Postheparin Plasma Lipoprotein Lipase and Hepatic Lipase

Postheparin Plasma Lipoprotein Lipase andHepatic Lipase in Diabetes Mellitus

Relationship to Plasma Triglyceride Metabolism

Esko A. Nikkild, M.D., Jussi K. Huttunen, M.D., and

Christian Ehnholm, M.D., Helsinki, Finland

SUMMARY

The activity of two triglyceride Upases was determined by animmunochemical method in the postheparin plasma of 60 diabeticpatients and of 47 age- and sex-matched nondiabetic control sub-jects. The results were related to the type of diabetes, to plasmatriglyceride and insulin concentrations, to removal of exogenous fatfrom the blood, and to turnover of VLDL-triglycerides. The meanpostheparin plasma lipoprotein lipase (LPL) activity was decreasedby 44 per cent (p < 0.001) in patients with untreated ketotic dia-betes and by 20 per cent (p < 0.01) in patients with untreated mildto moderate nonketotic early-onset diabetes. Insulin treatment ofketotic diabetes resulted in a rapid increase in the activity of LPLand decrease in serum triglyceride level, whereas sulfonylureatreatment of non-insulin-requiring diabetics did not significantlyinfluence the enzyme activity. In insulin-treated chronic diabeticsthe average postheparin plasma LPL activity was not differentfrom that of nondiabetic controls, but some of these patients hadhigh LPL values. In normolipidemic maturity-onset-type diabeticsthe LPL activity was within normal range, but in those havinghypertriglyceridemia the average LPL value was decreased by anaverage of 26 per cent (p < 0.01).

The LPL activity showed a significant negative correlation with

the logarithm of serum triglyceride concentration (r = -0.62) anda positive correlation with fractional removal of Intralipid (r =+0.64) and fractional turnover of VLDL triglyceride (r = +0.40).The activity of LPL was correlated to basal plasma insulin concen-tration in insulin-deficient diabetics (r = +0.34) but not in patientswith maturity-onset-type diabetes. The hepatic lipase (HL) activityof postheparin plasma was similar in diabetics and controls, withthe exception of hypertriglyceridemic maturity-onset diabetics,who had higher mean HL activity than the corresponding controlgroup (p > 0.01). The activity of HL was not related to triglycerideremoval but showed a significant correlation to VLDL-triglycerideproduction rate.

On the basis of these results it seems that a deficiency of LPLaccounts for a great deal of the elevation of serum triglyceride ininsulin-deficient human diabetes but has a smaller role in thepathogenesis of the hypertriglyceridemia that is associated withmaturity-onset diabetes. The latter abnormality is caused mainlyby an increased secretion of triglycerides into the blood eventhough a decreased LPL may contribute to development of hyper-lipemia in cases with gross elevation of serum triglycerides.DIABETES26:11-21, January, 1977.

Patients with juvenile- or adult-onset-type diabeteshave higher serum triglyceride levels than nondiabeticsubjects of corresponding age.1 The disturbance is

From the Third Department of Medicine, University CentralHospital, and Department of Serology and Bacteriology, Univer-sity of Helsinki, 00290 Helsinki 29, Finland.

Accepted for publication July 14, 1976.

usually manifested as an elevated plasma concentra-tion of very-low-density lipoprotein (VLDL), but insome cases fasting chylomicronemia is also present.23

The mechanisms responsible for diabetes-associatedhyperlipoproteinemia are variable depending, amongother things, on the degree of insulin deficiency, obe-sity, heredity, and diet.1 Even in the most simpleexperimental models of diabetes the elevation of

JANUARY, 1977 11

Dow

nloaded from http://diabetesjournals.org/diabetes/article-pdf/26/1/11/348997/26-1-11.pdf by guest on 27 January 2022

POSTHEPARIN PLASMA UPASES

serum triglyceride may be due to increased endoge-nous production of VLDL and chylomicrons,4'7 todefective removal of circulating triglycerides,8 or to acombination of both these mechanisms.9 Themechanisms are still more complicated in humandiabetes,10 where plasma insulin levels may be low,normal, or high and are associated with other hor-monal disturbances influencing triglyceride me-tabolism.

Several studies have suggested that a deficient re-moval of circulating triglycerides plays an importantrole in the pathogenesis of diabetic hyperlipemia inman.11"13 This view is supported by the knowledgethat insulin regulates the activity of lipoprotein lipase(LPL),* which is the rate-limiting enzyme in theperipheral breakdown of plasma triglycerides.14'16 Adecrease in the adipose tissue LPL activity has recentlybeen demonstrated in human diabetics.1718 On theother hand, assays of postheparin plasma lipolytic ac-tivity (PHLA) in diabetics have produced conflictingresults.11119"23 The PHLA is known to be composedof several lipolytic enzymes, which originate from dif-ferent tissues2425 and have different functions in thetriglyceride-removal process. Furthermore, the LPLactivities of different tissues are not equally sensitiveto insulin deficiency,142627 and therefore the activityof LPL in adipose tissue may not be representative ofthe over-all removal efficiency of circulating tri-glycerides. All these facts make it necessary toreevaluate the role of triglyceride removal in thepathogenesis of different forms of diabetic hyper-lipemia.

In the present study, LPL and a second triglyceridelipase, designated hepatic lipase (HL) according to itsprobable origin, have been assayed separately in post-heparin plasma of diabetic patients by a new specificimmunochemical method.28 To characterize the na-ture of hyperlipemia further, the lipase activities wererelated to plasma triglyceride concentration, to frac-tional removal of exogenous triglycerides, and toturnover of VLDL triglycerides.

*Lipoprotein lipase is defined as the triglyceride lipase of post-heparin plasma that is inhibited by high salt concentration andactivated by addition of serum. The other triglyceride lipase, herecalled hepatic lipase, is resistant to high salt concentration and isactive in the absence of added serum but is completely unactivatedby antibodies raised against purified salt-resistant triglyceridelipase from human postheparin plasma. Final evidence for thehepatic origin of this enzyme in man has not been presented, butit is absent from heparin perfusate of forearm.

