Analytical Biochemistry 451 (2014) 1824

Contents lists available at ScienceDirect

Analytical Biochemistry
journal homepage: www.elsevier.com/locate/yabio

Comparison of three common DNA concentration measurement

methods
Xiaofei Li 1, Yuhua Wu 1, Li Zhang, Yinglong Cao, Yunjing Li, Jun Li, Li Zhu, Gang Wu
Key Laboratory of Oil Crop Biology of the Ministry of Agriculture, Oil Crops Research Institute, Chinese Academy of Agricultural Sciences, Wuhan 430062, Peoples Republic of China

a r t i c l e

i n f o

Article history:
Received 3 December 2013
Received in revised form 23 January 2014
Accepted 24 January 2014
Available online 1 February 2014
Keywords:
DNA concentration measurement
UV absorption
PicoGreen dye
Diphenylamine reaction
Comparison

a b s t r a c t
Accurate measurement of DNA concentration is important for DNA-based biological applications. DNA
concentration is usually determined by the ultraviolet (UV) absorption, uorescence staining, and diphenylamine reaction methods. However, the best method for quality assurance of measurements is
unknown. Here, we comprehensively compared these methods using different types of samples. We
found that all three methods accurately determined the concentrations of high-purity DNA solutions.
After digestion of DNA samples, concentration measurements revealed that the PicoGreen dye method
was very sensitive to the degradation of DNA. The three methods displayed different anti-jamming ability
when contaminants such as transfer RNA (tRNA), protein, and organic chemicals were included in DNA
solutions. The diphenylamine reaction method gave the highest accuracy, with an average error of
approximately 10% between measured and true values. The PicoGreen dye method was inuenced by
tRNA and protein, and the UV absorption method was susceptible to all kinds of impurities. Overall,
the diphenylamine reaction method gave the most accurate results when DNA was mixed with contaminants, the PicoGreen dye method was most suitable for degraded DNA samples or DNA extracted from
processed products, and the UV absorbance method was best for evaluating the impurities in DNA
solutions.
2014 Elsevier Inc. All rights reserved.

Reliable methods to determine the precise concentration of

DNA are necessary for numerous biological applications, ranging
from traditional molecular biological manipulations, such as
restriction digest analysis, Southern blotting, and polymerase
chain reaction (PCR),2 to diagnostic techniques, including quantication of genetically modied organism (GMO) content of samples,
detection of DNA contamination in drug preparations produced from
recombinant organisms, and medical diagnosis of virus and cancer
[14].
There are three methods commonly used to determine DNA
concentration: ultraviolet (UV) absorbance, uorescence, and
diphenylamine reaction. Currently, the most common technique
to determine DNA concentration is the measurement of
absorbance of UV light at 260 nm [5]. The principle of the UV
Corresponding author. Fax: +86 27 86711573.
E-mail address: wugang@caas.cn (G. Wu).
These two authors contributed equally to this study.
2
Abbreviations used: PCR, polymerase chain reaction; GMO, genetically modied
organism; UV, ultraviolet; dsDNA, double-stranded DNA; EB, ethidium bromide; CRM,
certied reference material; CTAB, hexadecyl trimethyl ammonium bromide; SDS,
sodium dodecyl sulfate; tRNA, transfer RNA; BSA, bovine serum albumin; OD, optical
density; ANOVA, analysis of variance.
1

http://dx.doi.org/10.1016/j.ab.2014.01.016
0003-2697/ 2014 Elsevier Inc. All rights reserved.

absorbance method is that nucleic acids (DNA or RNA) contain

conjugated double bonds in their purine and pyrimidine rings
that have a specic absorption peak at 260 nm, the intensity of
which is proportional to the concentration of nucleic acid [6].
These physical characteristics allow the nucleic acid concentration
of a sample to be determined. The uorescence method uses a
uorometer and DNA-binding uorescent dye that specically
binds to double-stranded DNA (dsDNA) and then uoresces under
excitation [7]. Within a certain range, the uorescence intensity is
proportional to the amount of binding dye, so the DNA concentration of samples can be calculated by comparison with a standard
curve generated from reference solutions of known DNA concentration [7,8]. The principle of the diphenylamine reaction method
is that the cleavage of the glycosidic bond between a purine base
and deoxyribose generates purine base, 2-deoxyribose, and a
deoxy-pyrimidine nucleotide when DNA is heated under acidic
conditions. Then, 2-deoxyribose is dehydrated to form the
x-hydroxy-c-keto pentose, which can react with diphenylamine
to produce a blue substance with an absorption maximum at
595 nm [9,10]. Later, the conjugated polymer PFPBx was used to
determine DNA concentration but is currently not widely available [11].

