Arherosclerosis, 21 (1975) 235-244 /c:r Elsevier Scientific Publishing Company, Amsterdam
DIAGNOSIS MENT
OF
OF
FAMILIAL
HYPERCHOLESTEROLEMIA
STEROL
SYNTHESIS
A. K. KHACHADURIAN,
M. LIPSON
IN
AND
235
- Printed in The Netherlands
CULTURED
BY MEASURESKIN
FIBROBLASTS
F. S. KAWAHARA
Depurtlnent of Medicine, The College of’ Medicineand Dentistry of New Jersey, Rutgers Medicul School, Piscatawav, N.J. 088.54, and Department of Pediatrics, Northwestern University Medical School am1 The Clinical Research Center, The Children’s Memorial Hospital, Chicago, Ill. 60614 (U.S.A.) (Revised, received August 22nd, 1974) (Accepted September 25th, 1974)
SUMMARY
The incorporation of radioactive acetate into the digitonin precipitable fraction (cholesterol) was measured in monolayers of primary cultures of skin fibroblasts. Mean incorporation was increased approximately 20-fold in 4 subjects homozygous for familial hypercholesterolemia (FH) and 4-fold in 6 heterozygotes derived from the immediate
family of homozygotes.
Incorporation
was normal
in 4 subjects with Type
IV and V hyperlipoproteinemia. In cells that had been preincubated in lipid free medium, incorporation by cells from homozygotes was equal to controls, denoting a derangement in the feedback inhibition of cholesterol synthesis by medium lipids in FH. The activity
of the enzyme
3-hydroxy-3-methylglutaryl
paralleled the values obtained for sterol synthesis. The assay described could be useful in making lial hypercholesterolemia betalipoproteinemia.
Key words:
Cholesterol
and could
synthesis
possibly
- Familial
coenzyme
an “etiologic”
identify
variants
A reductase
diagnosis
of monogenic
hypercholestrrolernia
of famihyper-
- Feedback
in-
hibition - Fibroblasts
Supported by U.S. Public Health Grants RRO0199, HD 04252, GRSG-RR 05475, National Foundation-March of Dimes Grants No. 1283 and No. l-352, and CMDNJ-Rutgers Medical School Grants USPHS-PE15-01 and RR5576 27-1957. Presented in part at the meeting of the Society for Pediatric Research, Washington, D.C.. May 2, 1974. Reprint requests: A.K.K., Department of Medicine, CMDNJ-Rutgers Medical School.
A. K. KHACHADURIAN,
236
M. LIPSON,
F. S. KAWAHARA
INTRODUCTION
In previous
communicationslJ,
we reported
skin fibroblasts from patients homozygous cholesterol synthesis from acetate proceeds greater
than
in normal
fibroblasts.
that in suspensions
of cultured
for familial hypercholesterolemia at rates that are approximately
In cells from
heterozygotes,
(FH), 10 times
the synthetic
rates
were approximately 3 times faster but values overlapped with the normal. When fibroblasts are preincubated in medium containing delipidated calf serum, differences between patients and controls are abolished, suggesting a derangement in the feedback inhibition of cholesterol synthesis in FH. The present studies indicate that modification of the assay allows the identification of the heterozygote and suggests that it can be used to differentiate types of hyperlipoproteinemias.
