A new color reaction for the quantitation of serum cholesterol

=93

CLINICA CHIMICA ACTA

A NEK7

COLOR OF

REACTION SERUM

RONALD L. SEARCY

FOR

THE

QUA~TrTATION

CHOLESTEROL* AND

LOIS

M.

BERGQUIST

Departments of Pathology, Los Angeles Cowaty Osteopathic Hos@al; College of Osteo$athic Physicians and Surgeons, Los Angeles, Calif. (U.S.A .)

(Received August 3rd, rg5g)

Despite numerous variables the LIEBERMANN-BURCHARD color reaction is widely for serum cholesterol estimation. Many modifications of this method have been necessary since the reaction is dependent on temperature’, sensitive to light2, and produces a color which lacks stability3. Recently ZLATKIS ct al.4 have produced a stable chromogen by treating cholesterol with a Feel, reagent. However, this reaction may lack specificity. Proteins, tryptophans, bilirubin’, bromine*, vitamin A8, unsaturated fatty acidslo, hemoglobinll, and steroids12 form interfering colored complexes, many of which absorb significantly in the spectral region suggested by ZLATXIS et aE.4 for cholesterol quantitation. In order to remove interfering proteins HENLY~ modified the ZLATKIS technique by allowing serum to stand for at least 30 min with the FeCl,-acetic acid. Using a more complex procedure, CHIAMORI AND HENRY~~ promote protein precipitation by heating the mixture of serum and reagent to 60” with occasional shaking. In our hands, the complete removal of protein with the FeCl,reagent was often obtained only after vigorous shaking, warming or storage overnight. A new cholesterol color reaction and serum extraction technique have been investigated in order to eliminate errors associated with both the LIEBERMANN-BURCHARD and ZLATKIS procedures. An orange color was formed when cholesterol and a solution of FeSO, in glacial acetic acid were treated with H,SO,. This color could be quantitated at a wavelength less susceptible to interfering substances. A more rapid protein precipitation was obtained when serum was treated with acetone-ethanol rather than an acetic acid reagent. However, most procedures using this extraction technique require that the lipid solvent be removed prior to color development. Such evaporation is not only time-consuming but may lead to charring and possible mechanical technique to the loss of the precipitate. CRAWFORD]~ in applying this extraction ZLATKIS procedure found that the dried extract residues were not readily soluble in the FeCl, reagent, and to overcome this difficulty he resorted to a S-min incubation in a boiling water bath; this, however, lengthens the procedure since the mixtures had to be cooled to room temperature before the addition of the H,SQ,. These distinct disadvantages were obviated when it was found that the acetone-ethanol extracts could be successfully treated directly with the FeSO, reagent to produce the new color reaction. I_-* This work has been supported by a research grant from the United States Public Health used

Service, Project A-3213 and from the Attending Osteopathic Hospital.

Staff Association

of the Los Angeles County

Clin. Chim. Ada, 5 (1960) r9z-rgg

COLOR

REACTION

FOR SERUM

I93

CHOLESTEROL

MATERIALS

FeSO,-acetic

acid reagent

A saturated solution is prepared by adding several grams of Fe,SO. 7H,O to I 1 of reagent-grade glacial acetic acid. After a few minutes of mixing, a clear solution is obtained by filtration. The reagent is stable for many months at room temperature. Concentrated H,SO,

An analytical reagent grade is used. Digitonin

solution

A 0.5% solution is prepared by dissolving 1.0 g of digitonin in roe ml of absolute ethanol at 60”. After the addition of 2.0 ml of 10% acetic acid, the mixture is diluted to 200 ml with distilled water. Acetone-absolute

ethanol

Equal parts are mixed. Acetone-ether

Equal parts are mixed. Cholesterol standard

A 60 mg per IOO ml standard is prepared by dissolving 120 mg of purified cholesterol in 200 ml of acetone-ethanol. The solution is kept tightly stoppered and renewed frequently. METHODS