Evaluation of DNA quantitation methods / X. Li et al. / Anal. Biochem. 451 (2014) 1824

Over recent decades, uorescence methods have used various

dyes to measure the amount of dsDNA in a sample, including ethidium bromide (EB), Hoechst (bisbenzimide), ethidium homodimer1, and cyanine dyes (YOYO-1, YO-RPO-1, TOTO-1, and PicoGreen)
[8,1215]. These dyes possess different advantages and limitations.
In addition to binding dsDNA, EB and ethidium homodimer-1 dyes
also bind to both RNA and ssDNA with considerable background
uorescence and have a limited dynamic range [12,13]. Hoechst
33258 is an AT-selective binding dye that requires two different
concentrations of salt solution to detect dsDNA in the presence of
both RNA and ssDNA contaminants [14,16]. Cyanine dyes YOYO1, YO-PRO-1, and TOTO-1 also show marked uorescence enhancement on binding to RNA and ssDNA [15]. Compared with other
dyes, PicoGreen is more sensitive for low-concentration DNA solutions and shows the greatest dynamic range for selective quantitation of dsDNA in solution with little background uorescence [7,8].
Currently, DNA-based real-time PCR technology is the most
powerful method for implementation of the GMO labeling regulations enforced in the European Union (EU) and other countries
[17]. The European Commission recommends expressing the GM
content of a material as the percentage of the number of GM target
DNA sequences per target taxon-specic sequence [18]. Accurate
determination of DNA concentration is required through the entire
GMO detection process, including GMO certied reference material
(CRM) production, CRM application, establishment of detection
methods, and routine GMO quantitative detection [19,20]. Standard
ISO 21571 recommends the UV absorbance method as a routine
method to determine the quantity of DNA in solutions [5]. During
the preparation of matrix reference materials, the three methods
described above have all been used to measure DNA amount [21
23]. In some certication reports of EU matrix CRMs, both the UV
absorption and uorescence staining methods were used and it
was noted that the determined concentrations sometimes differed
considerably from each other [22,23]. In most articles on GMO
detection, researchers use UV spectrophotometric analysis without
considering that the contaminants in DNA solution, such as RNA,
protein, and guanidine, will contribute to the total absorbance at
260 nm, resulting in overestimation of DNA yield [24]. In routine
testing, the quality of DNA extracted from foodstuffs or processed
products is poorer than that used in standard material preparation
and the development of detection methods [25,26]. In addition,
quick, inexpensive DNA quantication methods such as UV absorption that are often used as the principal methods for evaluating the
quantity and quality of nucleic acids are prone to give anomalous
results in subsequent molecular biological research [7].
The previous studies have conrmed that the uorescent method using PicoGreen dye is more sensitive than the absorbance one,
particularly for low-concentration DNA solutions, and the use of
DNA-binding dyes allows more specic measurement of dsDNA
than do spectrophotometric methods [27,28]. However, these
three common DNA concentration determination methods have
not been compared in a comprehensive manner in the literature.
The current research was designed to evaluate and compare the
UV absorption, uorescence staining, and diphenylamine reaction
methods to allow a relatively precise method for determination
of DNA concentration in a wide variety of biological applications
to be recommended.

Materials and methods

Materials
Seeds of rapeseed (Brassica napus), maize (Zea mays), cotton
(Gossypium hirsutum), soybean (Glycine max), and rice (Oryza sativa) were collected by our own laboratory. Part of seeds were sowed

19

and cultivated at the appropriate temperatures in Sanyo MLR351H Versatile Environmental Test Chambers (Panasonic, Osaka,
Japan). One month after sowing, the fresh leaves were collected
for DNA extraction.
Salmon sperm DNA (Sigma, St. Louis, MO, USA) was used to prepare a high-quality DNA solution with a concentration of 100 ng/
ll. The chemicals hexadecyl trimethyl ammonium bromide
(CTAB), sodium dodecyl sulfate (SDS), transfer RNA (tRNA), bovine
serum albumin (BSA), phenol, chloroformisoamyl alcohol, ethanol, isopropanol, and NaCl were used as DNA contaminants to evaluate the anti-jamming capacity of different DNA concentration
estimation methods.
DNA extraction
All seeds and leaf samples were ground to ne powders in liquid nitrogen. The powders were freeze-dried in a lyophilizer
(FreeZone, Labconco, Kansas City, MO, USA). Seed powder
(100 mg) and leaf powder (50 mg) for each crop were weighed
with a calibrated balance before genomic DNA extraction. Four
commercial DNA extraction kitsDNeasy Plant Mini Kit (Qiagen,
Hilden, Germany), Wizard Genomic DNA Purication Kit (Promega,
Madison, WI, USA), DNA Extraction Kit for GMO Detection version
2.0 (Takara, Shiga, Japan), and GeneJET Plant Genomic DNA Purication Mini Kit (Fermentas, Vilnius, Lithuania)were used to extract genomic DNA from both seed and leaf powders according to
the instructions included with the kits.
DNA concentration analysis
The concentration of DNA in different solutions was measured
using the three methods of UV absorption, uorescent staining,
and diphenylamine reaction. To ensure accuracy, three replicate
readings of each sample were taken.
The UV absorbance was measured on a Nanodrop 2000 microvolume spectrophotometer (Thermo, Wilmington, DE, USA)
according to the manufacturers instructions. Based on the manufacturers guidelines, 1 TE buffer was rst used to zero the instrument (set reference), and then a 1.0- to 1.5-ll drop was taken
from the sample for measurement. Concentration values and
260/280 ratios were collected with the original instrument supporting software NanoDrop 2000/2000C installation version 1.4.1.
Fluorescent measurement was performed using a VersaFluor
uorometer (Bio-Rad, Hercules, CA, USA) with a 480-nm excitation
lter and a 520-nm emission lter according to the manufacturers
instructions. DNA standards (k DNA) and DNA samples were
stained using a Quant-iT PicoGreen dsDNA Assay Kit (Invitrogen,
Eugene, OR, USA). Before using PicoGreen dye to stain DNA, the
concentrated Me2SO solution containing PicoGreen dye was diluted 200-fold using 1 TE buffer to prepare 2 PicoGreen working
solution. The 50-fold diluted standard sample (2 ng/ll k DNA) and
experimental samples were mixed with PicoGreen dye working
solution with a volume ratio of 1:1 to prepare samples for testing
containing 1 PicoGreen dye. Here, 1 PicoGreen dye containing
no DNA was used to zero the instrument, and a standard sample
(1 ng/ll) was used to calibrate the reading of the instrument to
100 ng/ll. Then, the concentrations of experimental samples were
measured. Concentration values were recorded by the uorometers built-in software.
The diphenylamine reaction method was performed according
to the method of Abraham and coworkers [29] with minor modication. Calf thymus DNA (Sigma) was dissolved in NaOH (0.01 M)
to prepare DNA solutions (200 lg/ml). Six standard samples representing DNA amounts corresponding to 0 ml, 0.1 ml (20 lg), 0.2 ml
(40 lg), 0.3 ml (60 lg), 0.4 ml (80 lg), and 0.5 ml (100 lg) of DNA
solution were added to diphenylamine solution (1 ml) and then re-