FH from other
PATIENTS
Four patients homozygous for FH were selected according to established criteria based on high plasma cholesterol (626-800 mg per 100 ml), hyperbetalipoproteinemia, juvenile
xanthomatosis,
and genetic studies compatible
with a monogenic
heritancea-5. Six heterozygotes consisted of the parents of the homozygotes their first-degree relatives age 7 and 10 years (plasma cholesterol 294-324
mode of inand two of mg per 100
ml). Controls were age 1 month to 30 years and had plasma cholesterol of 145-l 88 mg per 100 ml. Subjects with other forms of hyperlipoproteinemias consisted of: (1) a male infant who presented with eruptive xanthomatosis, massive chylomicronemia with a plasma
triglyceride
of 8175 mg per 100 ml, plasma
cholesterol
of 670 mg per
100 ml and a lipoprotein phenotype V 6. His plasma postheparin lipolytic activity varied between 17 and 46% of normal; (2) the mother of this patient whose plasma cholesterol varied between 151 and 188 mg per 100 ml and triglycerides between 108 and 434 mg per 100 ml. The plasma lipoprotein pattern was type Ha, Ilb, IV and V, at various examinations; (3) a 35year-old male with Type IV hyperlipoproteinemia; (4) a 6-month-old female with Type IV hyperlipoproteinemia. METHODS
Primary cultures of skin fibroblasts were grown to confluency in 75 cm2 Falcon flasks in 5% COZ incubator7. The standard growth medium consisted of Eagle’s minimal medium (E.M.M.) containing 50 units per ml penicillin, 50 lug per ml streptomycine and 25 units per ml fungizone and 10 2) heat inactivated fetal calf serum (Grand Island Biological Company, Grand Island, N.Y.). Total cholesterol concentration in the fetal calf serum was 35 mg per 100 ml. The medium was changed every 3 days. Cells were subcultured by trypsinization after the growth was confluent and cells from one flask were seeded into 3 flasks. Cell cultures on the first day of confluency and in the 6th to 14th passages were used in all experiments. Delipidation of fetal calf
DIAGNOSIS OF FAMILIAL HY PERCHOLESrEROLEMlA
serum was done by extraction
with ice cold ether+thanol
ously. No cholesterol could Burchard reactio@. Human
be detected low density
ultracentrifugation8
and dialyzed
against
237 (1: 1, v/v) as described
previ-
in this preparation by the Liebermannlipoprotein was prepared by differential Krebs-Ringer
phosphate
buffer, pH 7.4. The
donor was a normolipidemic adult male. Twenty hours prior to the incubation with the radioactive precursor, the cultures were washed three times with 0.9 y0 sodium chloride solution and the medium replaced 1. Standard growth medium; 2. Delipidated medium. ed the fetal calf serum;
with one of the following
In this medium,
IO % delipidated
preincubation
fetal calf serum replac-
3. Mixtures of standard and delipidated medium; 4. Standard or delipidated medium plus various concentrations density lipoproteins (LDL). Following
the 20-h preincubation,
media:
of human
low
the cells were washed three times with saline
and the incubation medium was added. This consisted of 4 ml of Krebs-Ringer phosphate buffer, pH 7.4, containing 2 puCi of sodium, [2-‘W]acetate, specific activity 60 &i tration
per pmole (Amersham Searle, Arlington Heights, 111.). The substrate concenroutinely used was 0.165 pmole per flask, except for experiments shown in
Fig. 2 and Table 2 in which the acetate concentrations were 0.033 pmolc per flask (no carrier acetate added). Incubation was carried out for 2 h at 37“ C. In preliminary experiments. the effect of varying the time of incubation between 30 and 120 min as well as of adding 0.066 to 0.30 pmole carrier acetate per Aask were assessed. Reaction was stopped by adding 1 ml of 10 N NaOH. After incubation for 30 min at 37” C, the contents of the flask were quantitatively transferred to a test tube, mixed thorough11 and an aliquot removed for protein mcasurementz. Trapping of the COe, saponification, isolation of the digitonin precipitable fraction (DPF) and the fatty acid fraction (FAF) and the counting of the radioactivity was done according to standard techniques?,“. Isolation of cholesterol was also done by thin-iaqerchromatography on silica gel’” in several experiments. The activity of the enzyme 3-hydroxy-3methylglutaryl coenzyme A reductase (HMG CoA reductase, EC 1. 1. 34) was measured by the method of Brown, Dana and Goldstein 11. Cells used for these experiments were in the 6th to 14th passage and contluency was reached in all on the 6th day of subculture. RESULI S
There was a close agreement
between
the counts
in the digitonin
precipitable
fraction and in the thin-layer chromatographic spot corresponding to cholesterol. No effort was made to identify the nature of the component included in the “fatty acid fraction”. However, studies by thin-layer chromatography indicate that approximately half the counts in this fraction correspond to the long chain fatty spot obtained by hydrolysis of olive oil. The incorporation of acetate into the DPF was linear up to 120 min. Fig. 1 shows the effect of varying the concentration 01‘acetate on its incorporation into the
238
A. K. KHACHADURIAN, M. LIPSON, F. S. KAWAHARA
~2000 - /7--i 2
.z IOOJI
.o
D g
2-i
L-I
0 ,033
.I
,165 .2
.3 .33
Acetate, p mole per flask
Fig. 1. Effect of varying the concentration able fraction (DPF).
