An Enzymatic Approach to Lipoprotein
BERNARD W. STEELE, M.D., DONALD F. KOEHLER, M.S., MT,
KANTA KUBA, M.S., AND MIGUEL M. AZAR, M.D., PH.D.
Steele, Bernard W., Koehler, Donald F., Kuba, Kanta, and
Azar, Miguel M.: An enzymatic approach to lipoprotein quanti-
fication. Am J Clin Pathol 73: 75-78, 1980. Lipoprotein
cholesterol levels were determined without ultracentrifugation
by using modified enzymatic methods for cholesterol, high-
density-lipoprotein (HDL) cholesterol, and triglyceride and the
formula, low-density-lipoprotein (LDL) cholesterol = total
cholesterol HDL cholesterol - triglycerides/5. The methods
for cholesterol and triglyceride determinations were standard-
ized for accuracy and precision by the Center for Disease
Control's Lipid Standardization Laboratory, which monitored
this laboratory for 16 months. The lipoprotein cholesterol
values obtained correlated well with lipoprotein cholesterol
values determined at the Minnesota Lipid Research Clinic
Laboratory using ultracentrifugation. LDL cholesteroi deter-
mined at the Minneapolis Veterans Administration Hospital
Laboratories (Y axis) produced a curve with an intercept of
9.38 mg/dl, a slope of .977, standard error of the estimate (Sy.x)
of 8.8 mg/dl, and a correlation coefficient (r) of .983 (n = 32).
HDL cholesterol was Y = 0.998 X + .89 mg/dl, Sy.x = 1.6 mg/dl
(r = .984, n = 53), and very-low-density-lipoprotein (VLDL)
cholesterol was Y = 1.010 X - 1.32 mg/dl, Sy.x = 1.3 mg/dl
(r = .996, n = 54). (Key words: Very-low-density lipoprotein;
Low-density lipoprotein; High-density lipoprotein; M n + ! -
heparin precipitation; Cholesterol; Triglyceride.)
QUANTITATIVE LIPOPROTEIN ANALYSIS has
increased in popularity as an important factor in pre-
dicting the probability of coronary heart disease. Per-
sons with reduced high-density lipoprotein (HDL) or
increased low-density lipoprotein (LDL) levels7-11-17-19
are at increased risk of having coronary heart dis-
ease. Conversely, individuals with elevated HDL
levels, 1 0 " 1 3 1 8 as well as those with lower LDL levels, 10
are likely to be freer of coronary heart disease.
These findings, linking the incidence of coronary
heart disease with lipoprotein levels, have created a
demand for simple yet accurate and precise methods of
quantification.
The newer enzymatic methods for measurement of
cholesterol and triglyceride 2 " 514-20 meet this need by
eliminating extraction and harsh chemicals, and are
more rapid than older methods. 110 These methods
have also been shown to correlate with the long-
established Lipid Research Clinic methods. 22,24 To
Received October 10, 1978; received revised manuscript and
accepted for publication January 4, 1979.
Address reprint requests to Mr. Koehler: Laboratory Service,
Veterans Administration Medical Center, Minneapolis, Minnesota
55417.
0002-9173/80/0100/0075 $00.70 American Society of Clinical Pathologists
75
Quantification
Laboratory Service, Veterans Administration Hospital,
Minneapolis, Minnesota, and Department of Laboratory
Medicine and Pathology, University of Minnesota,
Minneapolis, Minnesota; Department of Pathology,
University of Miami/Jackson Memorial Hospital Medical
Center, Miami, Florida; Minnesota Lipid Research Clinic
Laboratory; Department of Medicine, University of
Minnesota, Minneapolis, Minnesota
document accuracy and long-term precision, the Lipid
Research Clinic laboratories are required to be stand-
ardized by the Center for Disease Control's (CDC)
Cooperative Cholesterol and Triglyceride Standardiza-
tion Program. We also participated in the CDC's pro-
gram, but used modified enzymatic methods for choles-
terol and triglyceride analyses. 14 - 20 These methods
were standardized for accuracy and precision by CDC
for a 16-month period, using the same criteria applied
to Lipid Research Clinic laboratories. To determine
HDL cholesterol, this laboratory used the standard
Lipid Research Clinic precipitation technic 8,13 and our
standardized enzymatic cholesterol method. 20
This report demonstrates that lipoprotein fractions
can be accurately and easily measured in the routine
clinical laboratory by use of the calculation developed
by Friedewald, Levy and Fredrickson 9 and enzymatic
methods.