20

Evaluation of DNA quantitation methods / X. Li et al. / Anal. Biochem. 451 (2014) 1824

acted for 1 h at 60 C in a reaction volume of 1.5 ml. Each reaction

was repeated three times. The optical density (OD) at 595 nm of
the standard samples was monitored using a spectrophotometer
(Lambda 25) with the software UV WINLAB version 2.85.04 (PerkinElmer, Ames, IA, USA). A standard curve was constructed by plotting DNA amounts against OD595 values using Excel software
(Microsoft, Seattle, WA, USA). Experimental samples were treated
the same as the standard samples, and their OD595 values were
measured. The amounts of DNA in experimental samples were calculated based on the OD595 values and constructed standard
curves.
Digestion of DNA
Extracted high-molecular-weight DNA samples were digested
with the cryonase cold-active nuclease (Takara), which can completely digest all types of DNA and RNA. Cryonase cold-active
nuclease (20 U/ll) was added to genomic DNA (1 ml) and then reacted for 12 h at 30 C. The digested DNA samples were used to
evaluate the background signals of different DNA concentration
determination methods.
Results and discussion
Concentration measurement of DNA samples extracted with kits
Four commercial plant genomic DNA extraction kits from different companies were used to extract DNA from the seeds and fresh
leaves of rapeseed, maize, cotton, soybean, and rice. The concentrations of solutions containing the extracted DNA were measured
using the uorescence, UV absorbance, and diphenylamine reaction methods. All experiments were repeated three times. The
mean values of three measurements are presented in Table 1.
For the same DNA sample extracted using the same method, the
DNA concentrations estimated by the different measurement
methods differed considerably. To judge whether differences exist
in the different DNA concentration measurement methods, a oneway analysis of variance (ANOVA) was performed. This showed
that the values obtained by different methods were signicantly
different from each other (P < 0.01) (Table 2). Further analysis
showed that the data obtained using the PicoGreen dye method
was quite different from those for UV spectrophotometric analysis
and the diphenylamine reaction method (P < 0.01). The values
measured using the UV absorbance and diphenylamine reaction
methods have a certain degree of similarity (P = 0.041), but the difference was still statistically signicant (P < 0.05). Therefore, it was
difcult to determine which method is the most accurate from the
above results.
During DNA extraction, each DNA extraction kit left some residues of chemical reagents, RNA, and protein in the DNA eluate [30],
which would interfere with the measurement of DNA concentration to different degrees. It was speculated that the dramatic differences in the estimated concentration of the same sample were
derived not only from the methods themselves but also from the
interference from impurities in the DNA samples.

