THE JOURNAL 0 1989 by The American Society for Biochemistry

OF BIOLOGICAL CHEMISTRY and Molecular Biology, Inc

Vol. 264, No. 30, Issue of October 25, pp. 17931-17938,1989 Printed in U.S.A.

Chylomicron Metabolism CHYLOMICRON UPTAKE BY BONE MARROW IN DIFFERENT ANIMAL SPECIES*

(Received for publication, May 23, 1989)

M. Mahmood HussainS, Robert W. Mahley, Janet K. Boyles, Peter A. Lindquist, Walter J. Brecht, and Thomas L. Innerarity From the Gladstone Foundation Laboratories for Cardiovascular Disease, Cardiovascular Research Institute, Departments of Pathology and Medicine, University of California, Sun Francisco, California 94140-0608

Previously it was shown in rabbits that 20-40% of the injected dose of chylomicrons was cleared from the plasma by perisinusoidal bone marrow macrophages. The present study was undertaken to determine whether the bone marrow of other species also cleared significant amounts of chylomicrons. Canine chylomi- crons, labeled in vivo with [14C]cholesterol and [3Hl retinol, were injected into marmosets (a small, New World primate), rats, guinea pigs, and dogs. Plasma clearance and tissue uptake of chylomicrons in these species were contrasted with results obtained in rab- bits in parallel studies. The chylomicrons were cleared rapidly from the plasma in all animals; the plasma clearance of chylomicrons was faster in rats, guinea pigs, and dogs compared with their clearance from the plasma of rabbits and marmosets. The liver was a major site responsible for the uptake of these lipopro- teins in all species. However, as in rabbits, the bone marrow of marmosets accounted for significant levels of chylomicron uptake. The uptake by the marmoset bone marrow ranged from one-fifth to one-half the levels seen in the liver. The marmoset bone marrow also took up chylomicron remnants. Perisinusoidal macrophages protruding through the endothelial cells into the marrow sinuses were responsible for the ac- cumulation of the chylomicrons in the marmoset bone marrow, as determined by electron microscopy. In con- trast to marmosets, chylomicron clearance by the bone marrow of rats, guinea pigs, and dogs was much less, and the spleen in rats and guinea pigs took up a large fraction of chylomicrons. The uptake of chylomicrons by the non-human primate (the marmoset), in associa- tion with the observation that triglyceride-rich lipo- proteins accumulate in bone marrow macrophages in patients with type I, 111, or V hyperlipoproteinemia, suggests that in humans the bone marrow may clear chylomicrons from the circulation. It is reasonable to speculate that chylomicrons have a role in the delivery of lipids to the bone marrow as a source of energy and for membrane biosynthesis or in the delivery of fat- soluble vitamins.

Chylomicrons play a major role in the transport of intesti- nally absorbed lipids. Their triglycerides are partially hydro- lyzed during circulation as a result of the action of endothelial

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ To whom correspondence should be addressed Gladstone Foun- dation Laboratories, P. 0. Box 40608, San Francisco, CA 94140-0608.

cell-bound lipoprotein lipase (1). The resulting remnants are cleared primarily by the liver through a process mediated by apolipoprotein E (2-6) and modulated by apolipoprotein C and phospholipids (3 ,5, 7). In rabbits, in addition to the liver, the bone marrow has been shown to be involved in the clearance of chylomicrons and chylomicron remnants (8). Chylomicrons, chylomicron remnants, and intestinally ab- sorbed cholesterol were taken up by the rabbit bone marrow, the lipoproteins were degraded, and the cholesterol was stored. The cells responsible for clearance by the bone marrow were identified as perisinusoidal macrophages (8). Furthermore, there is evidence that the bone marrow of rabbits takes up phospholipid liposomes and low density lipoproteins (9, 10).

The present study was undertaken to evaluate the involve- ment of the bone marrow in chylomicron metabolism in other species. The results show that the bone marrow not only of rabbits, but also of marmosets (a small, New World primate), is involved in chylomicron metabolism.

MATERIALS AND METHODS

Animals and Diets-Male New Zealand White rabbits (2.0-3.0 kg body weight, Animal West, Soquel, CA) were maintained on Purina Rabbit Chow 5315 (Purina Mills, St. Louis, MO) fed ad libitum. Marmosets were raised at the Gladstone Foundation Laboratories and fed canned marmoset diet (Zu/Preem, Hills Pet Products, To- peka, KS), supplemented with multiple vitamins, fruits, and rice cereal. Both male and female marmoset monkeys, weighing 300-400 g, were used in this study. Adult mongrel dogs (University of Califor- nia, San Francisco) were fed control Purina Dog Chow. Male Sprague- Dawley rats (Bantin and Kingman, Fremont, CA), weighing 200-400 g, were fed Rodent Chow 5001C (Purina). Hartley guinea pigs (Si- monsen Lab, Gilroy, CA), weighing 300-400 g, were maintained on a Guinea Pig Chow 5025C (Purina) diet. All animals were fasted overnight prior to metabolic studies.

