Assessing the Affect of RNA and cDNA Freeze-Thaw Cycling on Gene Expression using Microarray and RT-qPCR

1
Assessing the Affect of RNA and cDNA Freeze-Thaw
Cycling on Gene Expression using Microarray and
RT-qPCR Scott Tighe, Meghan Kohlmeyer, and Tim
Hunter UVM Microarray Facility and VCC DNA
Analysis Facility HSRF 305 Burlington, Vermont
05405

Abstract
Real-Time qPCR Results
Microarray Results
Results of the Real-Time qPCR analysis of
multiple freeze thaw cycling with RNA shows
insignificant changes in measured HPRT gene
expression from TP1 to any other time points
assessed. Using TP1 as the calibrator, the
largest change in gene expression observed was
between TP1 and TP6 which generated a delta Ct of
0.32. The delta Ct between TP1 and TP10 was
remarkably only 0.2.
Results of Agilent Bioanalyzer and microarray
analysis indicates little to no change of RNA
condition after all freeze thaw events as shown
below in RMA-derived MVA plots and RNA integrity
plots. MVA plots compare the fold change on the
Y-axis to the intensity log2 on the X-axis.
The routine handling of RNA for gene expression
studies often requires one or several freeze thaw
events prior to downstream synthesis reactions.
Because this is a common occurrence with many
RNAs used for gene expression studies in
microarray and Real-Time qPCR, it is necessary to
study the affects of freeze thaw cycling events
on both total RNA and cDNA. In this study, both
total RNA isolated from rat brain tissue and cDNA
prepared from HeLa cell RNA were frozen and
thawed from -20?C to room temperature several
times and subsequently analyzed using the Agilent
Bioanalyzer 2100, Affymetrix microarray
GeneChips, and Real-Time qPCR. Results from the
Bioanalyzer, microarray, and Real-Time qPCR on
total RNA indicate little to no affect on gene
signatures suggesting that the stability of RNA
to withstand repetitive freeze thaw episodes is
remarkable. Interestingly, the Real-Time qPCR
results for the freeze thaw cycling of cDNA
revealed a very small but noteworthy decrease in
gene detection suggesting that cDNA may be more
susceptible to these types of events. It should
be noted however, that these affects were nearly
insignificant and are highly dependent on
individual handling techniques and sample purity.
TP0 Control RNA
Methods
Microarray Study
RNA used for the microarray study was isolated
from rat brain tissue using Qiagens RNeasy kit
and evaluated using the Nanodrop
spectrophotometer and Agilent Bioanalyzer 2100
to ensure high quality and stored at -80 after
extraction. The freeze thaw strategy started with
6 micrograms of RNA being thawed to room
temperature for 8 hour and refrozen overnight.
This was repeated three additional times and 1.1
ug aliquots were removed for Bioanalyzer and
microarray analysis. After collecting all
samples, they were analyzed using the Agilent
Bioanalyzer and processed through the Affymetrix
standard target preparation synthesis protocol as
a group to eliminate synthesis variation. The
synthesized target was hybridized to RAE230A
GeneChips for 16 hours, stained, and scanned with
the GS3000 scanner. .
Results of the Real-Time qPCR analysis of
multiple freeze thaw cycling with cDNA shows
a slight decrease in measured HPRT gene
transcripts. The largest difference in measured
expression was between TP1 and TP10 which results
in a delta Ct of 0.54 representing a fold change
of 1.54.
Probe Set Name P-value Detection Fold change Probe designation
CLONEUI-R-Y0-ABI-C-04-0-UI 6537 0.067 P 5 1390368_at
PICCOLO (PRESYNAPTIC CYTOMATRIX PROTEIN) 0.037 P 4.5 1387399_at
PANCREATIC POLYPEPTIDE 0.334 P 4.4 1369196_at
EST197193 /CLONERKIBF80 0.870 P -4.3 1398637_at
Probe Set Name P-value Detection Fold change Probe designation
TRIADIN 10116 0.567 P 5.3 1370738_a_at
KINESIN-ASSOCIATED PROTEIN 3 0.601 P 5.2 1372963_at
CLONEUI-R-CA0-BKS-C-09-0-UI /FEAEST 0.828 P -4.7 1390693_at
G PROTEIN-COUPLED RECEPTOR 105 0.753 P -4.5 1370449_at
Sample Avg Ct Calibrator Delta Ct
TP1 24.78622 24.78622 0
TP2 24.94295 24.78622 0.156731
TP3 25.0743 24.78622 0.288072
TP4 25.0844 24.78622 0.298174
TP5 25.18659 24.78622 0.400365
TP6 25.19602 24.78622 0.409796
TP8 25.29049 24.78622 0.504264
TP10 25.3249 24.78622 0.538678
Freeze Thaw Schedule
Freeze Thaw Time Point TP Duration at Duration at Total Time Total Time
Freeze Thaw Time Point TP -80C 25C -80C 25C
TP0 NA NA NA NA
TP1 16hr 8hr 16hr 8hr
TP2 16hr 8hr 32hr 16hr
TP3 16hr 8hr 48hr 24hr
TP4 16hr 8hr 64hr 32hr
Probe Set Name P-value Detection Fold change Probe designation
MYOSIN IXA 0.533 P -3.3 1369707_at
TRANSFORMATION RELATED PROTEIN 63 0.533 P 3.3 1369108_at
SIMILAR TO CG11206-PA 0.366 P 3.3 1386497_at
SIMILAR TO ZINC FINGER PROTEIN 91 ISOFORM 1 0.919 P -3.3 1386512_at
CLONEUI-R-BT1-BNQ-E-11-0-UI /FEAEST 0.194 P 3.3 1377315_at
HOMEO BOX C8 0.828 M -3.3 1371281_at
Probe Set Name P-value Detection Fold change Probe designation
MULTIMERIN-SIMILAR TO A57384 0.696 P -3 1392053_at
MYOSIN HEAVY CHAIN, POLYPEPTIDE 6,7 0.303 P -3 1398248_s_at
TRIADIN 10116 0.303 P 3 1370739_x_at
CHEMOKINE (C-C) RECEPTOR 4 0.111 P 3 1369555_at
SIMILAR TO LRRGT(ribosomal L-protein) 0.149 P 3 1394330_at
T-BOX-18 (PREDICTED) 0.219 M 3 1379883_at
Gene Lists
Gene lists were generated for each time point
compared to the Time 0 TP0 control and sorted
for genes that represent a 2-fold change up or
down and had a P presences or M Marginal call
using the MAS algorithm regardless of P-value.