used to measure the DNA concentrations of the digested DNA solutions. The values obtained are listed in Table 3. These results show
that for the nuclease-digested samples, besides the samples extracted with the Takara kit, the data obtained by the PicoGreen
dye method was closest to the theoretical value of zero, followed
by the diphenylamine reaction and UV spectrophotometric
methods.
Digestion by cryonase cold-active nuclease destroyed the double-stranded structure of DNA molecules, producing 50 -phosphorylated dinucleotide, trinucleotide, and oligonucleotide products that
failed to bind to PicoGreen dye and so contributed less background
uorescence to the PicoGreen signal under excitation. Therefore,
the quantitative values obtained using the PicoGreen dye method
after digestion of the DNA solutions were close to zero. However,
the PicoGreen dye method measurements also revealed that the
concentration of digested DNA solutions extracted with the Takara
DNA Extraction Kit for GMO Detection version 2.0 did not decrease
to zero (Table 3). During DNA extraction, the Takara kit uses silicon
beads rather than column-based silica as an afnity matrix for
DNA, so after repeated centrifugation the nal DNA elution would
contain a small amount of residual absorbent material. The residual silicon beads might contribute background uorescence during
detection, resulting in measurement errors. When this was taken
into account, the data obtained using the PicoGreen method after
DNA digestion were close to the background.
The values measured using the diphenylamine reaction method
were not very close to zero (Table 3), probably because the di-, tri-,
and oligonucleotides released after digestion by the nuclease can
be further hydrolyzed into purine base, 2-deoxyribose, and
deoxy-pyrimidine nucleotides when heated under acidic conditions. The dehydrated product of 2-deoxyribose can react with
diphenylamine to generate a blue product that absorbs at a specic
wavelength of 595 nm. Therefore, the diphenylamine reaction
method was not able to accurately determine the concentration
of degraded DNA samples.
As shown in Table 3, the UV spectrophotometric analysis gave
the highest values, which were even higher than those measured
before digestion. In the UV absorbance method, DNA concentration
is estimated by measuring the absorbance at 260 nm. According to
the relationship that the absorbance value of 1 at 260 nm is equivalent to 50 lg/ml pure dsDNA, 40 lg/ml single-stranded DNA, and
33 lg/ml oligonucleotides [31], the absorbance at 260 nm would
increase with the degradation of DNA into oligonucleotides. Moreover, the added enzyme also absorbs UV light and so interferes
with the measurements [31]. Therefore, DNA degradation caused
the values measured by the UV absorbance method to increase
rather than decrease (Table 3).
DNA degradation is inevitable in DNA extraction and storage
[26]. For partially degraded DNA, such as commonly used samples
that have been stored at 4 C, repeatedly frozen/thawed, or extracted from the highly processed products, the PicoGreen dye
method is more suitable for accurate estimation of the amount of
dsDNA in degraded DNA samples than UV absorption or diphenylamine reaction.

Measurement of the concentration of high-quality DNA samples

Measurement of DNA digested by cryonase cold-active nuclease
An optimal DNA quantitation method should have sufcient
sensitivity for dsDNA molecules. When all of the DNA molecules
in experimental samples were removed, the determined value of
DNA concentration should be close to zero. To obtain background
data, cryonase cold-active nuclease was added to all 20 of the
DNA solutions extracted from seed samples to fully digest the
DNA molecules. Then, the three methods described above were

High-purity solid salmon sperm DNA was used to prepare a

DNA solution with a concentration of 100 ng/ll. The concentration
of this sample was determined by the three methods introduced
above, giving estimated values of 99.01 ng/ll using the PicoGreen
dye method, 99.94 ng/ll using the UV absorbance method, and
106.87 ng/ll using the diphenylamine reaction method (Table 4).
Minor differences in estimated values were observed among the
three methods.

21

Evaluation of DNA quantitation methods / X. Li et al. / Anal. Biochem. 451 (2014) 1824
Table 1
DNA concentration determination results obtained using three different methods.
Sample type

DNA extraction kit

DNA concentration determination method

a

Mean

SD

Seeds of rapeseed

Takara
Fermentas
Promega
Qiagen

88.14
10.82
112.90
36.52

77.13
16.37
331.53
49.93

69.15
19.01
135.35
35.84

78.14
15.40
193.26
40.76

9.53
4.18
120.27
7.95

12.20
27.15
62.23
19.50

Seeds of maize

Takara
Fermentas
Promega
Qiagen

113.20
117.98
122.66
31.34

155.53
156.27
788.60
31.13

151.93
153.02
612.30
18.09

140.22
142.42
507.85
26.85

23.47
21.23
345.04
7.59

16.74
14.91
67.94
28.26

Seeds of cotton

Takara
Fermentas
Promega
Qiagen

19.80
35.79
100.88
86.24

65.03
30.87
333.67
176.60

116.67
49.83
110.12
78.91

67.17
38.83
181.56
113.92

48.47
9.84
131.81
54.41

72.17
25.34
72.60
47.76

Seeds of soybean

Takara
Fermentas
Promega
Qiagen

110.51
13.94
19.67
30.86

229.73
52.83
616.27
705.47

137.93
66.05
109.29
14.63

159.39
44.27
248.41
250.32

62.44
27.09
321.71
394.25

39.18
61.19
129.51
157.50

Seeds of rice

Takara
Fermentas
Promega
Qiagen

59.02
32.20
19.78
28.80

203.37
29.27
130.53
197.23

79.30
36.01
93.74
24.20

113.90
32.49
81.35
83.41

78.14
3.38
56.41
98.60

68.61
10.40
69.34
118.21

Leaves of rapeseed

Takara
Fermentas
Promega
Qiagen

104.61
35.79
116.27
78.05

103.80
47.37
271.00
76.93

198.09
143.88
265.86
111.49

135.50
75.68
217.71
88.82

54.21
59.35
87.89
19.63

40.01
78.42
40.37
22.11

Leaves of maize

Takara
Fermentas
Promega
Qiagen

113.20
174.80
156.81
58.37

187.23
187.40
560.87
58.20

260.95
248.01
843.25
85.11

187.13
203.40
520.31
67.22

73.87
39.14
345.01
15.49

39.48
19.24
66.31
23.04

Leaves of cotton

Takara
Fermentas
Promega
Qiagen

77.73
39.05
135.66
97.63

85.83
69.10
469.47
198.27

235.41
225.33
172.96
156.21

132.99
111.16
259.36
150.70

88.79
100.01
182.91
50.54

66.76
89.97
70.52
33.54

Leaves of soybean

Takara
Fermentas
Promega
Qiagen

97.40
55.45
47.80
65.88

232.00
78.13
574.13
576.43

262.84
109.05
143.48
87.62

197.41
80.88
255.14
243.31

87.97
26.90
280.37
288.70

44.56
33.26
109.89
118.65

Leaves of rice

Takara
Fermentas
Promega
Qiagen

81.77
67.90
58.15
45.02

242.47
53.83
251.90
202.83

223.77
172.87
172.92
90.84

182.67
98.20
160.99
112.90

87.88
65.05
97.42
81.18

48.11
66.24
60.51
71.91

RSD (%)