DPF. Results indicate pmole per flask. Results corporation
of acetate on its incorporation
that saturation
with substrate
into the digitonin precipit-
occurs at a concentration
of 0.1
shown in Table 1 indicate that in fibroblasts from homozygotes, inof acetate into the DPF is approximately 20 times greater than normal,
while incorporation
into FAF
is equal in the two groups.
When the counts
in the
DPF are expressed as a ratio to the counts in the COZ, a marked difference is again apparent between homozygotes and controls, the values being 1.5 1 & 0.38 and 0.118 & 0.057, respectively. In fibroblasts
preincubated
in delipidated
medium
there
was no significant
difference in the incorporation of acetate into DPF between normal and homozygous cells, the mean (A SD.) counts per minute per mg protein being 16,950 & 3,827 and 19,090 f 5,583, respectively. Fig. 2 shows the effect of altering the ratio of delipidated to standard the preincubation mixture. In normal cells, 0.1 vol. of standard medium
serum in causes a
significant drop in the incorporation of acetate while in homozygotes a similar drop is seen with 0.4 vol. of standard medium. Table 2 shows the effect of adding increasing quantities of human LDL to delipidated preincubation medium as well as to standard TABLE 1 INCORPORATION FATTY n =
ACID
OF
RADIOACTIVE
FRACTION
(FAF)
number of measurements Homozygo
ACETATE
BY FIBROBLAST
tes
n = I3
INTO
THE
in quadruplicate;
FAF
Cofltrols n = 18
3868 i 2356 197 i 12 P ~< 0.001 6564 :F 3658 7121 1~ 4859 P N.S.
PRECIPITABLE
N.S. = not significant
(counts per min per mg proteitl & S.D.)
DPF
DIGITONIN
MONOLAYERS
FRACTION
(DPF)
AND
DIAGNOSIS
OF FAMILIAL
HYPERCHOLESTEROLEMIA
239
0
IO
‘0
9/
8,
I
$4
2
RATIO
OF
DELIPIDATED
$6 MEDIUM
TO
241 NATIVE
90
MEDIUM
Fig. 2. Incorporation of radioactive acetate into the digitonin precipitable fraction (DPF) in fibroblasts from a homozygote and control. Fibroblasts were preincubated in delipidated tmedium, standard medium, or their mixtures. Acetate concentration was 0.033 i&mole per flask. Mean counts per min in delipidated medium was 11010 for the homozygote and 12580 for the control.
TABLE 2 EFFECT
OF THE ADDITION
MEDIUM (EXPERIMENT l-ION OF ACETATE
INTO
OF HUMAN
LOW
I) OR TO STANDARD
DENSITY
LIPOPROTEIN
PREINCUBATI~N
(LDL)
TO DELIPIDATED
MEDIUM (EXPERIMENT
II) 0~
PREINCUBATION THE IiicoR~o~A-
DPF
The preincubation media consisted of 3.95 ml of delipidated medium I- 0.05 ml of buffer containing the given concentration of LDL cholesterol. Acetate concentration in the incubation medium was 0.033 /‘mole per flask.
(coutlts
Experittwtzt
mitt per tug prorein)
I
Delipidated medium (D.M.) D.M. i- LDL cholesterol, 0.05 mg/lOO ml” D.M. 1 LDL cholesterol, 0.25 mg/lOO ml D.M. t LDL cholesterol, 0.5 mg/lOO ml D.M. $ LDL cholesterol, 5 mg/lOO ml Standard medium E.~perirtwttf
pw
20700 14800 9050 3800 366 146
19600 21500
II6 96 88
3460 2130 3780
I8600 18800 4120 3260
II
Standard medium (KM.) S.M. + LDL cholesterol, S.M. L LDL cholesterol, iL Final concentration
10 mg!lOO ml 25 mg/lOO ml
of cholesterol
in preincubation
medium.
A. K. KHACHADURIAN,
240 TABLE
M. LIPSON,
F. S. KAWAHARA
3
EFFECT OF
VARYING
D,G,TON,N
PRECIPITABLE
THE
LIPOPROTEIN FRACTION
Preincubution medium
conrro1
Delipidated
24.0
CONCENTRATION
(DPF)
ON
IN A NORMAL
THE
AND
INCORPORATION
OF ACETATE
INTO
THE
HETEROZYGOTE
Heterozygofe
(counts per minper rngproteitl x 10 3; mean % S.D.)