Determination

of total cholesterol in serum

A I to IO dilution is made by adding 0.1 ml of serum to 0.9 ml of acetone-ethanol. Larger quantities are required if free cholesterol is to be determined. The serumsolvent mixtures are stoppered, mixed well and centrifuged at high speed for several minutes. A o.4-ml aliquot of the clear supernatant is placed in a calorimeter tube and diluted with 6.0 ml of the FeSO, reagent. A z.o-ml portion of H,SO, is blown into the mixture with force to obtain immediate uniformity. IO min are allowed to elapse before calorimetric measurement. The optical densities may be determined at 4go rnp on an instrument such as the Coleman Universal spectrophotometer or on the Klett calorimeter using the No. 50 filter (470 to 530 mr). All readings are made against a blank prepared with 0.4 ml of acetone-ethanol instead of serum filtrate. Determination

of free cholesterol in serum

A r.o-ml aliquot of serum extract is placed in a conical centrifuge tube and mixed well with 1.0 ml of digitonin solution. After at least a I h precipitation period the mixture is centrifuged. The supernatant is discarded and the precipitate is suspended in acetone-ether. The mixture is centrifuged again and the supernatant is carefully discarded. The cholesterol precipitate is then dissolved in 6.0 ml of the FeSO, reagent. Large amounts of digitonide may be put into solution rapidly by warming for a short period at 37”. The solutions are cooled to room temperature before mixing thoroughly Clin. Chim.

Acta, 5 (1960)

Igz-Igg

R. L. SEARCY, L. M. BERGQUIST

I94

with 2.0 ml of H,SO,. The reaction mixtures are transferred to calorimeter tubes and the optical densities are measured as described above. Standardization

technique

Dilutions of stock cholesterol are prepared as shown in Table I. Colorimeter tubes are used for the total cholesterol determinations and conical centrifuge tubes for the precipitation of digitonides. TABLE THE PREPARATION

OF STANDARD OF TOTAL

Stock cholesterol ml ..__.

AND

Acetoneethanol ml

0.1

0.4 0.3

0.2

0.2

0.3 0.4 ~____~

0.1

0.0

0.0

I

SOLUTIONS FREE

FOR

THE

DETERMINATION

CHOLESTEROL

Equivalent total cholesterol mg% ~__~._ Cl 150 300 450 600

Equivalent

free cholesterol mg% 0 60 I20

180 240

RESULTS ANDDISCUSSION Color formation

with cholesterol, its acetate and digitonide

Aliquots of free cholesterol, cholesterol acetate and cholesterol digitonide dissolved in acetone-ethanol were diluted with 6.0 ml of FeSO, reagent. Color was produced by mixing the solution with 2.0 ml of H,SO,. The absorption spectra of the various forms of cholesterol were then measured on a Coleman Universal spectrophotometer between 400 and 700 m,u (Fig. I). It was impossible to distinguish between the absorbence of the color produced by free cholesterol and its acetate. Both the free sterol and its ester gave a sharp absorption peak between 480 and 5oo m,u. The digitonide of cholesterol demonstrated a slight spectral displacement. The absorption maximum occurred near 490 rnp. This color difference is best explained by yellow reaction products derived from digitonin. ROSENTHALANDJUD” also observed that digitonin developed color when mixed with FeCl,, acetic acid and H,SO,; the digitonin chromogens exhibited an absorption maximum at 500 m,u but contributed minimal interference in the regions near 490 and 550 mp. Therefore, these wavelengths are best suited for the quantitation of the cholesterol color in the presence of this alkaloid. Despite the presence of digitonin, ROSENTHALANDJUD” were able to demonstrate that the cholesterol color followed Beer’s law. The e&2

of acid concentration

and time @on chromogen formation

Reaction mixtures were prepared with 180 pg of cholesterol so that the final H,SO, concentrations were 12.5, 25, 37.5 and 50%. After the addition of H,SO, the optical densities were measured at timed intervals on the Coleman Universal spectrophometer at 490 rnp (Fig. 2). The most intense color developed almost immediately in the solution containing 25% H,SO,. Increasing the acid concentration diminished color formation. With a 5 : 3 color reagent-H,SO, ratio only 60% of maximum color was produced. Likewise, color was reduced with an acid concentration below 25%. Clin. Chim. Acta, 5 (1960) 192-199

COLOR REACTIONFOR SERUM CHOLESTEROL

195

No significant chromogen was detected using a color reagent-H&SO, ratio of 7 : I (i.e. 12.5% H&O, in the final mixture). The proposed technique for cholesterol quantitation utilizes relatively small amounts of H,SO, for color’ production as compared to other procedures. The color reagent employed by CARR AND DREKTER~~contains 50% H&SO,, while that of ZAK et aLI has an acid level of 40%. The use of lower concentrations of acid is advantageous since charring is less likely. In addition, the handling of minimum amounts of this corrosive acid is always desirable. After the addition of H,SOd, color production was extremely rapid in all mixtures except at the 12.5% concentration. Maximum color intensity was effectively

b

0.2 -

400 440 480 520 560 600 640 680 720 Wavelength

I

I” t-np

Fig. I. The absorption curve of (a) free cholesterol or its acetate and (b) cholesterol digitonide.