Note. Methods: a, PicoGreen dye; b, UV absorption; c, diphenylamine reaction. DNA was extracted from the seeds and leaves of various plants using four different DNA
extraction kits. SD, standard deviation; RSD, relative standard deviation.

Table 2
Comparison of DNA concentration determination results obtained using different DNA measurement methods through one-way ANOVA.
V2

Spectrophotometry
Diphenylamine
PicoGreen dye

V1

41722.68
23417.90
1745.61

Spectrophotometry
41722.68

Diphenylamine
23417.90

PicoGreen dye
1745.61

N.S.
N.S.
N.S.

1.782a
N.S.
N.S.

23.901b
13.415b
N.S.

Note. F = V2/V1. N.S., not signicant.

a
Signicant at the 0.05 probability level with the critical value F0.05 = 1.70.
b
Signicant at the 0.01 probability level with the critical value F0.01 = 2.14.

The solid salmon sperm DNA samples also contain water, salt,
and other impurities, and the calibrated balance and volumetric
asks used were not sufcient to guarantee that the nominal concentration of each prepared sample is its true concentration. Both
the uorescence and diphenylamine reaction methods used as references calibrated DNA solution that had been stored at 4 C for a
period of time before use. During storage, DNA partially degrades,

resulting in a difference between the real concentration of a standard solution and its nominal value. The concentration of samples
was calculated by comparison with standard solutions, so it was
difcult to determine whether the minor differences in the measurement data originated from the accuracy of the method itself.
In general, the obtained concentration values of high-quality salmon sperm DNA solutions were similar across the three methods

22

Evaluation of DNA quantitation methods / X. Li et al. / Anal. Biochem. 451 (2014) 1824

Table 3
DNA concentration determination results after DNA was digested by cryonase cold-active nuclease.
Sample

DNA extraction kit

DNA concentration determination method

Seeds of rapeseed

Takara
Fermentas
Promega
Qiagen

24.61
1.31
0.24
0.05

95.90
69.20
1068.87
105.97

7.05
10.72
7.45
14.63

Seeds of maize

Takara
Fermentas
Promega
Qiagen

27.27
0.07
0.05
0.02

167.73
314.53
1109.77
72.67

6.53
1.30
46.93
42.30

Seeds of cotton

Takara
Fermentas
Promega
Qiagen

19.21
0.07
0.22
0.07

58.43
105.63
364.00
243.97

12.21
24.16
44.07
22.63

Seeds of soybean

Takara
Fermentas
Promega
Qiagen

27.04
0.02
0.98
0.09

253.20
184.53
664.93
732.57

47.99
27.63
55.43
15.95

Seeds of rice

Takara
Fermentas
Promega
Qiagen

18.97
0.03
0.11
0.03
6.02
10.48
109.82

212.40
143.47
156.50
212.43
316.84
319.77
102252.02

14.00
8.49
16.16
11.87
21.88
16.46
271.06

PicoGreen

Mean
SD
Variance

UV

Diphenylamine

Note. SD, standard deviation.

Table 4
DNA concentration determination results obtained for pure salmon sperm DNA and DNA mixed with chemical interferers.
No.

Chemical

Theoretical value (ng/ll)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22

Pure salmon sperm DNA

1% CTAB + DNA
0.25% CTAB + DNA
0.05% CTAB + DNA
0.01% CTAB + DNA
20% SDS + DNA
5% SDS + DNA
1% SDS + DNA
0.2% SDS + DNA
100 ng/ll tRNA + DNA
25 ng/ll tRNA + DNA
5 ng/ll tRNA + DNA
1 ng/ll tRNA + DNA
10000 ng/ll BSA + DNA
1000 ng/ll BSA + DNA
500 ng/ll BSA + DNA
100 ng/ll BSA + DNA
Phenol + DNA
Chloroformisoamyl alcohol + DNA
Ethanol + DNA
Isopropanol + DNA
1 mol/L NaCl + DNA

100
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50

PicoGreen dye

UV absorption

Diphenylamine

Tested value

Bias (%)

Tested value

Bias (%)

Tested value

Bias (%)

99.01
0.10
0.05
0.93
11.10
0.22
7.13
17.40
19.63
19.36
15.53
15.25
16.20
18.41
16.21
17.43
17.55
37.33
42.50
12.07
16.14
15.54

0.99
99.81
99.89
98.13
77.79
99.57
85.73
65.20
60.74
61.27
68.93
69.49
67.59
63.18
67.59
65.15
64.90
25.35
15.00
75.86
67.72
68.91