Methods and Materials
Subjects
Specimens were obtained from fasting subjects (12-
14 hours) screened at the Minnesota Lipid Research
Clinic. Blood was drawn using Vacutainer M tubes con-
taining EDTA (1.5 mg/ml) according to Lipid Research
Clinic protocol and stored at 4 C prior to analysis. 15
Triglyceride values for these subjects ranged from 76 to
447 mg/dl. None of the plasmas contained chylomicrons
after overnight refrigeration.
Analytic Procedures
Enzymatic Determinations. An Abbott Bichromatic
Analyzer (ABA-100, Abbott Laboratories, South
Pasadena, Cal. 91030) was used to assay for cholesterol
and triglyceride using enzymatic methods.
 STEELE ET AL. A.J.C.P. January 1980
76
Table I. Precision Data free glycerol values to correct for background reaction
rate. Triglyceride results were corrected for free
Target SD (mg/dl)
Value* Mean glycerol to obtain the final glycerol-corrected triglycer-
(mg/dl) (mg/dl) Overall Within-day ide concentrations, reported as mg/dl triolein.14
Lipid Research Clinic Determinations. Lipids were
Total cholesterol
Specimen Q3 154 152.04 2.68 0.70 assayed and lipoprotein fractions isolated at the Min-
Specimen 64 296 291.71 2.36 1.46 nesota Lipid Research Clinic Laboratory as described
Pool 179.85 2.23 in the Lipid Research Clinic's Procedure Manual."''
HDL cholesterol Plasmas and lipoprotein fractions were extracted
Specimen aQ2 54.lt 52.52 1.45 0.76 with isopropanol and treated with zeolite. The extracts
Pool 37.67 1.06
were then analyzed for cholesterol and triglyceride
LDL cholesterol using the Technicon AutoAnalyzer II (Technicon
Pool 117.61 2.47
Instruments Corporation, Tarrytown, N. Y. 10591).
VLDL cholesterol In order to obtain results comparable to the CDC refer-
Pool 24.61 0.39
ence cholesterol procedure, the Lipid Research Clinic
Triglyceride Laboratory uses an accurately labeled 1 serum cali-
Specimen Q3 69 61.77 1.99 1.72
Specimen 64 158 157.56 2.43 1.90 brator provided by the CDC as a secondary standard.
Pool 123.05 1.93 Cholesterol was determined colorimetrically with
Lieberman-Burchard reagent; triglyceride was de-
* Established at CDC with their reference methods.
t Value obtained from Minnesota LRC Laboratory. termined fluorimetrically with periodate and acetyl-
acetone reagents.
Specimens were assayed for cholesterol by a modifi- Plasma HDL Determinations. Plasmas at both lab-
cation of the method of Allain and Poon. 2 The ABA- oratories were treated with heparin and MnCI2 to pre-
100 was set to dispense a 1:51 ratio of specimen (10 /x\) pare alpha fractions by the method of Burstein and
to total reaction volume (510 /i\). "A Gent" choles- Samaille. 6 The alpha fractions were assayed for choles-
terol reagent (Abbott Laboratories) was reconstituted terol at each laboratory by their respective methods.
with 4 mmol/1 aqueous EDTA. Also, a sample of saline The uncorrected results obtained, multiplied by 1.09
solution, prepared with heparin and manganous chlo- to correct for reagent dilution, yielded the final results,
ride reagents, was assayed as an alpha-fraction reagent reported as mg/dl HDL cholesterol.
blank during test runs assaying for HDL cholesterol.
Crystalline cholesterol (99.9% pure, J. T. Baker
Chemical Co., Phillipsburg, N. J. 08865) dissolved in Table 2. Results Obtained in Maintaining Standardiza-
isopropanol was used as standard for each run made tion during the CDC Cooperative Cholesterol and
with the ABA-100.20 The occasional very lipemic Triglyceride Standardization Program
whole plasma was diluted 1:2 with isotonic saline so- Acceptable Value Value Obtained
lution prior to analysis.