Fig. a. The time-absorbence curves of cholesterol-FeSO, reagent mixtures containing (a) 25%. (b) 37.5%, and (c) 50% &SO,.

reached in 8 to 15 min using the other acid levels and was relatively constant during this period. The optical density varied only 0.006 units in this time interval when 25% sulfuric acid was used. Conversely, the lack of stability of the LIEBERMANN-BURCHARDcolor has been repeatedly demonstrateda, 17. CARR AND DREKTER*~using this color reaction observed that intensity changed continuously with time, and suggested that optical density measurements be made 15 to 25 min after the reaction is initiated. However, analysis of the time-absorbence curve reported by these investigators shows that the color intensity decreases 20 min subsequent to reagent mixing. On the other hand, ZAK’S methodX8, employing the FeCl, reagent, produces a stable color when reaction mixtures are allowed to stand at room temperature for 30 min. More recently ZAK et aLIs have suggested that readings can be made IO to 15 min after the reagents are combined. However, CRAWFORD’~ has noted that the optical density changed at a rate of 0.02 Cl&. Chim.Acta, 5 (1960) 192-199

R. L. SEARCY, L. M. BERGQUIST

196

units per min and did not reach a maximum for 30 min. Therefore, a half-hour color development period may be essential to obtain accurate measurements with the FeCl, reagent. The proposed technique using FeSO, not only produces a stable color, but color intensity reaches a maximum in only IO min. Cooling of the FeSO, reaction mixtures to room temperature is unnecessary. The e&et

of temperature

upon

color

formation

The temperatures of the reaction mixtures containing the various amounts of H,SO, were measured at timed intervals (Fig. 3). During the first IO set of mixing, the heat evolution reached a maximum. Furthermore, the reaction temperatures rose

l-l

4oLLLt 12.5

25.0

-

1

%

37.3

-

Sulfuric

40 :id

Fig. 3. The effect of increasing acid concentrations on maximum temperatures of FeSO, reaction mixtures.

80

160 200 240 /q cholesterol

120

Fig. 4. Standard curves obtained with cholesterolP FeSO, reagent mixtures on (a) the Coleman spectrophotometer at 4go rnp and (b) the Klett calorimeter using filter No. jo.

when increasing quantities of H,SO, were used. With the ~z.jo/b acid concentration a maximum reaction temperature of 47” was reached while a 75” maximum was obtained with 50% H,SO,. These data demonstrate that the acid concentration which produced the most intense color did not release the greatest amount of heat. Therefore, color production using the proposed method is not entirely temperature-dependent. On the other hand, it has been well established that the LIEBERMANN-BURCHARD reaction is extremely sensitive to temperature. COOK et al.” have found a sterol fraction which forms the LIEBERMANN-BURCHARD color rapidly at o0 and another which reacts readily at 25’. The LIEBERMANN-BURCHARDmixtures as prepared by CARR AND DREKTEP reached temperatures ranging from 50 to 60”; they were subsequently cooled and maintained at 25”. These authors suggest that this incubation must be regulated within I’ during color formation. ZAK and co-worker9 obtained a reaction temperature of 67’ when using the color reagent-H&GO, ratio of 3 : 2. In our hands ZAK’S reagents, used in a ratio of 5 : 3, produced a maximum temperature of approximately 70’. In the proposed method a temperature of 59” was reached with

COLORREACTION FOR SERUM CHOLESTEROL

I97

the optimal FeSO, reagent mixture. This is nearly 20% lower than the ZAK reaction temperature. This decreased heat production may be advantageous since ROBINSON AND PUGH~~ have reported charring at the acid-color reagent interface using ZAK’S technique. Moreover, the color which originates from charring may be difficult to detect in the presence of this purple color. It appears likely that the lower reaction temperature produced with the FeSO, reagent may significantly reduce the charring tendency. Optical density comparisons were made of FeSO, reagent mixtures held at 60” with those allowed to remain at room temperature during cholesterol color development. These results did not warrant the maintenance of a constant temperature during color formation. Similar findings have been made by ZAK with the FeCl, reagent, but LANGAN et aLa using the same color reaction suggest that a controlled temperature yielded more reproducible results. Reproducibility did not appear to be a function of temperature when FeSO, was employed. Color production

