296 GENERAL ANALYTICAL METHODS [16]

[16] P r o t e i n D e t e r m i n a t i o n in M e m b r a n e and Lipoprotein

Samples: Manual and Automated Procedures
By MARY ANN K. MARKWELL, S U Z A N N E M. HAAS,
N. E. TOLBERT, and L. L. BIEBER

A variety of methods have proved to be effective in estimating the

protein content of water-soluble samples. These include use of Kjeldahl
nitrogen values, 1 ultraviolet absorbance, 2'3 the biuret procedure, 4 the
Lowry adaptation 5 of the Folin-Ciocalteu procedure, 6 reaction with
fluorescamine,7 and the dye binding method of Bradford. 8 To adopt these
methods for use with water-insoluble systems, i.e., tissue homogenates,
subcellular membrane fractions, enveloped viruses, and lipoproteins, re-
quires extensive pretreatment to solubilize the sample or to remove mate-
rials that interfere with protein determination that are present in these
complex biological mixtures. A modification of the Lowry method was
introduced 9 to simplify protein determination in membrane and lipopro-
tein samples.
The procedure described here is based on this modified Lowry
method. By adding sodium dodecyl sulfate (SDS) to the alkali reagent,
samples can be assayed directly without prior solubilization or delipida-
tion. An increase in the copper tartrate concentration facilitates quantita-
tion of protein in the presence of sucrose and EDTA. 9 Color formation
depends mainly 1° on reduction of the Folin-Ciocalteu reagent by protein-
bound copper and proceeds in two distinct steps. In the first step the
protein sample is mixed with copper ions in alkaline medium. In the

1 R. Ballantine, this series, Vol. 3, p. 984.

2 0 . W a r b u r g and W. Christian, Biochem. Z. 310, 384 (1941); see also E. L a y n e , this series,
Vol. 3, p. 447.
a W. J. Waddell, J. Lab. Clin. Med. 48, 311 (1956).
4 A. G. Gornall, C. J. Bardawill, and M. M. David, J. Biol. Chem. 177, 751 (1949),
5 0 . H. L o w r y , N. J. R o s e b r o u g h , A. L. Farr, and R. J. Randall, J. Biol. Chem. 193, 265
(1951).
6 0 . Folin a n d V. Ciocalteu, J. Biol. Chem. 73, 627 (1927).
T p. Bohlen, S. Stein, W. Dairman, and S. Udenfriend, Arch. Biochem. Biophys. 155, 213
(1973).
s M. M. Bradford, Anal. Biochem. 72, 248 (1976).
9 M. A. K. Markwell, S. M. H a a s , L. L. Bieber, and N. E. Tolbert, Anal. Biochem. 87, 206
(1978).
~0 T h e color obtained in the absence o f copper has been attributed to the tyrosine and
t r y p t o p h a n contents o f the protein by L o w r y et al. ~

Copyright ~) 1981 by Academic Press, Inc.

METHODS IN ENZYMOLOGY, VOL. 72 All rights of reproduction in any form reserved.
ISBN 0-12-181972-8
[16] PROTEIN DETERMINATIONIN COMPLEXSYSTEMS 297

second step the phosphomolybdate-phosphotungstate (Folin-Ciocalteu)

reagent is reduced by protein-bound copper.

Manual Assay

Reagents
A: 2.0% Na~CO3, 0.4% N a O H , 0.16% sodium tartrate, 1.0% SDS
B: 4% CuSO4 • 5 HzO
C: Mix 100 parts of reagent A with 1 part of reagent B to form the
alkaline copper reagent.
D: Dilute Folin-Ciocalteu 2 N phenol reagent, 1:1 (v/v), with
deionized water
E: Dissolve crystalline bovine serum albumin (BSA) at 0.1 mg/ml in
deionized water.
Reagents A and B are stable indefinitely at room temperature. Reagent
A should be stored in a polyethylene bottle. The SDS may precipitate out
upon cooling below 20°, but will redissolve if the reagent is warmed briefly
before using. Because potassium dodecyl sulfate is insoluble, sodium tar-
trate rather than potassium-sodium or potassium tartrate is used in re-
agent A. Reagents C and D are made fresh daily. The undiluted Folin-
Ciocalteu reagent is stored at 4 °. Aliquots of the BSA standard solution
can be stored indefinitely at - 2 0 ° and thawed on the day of use.
Procedure. Mix the protein sample (volume of 1 ml containing 10-100
/xg of protein) with 3 ml of reagent C and incubate at room temperature for
a minimum of 10 min. Then add diluted Folin-Ciocalteu reagent D (0.3 ml)
to the 4-ml volume of copper-treated protein, vigorously mixing the con-
tents of each tube immediately after addition. Incubate the tubes for 45
min at room temperature and read at 660 nm against a reagent blank. The
concentration versus absorbance curve is generated by using 0 (reagent
blank), 0.20, 0.40, 0.60, 0.80, and 1.00 ml of the BSA standard as samples
diluted to a final volume of 1 ml.
The timing o f the first step, incubation of protein with reagent C, is not
critical. The reaction with copper ions is complete within 10 min, but can
be allowed to proceed for as long as 24 hr in capped tubes with no de-
crease in the final color. H o w e v e r , overnight incubation in alkali should
be avoided for samples with a high lipid content, such as isolated mem-
branes or lipoproteins, because several phospholipids as well as
arachidonic acid can develop appreciable color after autoxidation. H In the
second step maximum color development is produced by a 45-min incuba-
tion period; the color is stable for at least 40 additional minutes.