Range* SDt Mean SD
Specimens were assayed for triglyceride (i.e., tri-
olein) and corrected for free glycerol by a modification 118-132 (125) 7.00 122 1.36
of the method of Bucolo and David. 3 Each specimen 192-212 (202) 8.00 198 1.95
329-363 (346) 9.00 348 2.85
was assayed twice, once for triglyceride and once for 154-170(162) 7.00 165 3.45
free glycerol. Standards were prepared with reagent- 150-166(158) 7.00 156 2.98
grade glycerol (J. T. Baker Chemical Co.) and aqueous 294-326(310) 9.00 315 8.63
62-76 (69) 7.00 68 0.59
benzoic acid, 2g/l. The ABA-100 was set to dispense 186-206 (196) 7.00 194 2.62
a 1:51 ratio of specimen (10 /xl) to reaction volume (510 256-284 (270) 8.00 272 2.33
/AI). The three-vial triglyceride Stat Pak reagent, 47-65 (56) 7.00 49 2.74
Cat. #860263 and the glycerol Stat Pak reagent, Cat. 87-105(96) 8.00 90 2.57
#869253 (Calbiochem, La Jolla, Cal. 92037) were used. 230-248 (239) 10.00 236 3.72
63-81 (72) 7.00 65 5.49
After 10 min of reaction time, i.e., during the second 95-113 (104) 8.00 101 2.39
carousel revolution, 20 /LX.1 of glycerol kinase (vial " B " 216-234(225) 10.00 224 2.30
of the Stat Pak diluted to 1 ml) was added to each 36-54 (45) 7.00 40 1.68
116-134 (125) 8.00 128 6.90
reaction mixture at timed intervals. A result was printed 167-185 (176) 9.00 174 2.04
out for each specimen during the third revolution, i.e.,
the first print. Additional results for each specimen, * Target values (in parentheses) established at CDC with their reference method.
t Maximum allowable total SD as computed using a one-way analysis of variance
i.e., the second print values, were subtracted from all technic.
Vol. 73 No. I ENZYMATIC LIPOPROTEIN QUANTIFICATION 77
Plasma LDL Determinations. LDL cholesterol was -- 360 r
T>
estimated by this laboratory using the following for- \O )
mula: LDL cholesterol = total cholesterol - HDL E -
cholesterol - triglyceride/5. 9 i
The Minnesota Lipid Research Clinic Laboratory O 270 -
cc
determined LDL cholesterol by calculating the dif- LU
ference between the cholesterol in the infranate (D (
CO

> 1.006) of the ultracentrifuged plasma and the HDL LU

1
cholesterol from the alpha fractions. 16 The infranate O 180 -
was prepared by overlaying plasma with .15 M NaCl X
and centrifuging in a 40.3 rotor for 16 hours at 40,000
o
_l _
rpm in a Beckman Ultracentrifuge, Model L2-65B Q
_J
(Beckman Instruments, Inc., Fullterton, Cal. 92934). 90 -
Quality Control. Known controls supplied by the
o
r
CDC {i.e., Q3, 64 and aQ2) were assayed in quadrupli- <
cate in each of our weekly test runs. The precision ^
>-
M
data for one year were calculated for each of the CDC
controls. Target values for Q3 and 64 were supplied SO 90 180 270 360
by the CDC, forQ2 by the Minnesota Lipid Research
LRC LDL CHOLESTEROL (mg/dl)
Clinic. In addition, a serum pool was assayed singly
for each of the lipid fractions; data for 20 weekly test FIG. 1. LDL cholesterol determined by using the formula LDL
runs were obtained (Table 1). cholesterol = total cholesterol - HDL cholesterol - triglycerides/5
with enzymatic methods and by ultracentrifugation with Lipid Re-
search Clinic methods.
Results
glyceride values (r = .996) assayed enzymatically.
