Accepted Manuscript

Measurement of the total protein in serum by biuret method with uncertainty

evaluation

Kangle Zheng, Liqing Wu, Zhangjing He, Bin Yang, Yi Yang

PII: S0263-2241(17)30510-9
DOI: http://dx.doi.org/10.1016/j.measurement.2017.08.013
Reference: MEASUR 4911

To appear in: Measurement

Received Date: 12 April 2017

Revised Date: 4 July 2017
Accepted Date: 7 August 2017

Please cite this article as: K. Zheng, L. Wu, Z. He, B. Yang, Y. Yang, Measurement of the total protein in serum by
biuret method with uncertainty evaluation, Measurement (2017), doi: http://dx.doi.org/10.1016/j.measurement.
2017.08.013

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
 Measurement of the total protein in serum by biuret method with

uncertainty evaluation

Kangle Zheng1, Liqing Wu2, Zhangjing He1, Bin Yang2, Yi Yang1*

1
Beijing University of Chemical Technology, Beijing, China
2
National Institute of Metrology, Beijing, China
*Authors for correspondence
E-mail: yangyi@mail.buct.edu.cn
Address: No. 15 North Third Ring Road, Chaoyang District, Beijing, China

Abstract
As recommended by IFCC, the biuret method is generally used to determine the total protein as a
reference method. In order to achieve accurate measurement of the total protein by biuret method, a
detailed description of establishment, validation and uncertainty evaluation of this method is proposed
in the metrological view in this paper. Then the method was used to determine the total protein in
RELA serum samples and the results of 2015-RELA sample A and B were 73.9 g·L-1 and 58.9 g·L-1,
with the RSD of 1.18% and 1.07%, respectively. The uncertainty was evaluated in detail fully
following the Guide to the Expression of Uncertainty in Measurement, and the expanded uncertainty of
RELA sample A and B were 2.0 and 1.2 g·L-1, respectively. The results have good agreement with
those from other research institutions. The proposed procedure gives a good example to realize an
accurate reference method with strict metrological control, which can benefit not only total protein
measurement but also other biological measurement by reference methods.

Keywords
Biuret Method, Total Protein, RELA-Samples, Uncertainty

1. Introduction
The total protein content in human tissue fluid plays an important role on the state of health [1]. Serum
is the most important tissue fluid in human body, and total serum protein have a variety of functions,
such as maintaining the osmotic pressure of the blood vessel, maintaining the pH of plasma,
transporting various metabolites, regulating the physiological function of the transport material, and
has a close relationship with the immune function of the body, which can give us useful information
about our health condition. The level of the total serum protein reflects the loss of protein caused by
liver function and renal disease. Measurement of total serum protein can be used to monitor the
nutritional status of the body indirectly, and benefit diagnostic of certain diseases [2]. With the reasons
above, the total serum protein is a routine examination in clinical diagnostics.

Numerous measurement methods of total serum protein have been developed because of its importance.
Several methods, such as the Kjeldahl method [3], the biuret method [4, 5], the Lowry method [6], the
Bradford method [7] and the UV absorbance method [8] are widely used. Although the Kjeldahl
method is widely used, it is too tedious and time-consuming to be used as a routine assay procedure. The
Bradford method is easy to be operated with fewer reagents, but it cannot be used for different protein
determination because of the large deviation results. The Lowry method has a high sensitivity, but it is
subject to several interferences originating from the sample matrix and need to control the operation
time accurately. The UV absorbance method is simple to be used, because the total protein in serum
and other proteins solutions can be determinate directly by UV spectrometry and the method allows
almost complete sample reuse after the measurement. For all that, the UV absorbance method has
obvious weakness, because nucleic acids have almost the same absorption wavelength as protein,
which can disturb the protein measurement. Furthermore, the UV absorbance method has lower
sensitivity and its application is limited.

