Subscriber access provided by University of Otago Library
Article
An International Comparison of Enumeration-based Quantification of DNA Copy-concentration Using Flow Cytometric Counting and Digital Polymerase Chain Reaction Heebong Yoo, Sang-Ryoul Park, Lianhua Dong, Jing Wang, Zhiwei Sui, Jernej Pavši#, Mojca Milavec, Muslum Akgoz, Erkan Mozio#lu, Philippe Corbisier, Janka Matrai, Bruno Cosme, Janaina Japiassu de Vasconcelos Cavalcante, Roberto Becht Flatshart, Daniel Gerard Burke, Michael Forbes-Smith, Jacob L. H. McLaughlin, Kerry R. Emslie, Alexandra S. Whale, Jim Francis Huggett, Helen Parkes, Margaret Kline, Jo Lynne Harenza, and Peter M Vallone Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.6b03076 • Publication Date (Web): 10 Nov 2016 Downloaded from http://pubs.acs.org on November 12, 2016
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 27
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
1
An International Comparison of Enumeration-based Quantification of DNA Copy-
2
concentration Using Flow Cytometric Counting and Digital Polymerase Chain Reaction
3
Hee-Bong Yoo,1,2 Sang-Ryoul Park,1,2,∗ Lianhua Dong,3 Jing Wang,3 Zhiwei Sui,3 Jernej
4
Pavšič,4 Mojca Milavec,4 Muslum Akgoz,5 Erkan Mozioğlu,5 Philippe Corbisier,6 Matrai
5
Janka,6 Bruno Cosme,7 Janaina J. de V. Cavalcante,7 Roberto Becht Flatshart,7 Daniel Burke,8
6
Michael Forbes-Smith,8 Jacob McLaughlin,8 Kerry Emslie,8,∗ Alexandra S. Whale,9 Jim F.
7
Huggett,9 Helen Parkes,9 Margaret C Kline,10 Jo Lynne Harenza,10 Peter M. Vallone10
8 9
*Correspondences
10
(S-R Park)
[email protected]: correspondence on overall study
11
(K. Emslie)
[email protected]: correspondence on digital PCR techniques
12 13
1. Korea Research Institute of Standards and Science, Daejeon, Republic of Korea
14
2. University of Science and Technology, Daejeon, Republic of Korea
15
3. National Institute of Metrology, Beijing, P.R. China
16
4. National Institute of Biology, Ljubljana, Slovenia
17
5. TUBITAK UME National Metrology Institute, Kocaeli, Turkey
18
6. Institute for Reference Materials and Measurements, Joint Research Centre, European
19
Commission, Geel, Belgium
20
7. National Institute of Metrology, Quality and Technology, Xerém, Brazil
21
8. National Measurement Institute Australia, Lindfield, Australia
22
9. LGC, Teddington, UK
23
10. National Institute of Standards and Technology, Gaithersburg, USA
24 25
ACS Paragon Plus Environment
Analytical Chemistry
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
1
ABSTRACT
2
Enumeration-based determination of DNA copy-concentration was assessed through an
3
international comparison among national metrology institutes (NMIs) and designated
4
institutes (DIs). Enumeration-based quantification does not require a calibration standard
5
thereby providing a route to ‘absolute quantification,’ which offers the potential for reliable
6
value assignments of DNA reference materials, and International System of Units (SI)
7
traceability to copy number 1 through accurate counting. In this study, 2 enumeration-based
8
methods, flow cytometric (FCM) counting and the digital polymerase chain reaction (dPCR),
9
were compared to quantify a solution of the pBR322 plasmid at a concentration of several
10
thousand copies per microliter. In addition, 2 orthogonal chemical-analysis methods based on
11
nucleotide quantification, isotope-dilution mass spectrometry (IDMS) and capillary
12
electrophoresis (CE) were applied to quantify a more concentrated solution of the plasmid.
13
Although 9 dPCR results from 8 laboratories showed some dispersion (relative standard
14
deviation [RSD] = 11.8%), their means were closely aligned with those of the FCM-based
15
counting method and the orthogonal chemical-analysis methods, corrected for gravimetric
16
dilution factors. Using the means of dPCR results, the RSD of all 4 methods was 1.8%, which
17
strongly supported the validity of the recent enumeration approaches. Despite a good overall
18
agreement, the individual dPCR results were not sufficiently covered by the reported
19
measurement uncertainties. This finding suggests that some laboratories may not have
20
considered all factors contributing to the measurement uncertainty of dPCR, and further
21
investigation of this possibility is warranted.
22 23
Keywords: Enumeration-based quantification, DNA copy-concentration, flow cytometric
24
counting, digital PCR, chemical analytical quantification of DNA, national metrology
25
institute, international comparison
26 27
ACS Paragon Plus Environment
Page 2 of 27
Page 3 of 27
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
1
INTRODUCTION
2
The increasing application of quantitative molecular methods to measurements across several
3
sectors requires a robust mechanism for their quality control and quality assurance. This is
4
especially true where adequate analytical methods are required to provide accurate
5
measurements on which legislation can be based or clinical decisions can be made. However,
6
a lack of higher-order reference methods and materials is a major hindrance for deriving
7
traceability and measurement comparability, which impacts accreditation and regulatory
8
compliance.
9
In this study, we aimed to address this need by evaluating the performance of 2 different
10
enumeration-based, DNA-quantification methods as candidate reference methods that can be
11
applied without the requirement for calibration standards. Such methods would provide the
12
basis for a substantial increase in the range of molecular measurements that can rely on SI
13
traceability (to 1) through counting.
