FFPE DNA Quality: Impact on Downstream Performance in Molecular Diagnostic Assays
FFPE DNA Quality:
Impact on Downstream Performance in Molecular Diagnostic Assays
Jonathan M. Eibl, Julie A. Toplin, Chad D. Galderisi, Cindy S. Spittle, MolecularMD Corp., Portland, OR
Poster #: ID102
Figure 3a. Maxwell vs. MO BIO DNA extraction from different FFPE sample types (total DNA Yield). Overall, DNA yields were greater for l
d
h
ll
(
f
l
)
samples processed on the Maxwell CSC (30 of 34 samples, 88%).
60
30
20
400
300
200
100
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
Melanoma
GIST
Endometrial
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
NSCLC
Melanoma
Figure 4a. Optical densities (260/280) Maxwell vs. MO BIO DNA extraction from different FFPE sample types.
Place elution tubes
Add elution water
Spin FFPE Add digest sample to (5 min.)
cartridge
Figures 7a – 7e. Box plots representing sequence quality metrics using CodeCodon Aligner and Sequence Scanner software.
OD 260/230 Maxwell
2.5
OD 260/230 MO BIO
7a.
OD (260/230)
OD (260/280)
2
2.5
2
1.5
1.5
1
0.5
0.5
0
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
Melanoma
GIST
Endometrial
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
NSCLC
Melanoma
Figure 5a NRAS allele‐specific PCR for DNA extracted by both methods
Figure 5a. NRAS allele‐specific PCR for DNA extracted by both methods from melanoma FFPE; traces shown at right for samples with ∆Ct Values ˃ 2.0 (Samples 02, 10, 11) and no amplification/amplification (Sample 07).
Sample
MO BIO 01
Maxwell 01
MO BIO 02 Maxwell 02
MO BIO 03 Maxwell 03
MO BIO 04 Maxwell 04
MO BIO 05 Maxwell 05
MO BIO 06 Maxwell 06
MO BIO 07
Maxwell 07
MO BIO 08 Maxwell 08
MO BIO 09 Maxwell 09
MO BIO 10
Maxwell 10
MO BIO 11
Maxwell 11
MO BIO 12 Maxwell 12
Ct Value
∆ Ct Value (MO BIO – Maxwell)
33.04
31.93
35.42
33.25
29.93
29.56
32.09
31.43
32.50
31.77
34.64
32.70
No amplification
35.07
30.10
30.12
30.11
28.39
30.93
28.33
30.80
28.25
28.44
28.39
Enter sample data / run
GIST
Add FP4
Vortex
Add FP5
Vortex
NSCLC
7b.
7c.
7d.
7e.
MO BIO 02
Maxwell 02
1.11
Maxwell 02
2.17
F
W
D
0.37
0.66
R
E
V
MO BIO 07
Maxwell 07
Maxwell 07
F
W
D
0 73
0.73
1.94
N/A
R
E
V
MO BIO 10
Maxwell 10
MO BIO 02
‐0.02
FWD
Conclusions
1.72
REV
This comparison highlights the potential advantages of using an automated nucleic acid extraction platform for onco‐diagnostic
testing:
MO BIO 07
2.60
MO BIO 11
Maxwell 11
FWD
2.55
REV
0 05
0.05
• Standardized workflow with fewer hands‐on steps minimizing the likelihood of cross‐contamination
• Higher throughput for faster turn‐around‐time of results
• Higher
i h DNA yields,
i ld maximizing
i i i the
h potential
i l data
d
output from
f
li i d amounts off FFPE tumor tissue
limited
i
Figure 2. Representative specimen (Sample 02) requiring additional Zymo cleanup steps due to high melanin content (note red boxed area); H&E staining, 4X magnification.
Remove elution tube
• Higher quality DNA without extra purification steps, minimizing the potential need for repeat analysis
Acknowledgements
The authors would like to thank Jeannette Nussbaum for her valuable contributions.
contributions
For further information please contact [email protected] or visit www.molecularmd.com. MO BIO: post‐lysis steps 1 – 22 (Hands‐on time for 16 samples ‐ ~60 min.)
Heat Spin Transfer
(1 hr.)
(1 min.)
Endometrial
Figure 5b. Comparison of NRAS Sanger sequencing results for DNA extracted by Figure 5b Comparison of NRAS Sanger sequencing results for DNA extracted by
both methods from two melanoma samples, both of which required the Zymo
cleanup step for the MO BIO extraction.
Maxwell: post‐lysis steps 1 – 12 (Hands‐on time for 16 samples ‐ ~30 min.)
Place plungers
NSCLC
3
OD 260/280 Maxwell
OD 260/280 MO BIO
1
Lyse (incubate) Remove seals
Endometrial
3
Steps common to both extraction methods
Place cartridges
GIST
Figure 4b. Optical densities (260/230) Maxwell vs. MO BIO DNA extraction from different FFPE sample types.
4
3.5
Figure 1. Comparison of extraction method procedural steps and hands‐on time
Add Incubate RNase A
(5 min.)
NO SEQUENCE
N
500
NO SEQUENCE
N
% increase in DNA
A yield
DNA Yield (ugg)
600
0
Both extraction methods require the same upfront steps of scraping FFPE tissue, reagent
addition, and lysis/incubation, however, the Maxwell method requires fewer hands‐on,
liquid‐handling
liq id handling steps (Fig.