Compared with the above methods, the biuret method has the advantages of easy operation, excellent
precision and less interference. The principle of this method is that the peptide bond of amino acid
residues and two copper ions (Cu2+) can form a purple complex in alkaline condition, and the darkness
of the color is proportional to the content of protein in the sample. When it is compared with the
standard, treated in the same process, such as SRM 927, the protein content can be obtained. The
darkness of the color is only proportional to the number of peptide bond sand it has no obvious
relationship with the type of the protein composition, molecular weight and amino acid residue.
Although the sensitivity of the method is not good and the detection of limit is generally 0.2 ~ 1.7 g·L-1,
it is enough for the measurement of total protein in serum, whose concentration is 60 ~ 80 g·L-1 in
general. Therefore, the biuret method becomes the most popular routine method for the determination
of the total protein in serum, and it was also proposed as a candidate reference method for total serum
protein measurement to establish the traceability chain in clinical diagnostics [9, 10].

In order to measure the total serum protein accurately by the reference method, the method parameters
should be accurately and strictly followed. “Measurement traceability” is widely used by metrologies,
which refers to an unbroken chain of comparisons of any measurement to a known standard.
Calibration with a traceable standard can determine the bias and precision of specific measurement and
ensure the measurement accuracy. In order to ensure the accuracy of total serum protein measurement
in this research, both the physical parameters and the chemical parameters were well calibrated with
known traceable measurement standards or certified reference materials in advance. Then the method
performance was evaluated and applied in measurement of RELA samples. Good results were achieved
according to the feedback. It can be looked as a successful example of realization of an accurate
reference method with metrological techniques.

2. Materials and Methods

2.1 Chemicals
Potassium iodide (KI) and sodium hydroxide (NaOH) were purchased from Beijing Chemical Works
(Beijing, China). Potassium sodium tartrate (KNaC4H4O6·4H2O) and copper sulfate pentahydrate
(CuSO4·5H2O) were purchased from Alfa Aesar (Tianjin) Chemical Co., LTD. The certified reference
material of bovine serum albumin solution (SRM 927, 70.45±0.2 g·L-1) was purchased from the
National Institute of Standards and Technology (NIST, Gaithersburg, MD, USA). The certified
reference material of total protein in frozen human serum (GBW09187, 65.8±1.9 g·L-1) was purchased
from Beijing Aerospace Hospital (Beijing, China). The certified reference material of bovine serum
albumin solution (BW3627-1, 69.4±4.6 g·L-1) was developed by National Institute of Metrology (NIM,
Beijing, China). Two frozen dried human serum RELA samples A and B were provided by the
organizer. The water was purified by a Milli-Q system (Merck, Germany).

2.2 Instruments
UV-VIS spectrophotometer: PE, Lambda 950. Volumetric flask: 1L. Pipettes (200 µL, 1 mL): Gilson,
France. Balance: ME235S, 0.01 mg, German Satorius Company. Constant temperature incubation
apparatus: MTC-100, Hangzhou MIULAB Instrument Co. LTD.

2.3 Preparation of buffers and solutions

NaOH (6 mol·L-1): 60 g of NaOH was dissolved in 250 mL of ultrapure water. After cooling to room
temperature, it was stored in a sealed bottle at 4C.
Biuret reagent: 3.00 g of CuSO4·5H2O was dissolved in 500 mL of ultrapure water. Then 9.00 g of
KNaC4H4O6·4H2O and 5.0 g of KI were added. After completely dissolved, 100 mL of 6 mol·L-1
NaOH was added. Then it was diluted to 1 L with ultrapure water and stored in a sealed bottle at 4C.
Alkaline tartrate: 9.0 g of KNaC4H4O6·4H2O and 5.0 g of KI was dissolved in 0.8 L of ultrapure water.
After completely dissolved, 100 mL of 6 mol·L-1 NaOH was added. Then it was diluted to 1 L with
ultrapure water and stored in sealed bottle at 4C.