~1j. Eichberg and L. C. Mokrasch, Anal. Biochem. 30, 386 (1969).

298 GENERAL ANALYTICAL METHODS [16]

' I ' I ' I I I I I

E 0.500 --
E

~o
t.D
0.400
I--

~,, 0.500

Z
0 200
¢Y

o
0. I 0 0

o L I i J it J I i I
20 40 60 80 I00

BOVINE SERUM A L B U M I N , ,u.g

FIG. 1. Determination of protein (bovine serum albumin) by the modified Lowry method
as described under procedure for manual assay. Absorbance values are the average of
duplicate determinations and have been corrected for the absorbance of a reagent blank.

Range. When the described procedure is carried out at room tempera-

ture, a linear relationship is observed between the amount of protein and
final color up to an absorbance of approximately 0.45 at 660 nm (Fig. 1).
This corresponds to a protein concentration of 100/.tg in a total assay
volume of 4.3 ml. The sensitivity of the modified Lowry procedure can be
enhanced by reading the absorbance of samples with low protein concen-
trations at 750 nm. When limited amounts of material are available, similar
results can be obtained by reducing ehe assay volume 20-fold and using
microcuvettes.
The wavelength used routinely in the modified Lowry procedure (660
nm) was chosen as a compromise between increased absorption of the
final blue reduction product with longer wavelengths and the practical
limitations of most spectrophotometers. The absorption peak of the blue
chromophore extends through much of the visible spectrum in a broad
plateau and reaches maximum at 750 nm3 Samples with high protein
concentrations (100-400 /~g per assay volume) can be read at shorter
wavelengths, e.g., 500 nm, along with the proper BSA standards to assay
them without dilution. Alternatively if increased sensitivity is desired,
samples with low protein concentrations (less than l0/.tg/ml) can be read
at longer wavelengths, e.g., 750 nm, along with the proper BSA standards.
It should be noted, however, that not all spectrophotometers are equipped
to use these near-infrared wavelengths. By selecting the right wavelength
from a broad possible range (500-750 nm) the contribution of interfering
chromophores such as the pigments chlorophyll a and b to the final ab-
sorption value can be minimized.
[16] PROTEIN D E T E R M I N A T I O N IN COMPLEX SYSTEMS 299

Automated Assay
For routine assay of large numbers of samples the modified Lowry
procedure was automated using the Gilford System 3500 Computer-
Directed Analyzer. This system uses preprogrammed magnetic cards to
dilute samples, dispense reagents, aspirate reaction mixtures into the elec-
tronic temperature-controlled cuvette, and print out corrected values for
end-point absorbances or kinetic enzyme rates.

Reagents
A: 3.0% Na2COz, 0.6% NaOH, 0.24% sodium tartrate, 1.5% SDS
B: 6.0% CuSO4" 5 H~O
C: On the day of use, mix 100 parts of reagent A with 1 part of
reagent B to form the alkaline copper reagent.
D: On the day of use, dilute 1 part of Folin-Ciocalteu 2 N phenol
reagent with 14 parts of deionized water.
E: Standard BSA solutions of 0.2, 1.0, 2.0, 3.0, and 4.0 mg/ml are
used as 25-/xl samples containing 5-100/~g of protein.
Refer to the section on reagents for manual assay for further details on
the preparation and storage of these reagents.
Procedure. The instrumental conditions are as follows:
Temperature, 25°
Wavelength, 660 nm (tungsten lamp, no filter)
Sample volume, 0.025 ml
Dispenser A volume, 1.25 ml
The Gilford System 3500 General Endpoint Program # 1 (modes 1 and 3) is
used for this procedure. Set the dispensing tower in position 2. Load
dispenser A with the alkaline copper reagent C. In the first step the pipet-
ter diluter will deliver 25 tzl of each sample and 1.25 ml of reagent C into
each reaction cup. Load the sample cups of the macro size reaction strips
in the following order:
Cups 1 and 2: deionized water
Cups 3-7: BSA standards of 0.2, 1.0, 2.0, 3.0, and 4.0 mg/ml
Cups 8-12, 21-39; 41-56: samples
Cups 20 and 40: BSA standards of 2.0 mg/ml
The deionized water in cups 1 and 2 is used to prepare the reagent blank
whose absorbance will be subtracted automatically from each sample ab-
sorbance value. A BSA standard is used as every twentieth sample. Al-
though the transport accommodates 14 racks (56 samples) two or three
times this number can be processed easily in one run by replacing the
processed racks with new ones.
300 GENERAL ANALYTICAL METHODS [16]