99.94
58.37
31.80
33.67
50.80
61.57
54.83
51.50
49.50
118.27
68.30
48.47
50.57
142.80
62.53
52.83
49.40
21218.83
80.87
71.77
69.53
44.40

0.06
16.73
36.40
32.67
1.60
23.13
9.67
3.00
1.00
136.53
36.60
3.07
1.13
185.60
25.07
5.67
1.20
42337.67
61.73
43.53
39.07
11.20

106.87
91.62
59.84
51.58
54.68
55.31
49.25
51.62
50.93
44.48
46.98
42.94
45.09
49.80
47.82
59.14
49.95
54.93
86.26
51.07
47.65
39.86

6.87
83.24
19.68
3.16
9.35
10.62
1.50
3.25
1.86
11.03
6.04
14.11
9.82
0.40
4.36
18.28
0.11
9.87
72.51
2.15
4.71
20.27

and very close to the excepted values. Therefore, the quantitative

measurement of the concentration of salmon sperm DNA illustrates that all three of these methods can accurately estimate the
concentration of high-purity DNA and give uniformly accurate
values.
Interference of contaminants on DNA concentration measurements
Table 1 reveals that the measurement results obtained using the
different methods for the same sample were discordant, and the
variance across the three methods was far larger than that found
using the three methods to determine the concentrations of pure
DNA solutions. The extracted DNA solutions contain not only

DNA but also contaminants such as RNA, polypeptides, and residual extraction reagents. Systematic analysis of the inuences of
contaminants on different DNA determination methods is important for selecting an appropriate DNA quantitation method. A variety of potential interfering substances were tested in this study,
including residual compounds from the samples such as protein
(BSA) and RNA (yeast tRNA), extraction reagents commonly used
in a variety of plant genomic DNA extraction methods such as lytic
agent (CTAB and SDS), nucleic acid purication reagents (phenol
and chloroformisoamyl alcohol), and nucleic acid precipitant reagents (ethanol, isopropanol, and sodium chloride).
These organic compounds and extraction chemicals were dissolved into deionized water and serially diluted to several concen-

23

Evaluation of DNA quantitation methods / X. Li et al. / Anal. Biochem. 451 (2014) 1824
Table 5
Background signal with the presence of chemical interferers usually remaining in DNA extractions.
No.

Chemical

Concentration

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21

CTAB

1% (W/V)
0.25% (W/V)
0.05% (W/V)
0.01% (W/V)
20% (W/V)
5% (W/V)
1% (W/V)
0.20% (W/V)
100 ng/ll
25 ng/ll
5 ng/ll
1 ng/ll
10000 ng/ll
1000 ng/ll
500 ng/ll
100 ng/ll
Analytically pure
Analytically pure
Analytically pure
Analytically pure
1 mol/L

SDS

tRNA

BSA

Phenol
Chloroformisoamyl alcohol
Ethanol
Isopropanol
NaCl

trations that may exist in extracted DNA. The three DNA measurement methods were used to determine the background signal of
these possible contaminants. Each test was conducted three times.
The results are listed in Table 5. A previous study reported that the
contaminants sodium chloride and SDS weakened the PicoGreen
signal, whereas those of organic reagents such as phenol, ethanol,
and chloroform strengthened the PicoGreen signal [8]. In this
study, the PicoGreen dye method showed very weak background
signal to all of the contaminants except for 100 and 25 ng/ll tRNA,
giving measured values very close to zero. For the 100-ng/ll tRNA
solution, partially homologous tRNA may anneal to form doublestranded structures that can bind to PicoGreen dye, thereby producing a background signal. The diphenylamine reaction method
was obviously interfered with by chloroformisoamyl alcohol
and ethanol, whereas protein and RNA exerted relatively small effects compared with those of the other reagents. The results of the
UV absorption method were affected by most of the interfering
compounds and often gave relatively large error. In particular, phenol seriously interfered with the UV absorbance measurements,
increasing values considerably (Table 5). These results are consistent with a previous study nding that RNA, protein, and aromatic
organic compounds such as phenol interfered with the absorbance
measurements by absorbing light in the range of 260 to 280 nm
[7].
The prepared contaminant solutions (Table 5) were mixed with
100 ng/ll salmon sperm DNA solution with a volume ratio of 1:1.
Then, the concentration of the 2-fold diluted DNA solutions containing interfering substances were determined using the three
methods examined in this study. Each measurement was repeated
three times, and the mean values of three replicates are displayed
in Table 4. The diphenylamine reaction test gave the highest accuracy for most samples, with an average error of approximately 10%
between the measured and theoretical true values.
The results obtained using the UV absorbance method were
obviously overestimated in the presence of the aromatic organic
reagent phenol and high concentrations of tRNA and BSA, in line
with previous ndings [7]. For samples mixed with phenol, the
UV method displayed the largest measurement errors, suggesting
that the residual phenol seriously affected accurate quantitation
by the UV absorbance method. For serially diluted contaminants
CTAB, SDS, tRNA, and BSA, the test data were found to become

DNA concentration found by three methods

PicoGreen

UV

Diphenylamine

0.01
0.00
0.01
0.01
0.16
0.11
0.09
0.07
8.22
1.04
0.20
0.11
0.29
0.21
0.08
0.05
0.04
0.21
0.21
0.33
0.30