Each list of genes was submitted to the NCBI
DAVID database to determine the annotation. Data
are shown in each table above. Results
demonstrate that in all samples only 1 gene in
the TP1 sample had a significant P-value (0.037)
and fold change (4.5) for the gene
Piccolo-presynaptic cytomatrix protein. However,
when comparing all freeze thaw time points,
absolutely no common genes were affected between
samples!
Methods
Real-Time qPCR Study The intent of this study was
to mimic the sample processing occurring in the
facility for Real-Time qPCR request.
Investigators are often interested in measuring
new transcripts on a sample(s) previously run and
requires a freeze thaw cycle on either the RNA or
the cDNA. This requires a sample to completely
thaw to draw a subsequent aliquot for further
studies. Since all samples submitted to the
facility have experienced at least one freeze
thaw cycle, no TP0 was assessed. Further, this
study can address whether any targets are
compromised due to sample handling. RNA for
RTqPCR studies was isolated from HeLa cells using
Trizol and purified using Qiagens RNeasy
columns. All RNA was DNase-treated on-column
during the RNeasy extraction procedure. For cDNA
studies, 500ng total HeLa RNA was placed in 16
tubes. 8 samples were synthesized into cDNA
using the protocol for Superscript III
(Invitrogen). All 16 samples were frozen at -20C
and were freeze thawed 1-10 times according to
the schedule listed. Upon completion of the last
thaw, the RNA samples were synthesized into cDNA.
All gene expression was measured on an AB7900HT
Sequence Detection System employing HPRT
(hypoxanthine phosphoribyl transferease) as a
medium expressing housekeeping gene.
Freeze Thaw Schedule
Freeze Thaw Time Point TP cDNA cDNA RNA RNA
Freeze Thaw Time Point TP -20C 25C -20C 25C
TP0 NA NA NA NA
TP1 23.5hr 0.5hr 23.5hr 0.5hr
TP2 23.5hr 0.5hr 23.5hr 0.5hr
TP3 23.5hr 0.5hr 23.5hr 0.5hr
TP4 23.5hr 0.5hr 23.5hr 0.5hr
TP5 23.5hr 0.5hr 23.5hr 0.5hr
TP6 23.5hr 0.5hr 23.5hr 0.5hr
TP8 23.5hr 0.5hr 23.5hr 0.5hr
TP10 23.5hr 0.5hr 23.5hr 0.5hr
Total 235hr 5hr 235hr 5hr
Conclusion and Discussion Based on these
experiments in the UVM DNA Analysis and
microarray core laboratories using our standard
handling procedures, we observed insignificant
freeze thaw-related gene expression changes for
RNA when analyzed using Affymetrix microarray
analysis and Real-Time qPCR. Interestingly, when
cDNA was freeze thawed repeatedly, the measured
gene expression using Real-Time qPCR for HPRT was
slightly reduced each freeze thaw cycle with a
total reduced expression of 1.5 fold for all ten
freeze thaws. However, these results are highly
dependent on sample storage buffers, RNA and cDNA
purity, laboratory consumables, and individual
handling. As indicated by other experts in the
field, changes may be influenced by sample
concentration that is extremely low or high. We
did not investigate how concentration may affect
RNA or cDNA during freeze thaw cycling, instead
we chose concentrations that are routinely
encountered in our laboratory. Studies
investigating concentration affects maybe
considered in future work.
Acknowledgements The facility gratefully
acknowledges the Vermont Cancer Center, LCCRO,
Vermont Genetics Network, and NIH-NCRR for their
outstanding support for the use of reagents and
facilities. A special thanks goes to Russ Hovey
and Josie Trott for their inspirational debate on
freeze thawing mRNA which sparked the microarray
study.