14
Calibration standards for assuring the performance of a measurement method need to be
15
characterized by a suitable measurement method. In the case of trace amounts of DNA, a
16
calibration standard may be prepared by consecutive gravimetric dilutions of a concentrated
17
DNA stock that can be accurately determined by conventional chemical analysis.1-5 However,
18
the reliability of consecutive gravimetric dilutions of DNA solutions has yet to be confirmed,
19
and the manipulation of DNA materials may lead to their degradation through, for example,
20
shear damage.6 If possible, the value for a DNA reference material should be assigned using
21
methods that are capable of high-accuracy measurement across a DNA concentration range
22
that encompasses the concentration of the reference material. In the case of low or trace
23
concentrations of DNA, enumeration-based quantification may be a suitable approach for
24
absolute DNA quantification. If all target DNA molecules or sequences in a known volume
25
are successfully counted one by one (so called ‘exhaustive counting’), a calibrator commonly
26
needed to determine the relationship between the signal intensity and the sample
27
concentration is not required to determine the DNA concentration. Most simply, a single
28
target DNA molecule could be detected for enumeration using high-sensitivity, laser-induced
29
fluorescence detection provided that the target DNA is large enough to produce detectable
30
fluorescence with appropriate staining dyes. The sample solution should be sufficiently dilute
31
to result in singly separated DNA particles. This approach lead to the measurement principle
ACS Paragon Plus Environment
Analytical Chemistry
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 4 of 27
1
of flow cytometric counting (FCM counting) of target DNA particles, which was used in this
2
study.7,8 Alternatively, if target DNA sequences are distributed across a large number of
3
uniformly sized PCR chambers (partitions), they can be amplified by PCR and then
4
enumerated based on the number of PCR-positive partitions and the total number of partitions,
5
provided that a portion of the partitions contain no target DNA sequences. This detection
6
strategy is known as digital PCR (dPCR).9-13 Techniques of dPCR
7
studied not only as a superior alternative to real time PCR but also as an absolute
8
quantification method.14-19 Several NMIs recognized the potential advantage of these 2
9
enumeration strategies, and have developed or investigated intensely as candidate reference
have been actively
10
analytical procedures for absolute quantification of low-level DNA concentrations.7,8, 20-29
11
Although the FCM counting method follows a simple, straightforward measurement principle,
12
there could be measurement biases due to, for example, some DNA particles bypassing the
13
detection zone or multimeric appearances in the detection zone. These possibilities have
14
previously been thoroughly investigated and reduced. For example, multimeric appearances
15
in the detection zone were detected with accordingly larger signal intensities or enlarged
16
exposure times and comprised the largest component of the measurement uncertainty.8 Unlike
17
the FCM counting method, dPCR does not distinguish between partitions containing a single
18
copy of target DNA and partitions containing multiple copies of target DNA. The total
19
number of copies of target DNA across the partitions in dPCR is estimated by Poisson
20
statistics. This estimate has some uncertainty dependent upon the total number of partitions
21
and the proportion of partitions containing at least 1 copy of target DNA.24,25 As for all
22
analytical procedures, dPCR requires careful optimization. For accurate quantification by
23
dPCR, consistent amplification from a single copy of target DNA is required. For this, primer
24
sets, PCR reagents including polymerase and master mix, and thermal cycling conditions
25
need to be evaluated. Unintended melting of double-stranded DNA leads to overestimation
26
because the single strands are distributed independently across partitions, resulting in a higher
27
proportion of partitions containing amplifiable target sequences.27 Several actions required
28
for successful dPCR have been reported.20-29
29
For the acceptance of both FCM-based counting and dPCR as reference analytical procedures,
30
it is important to determine the level of comparability that is achievable among NMIs.
31
Because FCM counting and dPCR are fundamentally different measurement techniques,
32
comparability between these 2 enumeration methods would provide greater confidence in
ACS Paragon Plus Environment
Page 5 of 27
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
1
their performance. For this reason, an international comparison of enumeration-based
2
quantification of low level DNA was organized in this study.
3
In this comparison, a simple form of DNA was used as the test sample. Plasmid DNA (for
4
example, the pBR322 plasmid of 4.36 kb) can be obtained in a relatively pure form and is
5
detectable by both FCM counting and dPCR. Plasmid DNA can be measured in a circular
6
form (mostly in a supercoiled form) or after linearization with a restriction enzyme. To
7
investigate the applicability of these measurement techniques to both the structurally
8
hindered supercoiled form and the linear form, both forms of a plasmid DNA were tested.
9
Two chemical-analysis methods were also applied, namely high pressure liquid
10
chromatography (HPLC)-isotope-dilution mass spectrometry (IDMS)3,24 and capillary
11
electrophoresis with detection by ultraviolet absorbance (CE-UV)5, which quantify
12
nucleotides (or deoxynucleoside monophosphates [dNMPs]) that are enzymatically released
13
from the target DNA. These methods require substantially higher DNA concentrations than
14
do dPCR or FCM counting. Concentrated DNA samples were prepared, with the
15
gravimetrically determined concentrations. Samples were delivered to testing facilities across
16
several continents. The short-term stability of the test samples the under conditions expected
17
during transport was carefully evaluated after confirming their homogeneity. Participants in
18
the study were either NMIs or DIs of the Bioanalysis Working Group (BAWG) under the
19
Consultative Committee on the Amount of Substance (CCQM) for the Meter Convention.
20 21
EXPERIMENTAL
22
Comparison scheme, sample preparation, and delivery
23
All samples were prepared at the NMI of Korea (KRISS) using gravimetric dilutions,
24
according to the scheme shown in Figure 1. The mixed pool of pBR322 plasmid DNA
25
(#SD0041; Fermentas, Burlington, Canada) was divided into 2 parts. One part was diluted
26
approximately 100-fold in deionized water to generate a high level, intact sample of circular
27
plasmid DNA (Level I, type-C). The other part was linearized using the EcoRI restriction
28
enzyme (#ER0271; Fermentas), which resulted in approximately a 10-fold dilution of the
29
original material. The linearized plasmid was then further diluted 10-fold, resulting in
30
approximately a 100-fold dilution (Level I, type-L) from the original material.