Chylomicron and Chylomicron Remnant Preparation-Canine chy- lomicrons were obtained from a lymph fistula created at the left jugular vein essentially as described (11, 12). Mocha Mix (Presto Food Products, Industry, CA), cholesterol, and sucrose were fed to the dogs to induce chylomicron synthesis (8). Nascent chylomicrons were labeled in vivo with ['4C]cholesterol and [3H]retinol as described previously (8). Chylomicrons were isolated from the lymph by ultra- centrifugation (SW 28 rotor, 28,000 rpm, 90 min, 20 "C). In chylo- microns, more than 90% of the [3H]retinol and more than 70% of the ['4C]cholesterol were esterified, suggesting that the particles were primarily derived from the intestine.

Chylomicron remnants were prepared by circulating chylomicrons in hepatectomized rabbits for 30 min. A functional hepatectomy was performed by ligating the superior and inferior mesenteric and celiac arteries, along with the portal vein (8). Ligatures were also placed around the base of the liver lobes. Femoral arteries were also tied to decrease the bone marrow uptake of chylomicrons. Chylomicron remnants were isolated from the plasma of hepatectomized rabbits by ultracentrifugation (SW 28 rotor, 28,000 rpm, 2 h 45 min, 4 "C) (8).

In Vivo Metabolic Studies-The in vivo metabolic studies were performed on anesthetized animals that had been fasted overnight

17931

17932 Chylomicron Uptake by the Bone Marrow (15-18 h). The lipoproteins were injected into either a femoral, jugular, or ear vein. Blood samples were collected at designated time points from the contralateral femoral or ear artery or the jugular vein. Aliquots of plasma samples were counted for 10 min in a liquid scintillation counter (LS7500, Beckman Instruments). At the end of each experiment the animals were perfused at 80-90 mm of mercury pressure through the left ventricle. Rabbits were first perfused with about 500-600 ml of ice-cold minimal essential medium (GIBCO) and then with the same amount of ice-cold 3% glutaraldehyde in 0.1 M sodium cacodylate, pH 7.3. Other animals were perfused similarly, except for the amount of the medium and glutaraldehyde solution used 200-300 ml for marmosets and rats, 300-350 ml for guinea pigs, and 2500-3000 ml for dogs. The liver, bone marrow, spleen, kidney, lung, adrenals, heart, and adipose tissue were obtained at the end of the perfusion. Tissue slices (0.1-0.5 g) were digested in 0.5 ml of 6 N KOH, and [14C]cholesterol and [3H]retinol were extractedinto hexane as described (8). The hexane extracts were evaporated, suspended in 0.5 ml of 100% ethanol, and counted in the presence of 10 ml of nonaqueous scintillation mixture (Beckman).

For electron microscopy, the spleen and the bone marrow from the femur were subjected to post-fixation, dehydration, embedding, and sectioning as described (8). The sections were examined using a JEOL electron microscope (model 1OOCXII).

Calculation of the percentage of injected dose of lipoproteins re- maining in the plasma was based on the estimate that plasma volumes of rabbit, marmoset, dog, rat, and guinea pig constituted 3.5,3.5,4.35, 3.13, and 3.5% of body weight, respectively. The uptake of chylomi- crons/g of tissue was obtained from duplicate determinations. In the case of bone marrow, two samples were obtained from both the femur and tibia and the uptake values were averaged. The organ uptake was based on the actual weight of the organs, except for the bone marrow. The bone marrow was estimated to be 2.2% of the total body weight in dogs and rabbits (13-16), whereas in rats it was estimated to be 3.0% (17), and in guinea pigs 1.75% (18). The bone marrow weight in marmosets was estimated to be 2.2% (see "Discussion").

Tissue Distribution of Intestinally Absorbed ~'C]Choksterol-The rabbits received 25 pCi of [14C]cholesterol in 0.5 ml of corn oil followed by 20 ml of Mocha Mix (8). Rats and guinea pigs were anesthetized with methoxyflurane (Metofane, Pitman-Moore, Inc., Washington Crossing, NJ) and were given ['4C]cholesterol (2.0-2.5 pCi) in 0.6 ml of corn oil via gastric intubation. All animals were euthanized 4.5 or 6 h after the injection of radiolabeled cholesterol. The tissues were collected, then digested in 6 N KOH, and lipids were extracted as described above and in Ref 8.