with increasing

cholesterol concentrations

Amounts of cholesterol ranging from 20 to 240 pg were transferred to colorimeter tubes and diluted to 0.4 ml with acetone-ethanol. Color was developed in the usual manner and the optical densities were measured at 4go m,u on a Coleman Universal spectrophotometer. A similar standard curve was established on a Klett calorimeter using the No. 50 filter (470 to 530 mp). Fig. 4 demonstrates that Beer’s law was followed with these amounts of cholesterol,at the specified wavelengths. An attempt was made to quantitate the cholesterol color on the Klett using the No. 54 green filter (510 to 570 m,u) which is commonly available with the instrument, but a linear response could not be obtained with the cholesterol concentrations employed. It was possible with the proposed technique to measure accurately cholesterol concentrations below IOO pg. Therefore, it is practicable to use 0.1 ml or less of serum for the quantitation of cholesterol. A similar sensitivity was obtained by ZAK’~. However, it is difficult to obtain measurable color with such small quantities of cholesterol using the LIEBERMANN-BURCHARD technique. Therefore, methods based upon the LIEBERMANN-BURCHARD color reaction require as much as 1.0 ml of serum for cholesterol determination. Demonstratiop& of accuracy based Zcpolzcholesterol determination

in standard solutions

Purified cholesterol and its acetate were mixed in acetone-ethanol in quantities equivalent to a total of 200 to 300 mg%. These two sterols were combined so that 20% to 94% of the total cholesterol was free. The amounts of total and combined cholesterol were then determined by the proposed technique (Table II). An average difference of only 3% was found when comparisons were made between total cholesterol recoveries and the theoretical values. Individual differences were not lower than 5% or higher than 10% of the theoretical values. Amounts of free cholesterol recovered as digitonide varied from -1% to +IO% of the known quantities. The average error in the measurement of this cholesterol fraction amounted to less than 4%. These data demonstrate an accuracy well within the limits necessary for clinical or research analyses.

Clin. Chim. Acta,

5

(1960)lgz-lgg

R. Ix.SEARCY,

198

THE ANALYSIS ___.-_

OF CHOLESTEROL ~~~~es~e~ol

“_.___._

Sample number

L. M. BERCQUIST

AND

340

2

320

3

__~mgqb .--340 280

24"

ditference

200

280

260

334 278

-t-4 --i

3x2

-3

284

-+ *

288

-4 -1

240 240

“_ 3 .--3

200

0 -

6

120

240

211

3-3 -3

232 220

7 8

200

_ ---I

.._-5

272

160

2

303

272

5

.__ ._~

difference

free

. ..--.___ % -__ ?E_. 0 jr6

298 4

MIXTURES

~_. --__...______.to&l

300

ACETATE

~~oZeste~o1recovered

added

total w% ~_~ __.. . ^_ _._-.1

CHOLESTEROL

0 0

r9G

-2

220

-i-IQ

1.3

--.I

I?0

-tf)

"4

---5

120

0

80

212

_- 4

88

$10

-1

40

218 198

--I

78 44

-+-IO

214

f7

38

-3 -_$

Fifteen serum specimens, many of which were lipemic, icteric or hemolyzed, were randomly selected from hospital patients. Table III shows that the serum total cholesterols in this group ranged from approximately 80 to 360 m&G. The differences between duplicate total cholesterol analyses averaged less than 4 mg% while the di~tonide determinations were in even closer agreement. TABLE DUPLICATE

SERUM

TOTAL.

_ _-_l-_______

Total ckolestmol r,3gy1 WSC 3.3~~ 225 300

367 229 300

276

271

196

205

229

229

87 x58 330

79 158 327

253

198 284 I74

288 x74

AND

FREE

III CHOLESTEROL

DigemncF

mg% _..._-

DETERMINATIONS

Free cholesterol M’fi’l/o m.gy~

11

138

4

102

0

5 4 0

IXjiPrenre I?@$

141

3

95

99 97

3 2

83 87 59

85 83 55

2 4 4

37 59 55

32 63 53

5 4 2

26 51: 138

28

2

-I Ii4

0 4

65

5

59 120

I26

0

47

5”

4

Clin. Chim.