ACS Paragon Plus Environment
Analytical Chemistry
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
1
Both Level I, type-C and type-L were serially diluted by approximately 125,000-fold in a
2
gravimetric manner in 1× Tris-EDTA buffer to generate the Level II samples. Two hundred 1-
3
mL aliquots of Level II samples were prepared with both type-C and type-L samples, which
4
were immediately stored at −70ºC. Sample homogeneity was tested by FCM counting of 10
5
randomly chosen aliquots from both the Level II, type-C and type-L samples. Sample
6
stability was tested at 4 time intervals (1 day, 2 days, 5 days, and 10 days) and at 4 different
7
storage temperatures (−20ºC, 4ºC, 25ºC, and 45ºC). For each time and temperature, 2 aliquots
8
were randomly chosen for each type of Level-II aliquot and quantified at least 3 times by
9
FCM counting. Following the homogeneity and stability studies, each participating laboratory
10
was sent 5 vials of both Level II type-C and type-L samples packaged with cool-packs.
11 12 13
Figure 1. Schematic representation of the CCQM P-154 pilot study: absolute quantification of DNA
14 15
Participants and analytical methods. The participating laboratories and their analytical
16
methods are listed in Table 1. The full names of the participating institutes are as follows:
17
NMIA, National Metrology Institute of Australia; INMETRO, National Institute of Metrology,
18
Quality and Technology; NIM, National Institute of Metrology; IRMM, Institute of Reference
19
Materials and Measurements; KRISS, Korea Institute of Standards and Science; NIB,
20
National Institute of Biology; TUBITAK UME, National Metrology Institute of Turkey; LGC,
ACS Paragon Plus Environment
Page 6 of 27
Page 7 of 27
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
1
Laboratory of Government Chemist; NIST, National Institute of Standards and Technology.
2
Refer to Table 1 for the economy of each institute. Each of the applied analytical methods is
3
described in detail below. The details of dPCR protocol used by each participant are
4
described in the references shown in Table 1.
5 6 7
Table 1. Study participants and applied methods Institute* NMIA NMIA
Method d-PCR HPLC-IDMS (for Level I)
INMETRO
d-PCR
Slovenia Turkey United Kingdom
NIM IRMM KRISS KRISS NIB TŰBITAK-UME LGC
United States
NIST
Economy Australia Brazil China European Union Korea, Republic of
8
References
d-PCR d-PCR FCM counting CE-UV (for Level I) d-PCR d-PCR d-PCR
Instrumentation QX-100 (Bio-Rad) 6510 Q-TOF LCMS (Agilent) QX-100 (Bio-Rad) Biomark HD (Fluidigm) Biomark HD (Fluidigm) Home-made flow cytometry system G1600AX (Agilent) QX100 (Bio-Rad) Quant Studio3D (Thermo Fisher Sci.) Biomark HD (Fluidigm)
d-PCR
Biomark HD (Fluidigm) & QX100 (Bio-Rad)
31,32
36 24 38 33,34 28,37 8 5 30 35 21,39
*
Refer to the text for the detailed name of each participating institute.
9 10 11
dPCR
12
Three dPCR platforms were used in this comparative study (Table 1). Three participants used
13
only the BioMark platform, (Fluidigm; chip-based system), 1 participant used the
14
QuantStudio 3D platform (Thermo Fisher Scientific; chip-based system), and 3 participants
15
used only the QX100 platform (Bio-Rad; emulsion-based system). In addition, 1 participant
16
used both the BioMark and QX100 systems. One participant measured the Level 1, type-L
17
material using dPCR after the necessary dilution. Well-documented procedures for dPCR20,21
18
were suggested as standard protocols. Nevertheless, participants were allowed to choose their
19
own instrumentation and experimental protocols. An example of how each laboratory utilized
20
dPCR to quantify a DNA target is provided in Table 1 in the form of reference publications.
21
In addition, the details of dPCR conditions of each participant are provided in Table S-1A and
22
B. As the aim of this study was to assess the general comparability among the enumeration
ACS Paragon Plus Environment
Analytical Chemistry
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
1
methods, it was beyond the scope of this study to investigate whether specific experimental
2
conditions affected the quantification results.
3
FCM counting
4
Level-II samples were diluted 5-fold in a Tris-based running buffer containing 10% (v/v)
5
DMSO and 0.1× SYBRTM-Gold dye. Part of this diluted solution was introduced by pressure
6
into an approximately 1-m long, 50 µm × 50 µm square capillary tube (Polymicro
7
Technologies, Phoenix, AZ) for flow cytometry. The DNA particles were counted until the
8
mixture was completely driven out from the capillary, with fresh running buffer following the
9
sample solution. For this procedure, 0.07 MPa was applied for 2 min, with 20 kV applied to
10
the end of the capillary where the direction of the electric field was opposite to the direction
11
of the flow. The details of the experimental conditions are available elsewhere.8
12 13
Nucleotide analysis by CE-UV
14
Level-I samples were concentrated by approximately 10-fold by vacuum drying, and the
15
change in the sample mass following drying was gravimetrically determined. The
16
concentrated DNA was enzymatically digested to dNMPs using DNase I (PN 78311; USB,
17
Cleveland, OH) and phosphodiesterase I (PN 20240; USB) (Figure 1). The dNMPs were
18
separated and quantified by CE-UV using a G1600AX instrument (Agilent), with iothalamate
19
serving as the internal standard as described previously.5 Calibration of each nucleotide was
20
performed with nucleotide calibration-standard solutions that were determined by inductively
21
coupled plasma-optical emission spectroscopy (ICP-OES) in conjunction with purity
22
determinations by high-performance liquid chromatography-ICP-mass spectrometry (HPLC-
23
ICP-MS). The quantity of DNA was estimated by dividing the quantity of each dNMP by its
24
number in the DNA sequence, which resulted in 4 predictions for dAMP, dTMP, dGMP, and
25
dCMP. The average of these 4 values was assigned as the quantity of the 10-fold concentrated
26
Level-I DNA. Considering the gravimetric concentration factor and the dilution factor, the
27
quantity of Level-II DNA was also estimated.