RESULTS

Chylomicron Clearance by the Bone Marrow in Various Animals

Canine thoracic duct chylomicrons labeled in uiuo with ["C] cholesterol and 13H]retinol were injected into normal fasted rabbits, marmosets, rats, guinea pigs, and dogs. The plasma die-away and tissue uptake were determined. The metabolism of chylomicrons in rabbits served as the basis for comparison of the metabolism of these lipoproteins in other species. As summarized in Table I, the same chylomicron preparations were used in parallel in rabbits, marmosets, rats, guinea pigs, and dogs.

Rabbits-As established previously (81, chylomicrons were rapidly cleared from the plasma of rabbits and were taken up by both the liver and bone marrow. A typical plasma clearance and tissue distribution at 30 min after injection of the radio- labeled chylomicrons is shown in Fig. 1, A and E. Using four different preparations of chylomicrons (Table I), we deter- mined that 23-35% of the injected dose of chylomicrons was retained in the plasma at 20-30 min after injection and that the uptake was primarily by the liver (15-48%) and bone marrow (14-33%). Likewise, chylomicron remnants were cleared by both the liver and the bone marrow, and in a hepatectomized rabbit the bone marrow was the principal site responsible for chylomicron catabolism (Table I).

Marmosets-The plasma clearance and tissue uptake of chylomicrons in marmosets were compared with results ob-

tained in the rabbit. As shown in Fig. IC, 37 and 29% of the injected dose of ['4C]cholesterol- and [3H]retinol-labeled chy- lomicrons, respectively, remained in the plasma at 30 min. The liver accumulated 28% of the ['4C]cholesterol and 25% of the [3H]retinol. The marmoset bone marrow was respon- sible for the clearance of 13 and 9% of the ['4C]cholesterol- and [3H]retinol-labeled chylomicrons in the injected dose, respectively (Fig. lD) , whereas the spleen accumulated 2-7% of the injected chylomicrons. Other organs studied contained no more than 1% of the injected lipoproteins (Fig. I D ) .

As summarized in Table I, the amount of radiolabeled chylomicrons remaining in the plasma at 30 min ranged, in different experiments, from 29 to 46% of the injected dose, and the liver uptake varied from 23 to 28%. The bone marrow took up 6-13% of the injected chylomicrons, whereas the spleen had 2-7% of the injected chylomicrons. Furthermore, the chylomicron remnants were rapidly cleared from the plasma of marmosets (28 and 22% of the injected ['4C]choles- terol-labeled remnants remained in the plasma at 20 and 60 min, respectively). At 60 min, 29 and 15% of the injected dose of ['4C]cholesterol-labeled chylomicron remnants were within the liver and the bone marrow, respectively (Table I).

Even though the calculations of the uptake of chylomicrons by marmoset bone marrow were based on the assumption that the bone marrow constitutes 2.2% of the body weight (see "Discussion"), an estimate based on the values established for other animals (13-18), chylomicron uptake could be directly calculated as uptake/g of tissue rather than per organ. This uptake was then compared to the chylomicron uptake/g of tissue in the rabbit. As shown in Table 11, the chylomicron clearance/g of bone marrow in both the rabbit and marmoset was similar (0.5-1.1% of the injected dose/g). Likewise, rabbit and marmoset chylomicron uptake/g of bone marrow was approximately equal to the uptake/g of liver. However, as shown in Fig. 1D, marmoset liver (4.6 2 0.5% (n = 6)) cleared a larger fraction of chylomicrons compared to rabbit liver (2.8 f 0.5% ( n = 30)) because it constitutes a larger portion of the body weight. In contrast, the spleen of both animals takes up a large percentage of the injected dose/g (Table 11); however, the spleen is a relatively small organ in marmosets (0.1 & 0.03% of total body weight, n = 7) and rabbits (0.05 f 0.01%, n = 30) and thus only accounts for 1-7% of the clearance of chylomicrons/organ.

Rats-Chylomicrons were cleared more rapidly from rat plasma (Fig. 1E) as compared with rabbits (Fig. L4) and marmosets (Fig. 1C). At 30 min, only 5.0% of the ["c] cholesterol and 4.0% of the [3H]retinol of the injected dose were retained in the rat plasma. Liver uptake accounted for 84% of the ['4C]cholesterol- and 76% of the [3H]retinol- labeled chylomicrons. The spleen had the next highest uptake, with 8% of the ['4C]cholesterol and 5% of the [3H]retinol. In this experiment, the bone marrow accumulated 6% of the [14C]cholesterol and 2% of the [3H]retinol (Fig. 1F).