METHODS

Subjects. The material included 60 diabetic patientsand 47 age-matched healthy nondiabetic control sub-jects. Diabetes was diagnosed on the basis of fastingblood glucose exceeding 100 mg./lOO ml. Forty-seven patients were studied soon after the detection oftheir diabetes and before institution of any treatmentother than diet. Twenty-one of these patients wererestudied after diabetes had been brought under ade-quate control by insulin, glibenclamide, or phenfor-min (fasting blood glucose less than 125 mg./lOOml.)- Thirteen insulin-requiring juvenile diabeticswith variable duration of the disease were studied dur-ing their ordinary visits to the outpatient clinic.

For presentation of data the diabetic patients havebeen divided into four groups (table 1). Group 1 in-cluded 13 patients who had definite ketosis during thestudy (urinary Acetest + + +) and who subsequentlyrequired insulin treatment. None had a severeketoacidosis, however. Group 2 was composed of 14untreated patients who had nonketotic moderate ormild diabetes manifested before the age of 30. In 12 ofthese patients the diabetes responded well to dietarytreatment alone or in combination with oral an-tidiabetic drugs. Two patients of this group were notcontrolled without insulin in spite of absence ofketosis. Group 3 included 13 insulin-treated chronicdiabetics mentioned above. These patients had dia-betes of one to 18 years' duration and had been oninsulin treatment from the beginning. They were freeof signs of microvascular involvement or had onlyminimal retino- or neuropathy. None had proteinuriaor elevated serum creatinine level. Group 4 was com-posed of 20 patients with maturity-onset diabetes, allof whom responded to diet alone or to oral antidia-betic drugs. This group was further subdivided into anormoglyceridemic (TG < 200 mg./lOO ml.) and ahypertriglyceridemic subgroup (4a and 4b).

Control subjects were 47 nondiabetic healthy vol-unteers aged 18 to 56 years. They had normal serumlipid and fasting blood glucose levels. All were con-suming their ordinary diet and none were taking anydrugs. Since the postheparin plasma lipase activitiesare dependent on age and sex,29 the controls werematched with the diabetics in the way shown in table1. The diabetic patients of groups 2 and 3 had acommon control group.

Experimental design. All untreated diabetics werehospitalized and given a sucrose-free isocaloric dietwith 45 per cent of calories as carbohydrate and 35 per

12 DIABETES, VOL. 26, NO. 1

Dow


ESKO A. NIKKILA, M.D., AND ASSOCIATES

TABLE 1Clinical and laboratory data of diabetic and control subjects

Group

1

23

45

Ketotic diabetesDiabetic subjectsControl subjects

Early-onset diabetesModerate, untreatedChronic, insulin-treatedControl subjects (forII and III)

Maturity-onset diabetesSerum TG normalSerum TG elevatedControl subjects

N

1313

1413

14

91120

Sex

M/F

10/310/3

12/212/1

12/2

5/48/3

13/7

Age

16-4718-47

16-3018-31

18-32

32-5634-5531-56

RBW%*

96 ±3105 ± 4

99 ± 497 ± 4

100 ± 5

112 ± 11129 ± 7108 ± 5

Fasting BG

mg./lOO ml.

239 ± 14*79 ± 4

160 ± 18*164 ± 291:

77 ± 4

169 ± 19*211 ± 251:

82 ± 5

PlasmaIRI

MUVml.

5.9 ± 1.3—

11.8 ± 0.8—

—

15.9 ± 4.122.5 ± 4.1

—

Serum TG

mg./lOO ml.

234 ± 35*78 ± 8

116 ± 12*128 ± 21*

72 ± 7

112 ± 161240 ± 414*

94 ± 10

Serum cholesterol

mg./lOO ml.

253 ± 23*201 ± 10

235 ± l i t238 ± 16t

201 ± 14

212 ± 14401 ± 46*237 ± 18

*Relative body weight = per cent of ideal body weight taken from Geigy Tables.35 The values are mean ± S.E.M.tp < 0.05*p < 0.01 for the difference between diabetic/control.

cent as fat. The basal serum lipid and blood glucoselevels were determined on admission and thereafterevery second or third day. The heparin test was carriedout within three days of admission and the otherstudies (Intralipid test, endogenous triglyceride turn-over, glucose tolerance test) during the first hospitalweek, but never on the day following the heparin test.The insulin-treated juvenile diabetics and the controlsubjects were studied as outpatients. All tests werestarted between 8 and 9 a.m. after a fasting period of10 to 12 hours and before administration of insulin ordrugs.

Assay of postheparin plasma triglyceride lipase activities.Postheparin plasma was obtained five and 15 minutesafter a rapid intravenous injection of 100 I.U. of hepa-rin (Medica/Vitrum) per kilogram of body weight.This procedure allows measurement of both lipase ac-tivities at or near their maximum.28 Blood was col-lected into tubes containing 5 I.U. of heparin per ml.of blood and kept on ice. Plasma was separated bycentrifugation at 4° C. and stored frozen at —20° C.until assay, which was done within three weeks. Inseparate experiments it has been shown that no changeof the lipase activities occurs during storage of plasmaat —20° C. during several months.