28 29
Nucleotide analysis by HPLC-IDMS
ACS Paragon Plus Environment
Page 8 of 27
Page 9 of 27
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
1
Level-I, Type-L study materials were analyzed as they were received. A sample blend was
2
prepared by adding a mixture of stable isotopically labeled dNMPs to each of 5 100-mg
3
subsamples of the study material. A calibration blend was prepared by adding the same stable
4
isotopically labeled dNMP mixture to a mixture of reference dNMPs. Both the sample and
5
calibration blends were enzymatically hydrolyzed using a mixture of DNase I and
6
phosphodiesterase I. After hydrolysis, each blend was ultrafiltered through Pall Nanosep 3K
7
Ω membranes and the filtrate was analyzed by LC-MS on an Agilent 6510 QTOF system.
8
Details of the measurement system, reagents, and calculations were published previously.24
9 10 11
RESULTS AND DISCUSSION
12
Testing of comparison samples
13
The results of homogeneity testing using FCM counting for the study materials are shown in
14
Figure 2. For this test, 10 vials were randomly chosen from a batch of 200 vials. The relative
15
standard deviations (RSDs) of measurements for each vial ranged from 0.2 to 2.5% (n = 3).
16
The results of F-testing (df1 = 9, df2 = 20) were 0.35 and 0.89 for type-C and type-L samples,
17
respectively. Those values were substantially smaller than Fcritical value of 2.39 at α = 0.05.
18
Therefore, it was concluded that test samples were prepared homogeneously, within the
19
measurement uncertainty (~5%). It was necessary to avoid the potential problem of high
20
viscosity of the concentrated DNA stock solution while preparing homogenous samples.
21
Preparing Level-I samples after a ~100-fold dilution of the stock solution to ~5 ng/µL (Figure
22
1) led to satisfactory homogenization as expected, which was necessary to assure the
23
comparability of measurement results obtained by different methods. Agreement in the
24
measurement results for different vials obtained during measurement-repeatability testing by
25
both CE-UV and LC-IDMS demonstrated the homogeneity of the Level-I samples (data not
26
shown).
27
ACS Paragon Plus Environment
Analytical Chemistry
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
1 2 3 4 5
Figure 2. The results of homogeneity testing by FCM counting of Level-II samples: (A) typeC samples; (B) type-L samples. Each sample vial was measured 3 times, and 10 vials were measured simultaneously in exactly the same manner. Error bars indicate the CVs of 3 repeated measurements.
6 7
The short-term stability of the test samples during international delivery was examined by
8
measuring samples left at room temperature or at an elevated temperature (45ºC) for several
9
days. Measurement was performed by FCM counting, the results of which are presented in
10
Figure S-1. No significant reduction in the concentration of pBR322 was observed for up to
11
10 days. Instead, a few samples showed a slightly elevated level (not exceeding 3%, indicated
12
with red arrows) after 5 or 10 days, which was not directly linked to the storage temperature.
13
This was more likely due to the evaporation of water through loose caps. After this
14
observation, newly prepared samples were firmly sealed with great care, and no observable
15
changes beyond measurement uncertainty were observed after storage at 45ºC for 10 days
16
(data not shown). The pBR322 plasmid was stable in 1× TE buffer against chemical or
17
physical degradation for up to 10 days, indicating that accidental exposures to ambient
18
temperatures during sample delivery would not cause a problem. No considerable exposure to
19
a degrading environment during sample delivery was reported. All participating laboratories
20
(except for 1) received the test samples within 4 days. The laboratory in question received the
21
samples 7 days after shipment, but were stored in a refrigerator for some of the transit time.
22
ACS Paragon Plus Environment
Page 10 of 27
Page 11 of 27
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
1
Results obtained with the linearized pBR322 (type-L sample)
2
The reported results for the linearized pBR322 plasmid (type-L samples) are compared in
3
Table 2. One result by dPCR (result report 1) was more than an order of magnitude lower
4
than the other results. This result was reported by a laboratory relatively new to performing
5
dPCR. Subsequently, it was found that the calculation of the results did not include the
6
consideration of the dilution used to prepare the final test solutions. The dilution factor-
7
corrected result is shown as result report 5 in Table 2 and was relatively well aligned with the
8
other results. During the subsequent data processing, the initial result of the clearly identified
9
mistake (Result report 1) was excluded. For the enumeration methods, dilution of original
10
samples down to the proper concentration range (the linear dynamic range) for the selected
11
analytical method was frequently necessary. Considerations of such dilution factors when
12
calculating the results should not be neglected.
13
Table 2. Result reports for the linearized plasmid DNA sample. Unless noted, the samples were Level II, type-L. Results of Level-I samples were normalized to the concentrations of Level II, based on the dilution factor from Level I to Level II (refer to Table 3).