1 vol. delipidated
i_ 4.1
0.51
l_ 0.06
23.3
* 2.9
3.6
3: I.1
+ 1 vol. standard
1.07 1: 0.2
0.185 + 0.05
Standard
medium. Results again indicate that the drop in acetate incorporation occurs at lower concentrations of medium LDL cholesterol in controls. Increasing the concentration of lipoprotein cholesterol to 25 mg per 100 ml of preincubation medium (as compared to 3.5 mg per 100 ml in the standard medium) does not suppress acetate incorporation further. Addition of lipoprotein to the standard medium to a final cholesterol concentration further
of 50 mg per 100 ml in the preincubation medium again failed to suppress cholesterol synthesis in fibroblasts from homozygotes. Table 3 shows results of an experiment in which acetate incorporation was
measured after preincubation in standard, delipidated, or an equal mixture of standard and delipidated media. Cells from a heterozygote and control subject were used. Results again indicate a significant difference between the control and heterozygote in cells preincubated in lipid-containing media. Fig. 3 gives the individual values in 5 controls, 6 heterozygotes, and 4 patients with hyperlipoproteinemias experiments represents
other
the fibroblasts
than
familial
were preincubated
the mean of 4 measurements.
hypercholesterolemia.
in the standard
Results
indicate
medium
In all these and each value
that heterozygotes
can be
I@00 -
z ”
SQO-
:
600-
:
700-
.
600too-
:
I
.
400300-
l
too-
1
CONTROLS
8 HETEAOZY6OTES
TYPES II
1
Fig. 3. Incorporation of acetate into DPF in controls, subjects esterolemia and subjects with Type IV or V hyperlipoproteinemia. 4 flasks from the same subject. Cells from patients and controls incubated simultaneously to minimize experimental variables.
heterozygous for familial hypercholEach value represents the mean of were subcultured, preincubated, and
DIAGNOSIS
OF FAMILIAL
241
HYPERCHOLESTEROLEMIA
TABLE 4 3-HYDROXY-3-METHYLGLIJTARYL
COENZYME
A
REDUCTASE
ACTIVITY
IN HOMOZYGOTES,
HETEROZYGOTES
AND CONTROLS
Values (mean k S.D.) are expressed in picomoles of mevalonic acid produced per mg protein per min. II = number of measurements. -Subjects Pwincubation mediunr standard
delipidatrd
Controls, 12 7 5
2.8 k 1.1
15.7 :k 7.7
Heterozygotes, II = 3
7.2 _t 3.9
19.9 _?~3.6
Homozygotes, II ~ 5
14.4 1 7.4
20.3 im 6.0
identified by this test. A similar separation is obtained when the counts in the DPF are given as a ratio of the counts in the FAF, values for the controls ranging between 0.027 and 0.079 (mean _i: 0.015 S.D.) and for heterozygotes between 0.088 and 0.240 (mean 0.170 & 0.042 SD., P < 0.001). In subjects with Type IV or V hyperlipoproteinemia, the values are within the normal range. Table 4 shows the values obtained for HMG CoA reductase in experiments done in collaboration with Dr. H. M. Moon in our laboratory. Results indicate that in fibroblasts preincubated in the standard medium, the enzyme activity in heterozygotes is half the value of that in homozygotes, while in the controls it isapproximately 20%. In cells preincubated in the delipidated medium the values in homozygotes and heterozygotes are slightly higher than in controls, but the differences are not significant. DISCUSSION
The diagnosis of familial hypercholesterolemia is presently based on plasma lipoprotein phenotype (predominantly Type IIa and Type Ilb) and compatible genetic studies. The lipoprotein phenotype is not characteristic and, therefore, cannot differentiate FH from hyperbetalipoproteinemia resulting from various inherited or acquired disorders4,jJa. Although abnormalities in lipoprotein structure and turnover and in cholesterol turnover have been reported, in viw studies have thus far failed to elucidate the ultimate metabolic defect in familial hypercholesterolemiat2Ja. We had previously reported that feeding a high cholesterol diet for 3 days failed to suppress the incorporation of acetate by liver slices from 3 patients homozygous for FH. Acetate incorporation following a high cholesterol diet in these patients was was higher than in control subjects on a regular diet 9. A similar lack of inhibition noted in 3 additional homozygotes placed for 8 days on high cholesterol diets14. Our
A. K. KHACHADLJRIAN,
242
findings suggested a defect in the feedback
inhibition
metabolic basis of FH. Our recent experiments
with cultured dence in favor of this hypothesis lJ. Working and harvested
prior
to incubation
M. LIPSON,
of cholesterol
F. S. KAWAHARA
biosynthesis
as the
skin fibroblasts have added further eviwith skin fibroblasts grown to confluency
with radioactive
precursors,
we demonstrated
a
marked increase in cholesterol synthesis in cells from patients with FH. When the cells were preincubated for 20 h prior to biosynthetic studies in a medium containing delipidated calf serum instead of normal
serum, there was no difference
in synthetic
activity
between normals and FH. The results ofexperiments in which mevalonate was used as the precursor suggested that the defect is at a premevalonate step and, by inference from liver studiesl5, at the site of the enzyme converting 3-hydroxy-3-methylglutaryl coenzyme A to mevalonate. The studies reported present
studies, incubation
here confirm
and extend
with radioactive
the previous
findings.
