipment
for detecting RNA integrity.
We also detected the quality of DNA of tissues using four
methods, including DNA amplification by quantitative real-
time PCR and the degree of DNA methylation of 8 CpGs within
CpG Island A of RASSF1a. To our knowledge, this is the first
study that evaluated the impact on DNA methylation using
different storage conditions. At the DNA amplification level,
the analysis shows no significant differences between group D-
NO and three other groups in all genes, which indicates that
there is no influence of the four different storage methods on
DNA amplification. As a key epigenetic modification that
could contribute to tumor development, DNA methylation was
not systematically researched in the aspect that DNA methyl-
ation may be influenced by different storage conditions.
345
In addition, to explore the impact of different storage
methods on DNA methylation, we detected the degree of
DNA methylation of 8 CpGs within CpG Island A of
RASSF1a in a series of 20 frozen tissues with four storage
methods using a pyrosequencing technique. As shown in
Table 6, the methylation levels of group D-RL and D-NS are
varied in all five tissue samples. The methylation level of
group D-RL, for instance, is the lowest in the second and third
tissue samples, but the highest in the first and fourth tissue
samples, and the second highest only after group D-NS in the
fifth tissue samples. The methylation of D-NS is also unstable
and irregular in five tissue samples. However, group D-OCT
has a similar trend to group D-NO, and the methylations of
these two groups are always relatively stable compared with
other groups. The results of statistical differences also dem-
onstrate that there are no significant differences between the
group D-OCT and D-NO in the first and the third tissue
samples. Although there were significant differences between
group D-OCT and D-NO in the other three tissue samples, the
methylation level of D-OCT was the most similar to that of D-
NO compared with two other storage methods.
The reasons for this phenomenon are not clear, but we
speculate that the protective reagent could immerse and
permeate tissues and influence the detection results of meth-
ylation to some extent. However, OCT could create a rela-
tively closed space for tissues, and colloidal substances have
difficulty in influencing the internal environment of tissues.
Regarding proteins, only a few studies have focused on the
impact on protein quality using RNALater and those studies
were restricted to animal tissues.26,27 They had varying re-
sults: high-quality protein of arteries and veins from rat tis-
sues were preserved in RNALater; poor results were found in
liver and heart from gulf killifish tissues preserved in RNA-
Later compared with snap frozen tissues.
In this study, we found that the group P-RL yielded a poor
result where the IOD value was the lowest in two proteins, but
on the contrary, the group P-NO yielded the best results in two
proteins. The IOD values of group P-NS were also at a low
level in two proteins. However, group P-OCT yielded a similar
result with group P-NO in protein VEGF-C, but this phenom-
enon was not obvious in GAPDH and we did not identify more
evidence in other studies where protein quality was influenced
using OCT. Our results demonstrated that some protective
reagents might damage the protein of breast cancer tissues.
However, OCT also needs further experiments to verify if it is
suitable to protein preservation of tissue samples.
In summary, this study evaluated the RNA quality of tis-
sues that underwent different freeze-thaw times with different
storage methods, and the effect of tissues with different
storage methods on DNA amplification, DNA methylation
degree, and protein quality. Our results provide information
to researchers regarding RNA quality of tissue samples that
underwent freeze-thaw processes and also provide reference
values on DNA methylation and protein quality with different
storage conditions. Overall, it is difficult to recommend a
preservation method that could satisfy all the research pur-
poses and tissue types and maintain quality under repeated
freeze-thaw cycles.
As a standard of storing tissues, snap-freezing with no
protective reagent also has disadvantages, for example, the
RNA quality of tissues will be compromised by freeze-
thaw cycles. Although RNALater protects RNA quality of
tissues well even in extreme conditions, it may influence the
346
detection of DNA methylation and damages the protein
fraction. Normal saline is common and easily obtainable and
could be an alternative if there is no RNALater, but it has
the same effects as RNALater in terms of preserving DNA
and protein and is even worse than OCT when tissues un-
dergo three freeze-thaw cycles.
Based on this observation, we chose different ways to store
and handle tissue samples according to different research
purposes and tissue types. However, this approach may in-
volve high cost and is extremely complex. Considering budget
and efficiency, we aimed to develop a comprehensive method
to preserve tissues and balance diverse issues. The current
study may provide a possible method. OCT, which could
support organic organization structure, increases the continuity
of tissues and reduces buckling and fracture, and has been
widely used to embed tissue samples prior to freeze sectioning
on a microtome-cryostat. This process is undertaken to mount
slices or sections of a sample onto slides for analysis.
Our outcomes suggest that OCT not only protects RNA
from adverse conditions to some extent, but also ensures the
quality of the DNA and protein compared with other three
storing conditions. Therefore, tissue samples that are pre-
served by OCT not only can ensure the quality requirements of
pathology but also can be effectively and conveniently used
for providing frozen section with multiple slices without
damaging the quality of RNA, DNA (including detection of
methylation analysis), and protein, to some extent. Moreover,
compared with RNALater Solution, OCT is characterized by
low cost ($0.3/mL OCT vs. $6/mL RNALater). The presented
data should be considered as an attempt to further improve
tissue sample collection and storage, and demonstrate the
impact of collection, storage, and processing on epigenetics.