The selective assay of LPL and HL in postheparinplasma was carried out by the immunochemical pro-cedure described by Huttunen et al.28 This method isbased on the inactivation of hepatic lipase by a specificantiserum in the assay of LPL and on the use of highsalt concentration and omission of additional serum inthe assay of HL. The specificity of this procedure has

been checked.28 Lipoproteins do not interfere with theassay. For the assay of LPL activity 10 /xl. of posthepa-rin plasma was incubated for two hours at 4° C. with10 fA. of antiserum against HL (for preparation seeref. 28). Thereafter, 500 JU.1. of substrate mixture con-taining 3.2 mM of (acyl-l"14C)triolein-5 per cent-gum arabic emulsion, 40 mM Tris chloride buffer,pH 8.4, 0.1 M sodium chloride, 2.5 per cent FFA-poor human serum albumin, and 50 (xl. normalhuman serum was added and the tube was incubatedat 28° C. for 60 minutes. The released FFA was sepa-rated from glycerides by the liquid partition system ofBelfrage and Vaughan30 and its radioactivity deter-mined in a Packard liquid scintillation counter. Forthe assay of HL activity, 10 /xl. of postheparin plasmawas incubated at 28° C. for 60 minutes with a sub-strate mixture, which was similar to that used for LPLexcept that it contained 1.0 M sodium chloride and noserum. All assays were made in duplicate. Each seriesincluded two blanks containing saline instead of post-heparin plasma and two reference standards of post-heparin plasma taken from two normal subjects andkept frozen in small aliquots. The lipase activitieswere calculated from the difference of FFA radioactiv-ity between sample and blank, and they are expressedas ^moles FFA released per ml. of postheparin plasmaper hour. The lipase activities of each assay series werecorrected, if necessary, for the mean deviation of thetwo reference plasma Upases from their original activ-ity. The variation of the lipase activities of the refer-ence plasma from day to day was less than 5 per cent.Lipase assays were carried out from both five- and

JANUARY, 1977 13

Dow



fifteen-minute postheparin plasma samples, and thatgiving the highest lipase activity was used.

Intralipid test. The subjects were given 20 per centIntralipid (Vitrum) 0.1 gm./kg. b.w. as an intraven-ous infusion for 10 minutes, and blood samples weretaken before and immediately after the end of infusionand thereafter at 10, 15, 25, 30, 40, 50, and 60minutes. The plasma was separated and assayed fortotal triglycerides, the preinfusion (basal) triglyceridewas subtracted, and the half-life of the Intralipid tri-glyceride disappearance was obtained from asemilogarithmic plot of triglyceride versus time.

Measurement of VLDL-triglyceride turnover. This wascarried out by the endogenous 3H-glycerol labelingtechnique described in detail earlier.31 The subjectsreceived 150 fxCi. of 2-3H-glycerol intravenously, and14 blood samples were taken at one to nine hours afterthe injection. VLDL was separated from the plasma byultracentrifugation at density 1.006, and its tri-glyceride radioactivity was determined. Theradioactivity-decay curves were analyzed by a compu-ter to obtain fractional turnover rate constant. Theabsolute turnover rate (V) was obtained from the rateconstant (k) and serum concentration (S) of VLDL-TGby the formula V = k X S X plasma volume. Theplasma volume was measured with radioiodinated al-bumin.

Other methods. Blood glucose and serum triglyceridewere measured by Technicon AutoAnalyzer,32 serumcholesterol by the method of Huang et al.,33 andplasma insulin by radioimmunoassay.34

Materials. The (acyl-l-14C) triolein (s.a. 0.070fxCi./fAmole) and 3H-glycerol were purchased fromRadiochemical Centre, Amersham, England. Thetriolein was purified by thin-layer chromatographyand stored in benzene. Human serum albumin wasobtained from the Finnish Red Cross Blood Transfu-sion Service and was made FFA-poor by charcoaltreatment.

RESULTS

Plasma insulin and triglycerides of the diabetic patients.The clinical and laboratory data of the diabetic pa-tients and control subjects are presented in table 1.All the ketotic diabetics (group 1) had basal plasmainsulin levels at or below 10 ^tU./ml. in spite ofmarked hyperglycemia. The fasting serum tri-glyceride level was elevated in nine of the 13 patientswith ketosis. The patients with early-onset mild or mod-erate diabetes (group 2) had higher basal plasma insulin

concentration than the ketotic group, but the levelswere subnormal in relation to fasting blood glucose,and the early insulin response to intravenous glucoseload was absent or low (less than + 8 ju,U./ml.). One ofthe 14 patients in this group was obese and had anelevated serum triglyceride level. Among the nine pa-tients with maturity-onset diabetes and normal serumtriglyceride, three were obese. Of the 11 hyper-glyceridemic maturity-onset diabetics, eight had in-creased relative body weight. The basal plasma insulinlevel of maturity-onset-type diabetics ranged from 3to 43 A<,U./ml.

Postheparin plasma LPL in diabetic patients. The pa-tients with diabetic ketosis had significantly lowerpostheparin plasma LPL activity than matched non-diabetic controls (15.6 ± 1.9 vs. 23.6 ± 1.2 /xmol.FFA/ml./hr., p < 0.001, figure 1). Eight of the 13ketotic patients had abnormally low LPL values (lessthan mean —2 S.D. of controls), the lowest levelsbeing only 40 per cent of the mean activity of thecontrols. The patients with mild to moderate early-onset diabetes had also decreased mean LPL activityas compared with age-matched controls (18.7 ± 0.9vs. 23.5 ± 1.1 Atmol FFA/ml./hr., p < 0.01, figure1). However, most of these patients had LPL activityin the low-normal range; a subnormal value was ob-served only in three of 14 subjects. On the other hand,the insulin-treated juvenile diabetics, regardless of thedegree of their diabetic control, had either normal or

7- 40E

30

20

10

< 0.001 N.S. N.S. <O.O1

KETOTIC MODERATEEARLY-ONSET

INSULIN- MATURITY-TREATED ONSET

(NG) (HG)

FIG. 1. Postheparin plasma LPL activity of diabetic patients. Thehatched area indicates the mean ± 2 S.D. for LPL activityin age- and sex-matched nondiabetic controls. NG = nor-moglyceridemic, HG = hyperglyceridemic, N.S. = not sig-nificant. The numbers on the abscissa indicate the p valuefor the difference between diabetic and nondiabetic sub-jects.