As summarized in Table I, chylomicron clearance from the plasma was rapid in all rats studied. The clearance by the liver ranged from 54 to 88% of the injected dose. In contrast, the uptake by the bone marrow ranged from less than 1% to about 6% of the injected dose. The uptake by the spleen ranged from 5 to 9%. These studies suggest that the liver is the major organ for chylomicron clearance in rats. However, the uptake of chylomicrons by the bone marrow increased when chylomicron clearance was studied in a hepatectomized rat. In a hepatectomized rat, 61% of the ['4C]cholesterol- and 55% of the [3HJretinol-labeled chylomicrons remained in the plasma at 30 min, and the bone marrow accumulated 11% of the ['4C]cholesterol and 10% of the [3H]retinol (Table I). In

Chylomicron Uptake by the Bone Marrow 17933 TABLE I

Plasma and tissue distribution of chylomicrons and chylomicron remnants in different species Distribution of radiolabeled lipoproteins" Recoveryb

Species Lipoproteins Injected Plasma Liver Bone marrow Spleen dose ['HI ["C] [3H] [14C] 13H] ["C] [3H] ["Cl

['HI I'4Ccl

Rabbit Normal

Normal

Hepatectomized Marmoset

Normal

Normal

Rat Normal

Hepatectomized Guinea pig

Normal

DOE

Chylomicrons I' Chylomicrons I1 Chylomicrons 111 Chylomicrons IV Chylomicron

remnants Id Chylomicrons V

Chylomicrons I1 Chylomicrons I11 Chylomicrons IV Chylomicron

remnants VI'

Chylomicrons I Chylomicrons I Chylomicrons 111 Chylomicrons 111 Chylomicrons V

Chylomicrons I11 Chylomicrons 111

body wt

24 150 200 100 21

152

150 200 100 105

w l k g

33 24

200 200 15

200 200

22.8 23.1 24.2 26.2 31.3 23.9 34.1 35.3 15.6 22.9 25.8 14.5 16.2 51.1

47.1 48.4 0

35.0 44.0 45.8 25.4 28.8 36.9 25.3

22.0

4.1 53.5 0.2 81.5 7.6 8.2 56.0 3.7 4.5 15.7

54.5 61.3 0

11.0 11.7 35.6 13.3 14.3 30.7

% of injected dose

48.3 17.8 38.4 23.9 28.4 13.5 23.1 23.6

16.4

0 32.2

22.9 28.1 5.5 21.7 8.7 28.6

1.5 0.3

61.1 1.4 83.1 3.0 0 9.6

41.8 0.8 35.3 1.2

30.6 33.4 28.8 31.7

31.8

9.3 11.4 13.0 14.7

3.0 6.1

10.9

1.2 1.9

1.2 1.2 0.9 1.4 2.2

0.1

1.9 3.9

9.1 6.7 5.8 4.7 2.5

5.8 L0.4

2.5 67.5 1.8 76.2 1.7 65.3 2.1 63.1

87.6

0.2 83.1

2.4 3.6 78.5 7.1 67.7 2.4

70.0 95.0

7.4 77.0 1.7 87.7 3.0 76.1

7.0 55.5' 12.5 58.0'

115.0 106.2 95.8 83.1

86.8

71.6 91.2 86.3 81.0

83.5 101.9 86.3

64.8' 66.0'

1.6 64.0 76.8 Normal Chylomicrons I1 150 9.6 11.3 52.4 62.9 0.2 0.3 1.4 a Animals were fasted overnight and anesthetized before injection of chylomicrons. At 30 min, the animals were

* The sum of chylomicrons recovered from plasma, liver, bone marrow, spleen, kidney, lung, heart, and adrenal

e Chylomicron distribution in tissues was determined at 20 min, rather than 30 min.

perfusion fixed as described under "Materials and Methods," and the tissues were analyzed.

gland.

Chylomicron remnants were prepared by injecting chylomicrons into hepatectomized rabbits and allowing them to circulate for 30 min. Chylomicron remnants were isolated from the plasma of hepatectomized rabbits as described under "Materials and Methods." Triglyceride to cholesterol ratio in the isolated particles was 25.

e Chylomicron remnants were prepared as in Footnote d, but chylomicron remnant distribution was assessed at 60 min instead of 30 min.

'The low recovery of chylomicrons was due to high uptake by adipose tissue in guinea pig (see text).

contrast to the rat, the bone marrow from a hepatectomized rabbit cleared 32% of the injected dose (Table I). These experiments suggest that rat bone marrow has the potential to take up significant quantities of chylomicrons.

Guinea Pigs-In a typical experiment in guinea pigs, 12% of the [14C]cholesterol- and 11% of the [3H]retinol-labeled chylomicrons remained in the plasma at 30 min, while the liver accumulated 41% of the [14C]cholesterol and 35% of the [3H]retinol (Fig. 1, G and H ) . The spleen was the next major organ, accumulating 7 % of the [14C]cholesterol and 6% of the [3H]retinol. The bone marrow accounted for 1% chylomicron clearance. Other tissues studied took up 51% of the injected lipoproteins. As summarized in Table I, 11-14% of the in- jected chylomicrons in different studies remained in the plasma at 30 min. The spleen cleared 6-13%, and the bone marrow accumulated 1-2% of the injected chylomicrons. These studies suggest that guinea pig bone marrow does not play a major role in chylomicron metabolism, whereas the spleen may be involved in this process.