Acknowledgments
The research was performed at Laboratory of Molecular
Diagnosis of Cancer, West China Hospital, Sichuan Uni-
versity, Chengdu, P.R. China. The authors thank the State Key
Laboratory of Oral Diseases for support in analysis using
Agilent 2100 Bioanalyzer.
The abstract of part of this study was accepted by the
ISBER 2014 Annual Meeting & Exhibits entitled ‘‘The In-
fluence of Frozen Thawing on RNA Quality of Cancer-
Tissues with Different Preserving Methods.’’
Author Disclosure Statement
No conflicting financial interests exist.
References
1. Baker M. Biorepositories: Building better biobanks. Nature
2012; 486:141–146.
2. Bao WG, Zhang X, Zhang JG, et al. Biobanking of fresh-
frozen human colon tissues: Impact of tissue ex-vivo is-
chemia times and storage periods on RNA quality. Ann
Surg Oncol 2013;20:1737–1744.
3. Mutter GL, Zahrieh D, Liu C, et al. Comparison of frozen
and RNALater solid tissue storage methods for use in RNA
expression microarrays. BMC Genomics 2004;5:88.
4. Shabihkhani M, Lucey GM, Wei B, et al. The procurement,
storage, and quality assurance of frozen blood and tissue
biospecimens in pathology, biorepository, and biobank
settings. Clin Biochem 2014;47:258–266.
WANG ET AL.
5. Hong SH, Baek HA, Jang KY, et al. Effects of delay in the
snap freezing of colorectal cancer tissues on the quality of
DNA and RNA. J Korean Soc Coloproctol 2010;26:316–323.
6. Mager SR, Oomen MH, Morente MM, et al. Standard op-
erating procedure for the collection of fresh frozen tissue
samples. Eur J Cancer 2007;43:828–834.
7. Morente MM, Mager R, Alonso S, et al. TuBaFrost 2:
Standardising tissue collection and quality control proce-
dures for a European virtual frozen tissue bank network.
Eur J Cancer 2006;42:2684–2691.
8. Peiró-Chova L, Peña-Chilet M, López-Guerrero JA, et al.
High stability of microRNAs in tissue samples of com-
promised quality. Virchows Arch 2013;463:765–774.
9. Watson RW, Kay EW, Smith D. Integrating biobanks: Ad-
dressing the practical and ethical issues to deliver a valuable
tool for cancer research. Nat Rev Cancer 2010;10:646–651.
10. Myles R, Massett HA, Comey G, et al. Stakeholder re-
search on biospecimen needs and reactions to the devel-
opment of a national cancer human biobank by the National
Cancer Institute. JNCI Monographs 2011;2011:16–23.
11. Knox K, Kerr DJ. Establishing a national tissue bank for sur-
gically harvested cancer tissue. Brit J Surg 2004;91:134–136.
12. Hatzis C, Sun H, Yao H, et al. Effects of tissue handling on
RNA integrity and microarray measurements from resected
breast cancers. J Natl Cancer Inst 2011;103:1871–1883.
13. De Cecco L, Musella V, Veneroni S, et al. Impact of bios-
pecimens handling on biomarker research in breast cancer.
BMC Cancer 2009;9:409.
14. Ciotti P, Garuti A, Ballestrero A, et al. Reliability and repro-
ducibility of a RNA preamplification method for low-density
array analysis from formalin-fixed paraffin-embedded breast
cancer samples. Diagn Mol Pathol 2009;18:112–118.
15. Botling J, Edlund K, Segersten U, et al. Impact of thawing
on RNA integrity and gene expression analysis in fresh
frozen tissue. Diagn Mol Pathol 2009;18:44–52.
16. Ribeiro-Silva A, Zhang H, Jeffrey SS. RNA extraction
from ten year old formalin-fixed paraffin-embedded breast
cancer samples: A comparison of column purification and
magnetic bead-based technologies. BMC Mol Biol 2007;
8:118.
17. Strand C, Enell J, Hedenfalk I, et al. RNA quality in frozen
breast cancer samples and the influence on gene expression
analysis—A comparison of three evaluation methods using
microcapillary electrophoresis traces. BMC Mol Biol 2007;
8:38.
18. Micke P, Ohshima M, Tahmasebpoor S, et al. Biobanking
of fresh frozen tissue: RNA is stable in nonfixed surgical
specimens. Lab Invest 2006;86:202–211.
19. Jewell SD, Srinivasan M, McCart LM, et al. Analysis of the
molecular quality of human tissues: An experience from the
Cooperative Human Tissue Network. Am J Clin Pathol 2002;
118:733–741.
20. Hubel A, Spindler R, Skubitz APN. Storage of human
biospecimens: Selection of the optimal storage temperature.