Dow



increased postheparin plasma LPL activity (figure 1).The mean value was not different from that of non-diabetic subjects (25.8 ± 1.7 vs. 23-5 ± 1 . 1 /xmolFFA/ml./hr.). The patients with maturity-onset-typediabetes had an LPL activity not different from that ofthe nondiabetic subjects of corresponding age. How-ever, when the group was divided according to thepresence or absence of hypertriglyceridemia the sub-group with elevated triglyceride levels showedslightly reduced LPL activity as compared with con-trols (14.8 ± 1.6 vs. 20.0 ± 1.3 ^mol FFA/ml./hr.,p < 0.01, figure 1).

Postheparin plasma HL in diabetic patients. The HLactivities of the diabetic patients are shown in figure2. Most of the values were within normal range, and,aside from the patients with maturity-onset diabetes,the mean HL activity of the diabetics was not differentfrom the value of corresponding control groups. Thepatients with maturity-onset diabetes and hypertri-glyceridemia had elevated postheparin HL activity(31.1 ± 2.3 vs. 22.6 ± 1.8/xmol FFA/ml./hr., p <0.01, figure 2). In contrast to the LPL activity, theHL activity of postheparin plasma was not decreasedin any of the diabetic patients.

Relation of serum triglyceride to the postheparin plasmaLPL and HL activities. In untreated diabetic patientsthe logarithm of the fasting triglyceride level was in-versely correlated to postheparin plasma LPL activity(figure 3). The slope of log TG vs. LPL was less steepin maturity-onset-type diabetics than in the patientswith juvenile diabetes. In insulin-treated chronic

60

50

40

30

20

10

N.S. N.S. <aoi

FIG.

KETOTIC MODERATE INSULIN- MATURITY-EARLY-ONSET TREATED ONSET

(NG) (HG)

2. Postheparin plasma HL activity in diabetic patients. Thehatched area shows the mean ± 2 S.D. of matched non-diabetic controls. For explanations see the legend of figure1.

®

SERUM TRIGLYCERIDE.

500

ng/i00m

50

SERUM

100 200

TRIGLYCERIDE.

500

ng/i00ml

E 40

30

20

- 10

0.60

50 100 200 500 1000 2000

SERUM TRIGLYCERIDE mg/ i00ml

FIG. 3. Relationship between serum triglyceride concentrationand postheparin LPL activity in diabetic patients. Panel A:Untreated insulin-deficient early-onset diabetics; y = 44.9- 12.8 logx (r = - 0 .64 , p < 0.001). Panel B: Insulin-treated chronic diabetics (r = - 0 .24 , p > 0.10): Panel C:Untreated maturity-onset-type diabetic patients; y = 30.3- 5.5 logx(r = - 0 . 6 0 , p < 0 . 0 1 ) .

diabetics there was no correlation between the serumTG and the postheparin plasma LPL. However, manypatients of this group had disproportionately highserum triglyceride levels in relation to their posthepa-rin plasma LPL activity.

Of the 47 untreated diabetics, 21 had serum tri-glyceride levels above 200 mg./lOO ml. Only eight ofthese patients had subnormal LPL activity (less thanmean —2 S.D. of the corresponding control group),and five additional patients had LPL activity in a lownormal range. Of these 13 diabetics with moderate or

JANUARY, 1977 15

Dow



TABLE 2

Parameters of plasma triglyceride metabolism in untreated diabetic patients

Group

Ketotic*Moderate early-onsetMaturity-onset*Controls

Intralipid removal

2.46 ± 0.32§ (9)3.56 ± 0.21 (12)3.68 ± 0.16 (7)4.12 ± 0.35 (20)

VLDL-TGFractional turnover

hr."1

0.271 ± 0.071: (7)-0.344 ± 0.06 (7)0.366 ± 0.12 (7)0.432 ± 0.07 (14)f

Turnover ratemg./hr."'kg.M

10.8 ± 1.2 (7)7.6 ± 1.1 (7)

13.6 ± 1.3t(7)8.3 ± 0.7 (14)t

Mean ± S.E.M. Number of patients in parentheses.*Only patients with serum triglyceride less than 300 mg./lOO ml.tFrom the study of Nikkila and Kekki.31

tp < 0.05§p < 0.01

marginal LPL deficiency, eight were ketotic, two hadearly-onset moderate diabetes, and three werematurity-onset diabetics with hyperlipemia. Therewas no relationship between the serum triglyceridelevel and the postheparin plasma HL activity in any ofthe groups studied.

Removal of Intralipid in relation to postheparin plasmalipase activity. The fractional removal rate of intrave-nous Intralipid was subnormal in the patients withuntreated diabetic ketosis, but it was within normalrange in the diabetics with.less severe insulin defi-ciency (table 2). However, the test was not performedin patients with marked hypertriglyceridemia (> 300

FIG. 4.

5.0

4.0

3.0

2.0

1.0

r = 0.64

10 20 30

-1 u-1L I P O P R O T E I N L I P A S E , / i m o l F F A m f h

Relationship between Intralipid removal and postheparinplasma LPL activity in untreated diabetics of insulin-deficient (o) and of maturity-onset type (•). The slope is y= 1.23 + 0.11 x (r = +0.64, p < 0.01).

mg./lOO ml.) since the much enlarged endogenoustriglyceride pool decreases the rate of Intralipid clear-ance and no information is obtained on actual removalefficiency. The fractional elimination rate of In-tralapid was significantly correlated to postheparinplasma LPL activity (r = +0.64, p < 0.01, figure 4)but did not show any relationship to the HL activity (r= +0.02, p > 0.10).