The recovery of radiolabeled chylomicrons in the major tissues (liver, bone marrow, spleen, kidney, lung, heart, adre- nals, and plasma) of the guinea pig ranged from 56 to 66% of the injected dose (Table I). The low recovery of ['4C]choles- terol and I3H]retinol appears to have been caused in large part by the high level of accumulation of the radiolabeled chylomicrons in the adipose tissue of this animal. When 200

mg of chylomicron triglyceride was injected per kg of body weight, perinephric adipose tissue of guinea pigs accumulated 0.13-0.61% of the injected dose of ['4C]cholesterol and [3H] retinol/g of tissue at 30 min. Subcutaneous adipose tissue from the chest wall of guinea pigs accumulated 0.26-0.51% of the injected dose. In contrast, the perinephric adipose tissue of rats, rabbits, and marmosets contained 0.02-0.03%, -0.01%, and 0.09-0.15% of the injected dose/g of tissue, respectively. In the rat, the subcutaneous adipose tissue ac- counted for 0.01-0.02% of the injected dose. Assuming the fat content to be 7.1% of the body weight for guinea pigs (as has been reported in rats (19)), approximately 10-16% of the ["C] cholesterol and 6-15% of the [3H]retin~l injected would be accounted for in the adipose tissue. Therefore, adipose tissue may be responsible for much of the chylomicron catabolism in guinea pigs.

Dogs-At 30 min, the dog retained 11% of the ['4C]choles- terol- and 10% of the [3H]retin~l-labeled chylomicrons in the plasma (Fig. 11 and Table I). A t this time the liver had accumulated 63% of the ['4C]cholesterol and 52% of the [3H] retinol. The spleen took up the next highest amount, clearing about 1% of the injected chylomicrons. The bone marrow uptake was 0.2-0.3% of the injected dose (Fig. 1J and Table I). Other tissues studied accumulated less than 0.4% (Fig. 1J). These experiments suggest that the liver is by far the organ with the largest role in chylomicron catabolism in dogs.

17934 Chylomicron Uptake by the Bone Marrow

80 Rabbit B.

60 - [14 C]cholesterol 0 [31i]retinol ai:

Liver Bone Soleen Kidney Lung Adrenals Heart

FIG. 1. Representative plasma clearance and tissue uptake of ca- nine chylomicrons in the rabbit, marmoset, rat, guinea pig, and dog. Overnight-fasted animals were anesthe- tized, and canine chylomicrons were in- jected into the femoral, jugular, or ear vein. Plasma samples were obtained at designated time points. Animals were perfusion fixed 30 min after injection of the chylomicrons, and tissue uptake was determined. Pan& A and B represent plasma clearance and tissue uptake, re- spectively, in a rabbit. The rabbit re- ceived 150 mg of chylomicron triglycer- ide/kg of body weight. Panels C and D represent plasma clearance and tissue uptake, respectively, in a marmoset in- jected with 100 mg of triglyceride/kg of body weight. The plasma clearance and tissue uptake of chylomicrons in the rat are shown in panels E and F, respec- tively, whereas panels G and H represent plasma clearance and tissue uptake in the guinea pig. The rat and guinea pig received 200 mg of triglyceride/kg of body weight. The plasma clearance and tissue uptake in the dog are presented in panels I and J, respectively. The dog was injected with 150 mg of triglyceride/kg of body weight.

i+? 100

60

40

80 K Rat E.

20

10 59 49

a o 10 20 30 ul

10

0 IO 20 30

Minutes

; 2jfl d d Lwer Bone Ween Kidney Lung Adrenals Heart

8 , Marrow ’

60

II I-- rer Bone Spleen a 0% Kidney Lung Adrenals

w Marrow 0

0 80 Guinea Pig H.

# -

Liver Bone Spleen Kidney Lung Adrenals Heerl Marrow

Tissue Uptake of Intestinal Lipoproteins Induced by Feeding Unsaturated Fat and /WJCholesterol