acetate was done on the fibroblast
In the mono-
layers, omitting the trypsinization step required for harvesting of the cells and avoiding possible injury to cells. This modification, as well as a strict control of experimental conditions, has resulted in better reproducibility of the results and a sharper separation
of value between
lipid concentration medium
heterozygotes
in the preincubation
lipid in preventing
the induction
and controls. medium
indicate
of cholesterol
The studies
of the effect of
that maximum synthesis
effect of the
occurs at the lipid
concentration found in the standard medium and that differences between the patients and controls cannot be explained by small changes in lipid concentration of the preincubation medium. In the experiments described in Fig. 2 and Table 3, the acetate concentration was 0.033 pmole per flask which is lower than the concentration needed for maximum velocity. Therefore, the possibility that the results could be affected by differences in the pool sizes of acetate in the two cell lines has to be considered. Such a possibility is unlikely since similar differences in acetate incorporation between homozygotes and control cell lines were observed at saturating levels of acetate in cells preincubated in standard and delipidated media. The inverse relation between medium cholesterol and the rate of cholesterol synthesis by cultured human fibroblasts has been demonstrated repeatedlyle-ls. The studies of Brown, Dana and Goldstein lt indicate that this regulation is exerted by changes in the level of the enzyme 3-hydroxy-3-methylglutaryl coenzyme A reductase. Following the completion of the studies reported in the present paper, Goldstein and Brown1g,20 reported on their extensive studies on the defective regulation of the HMG CoA reductase in familial hypercholesterolemia. In homozygotes, low density lipoproteins (LDL) caused no inhibition ofenzyme activity while in controls there was more than 90% suppression. Heterozygotes had intermediate values. Our studies on HMG CoA reductase indicate a moderate prevention of induction with medium LDL in homozygotes as well. Differences in the design of experiments could be the reason for this difference as well as for the much lower levels of HMG CoA reductase activity in cells preincubated in delipidated medium in our experiments. However, our observations in patients as well as controls indicate that preincubation medium lipids exert a
DIAGNOSIS
OF FAMILIAL
243
HYPERCHOLESTEROLEMIA
greater effect on the incorporation
of acetate into sterols than on the activity of HMG
CoA reductase. These findings suggest the possibility that regulation of cholesterol synthesis may also occur at sites other than at the level of the enzyme HMG CoA reductase. Our findings
as well as those of Brown and Goldstein
assays could be used in making an “etiologic” olemia. It is possible that the assays described
indicate
that the fibroblast
diagnosis of familial hypercholesterwill identify “variants” of conditions
presently grouped under FH or familial Type II disease2. Our preliminary studies indicate that in some subjects heterozygous for “familial Type II disease” not derived from the families
of patients
with juvenile
xanthomatosis,
acetate incorporation
into
cholesterol and the activity of HMG CoA reductase are normal. Further standardization or modifications of the assays and their use in a larger number of subjects will determine the value of their clinical application in making an etiologic
diagnosis
of familial
hyperlipidemias.