Plasma VLDL-TG turnover in relation to postheparinplasma lipase activities. The mean plasma VLDL-TGproduction rate was slightly but not significantly in-creased in patients with ketotic diabetes and clearlyelevated in the patients with maturity-onset-typediabetes and moderate elevation of serum TG level(table 2). Cases with severe hypertriglyceridemia (>300 mg./lOO ml.) were not studied because of ex-tremely slow fractional turnover rate of VLDL-TG. Ofseven patients with moderate early-onset-type dia-betes who underwent the turnover study, only one hadincreased VLDL-TG production while three were at alow normal range. When calculated from the com-bined data of all diabetics, the fractional turnover rateof VLDL-TG correlated with postheparin plasma LPLactivity (r = +0.40, p < 0.05, figure 5) but not withthe HL activity. The production (turnover) rate ofVLDL-TG was not related to LPL but showed a weakpositive correlation with HL (r = +0.37, p < 0.05).

Relation of postheparin plasma lipase activities to obesityand to plasma insulin and blood glucose. There was nocorrelation between the relative body weight and theLPL or HL activities of postheparin plasma. Inmaturity-onset diabetics the plasma LPL and HL ac-tivities were similar in obese and nonobese subjects(16.9 vs. 16.5 Atrnol FFA./ml./hr. for LPL and 31.0vs. 25.5 fimoi FFA./ml./hr. for HL, respectively).

In insulin-deficient diabetics the LPL activity ofpostheparin plasma was significantly related to fasting

16 DIABETES, VOL. 26 , NO. 1

Dow


ESKO A. NIKKJLA, M.D., AND ASSOCIATES

.cO

O

ccLU>Ozcc

FIG. 5.

0.7

0.6

0.5

0.4

0.3

0.2

0.1

-

-

-

-

o

o

0

o

o

0

0 °

--o o °o

oo0

o

1

' 0

r = 0.40

10 20 3 0

L I P O P R O T E I N L I P A S E , ^ m o l FFA • ml"1- h"1

Relationship between the fractional turnover rate ofplasma VLDL-TG and postheparin plasma LPL activity indiabetic patients. The slope of the regression line is y =0.155 + 0.01 x (r = +0.40, p < 0.05).

plasma insulin levels (r = +0.34, p < 0.05). On theother hand, no such relationship was present in pa-tients with maturity-onset-type diabetes (figure 6).The HL activity was not related to plasma insulinvalues. The fasting blood glucose level did not show acorrelation with any of the lipase activities.

1cc -a.O

30

20

E 10

o r = 0.34 j

• r = 0.01 i

10 20 30 40

PLASMA INSULIN (BASAL) / lU/ml

FIG. 6. Relationship between fasting plasma insulin postheparinplasma LPL activity in ketotic or nonketotic insulin-deficient (o) and in maturity-onset-type (o) diabetic pa-tients. The correlation for all patients and for (o) is notsignificant. For (o) the regression is y = 12.9 + 0.48x (r =+ 0.34, p < 0.05).

Effect of antidiabetic treatment on postheparin plasmalipase activities. Ten diabetic patients who were studiedduring ketosis were restudied after an interval of 10 to20 days, during which period their diabetes had beenbrought under adequate control by insulin (fastingblood glucose > 125 mg./lOO ml.). In every subjectthe postheparin plasma LPL activity increased (figure7) and the serum triglyceride level decreased (meanchange from 204 ± 30 to 108 ± 19 mg./lOO ml.)

_ during the treatment. No consistent change occurredin the HL activity (21.9 ± 3.9 vs. 21.7 ± 3 . 8 jxmolFFA/ml./hr.). Seven non-insulin-requiring patientswere restudied after their diabetes had been ade-quately controlled with glibenclamide. The posthepa-rin plasma LPL increased during treatment in fourpatients, was not changed in two, and decreased inone. The difference between the mean LPL values be-fore and after treatment was not significant (figure 7).The serum triglyceride level and HL activity remainedunchanged during administration of glibenclamide.

? <

40

30

20

INSULIN

30

20

10

p <0.01

GLIBENCLAMIDIf

N.S.

UNTREATED TREATED UNTREATED TREATED

FIG. 7. Effect of antidiabetic treatment on postheparin plasma LPLactivity.

The effect of phenformin administration was tested infour diabetic patients, but no response of either LPL orHL was observed within two weeks in spite of a satis-factory control of hyperglycemia.

DISCUSSION

The results of the present study support the viewthat the hypertriglyceridemia of diabetic patients hasmore than one pathogenetic mechanism, which mayoperate simultaneously even in an individualpatient.1 9 1 0 The present data also explain some of thecontroversy that exists in the literature on the abnor-malities of PHLA, • tissue LPL, and triglyceride re-

JANUARY, 1977 17

Dow



moval in diabetes.11"23 Separation of the two post-heparin plasma triglyceride lipase activities has re-vealed that PHLA may be a poor indicator of thechanges of LPL and HL in diabetes, since the activitiesdeviate in opposite directions as was the case in hyper-lipemia associated with adult-onset diabetes. Since theactivity of HL is generally higher and more variablethan the activity of LPL, fluctuations of the formerenzyme easily mask substantial changes of LPL in theassay of PHLA. The poor correlation of plasma PHLAto tissue LPL activity has previously been demon-strated in alloxan-diabetic rats.27

A subnormal postheparin plasma LPL activity waspresent in most patients with untreated ketotic dia-betes and hypertriglyceridemia. When the diabeticketosis was not associated with hyperlipemia the LPLactivity was normal. On insulin treatment the LPLactivity was rapidly increased, and this was accom-panied by a fall in the serum triglyceride concentra-tion. These results are in agreement with the previousdemonstration of low PHLA in human insulin-deficient diabetes,11 and they strongly support theview that the hypertriglyceridemia accompanying se-vere insulin deficiency is at least partly caused by adecrease of activity of LPL1011 and concomitant im-pairment of TG removal.12

Since the lipase levels were determined from plasmawithdrawn relatively early after heparin injection, it isprobable that the present values of postheparin plasmaLPL reflect mainly the activity of the rapidly re leasa-ble LPL pool located at the vascular endothelial cells.This part of LPL rather than that located in actualtissues is primarily responsible for the breakdown andremoval of chylomicron and VLDL triglycerides.1936