2.4 Calibration and verification of the UV-VIS spectrophotometer

The biuret method is based on the absorbance measurement at specific wavelength. In order to follow
the reference method strictly and ensure the measurement accuracy, both the absorbance and the
wavelength were well calibrated and verified with known traceable filter standards prior to the sample
measurement. The absorbance measurement capability of the UV-VIS spectrophotometer was
calibrated and verified with the absorbance and wavelength Certified Reference Materials (CRMs,
GBW (E) 130225 and GBW (E) 130226), each filter was measured for 15 times and the data were used
to calculate the bias and repeatability of the absorbance or wavelength measurement.

2.5 Calibration of the pipettes

In this study, both the accuracy and the reproducibility of reagent transferring contributed a large
uncertainty to the final result. Therefore, the pipettes (1 mL, 200 µL) were examined carefully by
weighing a known amount of ultrapure water by a balance. Then the transferring volume was
calculated with density and the volume bias was obtained for uncertainty calculation.

2.6 Measurement Procedure

The measurement of total serum protein was fully followed the reference method (biuret method) in
this study. The absorbance change was measured by the UV-VIS spectrophotometer before and after
the biuret reaction. In order to correct the measurement blank, four absorbance values were measured
and used to calculate the total serum protein. The absorbance (A1) of alkaline tartrate soulution was
measured as reagent blank of alkaline tartrate soulution. The absorbance (A2) was measured with the
alkaline tartrate solution containing the same volume of the serum sample and the value of (A2-A1)
used as the sample blank. The absorbance (A3) was measured with the biuret reagent diluted by the
same volume of the ultrapure water as the sample, and it was used as the reagent blank. The sample
absorbance (A4) was measured after the biuret reagent adding to the serum, which was not corrected
with blanks. Because the reagent blank was A3i and the sample blank was (A2-A1), the change of the
corrected absorbance was A4-A3-(A2-A1). Finally, the total serum protein can be obtained by the
protein concentration and the change of the absorbance value (Formula (1)).

A4s  A3s   A2s  A1s 

cs  g  L1 ＝  cstd  g  L1  (1)
A4std  A3std   A2std  A1std 

where，
cs is the total protein in the sample;
A1s, A2s, A3s and A4s are the absorbance values of samples;
A1std, A2std, A3std and A4std are absorbance values of CRMs;
cstd is the total protein in SRM 927.

2.7 Evaluation of the method Performance

After calibration and verification of the physical parameters, the method performance was evaluated
prior to the RELA samples measurement. The measurement bias, repeatability, reproducibility, limit of
detection (LOD) and limit of quantitation (LOQ) were evaluated individually.
In order to evaluate the measurement bias, two CRMs (total serum protein, GBW09187; and bovine
serum albumin solution, BW3627-1) were analyzed with NIST CRM of bovine serum albumin (SRM
927) as standards. For each CRM, three sub-samples were taken and each sub-sample was measured
for three times. The same measurement procedure was repeated for three separated days. The bias can
be obtained by the Formula (2).

(2)
Where,
Cmeasurement is the average of data for one day;
CCRM is the standard value for CRM GBW09187 or BW3627-1.

For the repeatability, two CRMs (GBW09187, BW3627-1) were measured for nine times in one day.
And the data were brought into Formula (3) to calculate the relative standard deviation (RSD). To
investigate the reproducibility of the method, two CRMs (GBW09187, BW3627-1) were analyzed. For
each CRM, three sub-samples were taken and each sub-sample was measured for three times. The same
measurement procedure was repeated for three separated days. And the averages for each sub-sample
were brought into Formula (3) to calculate the RSD.

(3)

In order to calculate the LOD or LOQ, the maximum noise of the absorbance measurement had been
measured. Then, A4s-A3s-(A2s-A1s) in Formula (1) was replaced by 3 or 10 times of the maximum noise
to calculate the LOD or LOQ, respectively.