14 15
16
Result report
Method
Instrumentation
1 2 3 4 5 6 7 8 9 10 11 12 13
d-PCR d-PCR d-PCR d-PCR d-PCR ID/MS (Level I) d-PCR CE-UV (Level I) direct counting d-PCR d-PCR d-PCR (Lelevl I) d-PCR
Bio-Rad QX-100 Bio-Rad QX-100 Thermo Fisher QuantStudio 3D Fluidigm Biomark HD Bio-Rad QX-100 Agilent 6510 QTOF LC-MS Fluidigm Biomark HD Agielnt G1600AX Home-made Fluidigm Biomark HD Bio-Rad QX100 Bio-Rad QX100 Fluidigm Biomark HD
Value (cp/mg) 329 6640 6730 6990 7304 7550 7610 7680 7880 8060 8460 8840 9090
Uncertaint Coverage Factor y (cp/mg) 4.8 2.0 370 2.0 1050 2.0 560 2.6 37 2.0 220 2.0 240 2.2 500 2.0 490 2.0 550 2.0 250 2.0 450 2.0 920 2.0
17 18
While assessing the comparability of the applied methods, the mean of 9 dPCR results, 7747
19
copies/mg, was taken as the representative value for dPCR. The standard uncertainty of this
20
mean value was calculated to be 304 copies/mg (the standard deviation of 912 copies/mg
21
divided by the square root of nine, which was the number of independent results). Then, the
22
expanded uncertainty of the mean of the dPCR data set was calculated to be 687 copies/mg
ACS Paragon Plus Environment
Analytical Chemistry
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
1
based on the coverage factor of 2.26 (95% confidence level, degree of freedom = 8). This
2
representative result for dPCR is compared with the results of other methods in Figure 3. The
3
results of all 4 methods were in good agreement (RSD = 1.8%) within their measurement
4
uncertainties. Between 2 enumeration approaches performed without the use of calibration
5
standards, the relative deviation from their average value was only 0.9% (7747 vs. 7880
6
copies/mg). The close agreement of results obtained using methods with totally different
7
measurement principles strongly supported the validity of those methods. Therefore, the
8
validity of the new approaches of enumeration-based quantification of DNA copy-
9
concentration was confirmed through this comparison study, which was unique in terms of
10
employing multiple methods performed at multiple sites.
11 12 13 14 15 16 17
Figure 3. Comparison of the results obtained with type-L samples using different quantification methods. The results obtained for Level-I samples were normalized to the predictions of Level-II sample, based on the gravimetric dilution factors from Level I to Level II. The representative dPCR results are presented as the mean of all dPCR results, with an expanded uncertainty for the 95% confidence interval. Error bars shown for other methods are also their expanded uncertainties for the 95% confidence interval.
18
ACS Paragon Plus Environment
Page 12 of 27
Page 13 of 27
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
1
The observed agreement confirms the absence of a significant hidden bias in FCM counting
2
using unique homemade instrumentation. The possible underestimation of FCM-counting due
3
to some DNA particles straying away from the detection zone was confirmed to be
4
insignificant in this comparison study. Although FCM counting has a limitation in terms of
5
the lack of specificity for target DNA sequences, the method can be used as a credible
6
analytical method for quantifying pure DNA reference materials. FCM counting operates
7
under a straightforward measurement principle and enables real-time monitoring and
8
recording of raw counting signals (Figurer S-2), which facilitates the ready detection of errors
9
during measurement.
10 11
dPCR, the major enumeration approach adopted by several NMIs and DIs, was validated for
12
the first time through a comparison among multiple laboratories, in parallel with other
13
measurement methods. Although the dispersion of dPCR results was not insignificant (RSD =
14
11.8%), the distribution was rather symmetric to the mean value that was in close agreement
15
with the results of the other methods used. Therefore, dPCR in general is not likely to carry
16
substantial measurement bias. The dispersion of dPCR results could have been in part due to
17
the different experimental conditions optimized at each laboratory. As shown in many
18
reports,20-29 dPCR requires sophisticated optimization of experimental conditions to avoid
19
measurement biases. Each participant in this comparative study freely determined the
20
optimum measurement conditions, including the PCR priming sites, reagents, and
21
instrumentation, which was necessary to assess the comparability of dPCR techniques as
22
practiced at individual laboratories. Variations in dPCR protocols used among the participants
23
are presented in Table S-1. Briefly, 3 laboratories analyzed the samples using a single PCR
24
assay, 3 laboratories used 2 PCR assays, while 2 laboratories utilized either 4 or 8 separate
25
PCR assays. The PCR assays targeted sequences between nucleotides positions 748 and 2105
26
bp, or between 2900 and 4276 bp on the pBR322 plasmid (Accession Number J01749). At
27
this stage, the major causes of variation from the mean value are uncertain, and further
28
investigation on those causes is beyond the scope of this study. However, we are rapidly
29
gaining knowledge on dPCR performance characteristics,20-41 and comparability of dPCR
30
results among multiple laboratories may be significantly improved in the near future. At
31
present, the results of is study suggest an RSD of ~12% as the range of possible scattering of
32
dPCR results from multiple laboratories, if experimental conditions are not restricted. As
ACS Paragon Plus Environment
Analytical Chemistry
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
1
shown in Figure 4, some results were barely within the range of the representative dPCR
2
result, but with large measurement uncertainties, whereas some laboratories reported
3
substantially smaller uncertainties. Therefore, further investigations on the proper assessment
4
of the measurement uncertainty of dPCR are necessary.
5
6 7 8 9
Figure 4. Results of dPCR quantification for type-L samples aligned with their mean and its expanded uncertainty for the 95% confidence interval. The error bar of each result also indicates the reported expanded uncertainty for the 95% confidence interval.