This view is supported by the correlation found be-tween the early-peak postheparin plasma LPL activityand the fractional removal rate of exogenous fat fromthe blood in nondiabetic subjects.37 The results of thepresent study indicate that a similar relationship existsalso in diabetic patients. Thus, decrease of the en-dothelial LPL could account for much of the elevationof plasma triglyceride levels in those diabetic patientswho have a severe deficiency of endogenous insulin.On the other hand, in some patients with ketoticdiabetes the secretion of VLDL into plasma was in-creased, suggesting that the hypertriglyceridemiacould arise also from a combination of increased influxand decreased efflux of plasma TG. This combinedmechanism is present also in experimental diabetes.47

It is not known, however, whether the decrease ofendothelial LPL in the absence of insulin involves all

tissues or whether the abnormality is limited only toadipose tissue, where this enzyme is known to beinsulin-sensitive. 14"16>38>39 The effect of insulin onmyocardial LPL is still under dispute. The nonfunc-tional enzyme located in extravascular myocardiummay be insulin-sensitive,40 whereas total myocardialLPL has been reported to be insulin-independent.1439

The activity of LPL in different tissues of insulin-deficient human diabetics is not known.

The patients with moderate early-onset diabetesdiffer from the ketotic diabetics by having some en-dogenous insulin, but they are unable to increase insu-lin secretion in response to glucose. The abnormalitiesof triglyceride metabolism observed in this group areslight and correspond well with those that might beexpected in mild to moderate insulin deficiency. Theaverage postheparin plasma LPL activity was signifi-cantly lower than in age-matched nondiabetic con-trols, and the mean plasma triglyceride level wasslightly elevated. However, most of the diabetic pa-tients in this category had both LPL activity andserum triglyceride within normal range. Apart from onecase, the production rate of VLDL-TG was normal orlow. It is probable that a slight impairment of serumtriglyceride removal caused by insufficient insulin se-cretion is partly compensated for by a moderately re-duced VLDL synthesis. This kinetic pattern explainswhy these patients had normal or only moderatelyelevated serum triglyceride levels. Our results agreewith the previous observation of Elkeles et al.41 thatdiabetics with low insulin response have less hyper-triglyceridemia than those with normal or increasedinsulin.

A remarkable feature in the chronic insulin-treateddiabetics was the presence of elevated postheparinplasma LPL activity in several patients. In contrast toother diabetic patients, this group also did not showany correlation between the LPL activity and serumtriglyceride level. These observations suggest thatduring insulin administration the serum triglyceridelevel is mainly determined by the rate of triglycerideinflux rather than by the efficiency of its removal.This situation is perhaps explained by the fact thatinsulin stimulates both the synthesis of adipose tissueLpLi4.16.18.38,39 anc} t h e hepatic production of

VLDL.42 Even though there is no direct evidence forthe presence of insulin excess in these patients, it ispossible that the doses of insulin needed for control ofdiabetes may at least periodically producehyperinsulinism. We have found similar features oftriglyceride metabolism in patients with insulin-


Dow



secreting islet-cell adenomas: the fractional turnoverrate of VLDL-TG is high, its production rate is nor-mal or increased, and the level of serum triglyceride iseither low or normal.43 It has been reported recentlythat plasma LDL level is higher (type Ha hyperlipo-proteinemia is more frequent) in insulin-treated dia-betics than in nondiabetics or in other diabetic pa-tients and that the LDL of insulin-treated diabetics istriglyceride-rich—i.e., it contains more of the inter-mediate remnant particles than normal LDL.44 Thisfinding is consistent with the above kinetic pattern,where both the synthesis of VLDL and its catabolismto remnants and LDL are accelerated.

The patients with maturity-onset-type diabeteshave different abnormalities of serum triglyceridekinetics, and it is difficult or impossible to separatethe effects of diabetes from those caused by othermetabolic disturbances, such as obesity and familialhyperlipoproteinemia. In the present study most pa-tients with maturity-onset diabetes had postheparinplasma LPL activity within the same range as age-matched nondiabetic controls, while only exceptionalcases with severe hyperlipemia had somewhat low LPLvalues. On the other hand, many patients had elevatedserum triglyceride level in spite of a normal posthepa-rin plasma LPL activity. Still, in the whole group ofmaturity-onset diabetics the log triglyceride showed asignificant inverse .correlation with postheparinplasma LPL. Unlike the juvenile diabetics, the pa-tients with maturity-onset-type diabetes did not showany correlation between the basal plasma insulin leveland postheparin plasma LPL activity. It is thus clearthat in maturity-onset diabetes the LPL activity isusually not impaired and low activities, when present,are not caused by deficiency of insulin. Accordingly,the hyperlipemia associated with maturity-onset dia-betes is likely to be caused by an overproduction ofVLDL rather than by its deficient peripheral removal.This conclusion is supported by previous kineticstudies, which have demonstrated that the absoluteturnover rate of plasma triglyceride is increased inadult-onset diabetes.1013 Exceptions to this rule arefound in diabetic patients who develop gross hyper-lipemia with accumulation of both VLDL andchylomicrons into the blood.23 The present materialincluded four such patients (serum triglyceride above1,000 mg./ml.), and two of these had clearly subnor-mal LPL. In these instances the hypertriglyceridemiais caused by a combination of increased productionand defective removal of plasma triglycerides.Moreover, upon excessive increase of the plasma tri-

glyceride pool the removal system becomes over-loaded, and in this situation a "secondary" removaldefect aggravates the lipemia. These patients regularlyhave an impaired clearance of exogenous fat1213 and afat-induced hyperlipemia3 in spite of a moderate de-crease of LPL activity. The basic nature of themetabolic abnormality in these patients is not under-stood, but the possibility of a familial trait was notexcluded in the present cases.