2.8 RELA samples measurement

The RELA samples were taken from the refrigerators and kept in the room temperature. Then 5 mL of
ultrapure water was added and the samples were reconstituted for half an hour until they were
completely dissolved. Next, eight 15 mL centrifuge tubes were taken and numbered as A1s, A2s, A3s,
A4s and A1std, A2std, A3std, A4std. Then 5 mL alkaline tartrate solution was added to the tube A1s, A2s
and A1std, A2std, and 5 mL biuret reagent was added to tube A3s, A4s and A3std, A4std, respectively. After
that, 100 µL serum sample was added to tube A1s, A2s, respectively. 100 µL SRM 927 was added to
tube A1std, A2std. 100 µL ultrapure water was added to tube A3s, A3std. Then 100 µL serum sample was
added to tube A4s and 100 µL SRM 927 was added to tube A4std. After that, the solution was mixed
gently and thoroughly by inversion immediately. Incubation was carried out at 25 C for 60 minutes.
Finally, the solution absorbance of each tube was measured at 540 nm by UV-VIS spectrophotometer
with the air as the blank. The order of measurements is A1s, A2s, A3s, A1std, A2std, A3std, A4s and A4std.
The total protein in serum was further calculated according to Formula (1). The same cuvette was used
in all measurements and it was cleaned by ultrapure water and dried for 5 times after each analysis.

3. Results and Discussion

3.1 Calibration and verification of the UV-Spectrometer
The absorbance measurement capability of the UV-VIS spectrophotometer was calibrated and verified
with spectral neutral filter standards (GBW (E) 130225) with known absorbance at 540 nm. The
maximum measurement bias and repeatability were 0.0073 and 0.028% in the absorbance range from
0.2 to 1.5, respectively, which indicated the accuracy of the absorbance measurement. The holmium
glass wavelength standard (GBW (E) 130122) was used for calibration and verification of the
wavelength. The measurement bias and repeatability were only 0.3 nm and 5.910-6 at 537.1 nm,
respectively. Because the UV-VIS spectrophotometers were calibrated with metrological standards, the
traceability and accuracy were guaranteed; therefore, the related optical parameters required by the
biuret reference method were fully followed with confidence.

3.2 Calibration and verification of the pipettes

After calibration and verification of the physical parameters, the pipettes were used for liquid
transferring. In order to ensure the volume accuracy of the pipettes, they were calibrated and verified
by weighing known amount of ultrapure water, and the actual transferring volume was calculated based
on the weight and density of the water (Table 1). It was shown that the relative errors of 1 mL-pipette
and 100 µL pipette were 0.47% and 1.0%, and the repeatability were 0.09% and 0.15% for six repeats
of transferring, which indicated the good accuracy and repeatability of liquid transferring.
 Table 1 Verification of the pipettes
Water Density Relative
Pipette（µL） Mean(mg) Twater(℃) -1
VR(mL) RSD (%)
Mass(mg) (gm·L ) Error (%)
1002.08
1004.13
1001.77
1000 1002.76 20.68 998.056 1.0047 0.47 0.09
1002.56
1003.64
1002.38
101.03
100.62
100.87
100 100.835 19.38 998.325 0.1010 1.0 0.15
100.85
100.94
100.70