6.20
2.67
1.70
1.83
25.60
8.43
2.20
1.57
131.97
31.07
7.00
2.00
188.00
24.13
11.67
4.20
21304.83
440.47
7.07
11.87
1.30

0.32
2.89
4.54
5.55
7.37
4.06
0.82
0.31
1.27
0.68
0.55
0.47
11.44
9.14
6.28
4.53
5.32
12.25
24.58
3.68
6.14

closer to the expected values as the contaminant concentration

was reduced (Table 4).
The PicoGreen dye method showed serious measurement errors
in this part of the analysis, and all measured values were dramatically underestimated compared with theoretical values (Table 4).
We conrmed above that for the PicoGreen dye method, the tested
contaminants did not generate background signals. The mode of
binding between PicoGreen reagent and dsDNA is not known. We
speculate that these measurement errors did not arise from the
interaction of disruptors with the PicoGreen uorescent dye but
more likely resulted from the interference of contaminants with
the binding of dsDNA to PicoGreen dye. The test results also
showed that all of the listed contaminants, especially the lytic
agents (CTAB and SDS), considerably inuenced measurements obtained by the PicoGreen dye method. Low concentrations of CTAB,
SDS, tRNA, and BSA showed little inuence in the UV absorption
method but could interfere markedly with the PicoGreen method.
It is very difcult to completely remove RNA and protein in the actual DNA extraction process even after repeated washing and sedimentation steps. Therefore, it can be concluded that the PicoGreen
method is quite sensitive and fragile even though in theory it has
the highest specicity and quantitative accuracy for dsDNA.
The above analysis revealed that the three methods show different anti-jamming capacity to contaminants and give dramatically
different measurement results. Both the PicoGreen dye and UV
absorption methods are vulnerable to the effects of disruptors;
the PicoGreen dye method tends to underestimate DNA concentrations, whereas the UV absorption method tends to overestimate
them. The diphenylamine reaction method has the highest antijamming ability of the methods investigated and can give relatively accurate measurement values. During concentration measurement of DNA samples mixed with contaminants, the
existence of interfering substances should be considered and the
accuracy of measurement results should be carefully evaluated.
Conclusion
For high-purity DNA samples, the three methods of UV absorption, uorescence, and diphenylamine reaction can accurately estimate DNA concentration. The diphenylamine reaction method has
the highest anti-jamming ability to give the most accurate data

24

Evaluation of DNA quantitation methods / X. Li et al. / Anal. Biochem. 451 (2014) 1824

when the solution contains disruptors. The UV absorbance method

is easily interfered with by contaminants and fails to distinguish
dsDNA from single-stranded DNA and RNA; however, the ratio of
different wavelengths of UV absorption could be used to determine
the quality of a sample. The PicoGreen dye method is very sensitive
to impurities and DNA degradation and has the ability to distinguish intact DNA from degraded products, which is an advantage
over the other two methods. Therefore, for high-purity DNA that
has been stored for a period of time, the PicoGreen dye method
should be used whenever possible. Overall, to determine the
DNA concentrations of important samples, a variety of methods
with different principles should be used, and the errors that may
exist should be accounted for.
Acknowledgments
This work was supported by grants from the National Major
Special Project for the Development of Transgenic Organisms
(2013ZX08012-003, 2014ZX08012-002B, and 2011ZX08012-005)
and the National Natural Science Foundation of China (31271880).
References
[1] V. Webb, J. Davies, Antibiotic preparations contain DNA: a source of drug
resistance genes?, Antimicrob Agents Chemother. 37 (1993) 23792384.
[2] M. Miragliaa, K.G. Berdalb, C. Breraa, P. Corbisierc, A. Holst-Jensenb, E.J. Kokd,
H.J.P. Marvind, H. Schimmelc, J. Rentsche, J.P.P.F. van Ried, J. Zagonf, Detection
and traceability of genetically modied organisms in the food production
chain, Food Chem. Toxicol. 42 (2004) 11571180.
[3] E. Hyman, P. Kauraniemi, S. Hautaniemi, M. Wolf, S. Mousses, E. Rozenblum, M.
Ringnr, G. Sauter, O. Monni, A. Elkahloun, O.P. Kallioniemi, A. Kallioniemi,
Impact of DNA amplication on gene expression patterns in breast cancer,
Cancer Res. 62 (2002) 62406245.
[4] S.Y. Kim, R. Byrn, J. Groopman, D. Baltimore, Temporal aspects of DNA and RNA
synthesis during human immunodeciency virus infection: evidence for
differential gene expression, J. Virol. 63 (1989) 37083713.
[5] ISO 21571, FoodstuffsMethods of analysis for the detection of genetically
modied organisms and derived productsNucleic acid extraction, 2005.
[6] M. Ogur, G. Rosen, The nucleic acids of plant tissues: the extraction and
estimation of desoxypentose nucleic acid and pentose nucleic acid, Arch.
Biochem. 25 (1950) 262276.
[7] S.J. Ahn, J. Costa, J.R. Emanuel, PicoGreen quantitation of DNA: effective
evaluation of samples pre- or post-PCR, Nucleic Acids Res. 24 (1996) 2623
2625.
[8] V.L. Singer, L.J. Jones, S.T. Yue, R.P. Haugland, Characterization of PicoGreen
reagent and development of a uorescence-based solution assay for doublestranded DNA quantitation, Anal. Biochem. 249 (1997) 228238.
[9] P.K. Ganguli, A sensitive procedure for the estimation of deoxyribonucleic acid
by the diphenylamine reaction in the presence of cupric sulphate, Rev. Can.
Biol. 29 (1970) 339346.
[10] G.M. Richards, Modications of the diphenylamine reaction giving increased
sensitivity and simplicity in the estimation of DNA, Anal. Biochem. 57 (1974)
369376.
[11] B.J.W. Hong, W.L. Hemme, G.E. Keller, M.T. Rinke, G.C. Bazan, Conjugatedpolymer/DNA interpolyelectrolyte complexes for accurate DNA concentration
determination, Adv. Mater. 18 (2006) 878882.