Accuracy and Precision
LDL cholesterol values calculated from our choles-
The cholesterol and triglyceride procedures used in terol, triglyceride, and HDL results by use of the above
this study were standardized for accuracy and precision formula compared well (r = .983) with LDL cholesterol
by the CDC. Results obtained for unknown pools from values determined by subtraction of the HDL choles-
three quarters of the CDC surveillance phase of the terol in a Mn + 2 -heparin fraction from the cholesterol
standardization program are summarized in Table 2. of an ultracentrifuged infranate (D > 1.006) using the
Target (i.e., reference) values were established at the LRC methods (Fig. 1 and Table 3).
CDC for triglyceride by the CDC reference chromo-
tropic acid method (unpublished) and for cholesterol Discussion
by a modification of the method of Abell and associates. 1
The selection of a suitable lipoprotein quantification
method depends on several factors, including con-
Comparison
venience, precision, and agreement with reference
Results obtained with Lipid Research Clinic methods. The enzymatic methods for cholesterol and
methods were compared 12,23 (Table 3) with HDL cho- triglyceride as modified by us and the use of the for-
lesterol values determined by enzymatic assay of alpha mula developed by Friedewald and co-workers are suit-
fractions prepared with Mn +2 and heparin (r = .984), able for a clinical laboratory. The enzymatic methods
as well as total cholesterol values (r = .996) and tri- for cholesterol and triglyceride are clearly easier,

Table 3. Regression Analysis for Lipid and Lipoprotein Determinations by LRC Methods Using
Ultracentrifugation (X Axis) and Enzymatic Methods without Centrifugation (Y Axis)
Standard Error Correlation
Number Intercept Slope of the Estimate Coefficient

Total cholesterol 57 1.56 mg/dl 1.006 9.1 mg/dl .986

HDL cholesterol* 53 0.89 mg/dl 0.998 1.6 mg/dl .984
LDL cholesterol 32 9.38 mg/dl 0.977 8.8 mg/dl .983
VLDL cholesterol 54 -1.32 mg/dl 1.010 1.3 mg/dl .996
Triglyceride 54 -6.50 mg/dl 1.010 7.7 mg/dl .996
1
Previously reported. Clin Chem 22:98-101. 1976.
78 STEELE ET AL.
faster, and more convenient than the older methods
involving extraction, and can give accurate and precise
results, as documented by our participation in the CDC
Cooperative Cholesterol and Triglyceride Standard-
ization Program. It should be noted that the overall
standard deviations of cholesterol and triglyceride
values for the 12-month period were extremely low,
i.e., 2.36 mg/dl at 292 mg/dl for cholesterol and 2.43
mg/dl at 157 mg/dl for triglyceride.
While recently reported methods for HDL using
phosphotungstate precipitation15 or electrophoresis
coupled with an enzymatic cholesterol reagent8 appear
to be possibly more convenient, they do not have the
advantage of utilizing a reference technic, i.e., precip-
itation with manganous chloride and heparin. Also,
there is no backlog of experience that will tell whether
these methods are suitable for long-term studies or
whether the data derived by these methods can be
compared with those derived from studies using Lipid
Research Clinic methods. The disadvantages of these
methods include a coefficient of variation of 17% for the
electrophoresis technic8 and the detection of HDL
apoprotein in the precipitates using the phosphotung-
state method.21 Our Mn+2-heparin and enzymatic
methods for HDL are convenient, accurate, and pre-
cise, with an overall standard deviation for 12 months
of 1.45 mg/dl at 52.5 mg/dl.
For LDL determination, the use of the formula de-
veloped by Friedewald is the most convenient and the
most suitable technic of the various LRC methods for
the routine clinical laboratory. This technic with the
enzymatic methods for total cholesterol, HDL choles-
terol, and triglyceride will allow for accurate and pre-
cise fractionation of the lipoproteins. A clinical labo-
ratory should be able, using this approach, to give re-
liable patient data that can be interpreted with respect
to the findings of the major lipid studies.
Acknowledgments. The Clinical Chemistry Section of the Center
for Disease Control, especially Dr. Gerald Cooper, Ms. Carole
Winn, and Ms. Barbara Botero, permitted the use of their data
and offered helpful suggestions. Ms. Veronica Hill and Ms. Diane
Harrington, of the Minnesota LRC Laboratory, provided excellent
technical assistance.