REFERENCES
I KHATHADURIAN, A. K., Prospects for the prenatal diagnosis of familial hypercholesterolemia. Leh. Med. J., 26 (1973) 325. 2 KHACHADURIAN, A. K. AND KAWAHARA, F. S., Cholesterol synthesis by cultured fibroblasts: Decreased feedback inhibition in familial hypercholesterolemia, J. Lab. C/in. Med., 83 (1974) 7. 3 KHACHADURIAN, A. K., The inheritance of essential familial hypercholesterolemia. Atner. J. Med., 37 (1964) 402. 4 FREDRICKSON, D. S. AND LEVY, R. I., Familial hyperlipoproteinemia. In: J. B. STANBURY, J. B. WYNGAARDEN AND D. S. FREDRICKSON (Eds.), The Metabolic Busis of‘lnherited Disease, McGrawHill, New York, N.Y.. 1972. 5 KHACHADURIAN, A. K. AND UTHMAN. S. M., Experiences with the homozygous cases of familial hypercholesterolemia: A report of 52 cases, NW. Metahol., 20 (1973) 132. 6 HAMLY, C. A. AND KHACHADURIAN, A. K., Type V hyperlipoproteinemia in an infant, Pediat. Res.. 7 (1973) 387/159. 7 NADLER, H. L., CHACKO, C. M. AND RACHMELER, M., lnterallelic complementation in hybrid cells derived from human diploid strains deficient in galactose-l-phosphate uridyl transferase activity, Proc. Nat. Acad. Sri. U.S., 67 (1970) 976. 8 HAVEL, R. J., EDER, H. A. AND BRAGDON, J. H., The distribution and chemical composition of ultracentrifugally separated lipoproteins in human serum, J. C/in. Invest., 34 (1955) 1345. 9 KHACHADURIAN, A. K., Lack of inhibition of hepatic cholesterol synthesis by dietary cholesterol in cases of familial hypercholesterolemia, Lancer, ii (1969) 778. 10 GLOSTER, J. AND FLETCHER, R. F., Quantitative analysis of serum lipids with thin-layer chromatography, Clin. Chirn. Acta, I3 (1966) 235. 1 I BROWN, M. S., DANA, S. E. AND GOLDSTEIN, J. L., Regulation of 3-hydroxy-3-methylglutaryl coenzyme A reductase activity in human fibroblasts by lipoproteins, Proc. Nat. Acad. Sci. U.S., 70 (1973) 2162. I2 MYANT, N. B. AND SLACK, J., Type II hyperlipoproteinemia, C/in. Endocrinol. MetaboI., 2 (1973) 81. I3 LEES, R. S., WILSON, D. E., SCHONFELD, G. AND FLEET, S., The familial dyslipoproteinemias, Prop. Med. Genetics, 9 (1973) 237. 14 KHACHADURIAN, A. K., Unpublished. I5 DIETSCHY, K. M. AND WILSON, J. D., Regulation of cholesterol metabolism, New OIgl. J. Mccl., 282 (1970) 1128. 16 LENGLE, E., AND SMITH J. L., The effect of culturing sarcoma-180 cells with lipid-free serum, FPd. Aoc., 28 (1969) 688.
244 17 AVIGAN,J., WILLIAMS,C. D.
A. K. KHACHADURIAN,
M. LIPSON,
F. S. KAWAHARA
AND BLASS,J. P., Regulation of sterol synthesis in human skin fibroblast cultures, Biochin~. Bioph.~. Acta, 218 (1970) 381. 18 WILLIAMS,C. D. AND AVIGAN,J., In vitro effects of serum proteins and lipids on lipid synthesis in human skin fibroblasts and leukocytes grown in culture, Biochim. Biophys. Acta, 260 (1972) 413. identification of a defect in 19 GOLDSTEIN,J. L. AND BROWN,M. S., Familial hypercholesterolemia: the regulation of 3-hydroxy-3-methylglutaryl coenzyme A reductase activity associated with overproduction of cholesterol, Proc. Nat. Acad. Sci. U.S., 70 (1973) 2804. 20 BROWN, M. S., DANA, S. E., AND GOLDSTEIN,J. L., Regulation of 3-hydroxy-3-methylglutaryl coenzyme A reductase activity in cultured human fibroblasts, J. Biol. Chem., 249 (1974) 789.