Consideration was given to the possibility that the uptake of chylomicrons by the spleen of rats and guinea pigs and the lack of uptake by the bone marrow in these two species were secondary to the use of canine lipoproteins. To rule out this possibility, [ 14C]cholesterol was administered along with un- saturated fat by gastric intubation to label endogenous intes- tinal lipoproteins. Rabbits were studied in parallel for com- parison. The [W]cholesterol was used in these studies be- cause we had previously shown that it was retained within the bone marrow for several hours after uptake of the radio- labeled chylomicrons (8). Regardless of the potential compli-

cation that the [%]cholesterol could be reutilized for endog- enous lipoprotein production or transferred to other plasma lipoproteins, the tissue distribution of [YJ]cholesterol should provide at least a qualitative estimate of the in vivo sites of chylomicron catabolism. In these studies, animals were eu- thanized at 4.5 or 6 h after the administration of the [14C] cholesterol, and the uptake of the radiolabeled lipoproteins determined per g of tissue. As shown in Table III, the various animals absorbed different quantities of the radiolabel; how- ever, the data confirmed the observations obtained with in- jected canine chylomicrons. In rats and guinea pigs the liver took up a much larger fraction of the [14C]cholesterol-labeled intestinal particles/g of tissue (Table III). As shown in Table III, the ratio of liver to bone marrow uptake was at least 2-

Chylomicron Uptake by the Bone Marrow 17935

TABLE I1 Comparison of the chylomicron uptakelg of tissue

in the rabbit and marmoset An overnight-fasted rabbit and an overnight-fasted marmoset were

anesthetized, and chylomicrons were injected a t a dose of 150 mg of triglyceride/kg of body weight. The animals were perfusion fixed at 30 min, and chylomicron uptake/g of tissue was determined. The values represent the average of two determinations for each tissue, except for bone marrow. The bone marrow values are the average of four determinations, two each from the femur and tibia.

Rabbit Marmoset

[3H]Retinol [‘4C]Cholesterol [3H]Retinol [‘4C]Cholesterol % of injected dose/g of tissue

Tissue ___

Liver 0.45 0.72 1.04 1.27 Bone marrow 0.52 0.72 0.70 1.11 Spleen 1.02 1.56 2.40 6.09 Kidney 0.01 0.01 0.04 0.03 Lung 0.08 0.11 0.16 0.35 Adrenals 0.03 0.17 0.16 0.45 Heart 0.04 0.04 0.14 0.24 Adipose tissue 0.01 0.01 0.05 0.08

TABLE 111 Tissue distribution of [’4C]cholesterol after i n uiuo absorption in the

rabbit, rat, and guinea pig The animals received [‘4C]cholesterol with unsaturated fat by

gastric intubation as described under “Materials and Methods.” They were euthanized a t 4.5 or 6 h, and tissue distribution of intestinally absorbed (’4C]cholesterol~ was assessed (see “Materials and Meth- ods”). Each set of values represents a separate animal.

[“CjCholesterol Time of ___ Liverbone Liver/spleen analysis Liver I h e Spleen marrow ratio ratio

marrow

(dpm X lO-’)lg of tissue Rabbit

4.5 h 1.55 0.55 3.16 2.8 0.5 0.30 10.11 0.37 2.7 0.8

6.0 h 1.02 1.21 2.72 0.8 0.4 0.62 0.20 0.59 3.1 1.1

4.5 h 18.70 1.93 8.85 9.7 2.1 6.0 h 19.48 1.19 17.27 16.4

22.91 2.30 20.87 1.1

10.0 1.1

4.5 h 0.69 0.12 0.67 6.0 h

5.8 4.41 0.37 4.31

1 .o 11.9 1.0

Rat

Guinea pig

fold and as much as 20-fold greater in rats and guinea pigs than in rabbits. On the other hand, uptake of [‘4C]cholesterol- labeled intestinal lipoproteins by the spleen (when expressed as the ratio of liver to spleen uptake/g of tissue) was very similar for rats, guinea pigs, and rabbits (Table 111). However, when the uptake was calculated per organ, there was a greater percentage of uptake by the spleen because of the relatively larger size of the spleen in rats (0.22 f 0.04% of the total body weight, n = 16) and guinea pigs (0.15 k 0.02%, n = 6). By comparison with the size of the rabbit spleen, the spleens of rats and guinea pigs are 3- to 4-fold larger (based on percent of body weight). Thus the results described in these feeding studies establish that native lipoproteins, in addition to iso- lated chylomicrons, do deliver significant amounts of radio- labeled lipoproteins to the spleen but much less to the bone marrow of rats and guinea pigs.

Electron Microscopic Examination of Bone Marrow and Spleen for Chylomicron Uptake

As described previously (8), chylomicrons were visualized within perisinusoidal macrophages in the bone marrow of rabbits after the intravenous injection of these lipoproteins.

Perisinusoidal macrophages in rabbit bone marrow protrude through the endothelium and have access to the marrow sinuses. A very similar electron microscopic image was ob- served within the bone marrow of marmosets. As shown in Fig. 2 A , chylomicron size particles accumulated within peri- sinusoidal macrophages of the marmoset bone marrow. As in the rabbit, the bone marrow of marmosets displayed macro- phages that extended through the endothelium into the si- nuses (Fig. 2B) and accumulated chylomicron size lipopro- teins. These observations are consistent with the biochemical data demonstrating that the bone marrow of marmosets and rabbits represents a major site for chylomicron catabolism.