References
1. Abell LL, Levy BB,Brodi BB,etal: A simplified method for the
estimation of total cholesterol in serum and demonstration of
its specificity. J Biol Chem 195:357-366, 1952
2. Allain CC, Poon LS, Chan CSG, et al: Enzymatic determination
of total serum cholesterol. Clin Chem 20:470-475, 1974
A.J.C.P. January 1980
3. Bucolo G, Yabut J, Chang T: Mechanized enzymatic determina-
tion of triglyceride in serum. Clin Chem 21:420-424, 1975
4. Bucolo G, McCroskey R, Whittaker N: Lipase-triggered kinetic
assay of serum triglyceride. Clin Chem 21:424-426, 1975
5. Bucolo G, David H: Quantitative determination of serum tri-
glyceride by the use of enzymes. Clin Chem 19:476-482, 1973
6. Burstein M, Samaille J: Sur un dosage rapide du cholesterol
lie aux a- et aux /3-lyopropteines due serum. Clin Chim Acta
5:609, 1960
7. Castelli WP: CHD risk factors in the elderly. Hosp Pract 11:
113-21, 1976
8. Cobb SA, Gamblin JM: Enzymatic determination of cholesterol
in serum lipoproteins separated by electrophoresis. Clin
Chem 24:1025, 1978
9. Friedewald WT, Levy RI, Fredrickson DS: Estimation of the
concentration of low-density lipoprotein cholesterol in plasma,
without use of the preparative ultracentrifuge. Clin Chem 18:
499-502, 1972
10. Gluek CJ, Fallat RW, Spadafora M, et al: Longevity syndromes.
Circulation 52 (suppl 11): 272, 1975
11. Gordon T, Castelli WP, Hjortland MC, et al: High density
lipoprotein as a protective factor against coronary heart dis-
ease. Am J Med 62:707-714, 1977
12. Handbook of Chemistry and Physics. 47th edition. Edited by
RC Wrast, SM Selby. Chemical Rubber Co.. Cleveland, Ohio,
p A-244, 1966
13. Havel RJ: Lipids and atherosclerosis. Cardiol Res Center Bull
15:93-97, 1977
14. Koehler DF, Steele W, Azar MM, et al: An enzymatic triglycer-
ide method that is suitable for long term population studies.
Clin Chem 24:326-329, 1978
15. Lopes-Virella MF, Stone P, Ellis S, et al: Cholesterol determina-
tion in high density lipoproteins separated by three different
methods. Clin Chem 23:882-884, 1977
16. Manual of Laboratory Operations, Lipid Research Clinics
Program, LLipid and Lipoprotein Analyses. National Heart
and Lung Institute, National Institutes of Health, Bethesda,
Md., 1974. U.S. Government Printing Office, Washington,
D. C. 20402
17. Miller GJ, Miller NE: Plasma high-density lipoprotein con-
centration and development of ischaemic heart disease.
Lancet 1:16-19, 1975
18. Nikkila EA: Serum high-density lipoprotein and coronary
heart disease. Lancet 2:320, 1976
19. Rhoads GG, Gulbrandsen CL, Kagan A: Serum lipoproteins
and coronary heart disease in a population study of Hawaii
Japanese men. N Engl J Med 294:293-298, 1976
20. Steele BW, Koehler DF, Azar MM, et al: Enzymatic determina-
tion of cholesterol in high-density lipoprotein fractions pre-
pared by a precipitation technique. Clin Chem 22:98-101,
1976
21. Warnick GR, Albers JJ: A comparison of current methods for
high density lipoprotein cholesterol quantification. Clin Chem
24:1025, 1978
22. Wentz PW, Cross RE, Savory J: An integrated approach to lipid
profiling: Enzymatic determination of cholesterol and tri-
glycerides with a centrifugal analyzer. Clin Chem 22:188-192,
1976
23. Westgard J O , Hunt MR: Use and interpretation of
common statistics, test in method-comparison studies. Clin
Chem 19:49-57, 1973
24. Witte DL, Arrett DA II, Wycoff DA: Evaluation of an enzymatic
procedure for determination of serum cholesterol with the
ABA-100. Clin Chem 20:1282-1286, 1974