Biopreserv Biobank 2014;12:165–175.
21. Buś MM, Allen M. Collecting and preserving biological
samples from challenging environments for DNA analysis.
Biopreserv Biobank 2014;12:17–22.
22. Avraham A, Cho SS, Uhlmann R, et al. Tissue specific
DNA methylation in normal human breast epithelium and
in breast cancer. PLoS One 2014;9:e91805.
23. You JS, Jones PA. Cancer genetics and epigenetics: Two
sides of the same coin? Cancer Cell 2012;22:9–20.
24. Hanahan D, Weinberg RA. Hallmarks of cancer: The next
generation. Cell 2011;144:646–674.
RNA, DNA, AND PROTEIN IN FREEZE-THAW CANCER TISSUE
25. Hansen KD, Timp W, Bravo HC, et al. Increased methyl-
ation variation in epigenetic domains across cancer types.
Nat Genet 2011;43:768–775.
26. Abbaraju NV, Cai Y, Rees BB. Protein recovery and
identification from the gulf killifish, Fundulus grandis:
Comparing snap-frozen and RNAlater(R) preserved tissues.
Proteomics 2011;11:4257–4261.
27. Rodrigo MC, Martin DS, Redetzke RA, et al. A method for
the extraction of high-quality RNA and protein from single
small samples of arteries and veins preserved in RNAlater.
J Pharmacol Toxicol Methods 2002;47:87–92.
28. Agathanggelou A, Cooper WN, Latif F. Role of the Ras-
association domain family 1 tumor suppressor gene in hu-
man cancers. Cancer Res 2005;65:3497–3508.
29. Yan PS, Shi H, Rahmatpanah F, et al. Differential distri-
bution of DNA methylation within the RASSF1A CpG is-
land in breast cancer. Cancer Res 2003;63:6178–6186.
30. Liu L, Yoon JH, Dammann R, et al. Frequent hy-
permethylation of the RASSF1A gene in prostate cancer.
Oncogene 2002;21:6835–6840.
31. Lo KW, Kwong J, Hui AB, et al. High frequency of pro-
moter hypermethylation of RASSF1A in nasopharyngeal
carcinoma. Cancer Res 2001;61:3877–3881.
32. Agathanggelou A, Honorio S, Macartney DP, et al. Me-
thylation associated inactivation of RASSF1A from region
3p21.3 in lung, breast and ovarian tumours. Oncogene
2001;20:1509–1518.
33. Malpeli G, Amato E, Dandrea M, et al. Methylation-asso-
ciated down-regulation of RASSF1A and up-regulation of
RASSF1C in pancreatic endocrine tumors. BMC Cancer
2011;11:351.
34. Colella S, Shen L, Baggerly KA, et al. Sensitive and
quantitative universal Pyrosequencing methylation analysis
of CpG sites. Biotechniques 2003;35:146–150.
35. Tost J, Gut IG. DNA methylation analysis by pyr-
osequencing. Nat Protoc 2007;2:2265–2275.
36. Strand C, Enell J, Hedenfalk I, et al. RNA quality in frozen
breast cancer samples and the influence on gene expression
347
analysis—A comparison of three evaluation methods using
microcapillary electrophoresis traces. BMC Mol Biol 2007;
8:38.
37. Fleige S, Pfaffl MW. RNA integrity and the effect on the
real-time qRT-PCR performance. Mol Aspects Med 2006;27:
126–139.
38. Kap M, Oomen M, Arshad S, et al. Fit for purpose frozen
tissue collections by RNA integrity number-based quality
control assurance at the Erasmus MC Tissue Bank. Bio-
preserv Biobank 2014;12:81–90.
39. Schroeder A, Mueller O, Stocker S, et al. The RIN: An
RNA integrity number for assigning integrity values to
RNA measurements. BMC Mol Biol 2006;7:3.
40. Keller M, Kralisch S, Rohde K, et al. Global DNA methylation
levels in human adipose tissue are related to fat distribution
and glucose homeostasis. Diabetologia 2014;57:2374–2383.
41. Lisanti S, Omar WAW, Tomaszewski B, et al. Comparison
of methods for quantification of global DNA methylation in
human cells and tissues. PLoS ONE 2013;8:e79044.
42. Dammann R, Yang G, Pfeifer GP. Hypermethylation of the
cpG island of Ras association domain family 1A (RASS-
F1A), a putative tumor suppressor gene from the 3p21.3
locus, occurs in a large percentage of human breast cancers.
Cancer Res 2001;61:3105–3109.
43. Imbeaud S. Towards standardization of RNA quality assess-
ment using user-independent classifiers of microcapillary
electrophoresis traces. Nucleic Acids Res 2005;33:e56.
Address correspondence to:
Dr. Yan-Ping Wang
Laboratory of Molecular Diagnosis of Cancer
West China Hospital
Sichuan University
Chengdu 610041
People’s Republic of China
E-mail: wyanping@scu.edu.cn