Transcript Slide 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.
RNA FREEZE-THAW -20
TP0 Control RNA
26
25.5
CT VALUE
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 RNA’s used for gene expression studies in microarray and RealTime 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 -20೦C 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.
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.
TP1
TP3
25
24.5
24
23.5
23
Methods
#1
#2
#3
#4
#5
#6
#8
#10
NUMBER FREEZE-THAWS
Microarray Study
RNA used for the microarray study was isolated from rat brain tissue using
Qiagen’s 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.
6ug of rat brain
RNA -80C
Leave 1 ug at
-80c as
control
Probe Set Name
Control
CLONE=UI-R-Y0-ABI-C-04-0-UI 6537
25c 8hr
Remove 1ug, run
bioanalyzer and
freezer for
microarray
Remaining
RNA to -80C
for 16hr
.
Remaining
RNA to -80C
for 16hr
Detection
Fold change
Probe designation
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 /CLONE=RKIBF80
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
CLONE=UI-R-CA0-BKS-C-09-0-UI /FEA=EST
0.828
P
-4.7
1390693_at
G PROTEIN-COUPLED RECEPTOR 105
0.753
P
-4.5
1370449_at
Time Point 1
TP4
TP2
25C 8hrs
Remove 1ug, run
bioanalyzer and
freezer for
microarray
P-value
Time Point 2
Freeze Thaw Schedule
25C 8hrs
TP0
TP1
TP2
TP3
TP4
Duration at
Total Time
-80C
+25C
-80C
+25C
NA
NA
NA
NA
16hr
16hr
16hr
16hr
8hr
8hr
8hr
8hr
16hr
32hr
48hr
64hr
Remove 1ug, run
bioanalyzer and
freezer for
microarray
Remaining
RNA to -80C
for 16hr
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
Time Point 3
25C 8hrs
P-value
Detection
Fold
change
MYOSIN IXA
0.533
P
-3.3
1369707_at
TRANSFORMATION RELATED PROTEIN 63
0.533
P
3.3
SIMILAR TO CG11206-PA
0.366
P
SIMILAR TO ZINC FINGER PROTEIN 91 ISOFORM
1
0.919
CLONE=UI-R-BT1-BNQ-E-11-0-UI /FEA=EST
HOMEO BOX C8
Probe Set Name
8hr
Remove 1ug, run
bioanalyzer and
freezer for
microarray
16hr
Time Point4
24hr
32hr
cDNA FREEZE-THAW -20
P-value
Detection
Fold
change
Probe designation
MULTIMERIN-SIMILAR TO A57384
0.696
P
-3
1392053_at
1369108_at
MYOSIN HEAVY CHAIN, POLYPEPTIDE 6,7
0.303
P
-3
1398248_s_at
3.3
1386497_at
TRIADIN 10116
0.303
P
3
1370739_x_at
CHEMOKINE (C-C) RECEPTOR 4
0.111
P
3
1369555_at
P
-3.3
1386512_at
SIMILAR TO LRRGT(ribosomal L-protein)
0.149
P
3
1394330_at
0.194
P
3.3
1377315_at
T-BOX-18 (PREDICTED)
0.219
M
3
1379883_at
0.828
M
-3.3
1371281_at
Probe designation
Probe Set Name
Gene Lists
25.5
25
24.5
24
23.5
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.
Methods
26
CT VALUE
Freeze Thaw
Time Point [TP]
Sample
23
#1
#2
#3
#4
#5
NUMBER FREEZE-THAWS
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!
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 Qiagen’s 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
RNA
Conclusion and Discussion
-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
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.
#6
#8
#10