(Fig 1).
1) Additional Zymo
Z mo cleanup
clean p steps were
ere required
req ired for 2 manually‐
man all
extracted samples due to visible melanin in the FFPE tissue/eluted DNA (Fig. 2; Samples 02
and 07). DNA yields were greater for samples processed on the Maxwell CSC (30 of 34 total
samples, 88%; 1.9‐fold increase overall; Figs. 3a and 3b). Optical densities for both methods
were largely congruent for the 260/280 ratio (Fig. 4a) but more variable for the 260/230
ratio (Fig. 4b). In qPCR, 11 of 12 (92%) of the Maxwell‐extracted melanoma samples crossed
the threshold 1.3
1 3 Ct sooner on average
a erage than the corresponding manually‐extracted
man all e tracted
templates (Fig. 5a), and one manually‐extracted, Zymo‐purified sample failed to amplify
(Sample 07). Two Maxwell‐extracted samples (Samples 10 and 11) crossed the threshold
markedly sooner (~2.5 Ct). Similarly, all Maxwell‐extracted melanoma DNA isolates
produced robust bidirectional sequence, whereas two manually‐extracted isolates failed to
yield NRAS sequence despite the inhibitor‐removal step (Fig. 5b). Aside from the sequence
f il
failures
off these
th
t
templates
l t and
d one additional
dditi
l template
t
l t which
hi h exhibited
hibit d elevated
l t d baseline
b li
noise, NRAS codon 61 mutation‐detection results for the remaining nine manually‐extracted
eluates were identical to the Maxwell‐extracted samples (data not shown). Templates from
both extraction methods exhibited higher median Phred scores and narrower interquartile
ranges (IQR) for reverse traces than for forward traces (Figs. 6a and 6b); this characteristic is
attributed to the proximal location of the forward primer relative to the NRAS codon 61
h t t Sequences
hotspot.
S
f
from
M
Maxwell‐extracted
ll t t d FFPE DNA exhibit
hibit higher
hi h mean Phred
Ph d scores
and less variability (insets; Figs. 6a and 6b). Overall, metrics generated using other software
packages (i.e., quality values, trace scores, quality value 20+, continuous‐read length, and
signal‐to‐noise ratio) illustrate equivalent or better results for the Maxwell‐extracted
templates especially in regards to variability (see IQRs for boxplots; Figs. 7a –7e).
Cool to RT
Figure 6b. Box plot graph of Phred scores for Sanger sequences (see Figure 5; reverse traces; 51‐bp window; melanoma samples). 10
Results
Add reagents
700
MO BIO 40
FFPE tissue from melanoma, GIST, endometrial, and NSCLC tumor types was obtained from
equivalent numbers of slides for each extraction method. Manual extractions were
performed using the MO BIO BiOstic FFPE Tissue DNA Isolation Kit followed in some
instances by an additional purification step using the Zymo Research OneStep PCR Inhibitor
Removal Kit to rid DNA isolates of visible melanin. Automated extractions were carried out
using the Maxwell CSC DNA FFPE Custom Kit on the Maxwell CSC instrument with no
additional PCR inhibitor removal steps. DNA eluate yields and optical densities were
determined via spectrophotometry. Equivalent input DNA amounts from melanoma
samples were assayed using both NRAS allele
allele‐specific
specific qPCR (25 ng) and NRAS Sanger
sequencing (200 ng) on the BIORAD CFX platform and Applied Biosystems 3730 DNA
Analyzer, respectively. NRAS qPCR data were reported as delta‐Ct values to determine qPCR
differences. Sanger sequencing Phred scores were analyzed for the 51‐base pair window
surrounding the NRAS codon 61 hotspot to quantitatively determine sequence‐trace quality.
Quality metrics were further evaluated using the CodonCode Aligner and Sequence Scanner
software packages.
packages
Figure 6a. Box plot graphs of Phred scores for Sanger sequences (see Figure 5; forward traces; 51‐bp window; melanoma samples). 800
Maxwell 50
Materials & Methods
Scrape FFPE Figure 3b. Percent increase in DNA yield by Maxwell vs. MO BIO in different FFPE sample types (1.9‐fold or 190% increase overall). NO SEQUENCE
Personalized medicine for oncology patients is driven by mutation profiling of tumors.
Improved
robust DNA extraction methods that produce
high
p
p
g quality
q
y FFPE DNA template
p
suitable for complex molecular diagnostic testing are needed. In addition to DNA
fragmentation and protein‐DNA cross linking, some FFPE tissues contain inhibitors and/or
quenching substances (e.g. melanin) that adversely affect downstream performance in PCR
and sequencing assays. In this study, we provide a side‐by‐side evaluation of two DNA
isolation methods from FFPE tissue of various tumor types: a manual, column‐based
)
method ((MO BIO)) and the Maxwell CSC automated p
platform ((Maxwell).
NO SEQUENCE
Introduction
Spin
Apply to column
Spin (1 min.)
Transfer spin filter
Add FP6 (continue)
Spin (1 min.)
Add FP7 (1 of 2)
Spin (1 min.)
Add FP7 (2 of 2)
Spin (1 min.)
Transfer column
Spin (2 min.)
Add FP8
Incubate Spin (5 min.)
(1 min.)
*(28 total steps if Zymo needed)
* ZYMO cleanup (as needed)