To start the assay insert General Endpoint Program #1 into the card
reader, select mode I (dispensing mode), and depress the run button. The
samples will be diluted automatically in sequence with reagent C in dis-
penser A. After reagent C has been added to the last sample, the transport
will automatically stop. Allow the samples to incubate for a minimum of
10 min. During this time r e m o v e the sample cups, clean out dispenser
syringe A, and replace reagent C with reagent D at dispenser A. After 10
or more minutes restart General Endpoint Program # 1 in mode 1. The
diluted Folin-Ciocalteu reagent D (1.25 ml) will be added automatically to
each reaction cup in sequence. Incubate the samples for 45 min, timing
from the addition of reagent D to the first sample. The color formed will be
stable for at least 40 additional minutes after the 45-min incubation period.
To read the samples, restart General Endpoint Program # 1 in mode 3
(reading mode) and enter a factor o f 1.0. The results will be corrected
automatically for the reagent blank and are printed out in absorbance
units. The linear range extends to an absorbance of approximately 0.8,
which corresponds to 100/~g of protein in a 2.5-ml reaction volume. Sam-
ples with absorbances less than 0.025 or greater than 0.8 after correction
for reagent blank should be reassayed after adjusting the sample size to
bring it within the linear range o f the assay. Changing the sample size from
25/xl to 5/zl or 100/xl will not significantly change the total assay volume
and need not be corrected for. This automated assay system has been
proved to be effective for the determination o f protein concentration in
subcellular or lipoprotein fractions separated by sucrose gradient cen-
trifugation, in column fractions during enzyme purification, and in clinical
blood and urine samples.

TABLE I
COMPARISON OF MODIFIED AND ORIGINAL LOWRY PROCEDURES FOR
PROTEIN DETERMINATION IN MEMBRANE AND LIPOPROTEIN SAMPLES a

Sample Lowry procedure Pretreatment A (at 660 nm)

Rat liver Modified None 0.142

homogenate Original Alkali solubilizationb 0.141
Mitochondria Modified None 0. l 13
Original Alkali solubilizationb 0.112
Lipoprotein Modified None 0.296
Original Delipidationc 0.285

a Data from Markweil e t al. 9

Protein was determined after overnight solubilization in 1 N NaOH.
c Protein was determined after extraction of the low-density lipoprotein sample
with ether.
[16] PROTEIN DETERMINATION IN COMPLEX SYSTEMS 301

Use with Complex Biological Systems

The effectiveness and rapidity of this modified Lowry procedure as
compared to the original Lowry procedure 5 for assaying complex biolog-
ical systems is shown in Table I. The absorbance values of liver homoge-
hates and mitochondrial and lipoprotein fractions as determined in the
modified procedure with no pretreatment are the same as those deter-
mined in the original procedure after the recommended extensive alkali
solubilization or delipidation. The inclusion of SDS as a solubilizing agent
in the modified procedure facilitates assay of samples prepared in various
ways: freshly prepared, frozen, lyophilized, acid-precipitated, or ex-
tracted with lipid solvents and dried at 100°. Even with soluble samples, it
offers the advantage of preventing the buildup of denatured protein inside
the cuvette when a large number of samples are read in succession.

Comparison with Dye Binding Method

The dye binding method introduced by Bradford 8 was investigated as
an alternative to the modified Lowry method for rapid determination of
protein in membrane and lipoprotein samples. Color formation by the
dye-protein complex was measured 15 min after reagent addition as rec-
ommended for precise determinations .8 The dye binding method produced
lower protein estimates than the modified Lowry method for all the sys-
tems examined (Table II). The smallest difference (29%) occurred with
low density lipoprotein samples; the greatest difference (150%) occurred
with myxovirus samples. The A : B ratios of 1.52 for total cell samples and

TABLE II
COMPARISON OF MODIFIED LOWRY AND DYE BINDING METHODS FOR PROTEIN
DETERMINATION IN MEMBRANE AND LIPOPROTEIN SAMPLES

Protein concentration (mg/ml)