10 11
The results of orthogonal chemical analysis (HPLC-IDMS and CE-UV), normalized using the
12
gravimetric dilution factors from Level I to Level II, deviated from the average of all methods
13
only by −2.1% and −0.4%, respectively (Table 2). The originally reported data and their
14
conversions to the Level-II concentrations are shown in Table 3, which also includes a dPCR
15
result. Even with a dilution of ~1/125,000, significant deviations were not evident, which
16
strongly suggested that gravimetric dilution of a plasmid solution can be performed reliably.
17
Note that dilution was performed in only 2 steps to simplify the procedure (Figure. 1). Unlike
ACS Paragon Plus Environment
Page 14 of 27
Page 15 of 27
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
1
the enumeration methods, the chemical-analysis methods were calibrated with calibration
2
standards, which were nucleotide standard solutions quantified by ICP-OES to measure the
3
quantity of phosphorus in the nucleotides. The credibility of the nucleotide-measurement
4
methods was also demonstrated by the good agreement with enumeration-based methods,
5
along with the validity of gravimetric dilution.
6 7 8 9
10
Table 3. Report results for a Level-I, type–L sample and the predicted Level-II concentrations based on the dilution factor from Level I to Level II Result report 6 8
HPLC-IDMS CE-UV
12
dPCR
Method
Report data (copies/mg) (9.54 ± 0.28) x 108 (9.70 ± 0.63) x 108 (1.116 ±0.057) x109
Dilution factor* 7.918 x 10-6
Prediction of Level II (copies/mg) 7550 ± 220 7680 ± 500 8840 ± 450
* The uncertainty of dilution factor was ignored as the weighing uncertainties were relatively negligible.
11 12
Results for the circular pBR322 plasmid (type-C sample)
13
The results of the analysis of intact, circular pBR322 plasmid DNA (type-C samples) are
14
shown in Table 4. Three out of 5 dPCR results were substantially lower than the results of
15
FCM counting and CE-UV, which were not susceptible to the target-DNA structure during
16
quantification (internal communication with the KRISS laboratory). One dPCR result (result
17
report 4) was obtained after on-site linearization of the type-C sample, which dramatically
18
raised the result to ~80% of that of FCM counting.
19
The possibility of underestimating circular DNA concentrations by dPCR was revealed by the
20
fluorescence responses observed during real-time PCR. As shown in Figure S-3, the
21
fluorescence response curves for circular DNAs were markedly less robust compared to those
22
observed with linearized DNA, which implies that a substantially higher failure rate may
23
occur during PCR amplification of circular DNA. A substantial portion of circular DNA is in
24
the form of compactly packed supercoils (Figure S-4), which could structurally hinder PCR
25
primers from accessing the target sequences. Interestingly, 1 participant reported a dPCR
26
result (result report 6) that closely agreed with the FCM-counting result. The particular dPCR
27
conditions used for this exceptional result featured a specific PCR master mix combined with
ACS Paragon Plus Environment
Analytical Chemistry
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 16 of 27
1
a fluorescent probe.33 This finding raises the possibility of accurate quantification of circular
2
DNAs by dPCR without linearization.
3 4 5 6 7
8
Table 4. Results for the circular pBR322 plasmid (type-C sample). The reported values are listed together with the expanded uncertainty and coverage factor to provide the 95% confidence level. The result for each Level-I sample was normalized to that predicted for Level II, using the gravimetric dilution factor from Level I to Level II. Result report
Method
Instrumentation
Value (cp/mg)
1 2 3 4 5 6
d-PCR d-PCR d-PCR d-PCR CE-UV d-PCR
Thermo Fisher QuantStudio 3D Bio-Rad QX-100 Fluidigm Biomark HD Fluidigm Biomark HD Agielnt G1600AX Fluidigm Biomark HD
2690 2710 4140 6980 8200 8680
Expanded uncertainty 250 180 330 530 700 620
7
direct counting
Home-made
8700
500
Coverage factor 2.0 2.0 2.6 2.0 2.0 2.0 2.0
9 10 11
Other noteworthy findings
12
Estimation of the sample volume or partition volume. During FCM counting, the sample
13
volume is typically about 1.5 µL. Accurate determination of this small volume was confirmed
14
only after demonstrating satisfactory comparability among several different measurement
15
methods.8 For dPCR-based methods, an accurate estimation of the partition volume is also
16
critically important. While calculating the DNA concentration, the average number of
17
counted entities per partition is divided by the partition volume. An inaccurate determination
18
of the partition volume directly leads to an error in the final concentration result. This was
19
exemplified by a recent report in which a measurement bias was only removed after
20
correcting the partition volume to the self-measured value.28 The partition volume in dPCR is
21
often less than 1 nL, and accurate determination of this volume is challenging. In this
22
comparison study, some participants measured the partition volume together with its
23
associated uncertainty, while other participants used the manufacturer-defined partition
24
volume, which has no stated uncertainty. The information on the partition volume provided
25
by the manufacturers was not sufficient to affirm the accuracy and evaluate the related
26
uncertainty. One participant demonstrated a dramatic difference in the final result: −18% or
ACS Paragon Plus Environment
Page 17 of 27
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
1
+27% of the reference value, depending upon whether the partition volume was as provided
2
by the manufacturer or self-measured, respectively. It is likely that variation in determining
3
the partition volume contributed to the observed variation in the reported results. Therefore,
4
the use of accurately determined (self or given) sample or partition volumes should be
5
considered to ensure accuracy and comparability of enumeration-based quantification
6
methods.
7
For this study, participating laboratories were required to report the results in terms of
8
copies/mg, rather than the more common unit of copies/µL. Glycerol is a common ingredient
9
in PCR master mixes and has a density of 1.261 g/cm3, making the density of the PCR master
10
mix larger than the density of the DNA solution. Depending on whether laboratories mixed
11
the DNA sample and dPCR mix gravimetrically or volumetrically, it was necessary for the
12
laboratory to account for the difference in density between these 2 solutions when reporting
13
the results in copies/mg. During FCM counting, the density of the final sample solution was
14
1.02 due to the presence of 10% of DMSO, which has a density of 1.10 g/cm3 at room
15
temperature.