Pykalisto, Smith, and Brunzell have recently re-ported data showing that the in-vitro heparin-releasable LPL activity of adipose tissue is subnormalin patients with untreated maturity-onset diabetes.18

Most of these patients had elevated basal triglycerideand insulin levels. These results are not necessarilyinconsistent with our observations since, as em-phasized above, the early postheparin plasma LPL re-sponse reflects the sum activity of endothelial LPL ofmany tissues, not merely -of adipose tissue. Thus, alow LPL activity in adipose tissue may well be com-pensated for by an increased activity in other capillarybeds, such as those in skeletal muscle and myocar-dium. This possibility is also supported by the recentresults of Brunzell, Porte, and Bierman, who found,in agreement with our present data, that the early-peak PHLA response is normal in diabetichyperlipemia.23 On the other hand, many of the pa-tients studied by Pykalisto and co-workers18 hadmarked familial hyperlipemia, and it is therefore pos-sible that they represent the special group of hyper-lipemic diabetics discussed above. Thus, the apparentdiscrepancy between the results of the two studiescould be entirely due to different principles of selec-tion of cases.

The increase of postheparin plasma HL activity inhyperlipemic diabetics is consistent with our previousdemonstration that a positive correlation exists be-tween serum triglyceride level and HL activity innondiabetic obese subjects.37 We do not have anyadequate explanation for this relationship, but itseems that high hepatic lipase activities of postheparinplasma are particularly associated with an increasedproduction of VLDL.

ACKNOWLEDGMENTS

The authors are grateful to M. Kekki, M.D., forcarrying out the computer analyses for the turnoverstudies and to Mrs. Marja-Leena Suovirta, Sirkka-LiisaRuneberg, and Seija Lehtonen for skillful technicalassistance.

JANUARY, 1977 19

Dow



This study has been aided by grants from the Fin-nish State Medical Research Council (Academy of Fin-land), from Nordisk Insulinfond, and from the Fin-nish Cultural Foundation. Two of the authors(E.A.N. and C.E.) are senior investigators of the Fin-nish State Medical Research Council.

REFERENCES

'Nikkila, E. A.: Triglyceride metabolism in diabetes mellitus.Prog. Biochem. Pharmacol. 8:271-99, 1973-

2Adlersberg, D., and Wang, C.-I.: Syndrome of idiopathichyperlipemia, mild diabetes mellitus, and severe vascular damage.Diabetes 4:210-18, 1955.

3Bagdade, J. D., Porte, D., Jr., and Bierman, E. L.: Diabeticlipemia. N. Engl. J. Med. 276:427-33, 1967.

4Woodside, W. F., and Heimberg, M.: Hepatic metabolism offree fatty acids in experimental diabetes. Isr. J. Med. Sci.8:309-16, 1972.

5Schnatz, J. D., Formaniak, J. M., and Chlouverakis, C: Theorigin of hypertriglyceridemia in streptozotocin diabetic rats.Diabetologia 8:125-30, 1972.

6Balasse, E. O., Bier, D. M., and Havel, R. J.: Early effects ofanti-insulin serum on hepatic metabolism of plasma free fattyacids in dogs. Diabetes 21:280-88, 1972.

7Steiner, G., Poapst, M., and Davidson, J. K.: Production ofchylomicron-like lipoproteins from endogenous lipid by intestineand liver of diabetic dogs. Diabetes 24:263-7 1, 1975.

8Basso, L. V., and Havel, R. J.: Hepatic metabolism of freefatty acids in normal and diabetic dogs. J. Clin. Invest.49:537-47, 1970.

9Reaven, E. P., and Reaven, G. M.: Mechanisms for develop-ment of diabetic hypertriglyceridemia in streptozotocin-treatedrats. Effect of diet and duration of insulin deficiency. J. Clin.Invest. .54:1167-78, 1974.

l0Nikkila, E. A., and Kekki, M.: Plasma triglyceride transportkinetics in diabetes mellitus. Metabolism 22:1-22, 1973.

"Bagdade, J. D., Porte, D., Jr., and Bierman, E. L.: Acuteinsulin withdrawal and the regulation of plasma triglyceride re-moval in diabetic subjects. Diabetes 77:127-32, 1968.

12Lewis, B., Mancini, M., Mattock, M., Chait, A., and Fraser,T. R.: Plasma triglyceride and fatty acid metabolism in diabetesmellitus. Eur. J. Clin. Invest. 2:445-53, 1972.

I3Kissebah, A. H., Adams, P. W., and Wynn, V.: Interrela-tionship between insulin secretion and plasma free fatty acid andtriglyceride transport kinetics in maturity onset diabetes and theeffect of phenethylbiguanide (phenformin). Diabetologia70:119-30, 1974.

14Kessler, J. I.: Effect of diabetes and insulin on the activity ofmyocardial and adipose tissue lipoprotein lipase of rats. J. Clin.Invest. 42:362-67, 1963-

15Schnatz, J. D., and Williams, R. H.: The effect of acuteinsulin deficiency in the rat on adipose tissue lipolytic activity andplasma lipids. Diabetes 72:174-78, 1963-

16Wing, D. R., Salaman, M. R., and Robinson, D. S.:Clearing-factor lipase in adipose tissue. Factors influencing theincrease in enzyme activity produced on incubation of tissue fromstarved rats in vitro. Biochem. J. 99:648-56, 1966.

17Persson, B.: Lipoprotein lipase activity of human adiposetissue in health and in some diseases with hyperlipidemia as a

common feature. Acta Med. Scand. 793:457-62, 1973.18Pykalist6, O. J., Smith, P. H., and Brunzell, J. D.: Deter-

minants of human adipose tissue lipoprotein lipase. Effect of dia-betes and obesity on basal- and diet-induced activity. J. Clin.Invest. 36:1108-17, 1975.

19Robinson, D. S.: The function of the plasma triglycerides infatty acid transport. In Comprehensive Biochemistry. Florkin,M., and Stotz, E. H., Eds. Amsterdam, Elsevier Publ. Co.,1970, vol. 18:51-116.