[12] M.D. Dutton, R.J. Varhol, D.G. Dixon, Technical considerations for the use of
ethidium bromide in the quantitative analysis of nucleic acids, Anal. Biochem.
230 (1995) 353355.
[13] J. Markovitz, B.P. Roques, J.B. LePecq, Ethidium dimer: a new reagent for the
uorimetric determination of nucleic acids, Anal. Biochem. 94 (1979) 259
264.
[14] C. Labarca, K. Paigen, A simple, rapid, and sensitive DNA assay procedure, Anal.
Biochem. 102 (1980) 344352.
[15] H.S. Rye, J.M. Dabora, M.A. Quesada, R.A. Mathies, A.N. Glazer, Fluorometric
assay using dimeric dyes for double- and single-stranded DNA and RNA with
picogram sensitivity, Anal. Biochem. 208 (1993) 144150.
[16] F.G. Loontiens, R. Regenfuss, A. Zechel, L. Dumortier, R.M. Clegg, Binding
characteristics of Hoechst 33258 with calf thymus DNA, poly[d(A-T)], and
d(CCGGAATTCCGG): Multiple stoichiometries and determination of tight
binding with a wide spectrum of site afnities, Biochemistry 29 (1990)
90299039.
[17] A. Holst-Jensen, Testing for genetically modied organisms (GMOs): past,
present, and future perspectives, Biotechnol. Adv. 27 (2009) 10711082.
[18] European Commission, Technical guidance in particular about sampling (incl.
unit of measurement for percentage of DNA), EC recommendation No. 787/
2004.
[19] S. Trapmann, H. Schimmel, G.N. Kramer, E.G. Van, J. Pauwels, Production of
certied reference materials for the detection of genetically modied
organisms, J. AOAC Int. 85 (2002) 775779.
[20] W. Broothaerts, P. Corbisier, H. Emons, H. Emteborg, T.P. Linsinger, S.
Trapmann, Development of a certied reference material for genetically
modied potato with altered starch composition, J. Agric. Food Chem. 55
(2007) 47284734.
[21] Institute for Reference Materials and Measurements, The certication of
reference materials of cotton seed powder with different mass fractions of the
cotton
event
281-24-236  3006-210-23,
<http://www.erm-crm.org/
ERM_products/search/reports/BF422.pdf>.
[22] Institute for Reference Materials and Measurements, The certication of
reference materials of dry-mixed maize powder with different mass fractions
of NK603 maize, <http://www.erm-crm.org/ERM_products/search/reports/
BF415.pdf>.
[23] Institute for Reference Materials and Measurements, The certication of drymixed soya powder with different mass fractions of Roundup Ready soya,
<http://www.erm-crm.org/ERM_products/search/reports/BF410k.pdf>.
[24] S. Folloni, G. Bellocchi, A. Prospero, M. Querci, W. Moens, M. Ermolli, G. Van
den Eede, Statistical evaluation of real-time PCR protocols applied to quantify
genetically modied maize, Food Anal. Methods 3 (2010) 304312.
[25] N. Gryson, Effect of food processing on plant DNA degradation and PCR-based
GMO analysis: a review, Anal. Bioanal. Chem. 396 (2010) 20032022.
[26] F. Moreano, U. Busch, K.H. Engel, Distortion of genetically modied organism
quantication in processed foods: inuence of particle size compositions and
heat-induced DNA degradation, J. Agric. Food Chem. 53 (2005) 99719979.
[27] E. Romppanen, K. Savolainen, I. Mononen, Optimal use of the uorescent
PicoGreen dye for quantitative analysis of amplied polymerase chain reaction
products on microplate, Anal. Biochem. 279 (2000) 111114.
[28] P. Murakami, M.T. McCaman, Quantitation of adenovirus DNA and virus
particles with the PicoGreen uorescent dye, Anal. Biochem. 274 (1999) 283
288.
[29] G.N. Abraham, C. Scaletta, J.H. Vaughan, Modied diphenylamine reaction for
increased sensitivity, Anal. Biochem. 49 (1972) 547549.
[30] C. Peano, M.C. Samson, L. Palmieri, M. Gulli, N. Marmiroli, Qualitative and
quantitative evaluation of the genomic DNA extracted from GMO and nonGMO foodstuffs with four different extraction methods, J. Agric. Food Chem. 52
(2004) 69626968.
[31] J. Sambrook, D.W. Russell, Molecular Cloning: A Laboratory Manual, 3rd ed.,
Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 2001.