A MICRO BIURET METHOD F O R PROTEIN DETERMINATION

DETERMINATION OF TOTAL PROTEIN JN CEREBROSPINAL FLUID

BY J. GOA
From the Ccntral Laboratory, Sahlgren’s Hospital, Gothcnbicrg and the Dcpartmcnt of Clinical
C’hrinistry. A1 cdical School in Gothorburg, Szrvdrn
(Received for publication Dec. 10, 1952)

Studies on the protein components in bio- position of the reagents, amount of protein
Scand J Clin Lab Invest Downloaded from informahealthcare.com by McGill University on 09/27/13

logical fluids often require determination of to be determined and choice of wavelength

small amounts of protein. The quantitative for reading. The method is here applied to
methods available for this purpose are determination of total protein in cerebro-
usually time-consuming or their sensitivity spinal fluid, but can also be used for other
is not satisfactory. protein determinations. A micro procedure
One of the most convenirnt methods for is described for quantities of 0.1 to 2.0 mg
the determination of proteins is the biuret protein and an ultramicro procedure for
method. Of the numerous modifications of quantities of 10 to 400 micrograms.
this principle none of them are worked out
for micro estimation, i. e . estimation of about
For personal use only.

MICROMETHOD
0.1 mg of protein. The reason is, at least in
(0.1-2.0 mg protein)
part, that with the photometers commonly
Reagents and apparatus
used the readings were limited to the visible
part of the spectrum. Biuret has an absorp- 1. 3 per cent solution of sodium hydroxide.

tion maximum at 560 m p and in all the 2. 11 per cent solution of trichloracetic acid.
3. Benedict’s reagent.
biuret methods the readings are made in this
Dissolve in a 1 litre beaker 173 g sodium
region. Minimum absorption for the biuret
citrate (purum) and 100 g sodium carbonate
color is at 450 m p and with shorter wave- (sicca) by heating, boiling must be avoided.
lengths the absorption increases again as Filter the solution warm.
shown by Mehl (1945). In the region of Dissolve in another beaker 17.3 g cupric
400-320 m p the increase is moderate, sulphate (CuSO4. 5 H z 0 ) in about 100 ml of
thereafter the absorption increases to in- distilled water by heati‘ng. Mix the two solu-
finity. Readings of the biuret color at tions in a 1 litre volumetric flask, cool and fill
to the mark with distilled water. Keep the solu-
330 m p follow the Lambert-Beers law and
tion in a dark bottle with rubber stopper.
as the light absorption at this wavelength
4. Photometer. I n this investigation a Beckman
is far stronger than at 560 mp, small spectrophotometer model B and 1 cm corex
amounts of biuret can be distinguished. cuvettes are used. Other spectrophotometers
The method now to be described differs which permit readings a: 330 m!c might also
from other biuret methods only in the com- be used.

218
 A MICRO BIURET METHOD FOR PROTEIN DETERMINATION 219

Procedure
The total protein content of cerebro-
spinal fluid cannot be determined directly
without previous precipitation because some
of the non-protein nitrogen compounds also
give a biuret color. The precipitation is car-
ried out in a final concentration of 8.8 per 50 100 150 200
MS PtI CENT PROTEIN
cent trichloracetic acid according to IzikowitL
( 1941).
Fig. 1. Standard curve. Readings a t 330 m!c and
1 cm corex cuvettes. 30 values obtained from dilu-
Scand J Clin Lab Invest Downloaded from informahealthcare.com by McGill University on 09/27/13

Mix in a conical centrifuge tube 1 ml of cere- tions of six different sera are plotted.
brospinal fluid with 4 ml of 11 per cent trichlor-
acetic acid. Let stand for about ten minutes at
The correlation between the nitrogen con-
room temperature and centrifuge for 15 minutes at
2500 rimin. Discharge the supernatant fluid by in- tent and the biuret color was linear and in-
verting the tube slowly and let the tube drain off dependent of the alb./glob. ratio. The latter
for some minutes on a filter paper. Dissolve the was determined by paper electrophoresis
precipitate in 4 ml of 3 per cent sodium hydroxide, (Goa 1951) and varied from 0.11 to 1.23.
add 0.2 ml of Benedict’s reagent and mix. Prepare
a blank by mixing 4 ml of 3 per cent sodium
The standard curve was calculated by the
hydroxide with 0.2 ml of Benedict’s reagent.. Read method of least squares. The equation of the
the optical density against the blank after 15 minu- curve is:
For personal use only.