16 17
Platform dependence of dPCR results. The 2 main instrument platforms used in this
18
comparison, chip-based dPCR from Fluidigm and droplet dPCR from Bio-Rad, did not show
19
a significant difference in their averaged results, which were 7940 copies/mg vs. 7980
20
copies/mg, respectively. Therefore, the choice of instrument between these 2 platforms could
21
be a minor factor affecting the comparability of dPCR results. However, a large variation in
22
intra-laboratory measurement results was observed using the newer platform from Thermo
23
Fisher Scientific, which may have been due to the large uncertainty in determining the
24
partition volume.
25 26
CONCLUSIONS
27
In this study, the comparability of enumeration-based quantification of plasmid DNA by
28
FCM counting and dPCR was evaluated through an international comparison among national
29
metrology institutes and equivalent organizations. Tight agreement (0.9% deviation from the
30
average value) between the result of FCM counting and the mean of the dPCR results was
ACS Paragon Plus Environment
Analytical Chemistry
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
1
demonstrated for linearized plasmid DNA, which is impressive considering that the
2
measurements were made without the help of calibration standards (absolute quantification).
3
Furthermore, the results of the enumeration methods also closely agreed with the results of 2
4
orthogonal chemical analysis methods (HPLC-IDMS and CE). The RSD of all 4 methods was
5
1.8%. Therefore, all 4 measurement approaches were confidently validated through this
6
comparison, which was unique in terms of the performance of multiple measurements at
7
multiple sites using multiple approaches. Regarding individual dPCR results, however, the
8
dispersion of the results was not insignificant (RSD = 11.8%). The scattering was not fully
9
accounted for by the reported measurement uncertainties, which suggests the need for further
10
improvement and investigation. As expected, dPCR underestimated the concentration of the
11
circular plasmid, although a possible resolution for this problem was suggested. The
12
importance of accurately determining sample volumes and partition volumes in enumeration-
13
based quantification was also addressed in this study.
14
The results of this study demonstrated a rational basis for the direct counting of nucleic acid
15
(plasmid) copy number, validated by orthogonal approaches, strongly supported the validity
16
of these enumeration approaches for absolute DNA quantification, and allow for claims of SI
17
traceability to 1 through counting.
18 19
Acknowledgements
20
The cost for the preparation, characterization, and delivery of the test samples of this
21
international comparison was fully covered by the KRISS internal research fund,
22
Establishment of measurement standards for quality of life.
23 24
Conflict of Interest Disclosure
25
The authors declare no competing financial interest’.
26 27 28
ACS Paragon Plus Environment
Page 18 of 27
Page 19 of 27
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
1
References
2
(1) Yang, I.; Han, M.-S.; Yim, Y.-H.; Hwang, E.; Park, S.-R. Anal. Biochem. 2004, 335, 150-
3
161.
4 5
(2) Holden, M. J.; Rabb, S. A.; Tewari, Y. B.; Winchester, M. R. Anal. Chem. 2007, 79,
6
1536-1541.
7 8
(3) O'Connor, G.; Dawson, C.; Woolford, A.; Webb, K. S.; Catterick, T. Anal. Chem. 2002, 74,
9
3670-3676.
10 11
(4) Donald, C. E.; Stokes, P.; O'Connor, G.; Woolford, A. J. J. Chromatogr. B Anal. Technol.
12
Biomed. Life Sci. 2005, 817, 173-182.
13 14
(5) Hong, N.-S.; Shi, L.-H.; Jeong, J.-S.; Yang, I.; Kim, S.-K.; Park, S.-R. Anal. Bioanal.
15
Chem. 2011, 400, 2131-2140.
16 17
(6) Yoo, H.-B.; Lim, H.-M.; Yang, I.; Kim, S.-K.; Park, S.-R. J. Biophys. Chem. 2011, 2,
18
102-111.
19 20
(7) Lim, H.-M.; Yoo, H.-B.; Hong, N.-S.; Yang, I.; Han, M.-S.; Park, S.-R. Metrologia 2009,
21
46, 375-387.
22 23
(8) Yoo, H.-B.; Oh, D.; Song, J.-Y.; Kawaharasaki, M.; Hwang, J.; Yang, I.; Park, S.-R.
24
Metrologia 2014, 51, 491-502.
25 26
(9) Sykes, P. J.; Neoh, S. H.; Brisco, M. J.; Hughes, E.; Condon, J.; Morley, A. A.
27
BioTechniques 1992, 13, 444–449.
28 29
(10) Pohl, G.; Shih, I. M. Expert Rev Mol. Diagn. 2004, 4, 41–47.
30 31
(11) Pekin, D.; Skhiri, Y.; Baret, J. C.; Le-Corre, D.; Mazutis, L.; Ben-Salem, C.; Millot, F.;
32
El Harrak, A.; Hutchison, J. B.; Larson, J. W.; Link, D. R.; Laurent-Puig, P.; Griffiths, A. D.;
ACS Paragon Plus Environment
Analytical Chemistry
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
1
Taly, V. Lab Chip 2011, 11, 2156–2166.
2 3
(12) Hindson, B. J.; et al. Anal. Chem. 2011, 83, 8604–8610.
4 5
(13) Baker, M. Nature Methods 2012, 9, 541–544.
6 7
(14) Kim, T. G.; Jeong S.-Y.; Cho, K.-S. Biotechnology Reports 2014, 4, 1-4.
8 9
(15) Drandi, D.; et al. J. Molecular Diagnostics 2015, 17, 652-660.