3.3 Evaluation of the method Performance

After calibration and verification of the physical parameters, the method performance was evaluated. In
order to keep the traceability for total serum protein measurement, the NIST SRM 927 was used as the
standard. Before measuring the RELA samples, both CRMs of total serum protein (GBW09187) and
bovine serum albumin (BW3627-1) were used to evaluate the measurement bias, repeatability,
reproducibility, LOD and LOQ. For each CRM, three sub-samples were taken and each sub-sample
was measured for three times. The same measurement procedure was repeated for three separated days
(Table 2). The average was calculated and the bias was -0.52 ~ 0.08 g·L-1 and -1.25 ~ -0.68 g·L-1 for
total serum protein and bovine serum albumin measurement, respectively. Each result was located
within the uncertainty range of the CRM GBW09187 (65.8±1.9 g·L-1) and BW3627-1 (69.4±4.6 g·L-1),
indicating good measurement accuracy.
The within-day repeatability was 0.53% ~ 0.88% and 0.76% ~ 1.89% for the measurement of CRM
GBW09187 and BW3627-1, respectively; and the between-day reproducibility was 0.61% and 1.90%,
respectively, which indicated the good method variance.
The maximum noise of absorbance measurement was about 0.729×10-3, and the LOD and LOQ were
calculated as 0.4 g·L-1and 1.32 g·L-1, respectively, which can meet the requirement of serum total
protein measurement though the LOD and LOQ were high.
 Table 2 Evaluate the bias and method variance by GBW09187 and BW3627-1
Total protein Standard
CRM Day -1
MeanD(g·L-1) Bias Mean(g·L-1) RSD (%)
content(g·L ) Value(g·L-1)
64.83
1 65.05 65.28 -0.52
65.96
65.90
GBW09187 2 65.72 65.70 -0.10 65.57 65.8±1.9 0.61
65.49
65.89
3 65.49 65.72 -0.08
65.79
67.76
1 69.42 68.51 -0.89
68.34
65.69
BW3627-1 2 68.96 68.15 -1.25 68.46 69.4±4.6 1.90
69.79
69.50
3 68.89 68.72 -0.68
67.76

3.4 RELA sample measurement

After the method performance evaluation, the RELA sample A and B were analyzed. Three
sub-samples were taken from each RELA sample and each sub-sample was analyzed for three repeats
in one day. The process was repeated for three continued days, which were shown in Table 3. The
average of RELA sample A was 73.9 g·L-1 with the within-day repeatability of 1.08% ~ 1.74%. The
average of RELA sample B was 58.9 g·L-1 with the within-day repeatability of 0.69% ~ 1.37%. The
overall reproducibility was 1.18% or 1.07% for sample A or B, respectively, indicating good
measurement precision.
 Table 3 Measurement results of RELA sample A and B
Total protein
Sample Day Sub-sample MeanD (g·L-1) Mean (g·L-1) RSD (%)
content (g·L-1)
1 72.15
1 2 73.92 73.71
3 75.08
4 72.63
A 2 5 73.70 73.93 73.87 1.18
6 75.47
7 72.98
3 8 73.99 73.92
9 74.81
1 57.70
1 2 58.86 58.70
3 59.53
4 58.47
B 2 5 58.60 58.80 58.87 1.07
6 59.33
7 58.49
3 8 59.13 59.11
9 59.70

3.5 Uncertainty evaluation

The uncertainty evaluation was shown with RELA sample A as an example. The measurement
uncertainty came from the weighing process, liquid transferring, chromogenic reaction, absorbance
measurement, protein CRM and the whole method variation.

Based on the mathematical measurement model shown in Formula (1), there were nine variables and
the typical value of each variable was shown in Table 4.

Table 4 The typical value of the nine variables

A1s A2s A3s A4s A1std A2std A3std A4std cstd (g·L-1)

0.0332 0.0513 0.1306 0.4771 0.0334 0.0375 0.1307 0.5198 70.45

In order to calculate the uncertainty components, the sensitivity factor of each variable was calculated
firstly, which was shown in the following:
 cstd
c A4s  c A1s   182.99
A4std  A3std  A2std  A1std
cstd
c A3s  c A2s    182.99
A4std  A3std  A2std  A1std
( A4s  A3s  A2s  A1s )cstd
c A4std  c A1std    156.09
( A4std  A3std  A2std  A1std ) 2
( A4s  A3s  A2s  A1s )cstd
c A3std  c A2std   156.09
( A4std  A3std  A2std  A1std ) 2
A4s  A3s  A2s  A1s
ccstd   0.85299
A4std  A3std  A2std  A1std

(1) Uncertainty from the chromogenic and the whole method variation
It is hard to evaluate the uncertainty component from the chromogenic reaction and the whole method
variation by type-B uncertainty evaluation method. Therefore, the type-A method was used and the
standard deviation of RELA sample A measurement was used to calculate the uncertainty component.