Sample Modified Lowry (A) Dye binding (B) Ratio A : B

MDBK cells" 4.80 3.15 1.52

Low-density lipoprotein 11.85 9.20 1.29
Myxovirus:
Untreated 0.50 0.20 2.50
Freeze-thawed~ 0.49 0.28 1.75

a Confluent monolayers of Madin-Darby bovine kidney (MDBK) epithelial cells were

homogenized in 5 mM sodium phosphate buffer at pH 7.2.
b Samples of influenza A virus were assayed after being rapidly frozen and thawed three
times.
302 GENERAL ANALYTICAL METHODS [16]
T A B L E III
EFFECT OF SUCROSE, E D T A , DETERGENTS, AND
UREA ON COLOR FORMATION IN THE MODIFIED
LOWRY PROCEDURE

A (at 660 nm)

Addition a of 40/zg of BSA

None 0.184
0.1 M sucrose, 2.5 m M E D T A 0.183
1% Triton X-100 0.183
1% Nonidet P-40 0.188
I% Sodium cholate 0.186
1% Sodium deoxycholate 0.182
1% Empigen BB/P 0.187
1% Empigen B T 0.183
4 M Urea 0.250
0.187 b

a Reagents were added to achieve the final concen-

tration indicated in a l-ml sample containing 40
/~g o f bovine s e r u m albumin.
b T h e sample a b s o r b a n c e value was corrected by
subtracting the absorbance o f an individual re-
agent blank containing 4 M urea.

2.50 for membrane enveloped virus samples are in good agreement with
those obtained in previous comparisons of these two or similar meth-
ods. 12a3 This consistent discrepancy between protein values obtained by
the two methods does not appear to be due to the presence of small
peptides that react with the Lowry reagents, lz The influence of other
factors, such as accessibility of binding sites on the proteins to the dye, is
suggested by the 40% increase in protein value after the virus sample is
frozen and thawed three times. This procedure is known to disrupt the
membrane of enveloped viruses. 14 The protein values as determined by
the modified Lowry method did not change after freeze-thaw treatment of
the virus sample.

Effect of Sucrose, EDTA, Detergents, and Urea on Color Formation

The modified Lowry procedure includes SDS in the alkali reagent and
an increased concentration of copper tartrate. These additions allow assay

12 F. Chiappelli, A. Vasil, and D. F. Haggerty, Anal. Biochem. 94, 160 (1979).

13 M. A. K. Markwell and C. F. Fox, J. Virol. 33, 152 (1980).
14 M. H o m m a , K. Shimizu, Y. L. Shimizu, and N. Ishida, Virology 71, 41 (1976).
[17] rATTY ACID SYNTHASEASSAY 303

of samples containing some reagents that interfere with protein determina-

tion by the original L o w r y method. Inhibition of color development by 0.1
M sucrose and 2.5 m M EDTA was abolished by using the modified
method 9 (Table III). Color development also proceeded normally in the
presence of a number of detergents commonly used in the extraction and
purification of membrane proteins, such as the nonionic polyoxyethylene
alcohols (Triton X-100 and Nonidet P-40), anionic bile salts (sodium cho-
late and deoxycholate), and zwitterionic betaines (Empigen BB/P and
Empigen BT). Because no additional color is generated with these deter-
gents, an individual reagent blank for each one is not necessary. This is
not the case, however, with 4 M urea. With this reagent a correction for
the increased final color is obtained by subtracting the absorbance of an
individual reagent blank containing 4 M urea. Not all effects of interfering
chemicals can be corrected in this manner. Sulfhydryl compounds, such
as dithioerythritol, 2-mercaptoethanol, and glutathione, produce an in-
tense color because of their reducing powers. 15 To r e m o v e these sub-
stances or to concentrate dilute protein solutions, protein is quantitatively
precipitated by the combined use of sodium deoxycholate and
trichloroacetic acid. TM The precipitated protein is then dissolved in reagent
C, and the assay proceeds as usual.
~ c. G. Vallejo and R. Lagunas, Anal. Biochem. 36, 207 (1970).
~GA. Bensadoun and D. Weinstein,Anal. Biochem. 70, 241 (1976).

[17] Assay of Fatty Acid Synthase Using a

Bicyclic Dione as Substrate
By ALAN H. ULLMAN and HAROLD B. WH1TE III

Reactions
0 0
II II
R--C--CH2--C-S--Enz + NADPH + H +--+
OH 0
I II
R - - C - - CHz---- C - - S - - E n z + NADP + (1)
I
H

Copyright (D 1981 by Academic Press, Inc.

METHODS IN ENZYMOLOGY. VOL, 72 All rights of reproduction in any form reserved.
ISBN 0-12-181972-8