The sites of chylomicron uptake by the spleen of the various animals were determined. Despite the fact that the spleen of rabbits and marmosets accounted for only low levels of uptake of chylomicrons, chylomicron size particles could be identified within splenic macrophages in rabbits and marmosets (Fig. 3, A and B ) , guinea pigs, and dogs (data not shown). Likewise, chylomicron size lipoproteins were abundant within macro- phages (Fig. 3C) and entrapped within the interstitial spaces (Fig. 3 0 ) of the rat spleen. It was not possible to determine the relative percentage of particles within the cells and trapped between cells. These results are consistent with the biochemical observation demonstrating that chylomicrons are taken up by the spleen.

DISCUSSION

In a previous study (81, rabbit and canine chylomicrons were shown to be cleared almost entirely by the liver and bone marrow of rabbits. In the present study, canine thoracic duct chylomicrons, labeled in uiuo with [’4C]cholesterol and [3H] retinol, were used to study uptake by the bone marrow in the marmoset, rat, guinea pig, and dog. Approximately two-thirds of the injected chylomicrons were cleared from the circulation of these animals within 30 min after intravenous injection. This rapid clearance of chylomicrons is consistent with results from previous studies reported for rats, rabbits, and dogs (12, 19-29). In dogs, rats, and guinea pigs, the chylomicrons cleared from the plasma were recovered primarily in the liver. The observation that the liver is the primary organ responsi- ble for chylomicron clearance is consistent with other studies in dogs and rats (12, 19, 21, 25-29). However, in rabbits and marmosets the bone marrow also cleared a significant fraction of chylomicrons, whereas in rats, guinea pigs, and dogs the bone marrow was less active.

The calculation of the uptake of chylomicrons by bone marrow is based on the total organ weight of the marrow. This value has been very carefully and exhaustively deter- mined in several species but not in marmosets or other non- human primates. In rabbits, bone marrow represents 2.2- 2.3% of the body weight (13-15), whereas it ranges from 1.9 to 2.4% in dogs (16), 3.0% in rats (17), and 1.75% in guinea pigs (18). We have used 2.2% as the factor for conversion of data from values/g of organ to values/organ in the marmoset. This estimation was based on the following reasons. Marmo- sets have a total body weight approximately equal to that of rats and guinea pigs, and as noted, bone marrow in these two animals averages 1.75-3.0% of the body weight. Furthermore, based on values obtained per g of bone marrow in rabbits and marmosets (Table 111, both the rabbit and marmoset marrow take up very similar amounts of radiolabeled chylomicrons. In addition, the uptake of [‘4C]cholesterol and [3H]retinol by bone marrow was similar to that of the liver (per g) in these animals. Finally, in rabbits, the bone marrow weight is equal to approximately two-thirds of the weight of the liver (15). If this ratio is similar in the marmoset, the actual bone marrow

17936

FIG. 2. Electron micrographs of the bone marrow from marmosets injected with chylomicrons. Thirty min after injection of the chylomicrons, the animals were perfusion fixed as de- scribed under “Materials and Methods.” A, a marmoset was injected with 200 mg of chylomicron triglyceride/kg of body weight. Among the marrow cells a large macrophage containing numerous 1000- to 2000-A particles resembling chylomi- crons can be seen. The particles are present in intracellular endosomal vesi- cles. This macrophage also contains a large lipid droplet (LD). X3200. B, a perisinusoidal macrophage protruding through the endothelium (the protrusion is indicated by the arrows) into a blood sinus of the marrow. The macrophage projection contains numerous chylomi- cron size particles. RBC, red blood cell. X3200.

Chylomicron Uptake by the Bone Marrow

”I

weight would be a t least 3% of the body weight. For all of these reasons we have used 2.2% as a reasonable estimate.

Macrophages were responsible for the uptake of chylomi- crons in the bone marrow of rabbits and marmosets. Specifi- cally, the uptake was restricted primarily to perisinusoidal macrophages, which had access to the circulating blood through processes extending into the marrow sinuses. How- ever, in some other organs, macrophages have access to the plasma but do not take up chylomicrons (e.g. the Kupffer cells in the rabbit liver do not (8)). Furthermore, even though rabbit bone marrow macrophages have access to all circulating lipoproteins, only triglyceride-rich lipoproteins are taken up (8). Previously it has been reported that human bone marrow accumulates large quantities of triglyceride-rich lipoproteins present in the plasma of patients with type I, 111, or V hyperlipoproteinemia (30).