Acta,

_

5 (1960)

. r92-~99

COLORREACTIONFOR SERUH CHOLESTEROL Determination

I99

of normal values

Blood samples were obtained from 200 apparently healthy blood donors. Using the proposed technique the serum total cholesterol determinations ranged from 88 to 354 mg”/bwith an average of 177 mg%. In reviewing current techniques, COOK” points out the general lack of agreement as to the normal range for serum total cholesterol. MERTENS AND ALRERS~~report normal values from IOO to 250 mg% while SPERRY ANDSCHOENHEIMER~~ found normals to range from about 150 rng% to above 350 mg”/b. The amounts of free cholesterol in the 200 serum samples were also quantitated as the digitonide. This fraction averaged 53 mg% with a range of 18 to 154 mg%. This average free cholesterol value obtained is in close agreement with that of 50 mg% reported by BoYD~~. SZJMMARY A new color has been formed by mixing cholesterol with FeSOp, glacial acetic acid and H,SO,. This reaction provides a sensitive method for the quantitation of total and free cholesterol using 0.1 ml or less of serum. Proteins may be removed from serum by extraction with acetone--ethanol. These extracts may then be treated directly with the color reagents thereby eliminating time-consuming evaporation procedures. The cholesterol color is stable and can be produced rapidly with small quantities of H,SO,, Color formation occurs at a relatively low temperature which need not be rigidly controlled. The use of low concentrations of H,SO, and the production of minimal amounts of heat make the serum extracts less susceptible to charring. The accuracy and precision obtained with the proposed procedure validates its use for either clinical or research purposes. The serum total and free cholesterol levels of zoo blood donors measured with the new color reaction were found to agree well with normal values reported by others. REFERENCES I A. SAIFER AND 0. F. KAMMERER,J. Biol. Chem., 164 (‘946) 657. 2 J. G. REINHOLD, Am. J. Clin. Pathol., 6 (1936) 31. 3 P. G. SCHUBE, J. Lab. Clin. Med., 18 (1932) 306. 4 F. ZLATKIS, B. ZAK AND A. J. BOYLE, J. Lab. Clin. Med., 41 (1953) 486. 6 A. A. HENLY, Analyst, 82 (1957) 286. 6 F. G. HOPKINS AND S. W. COLE, Proc. Roy. Sot. London, 3, 68 (1901) 21. 7 I. MACINTYRE AND M. RALSTON, Biochem. J., 56 (1954) XLIII. * E. W. RICE AND D. B. LUKASIEWICZ, Clin. Chem., 3 (1957) 160.

9 L. J. KINLEY AND R. F. KRAUSE, PYOC.Sot. ExptZ. Biol. Med., 99 (1958) 244. 10 D. N. RHODES, Biochem. J., 71 (1959) 26. I1 H. L. ROSENTHAL AND L. JUD, J. Lab. Clin. Med., 51 (1958) 143. 12 F. BISCHOFF AND J. G. TURNER, Clin. Chem., 4 (1958) 300. 18 N. CHIAMORIAND R. J. HENRY, Am. J. Clin. Pathol., 31 (1959) 305. 14 N. CRAWFORD,Clan. Chim. Arta, 3 (1958) 357. 15 J. J. CARR AND I. J. DREKTER, C&n. Chem., 2 (x956) 353. fe B. ZAE, R. C. DICEENMAN, E. G. WHITE, H. BURNETT AND P. J. CHERNEY, Am. J. Cl&. Pathol., 24 (1954) 1307. I7 R. P. COOK,ChoEesterol:Chemistry, Biochemnistvy and Pathology, Academic Press, New York, x958. Is B. ZAK, Am. J. C&n. Pathol., 27 (1957) 583. 19 B. ZAK. D. A. Luz AND M. FISHER, Am. J. Med. Technol., 23 (1957) 283. 1o L. G. ROBINSON AND E. R. PUGH, U.S.Armed Forces Med. J., g (1958) 501. *I T. A. LANGAN, E. L. DURRUM AND W. P. JENCKS, J. Clin. Invest., 34 (1955) 1427. z2 E. MERTENS AND G. ALBERS, Z. physiol. Chem., 293 (1953) 244. 2s W. M. SPERRY AND R. SCHOENHEI~ER,J. Biol. Chem., I 10 (1935) 655. 24 E. M. BOYD, J. Biol. Chem., 143 (1942) 131. C&n. Chim. Acta, 5 (1960) rgz-rgp