s 1.13
um    0.38 g  L1 (4)
9 3

(2) Uncertainty from weighing

The imprecision of the balance used is 0.0001 g, therefore the uncertainty can be calculated as Formula
(4) with rectangle assumption and each weighing uncertainty is listed in Table 5. All the weighing
uncertainties were very small and they were already covered by uncertainty evaluation of method
variation, therefore the weighing uncertainties were not combined further.
0.0001
ubalance   5.77 105 g
3

Table 5 The uncertainty from each weighing

Weighing Value /g u(x) /g ur(x)
-5
NaOH 60.0237 5.7710 9.610-7
CuSO45H2O 3.0026 5.7710-5 1.910-5
KNaC4H4O6·4H2O 9.0053 5.7710-5 6.410-6
KI 5.0243 5.7710-5 1.110-5

(3) Uncertainty from liquid transferring

The relative errors of 1 mL pipette and 100 µL pipette were 0.47% and 1.0%, respectively. Therefore,
the relative uncertainties were 0.19% and 0.41% with the triangle assumption. However, these
uncertainties evaluation of the whole method variation, therefore, these uncertainties were not
combined further.

(4) Uncertainty from the protein CRM

The NIST SRM 927 was used as the standard during the measurement. The expanded uncertainty of
this CRM is 0.2 g·L-1 with a coverage factor of 2. Therefore, the uncertainty component from the
protein CRM can be calculated as following:
U 0.2
uCRM    0.1 g  L1
k 2

Based on the mathematical model and sensitivity factor, the uncertainty component contributed by
protein CRM was calculated as following:
c
us  ustd  ccstd ustd  0.085 g  L1
cstd

(5) Uncertainty from the absorbance measurement

The “Calibration and verification of the spectrometer” showed that the maximum bias was 0.0073,
therefore the related uncertainty component was estimated with rectangle assumption as following:
0.0073
u Abs   0.0042
3

The same spectrometer was used to measure all the absorbance in this study, therefore, strong
correlation exists among the variables and correlation factor of 1 was assumed. Based on Formula (1),
the uncertainty came from the absorbance measurement was calculated as following:

c A2 1s u A2 1s  c A2 2s u A2 2s  c A2 3s u A2 3s  c A2 4s u A2 4s  c A2 1std u A2 1std  c A2 2std u A2 2std  c A2 3std u A2 3std  c A2 4std u A2 4std

uA  4,4 4,4 4,4
2 
i 1, j 1,i  j
c Ais c Ajs u Aisu Ajs  2 
i 1, j 1,i  j
c Aistd c Ajstd u Aistd u Ajstd  2 
i 1, j 1
c Ais c Ajstd u Aisu Ajstd

 0.91 g  L1

(6) Combined and expanded uncertainty

The combined uncertainty was calculated as following:

uc  um2  us2  u A2  0.99 g  L1

The expanded uncertainty can be calculated with a coverage factor of 2:

U =2  uc  2.0 g  L1

Therefore, the result of RELA sample A can be expressed as (73.9 ± 2.0) g·L-1. Similarly, the
uncertainty of RELA sample B was evaluated and the measurement result can be express as (58.9 ±1.2
g·L-1).

3.6 Feedback
The feedback from the RELA organizer was shown in Fig.1 and the lab-code 092 indicated our results.
Our results were overlapped with other results from 13 labs among total 17 participants, demonstrating
the accuracy and reliability of our results.
 Fig.1 Results of total serum protein measurement (RELA 2015)

3.7 Highlights of this work

Biological measurement is very complicated and it is different from traditional measurement. Not all
the biological measurement result can be traced to SI unit when the measurement cannot be clearly
defined or for activity measurement. In this case, a reference method is generally defined and it is used
to harmonize such measurement results. Therefore, in order to make the measurement defined by
reference method comparable and accurate, the reference method should be strictly followed, which
means that all the parameters required by the reference method should be traced to their metrological
standards to ensure the correct and comparable parameters. Here is an example of how to realize an
accurate and reliable measurement based on absorbance measurement. There are several reference
methods based on the principle of absorbance measurement, such as the IFCC reference methods for
serum enzyme activity measurements. This work can also benefit these measurements based on
absorbance measurement defined by reference methods, which have a great significance in metrology.