20Jones, D. P., Plotkin, G. R., and Arky, R. A.: Lipoproteinlipase activity in patients with diabetes mellitus, with and with-out hyperlipemia. Diabetes 73:565-70, 1966.

21Perry, W. F.: Heparin-activated lipase in diabetic, athero-sclerotic and healthy subjects. Clin. Chim. Acta 76:189-94,1967.

22Shigeta, Y., Nakamura, K., Hoshi, M., Kim, M., and Abe,H.: Effect of dextran sulfate on plasma lipoprotein lipase activityin obese subjects and diabetic patients. Diabetes 76:238-41,1967.

23Brunzell, J. D., Porte, D., Jr., and Bierman, E. L.: Reversi-ble abnormalities in postheparin lipolytic activity during the latephase of release in diabetes mellitus (postheparin lipolytic activityin diabetes). Metabolism 24:1123-38, 1975.

24Krauss, R. M., Windmueller, H. G., Levy, R. I., and Fred-rickson, D. S.: Selective measurement of two different triglyceridelipase activities in rat postheparin plasma. J. Lipid Res.74:286-95, 1973.

25Fielding, C. J.: Further characterization of lipoprotein lipaseand hepatic post-heparin lipase from rat plasma. Biochim.Biophys. Acta 280:569-78, 1972.

26Gries, F. A., Potthoff, S., and Jahnke, K.: The effect ofinsulin on the uptake of radioactive labelled plasma triglyceridesby rat tissue in vivo. Diabetologia 3:3 11-17, 1967.

27Elkeles, R. S., and Williams, E.: Post-heparin lipolytic ac-tivity and tissue lippprotein lipase activity in the alloxan-diabeticrat. Clin. Sci. Mol. Med. 46:661-64, 1974.

28Huttunen, J. K., Ehnholm, C., Kinnunen, P. K. J., andNikkila, E. A.: An immunochemical method for selective meas-urement of two triglyceride lipases in human postheparin plasma.Clin. Chim. Acta 63:335-47, 1975.

29Huttunen, J. K., Ehnholm, C., Kinnunen, P. K. J., andNikkila, E. A.: Postheparin plasma lipoprotein lipase and hepaticlipase in normal subjects and in patients with hypertri-glyceridemia. Clin. Sci. Mol. Med. 30:249-60, 1976.

30Belfrage, P., and Vaughan, M.: Simple liquid-liquid parti-tion system for isolation of labeled oleic acid from mixtures withglycerides. J. Lipid Res. 70:341-44, 1969.

31Kekki, M., and Nikkila, E. A.: Turnover of plasma total andvery low density lipoprotein triglyceride in man. Scand. J. Clin.Lab. Invest. 33:171-79, 1975.

32Kessler, G., and Lederer, H.: Fluorometric measurement oftriglycerides. In Automation in Analytical Chemistry. Skeggs, L.T., Ed. New York, Mediad, 1966, p. 341.

33Huang, T. C., Chen, C. P., Wefler, V., and Raftery, A.: Astable reagent for the Liebermann-Burchard reaction. Applicationto rapid serum-cholesterol determination. Anal. Chem.33:1405-08, 1961.

34Herbert, V., Law, K.-S., Gottlieb, C. W., and Bleicher, S.J.: Coated charcoal immunoassay of insulin. J. Clin. Endocrinol.Metab. 23:1375-84, 1965.

35Documenta Geigy. Scientific Tables, 6th Edition. Basle, J.R. Geigy, S. A., I960, p. 588.

20 DIABETES, VOL. 2 6 , NO. 1

Dow


ESKO A. NIKKILA, M.D. , AND ASSOCIATES

36Blanchette-Mackie, J., and Scow, R. O.: Sites of lipoproteinlipase activity in adipose tissue perfused with chylomicrons. Elec-tron microscope cytochemical study. J. Cell. Biol. 57:1-25,197 1.

37Huttunen, J. K., Ehnholm, C. Nikkila, E. A., and Ohta,M.: Effect of fasting on two postheparin plasma triglycerideUpases and triglyceride removal in obese subjects. Eur. J. Clin.Invest. 5:435-45, 1975.

38Nikkila, E. A., and Pykalisto, O.: Regulation of adiposetissue lipoprotein lipase synthesis by intracellular free fatty acid.LifeSci. 7: 1303-09, 1968.

39Borensztajn, J., Samols, D. R., and Rubenstein, A. H.:Effects of insulin on lipoprotein lipase activity in the rat heart andadipose tissue. Am. J. Physiol. 223:1271-75, 1972.

40Aktin, E., and Meng, H. C : Release of clearing factor lipase(lipoprotein lipase) in vivo and from isolated perfused hearts of

alloxan diabetic rats. Diabetes 27 :149-56, 1972.41Elkeles, R. S., Lowy, C , Wyllie, A. H., Young, J. L., and

Russell Fraser, T.: Serum insulin, glucose and lipid levels amongmild diabetics in relation to incidence of vascular complications.Lancet 7:880-82, 1971.

42Topping, D. L., and Mayes, P. A.: The immediate effects ofinsulin and fructose on the metabolism of the perfused liver.Changes in lipoprotein secretion, fatty acid oxidation and esterifi-cation, lipogenesis and carbohydrate metabolism. Biochem. J.726:291-311, 1972.

43Nikkila, E. A., and Kekki, M.: Triglyceride turnover inpatients with hyperinsulinism due to islet cell adenoma. In prep-aration.

44Schonfeld, G., Birge, C , Miller, J. P., Kessler, G., andSantiago, J.: Apolipoprotein B levels and altered lipoprotein com-position in diabetes. Diabetes 23:827-34, 1974.

JANUARY, 1977 21

Dow


Documents

Postheparin Plasma Lipoprotein Lipase and Hepatic Lipase