10 11 12
(16) Coudray-Meunier, C.; Fraisse, A.; Martin-Latil, S.; Guillier, L.; Fach, P.; Pérelle, S. Int. J. Food Microbiol. 2015, 201, 17-26.
13 14
(17) Köppel, R.; Bucher, T. Eur. Food Res. Technol. 2015, 241, 427-439.
15 16
(18) Köppel, R.; Bucher, T.; Waiblinger, H.-U. Eur. Food Res. Technol. 2015, 241, 521-527.
17 18 19
(19) Taylor, S. C.; Carbonneau, J.; Shelton, D. N.; Bovin, G. J. Virol. Methods 2015, 224, 5866.
20 21
(20) Bhat, S.; Curach, N.; Mostyn, T.; Bains, G. S.; Griffiths, K. R.; Emslie, K. R. Anal.
22
Chem. 2010, 82, 7185–7192.
23 24
(21) Sanders, R.; Huggett, J. F.; Bushell, C. A.; Cowen, S.; Scott, D. J.; Foy, C. A. Anal.
25
Chem. 2011, 83, 6474–6484.
26 27
(22) Whale, A. S.; Huggett, J. F.; Cowen, S.; Speirs, V.; Shaw, J.; Ellison, S.; Foy, C. A.;
28
Scott, D. J. Nucleic Acids Res. 2012, 40, e82.
29 30
(23) Sanders, R.; Mason, D. J.; Foy, C. A.; Huggett, J. F. PLoS One 2013, 8, e75296.
31
ACS Paragon Plus Environment
Page 20 of 27
Page 21 of 27
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
1
(24) Burke, D. G.; Dong, L.; Bhat, S.; Forbes-Smith, M.; Fu, S.; Pinheiro, L.; Wang, J.;
2
Emslie, K. R. Anal. Chem. 2013, 85, 1657–1664.
3 4
(25) Bhat, S.; Herrmann, J.; Armishaw, P.; Corbisier, P.; Emslie, K. R. Anal. Bioanal. Chem.
5
2009, 394, 457-467.
6 7
(26) Devonshire, A. S.; Honeyborne, I.; Gutteridge, A.; Whale, A. S.; Nixon, G.; Wilson, P.;
8
Jones, G.; McHugh, T. D.; Foy, C. A.; Huggett, J. F. Anal. Chem. 2015, 87, 3706-3713.
9 10
(27) Bhat, S.; McLaughlin, J. L.; Emslie, K. R. Analyst 2011, 136, 724-732.
11 12
(28) Corbisier, P.; Pinheiro, L.; Mazoua, S.; Kortekaas, A. M.; Chung, P. Y.; Gerganova, T.;
13
Roebben, G.; Emons, H.; Emslie, K. Anal. Bioanal. Chem. 2015, 407, 1831-1840.
14 15 16
(29) Féllix-Urquidez, D.; Pérez-UrquizaM.; Valdez Torres, J. B.; León-Félix, J.; GarciaEstrada, R.; Acatzi-Silva, A. Anal. Chem. 2016, 88, 812-819.
17 18
(30) Pavšič, J.; Žel, J.; Milavec, M. Bioanal. Chem. 2015, 408, 107-121.
19 20
(31) Kline, M. C.; Romsos, E. L.; Duewer, D. L. Anal. Chem. 2016, 88, 2132-2139.
21 22
(32) Duewer, D. L.; Kline, M. C.; Romsos, E. L. Anal. Bioanal. Chem. 2015, 407, 9061-9069.
23 24
(33) Dong, L.; Yoo, H.-B.; Wang, J.; and Park, S.-R. Scientific Reports 2016, 6, 24230.
25 26
(34) Dong, L.; Meng, Y.; Sui, Z.; Wang, J.; Wu, L.; Fu, B. Scientific Reports 2015, 5, 13174.
27 28
(35) Nivedita, Majumdar.; Thomas, Wessel.; and Jeffrey, Marks. PLoS One 2015, 10,
29
e0118833
30 31
(36) Pinheiro, L. B.; et al. Anal. Chem. 2012, 84, 1003-1011.
32 33
(37) White, H.; et al. Leukemia. 2015, 29, 369-376.
ACS Paragon Plus Environment
Analytical Chemistry
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
1 2
(38) Flatschart, R. B.; Almeida, D. O.; Heinemann, M. B.; Medeiros, M. N.; Granjeiro, J. M.;
3
Folgueras-Flatschart, A. V. J. Physics. Conference Series 2015, 575, 12038-12041.
4 5
(39) Whale, A. S.; Cowen, S.; Foy, C. A.; Huggett, J. F. PLoS One 2013, 8, e58177.
6 7
(40) White III, R. A.; Blainey, P. C.; Fan, C. H.; Quake, S. R. BMC Genomics 2009, 10, 116.
8 9 10
(41) Hindson, C. M.; Chevillet, J. R.; Gallichotte, E. N.; Ruf, I. K.; Hindson, B. J.; Vessella, R. L.; Tewari, M. Nature Methods 2013, 10, 1003-1005.
11 12
ACS Paragon Plus Environment
Page 22 of 27
Page 23 of 27
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
1
Analytical Chemistry
For TOC only
2 3 4 5 6
ACS Paragon Plus Environment
Analytical Chemistry
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
132x78mm (96 x 96 DPI)
ACS Paragon Plus Environment
Page 24 of 27
Page 25 of 27
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
132x53mm (96 x 96 DPI)
ACS Paragon Plus Environment
Analytical Chemistry
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
63x48mm (96 x 96 DPI)
ACS Paragon Plus Environment
Page 26 of 27
Page 27 of 27
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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
119x94mm (96 x 96 DPI)
ACS Paragon Plus Environment