4. Conclusion
In this study, accurate and comparable measurement of total serum protein was achieved by fully
following the reference method via metrological techniques. This is a good example to use the
metrology to achieve accurate and comparable measurement. All the apparatus in the experiment were
calibrated and verified carefully with traceable metrological standards and the method performance
was evaluated with CRMs thoroughly. Therefore, the parameters in the reference method can be
followed strictly to ensure the accuracy and comparability of the measurement. Only a small number of
biological measurements can be traced to SI unit and there are still numerous biological measurements
are defined and traced to reference methods nowadays. Therefore, it is can be expanded to realize
accurate and comparable biological measurements according to other reference methods.

Acknowledgement
This work was supported by National Key Scientific Apparatus Development Project
(2012YQ09019708).
References
[1]K. Murakami, T. Yonezawa, N. Matsuki, Synovial fluid total protein concentration as a possible
marker for canine idiopathic polyarthritis, J Vet Med Sci, 77 (2015) 1715-1717
[2]A. Koyama, M. Hashimoto, H. Tanaka, N. Fujise, M. Matsushita, Y. Miyagawa, Y. Hatada, R.
Fukuhara, N. Hasegawa, S. Todani, K. Matsukuma, M. Kawano, M. Ikeda, Malnutrition in Alzheimer's
Disease, Dementia with Lewy Bodies, and Frontotemporal Lobar Degeneration: Comparison Using
Serum Albumin, Total Protein, and Hemoglobin Level, PloS one, 11 (2016)
[3]V. Chromy, B. Vinklarkova, L. Sprongl, M. Bittova, The Kjeldahl Method as a Primary Reference
Procedure for Total Protein in Certified Reference Materials Used in Clinical Chemistry. I. A Review
of Kjeldahl Methods Adopted by Laboratory Medicine, Crit Rev Anal Chem, 45 (2015) 106-111
[4]K.P. Shrivastaw, S.S. Jhamb, A. Kumar, Quantitation of the protein content of diphtheria and
tetanus toxoids by the Biuret method during production of combined vaccines, Biologicals, 23 (1995)
61-63
[5]Z.L. Liu, J.H. Pan, A practical method for extending the biuret assay to protein determination of
corn-based products, Food Chem, 224 (2017) 289-293
[6]N. Lee, S. Shin, H.J. Chung, D.K. Kim, J.M. Lim, H. Park, H.J. Oh, Improved quantification of
protein in vaccines containing aluminum hydroxide by simple modification of the Lowry method,
Vaccine, 33 (2015) 5031-5034
[7]K. Grintzalis, C.D. Georgiou, Y.J. Schneider, An accurate and sensitive Coomassie Brilliant Blue
G-250-based assay for protein determination, Anal Biochem, 480 (2015) 28-30
[8]J.E. Noble, Quantification of Protein Concentration Using UV Absorbance and Coomassie Dyes,
Methods Enzymol, 536 (2014) 17-26
[9]B.T. Doumas, D.D. Bayse, R.J. Carter, T. Peters, Jr., R. Schaffer, A candidate reference method for
determination of total protein in serum. I. Development and validation, Clin Chem, 27 (1981)
1642-1650
[10]B.T. Doumas, D.D. Bayse, K. Borner, R.J. Carter, F. Elevitch, C.C. Garber, R.A. Graby, L.L.
Hause, A. Mather, T. Peters, Jr., R.N. Rand, D.J. Reeder, S.M. Russell, R. Schaffer, J.O. Westgard, A
candidate reference method for determination of total protein in serum. II. Test for transferability, Clin
Chem, 27 (1981) 1651-1654
Highlights
· A biuret method to measure the total protein in serum.
· Instruments were certificated and the data used in calculation of uncertainty.
· Two known concentration standard protein to evaluate the method performance.
· Two serum samples given by RELA were measured.
· The calculation of the uncertainty here developed is detailed.

Graphical abstract