tes in a Beckman spectrophotometer model B at

330 mu and 1 cm corex cuvettes.
1 cm
D330 m p
= 0.02 + 0.0029 P
The result in mg per cent is read directly 1 cm the optical density at 330 m p
where 330 m u
on a standard curve or is calculated by means
and a light path of 1 cm and P is nig per
of the equation:
cent protein in the sample. In order to cal-
F‘=345 D - 7 culate the protein content from the optical
where P is mg per cent protein and D the density the equation can be written :
optical density. P = 3.45 D -0.07
Where P is mg protein in the sample.
Standard curve
To get a standard curve the nitrogen con- ULTR.4 M I C R O M E T H O D
tent of six sera was determined by the micro- (10 to 400 micrograms protein)
Kjeltlahl method and the usual factor of 6.25 With the method described, amounts of
was used to convert the nitrogen to protein. protein from 0.1 to 2.0 mg can be deter-
Corrections were made for the non-protein mined. If it is necessary to determine smaller
nitrogen. amounts the sensitivity of the method can be
Dilutions 1 : 100 to 1 : 500 of the sera in increased by using microcuvettes and carry-
0.9 per cent sodium chloride were analyzed ing out the reaction in a volume of 0.5 ml.
and the readings were plotted in a diagram I n this way amounts of protein from 10 to
against the protein content. (Fig. 1.) 400 micrograms can be determined.
 220 J. GOA

Rcagents and apparatus

1. 26 per cent trichloracetic acid.
2. blicrocuvettes of the dimensions 2.5 X 10
x 25 mm. Otherwise as for the micromethod.

Procedure
1 hen the protein in very diluted solutions,
as in cerebrospinal fluid, is to be determined,
precipitate the protein by half the sample Fig.2. Absorbtion spectra of the biuret color, the
volume of 26 per cent trichloracetic acid in reagents and protein, all in equivalent amounts.
Scand J Clin Lab Invest Downloaded from informahealthcare.com by McGill University on 09/27/13

Biuret ___ , reagents - --, protein . . . . . . . .

small conical centrifuge tubes containing
about 2 ml. For example: 0.2 ml of cerebro- DU and ultraviolet accesories were used and
spinal fluid and 0.1 ml of 26 per cent tri- in the region of 320 to 700 m p the readings
chloracetic acid. Centrifuge for 15 minutes were also carried out with Beckman spectro-
at 2500 r/min., then let the supernatant fluid photometer model B. The solution of biuret
drain off by inverting the tubes. Droplets on consists of 4 ml of 3 per cent sodium
the wall of the centrifuge tubes are removed hydroxide, 1 mg of precipitated protein and
by means of a capillary pipette and suction. 0.2 ml of Benedict's reagent. The solution
Dissolve the precipitate in 0.5 ml of a mix- for the absorbtion spectrum of the reagents
ture of 4 ml 3 per cent bodium hydroxide
For personal use only.

consists of 4 ml of 3 per cent sodium

and 0.2 ml of Benedict's reagent. After 15 hydroxide solution and 0.2 ml of Benedict's
minutes the solutions are transferred into reagent. The solution for the protein absorb-
the micro-cuvettes using a capillary pipette tion spectrum is the same as that for the
and read at 330 m p against the mixture of biuret but with 0.2 ml of water instead of
sodium hydroxide and Benedict's reagent as Eenedict's reagent.
a blank. The cuvettes must be carefully Fig. 2 shows the absorbtion spectra. The
centered. biuret color has an absorbtion maximum at
A standard curve was performed as for 560 m p within the visible part of the spec-
the micromethod. The equation for the stan- trum but a far stronger light absorbtion in
dard curve of the ultra micromethod is : the region of 330 m p and lower wavelengths.
D = 0.03 4- 2.19 P The absorbtion of the copper-NaOH-mix-
where D is the optical densit$ and P is mg ture has a maximum at 630 mp, for lower
of protein in the sample. wavelengths it follows the biuret absorbtion
but is lower. The absorbtion of the protein
EXPERIMENTAL is negligible for wavelengths higher than
Absorption spectra. Selection of wavelength 320 mp, increasing at lower wavelengths and
Determinations of the absorbtion spectra reaching a maximum at 280-290 mp. The
of biuret, protein and reagents are all carried wavelength 330 m p has therefore been selec-
out in a volume of 4.2 ml and read against ted as most convenient for reading the biuret
water. A Beckman spectrophotometer model color. Fig. 2.
 A MICRO B I t i R E T METHOD F O R PROTEIN DETERMINATION 22 1

the biuret color with increasing temperature,

Ob but that the ordinary variations in room
' 1 . 8 ~ ~
temperature are of negligible influence.
05
./- Fig. 3 shows some experiments carried
. 04. out in order to follow the color development.
The procedure is the same as that for the

;
determination of total protein. As seen, the

y,; ,Q,5"6. ,
color is practically fully developed in 15
niinutes at room temperature and a slow in-
crease in density follows during the first
Scand J Clin Lab Invest Downloaded from informahealthcare.com by McGill University on 09/27/13

5 10 15 20 25 30 MINUTES hour, more marked for high protein content

than for low.
Turbidity often develops in the samples
Fig. 3. Color development as a function of time.
The vertical dotted line indicates 15 minutes, when when standing. To avoid this error the
- as seen - the color is fully developed. author has tried different reagents and
varied the alkalinity of the solutions. The
turbidity is not caused by the copper reagent
Ucvclopmcnt and stability of the biurct color as it develops even in pure protein-sodium
Previous investigations on this problem hydroxide solutions. 'With decreasing alkal-
have all been concerned with the visible part inity the development of the turbidity is
For personal use only.

of the spectrum. Robinson Sr Hogden (1940) delayed and with 3 per cent sodium hyd-
have found the density values at 560 nip, roxide as in the described method the solu-
measured at various time intervals up to 48 tions remain clear for several hours. The
hours, to be nearly constant. Measured at nature of the turbidity is not known. I t is
700 n i p the density value decreases. This is not a protein, it contains phosphorus and it
in agreement with the measurements of is soluble in alcohol. Possibly it is a phos-
Jesserer (1936) and Sizer (1937). These pholipid liberated from the lipoproteins as
authors suggest that the density at 560 inp assumed by Lewin gL Braucr (1951).
which is stable, is due to the biuret color and
that the decrease in density at 700 n i p is
due to loss of cupric hydroxide from the Reproducibility of the imfhod
system. Twelve determinations on two cerebro-
Hinsberg & Gleiss (1950) developed the spinal fluids, one with low and one with
color in a water bath at 37' C during 30 high total protein content were carried out.
minutes and Lehmann (1944) got a nearly The first series gave an average of 38 mg
stable color after 10 minutes at room tem- per cent, lowest value 37 mg per cent, highest
pertaure. The biuret color as subject to value 40 mg per cent and a standard devia-
temperature variations has been investigated tion of k 1.3 nig per cent protein. The
by Josephson 8r AndurPn (1949). ITsing second series gave an average of 84 mg per
their method they found a small increase of cent, lowest value 52 mg per cent, highest
 222 J. GOA

value 86 mg per cent and a standard devia- Izikowitz, S. : hlethodological and clinical studies
tion of z k 1.5 mg per cent protein. As seen, on total protein, globulin and albumin concen-
trations in lumbar fluid. Stockholm 1951.
the standard deviation of estimate is of the
Jesserer, H. : Studien zur Biuretreaktion. Biochem.
same order of magnitude for low as well as Ztschr. 287, 71, 1936.
for high protein content. Josephson, B. & AndurCn, H.: A micromethod
for the colorimetric detemination of total
serum proteins and of serum albumin and glo-
SUMMARY
bulin. Acta Paediat. 38, 335, 1949.
A rapid and convenient colorimetric me- Lehmann, J. : Eiikel biuretmetod for fraktionerad
thod for the determination of small amounts proteinbestamning i lumbalvatska, ex- och
of protein, based on the biuret reaction and transudat samt urin med anvandande av Pul-
Scand J Clin Lab Invest Downloaded from informahealthcare.com by McGill University on 09/27/13

frich Stufenfotometer. Nord. med. 25, 604,

readings at 330 mp, is described. The
1945.
“micromethod” allows the determination of Lewin, R. & Brauer, R. W.: The biuret reaction
10 to 200 mg per cent protein with an for the determination of proteins. An im-
acurassy of 1.5 mg per cent. An “ultra proved reagent and its application. J. Lab. &
micro” modification allows determinations Clin. Med. 38, 474, 1951.
Mehl, J. W.: The biuret reaction of proteins in the
of 10 to 400 micrograms protein.
presence of ethylene glycol. J. Biol. Chem.
The method is applied for the determina- 157, 173, 1945.
tion of total protein in 1 ml of cerebrospinal Robinson, H. W. & Hogden, C. G.: The biuret
fluid. reaction in the determination of serum proteins.
J. Biol. Chem. 135, 707, 1940.
For personal use only.

REFERENCES Seizer, I. W.: The biuret and ninhydrin tests for

Hinsberg, K. & Gleiss, J.: Uber die Bestimmung proteins as measured with Hardy’s spectro-
der Proteine in1 Liquor Cerebrospinalis. Kliu. photometer. Proc. SOC.Exper. Biol. & Med.
Wchnschr. 28, 444, 1950. 37, 107, 1937.