In contrast to rabbits and marmosets, the bone marrow of dogs and guinea pigs does not appear to be involved to a significant extent in the clearance of chylomicrons (less than 2% of the injected dose). It is possible that the bone marrow of these species lacks the ability to take up chylomicrons because the bone marrow macrophages in these species do not have easy access to large lipoprotein particles in the blood sinuses. Alternatively, in these species chylomicrons may not be required in the bone marrow because other lipoproteins or transport molecules fulfill the same functions as chylomi- crons. In addition, it is possible that the bone marrow of the rabbit and marmoset performs a function requiring chylomi- crons and chylomicron remnants that is served by another organ in guinea pigs and dogs, e.g. the spleen.

Rat bone marrow appears to take up chylomicrons at a level below that of rabbit and marmoset marrow and above that of guinea pig and dog marrow. Although the liver cleared the highest amounts of chylomicrons in the rat, the bone marrow did take up 1-6% of the injected dose. Small amounts of chylomicrons have been shown to be cleared by the rat bone marrow (27). Furthermore, in the hepatectomized rat, the chylomicron uptake by bone marrow increased to 10-11% of the injected dose. Previous studies have demonstrated that chylomicrons are cleared from the plasma of hepatectomized rats, rabbits, and dogs (8, 21, 29, 31, 32). In the rabbit, the uptake of chylomicrons by bone marrow was almost quanti- tative (Table I and Ref. 8). The rat bone marrow does have

at least a limited ability to take up chylomicrons in a hepa- tectomized animal. In high concentrations and with the long circulation time that occurs in hepatectomized rats, the chy- lomicrons may cross the bone marrow endothelium and reach the macrophages.

Several caveats concerning these data should be mentioned. First of all, in many of our studies we have injected high levels of chylomicrons or fed a fat-rich diet to induce a significant concentration of intestinal lipoproteins. Even though these levels may seem large, one needs to recall that humans absorb very large quantities of lipids (several grams) after a typical meal. However, in other studies we have attempted to admin- ister much lower levels of chylomicrons, to infuse whole lymph slowly, or to feed only the low-fat animal chow containing the radiolabeled cholesterol (8). In all of these studies with rab- bits, the bone .marrow took up significant quantities of chy- lomicrons or intestinally absorbed lipids. Second, even though all of our data are consistent with chylomicrons and chylo- micron remnants being principally involved in the delivery of the radiolabeled cholesterol and retinol to the bone marrow, we cannot entirely rule out the possibility that some of the cholesterol might be transferred to other lipoproteins (espe- cially in the rabbit and marmoset, which have cholesteryl ester transfer protein) or reutilized in the synthesis of endog- enous lipoproteins and be taken up by the bone marrow. Studies are in progress using an isolated hind-limb perfusion system for direct measurement of bone marrow uptake of lipoproteins.

The physiological importance of the uptake of chylomicrons by the bone marrow is not known. However, there are several possibilities. First, chylomicrons and chylomicron remnants may serve to deliver lipids or fat-soluble vitamins to the bone marrow. Fatty acids may serve as an energy source and cholesterol may be used in membrane biosynthesis during hematopoiesis. Fat-soluble vitamins are transported by lipo- proteins, specifically chylomicrons, which may deliver these vitamins to the marrow. For example, vitamin E is known to play a key role in stabilizing red blood cell membranes. In fact, autohemolysis can be corrected in some patients by administering vitamin E (33). Furthermore, it has been sug- gested that large oral doses of vitamin A can be used to maintain remission in children with acute myeloid leukemia (34). The possible importance of chylomicrons in this process

Chylomicron Uptake by the Bone Marrow 17937 I , . . . . .

FIG. 3. Electron micrographs of the spleen of different species 30 min after the injection of chylomi- crons. A, the spleen of a rabbit given an injection of 150 mg of chylomicron tri- glyceride/kg of body weight has macro- phages that contain chylomicron size particles. X3450. B, the spleen of a mar- moset injected with 100 mg of chylomi- cron triglyceride/kg of body weight has macrophages that contain intracellular chylomicron size particles and lipid droplets (LD). X3450. C and D, the spleen of a rat injected with 33 mg of chylomicron triglyceride/kg of body weight contains intracellular chylomi- cron size particles within a macrophage (panel C ) and trapped in the intracellu- lar space between cells (panel D). X3450.

was suggested in i n vitro studies in which vitamin A delivered to promyelocytic leukemic cells on chylomicrons was the most effective means of inducing differentiation and inhibiting proliferation of these cells (35) .

A second possible role for chylomicron uptake by the bone marrow is to maintain adipocyte lipid stores. The turnover of fatty acids in bone marrow adipocytes of rabbits is five times higher than in extramedullary adipocytes (36); thus, the adi- pocytes appear to play a dynamic role in the bone marrow. Furthermore, there is an inverse relationship between high hematopoietic cell synthetic activity and adipocyte triglycer- ide content (37). The triglyceride may be mobilized as an energy source or may simply provide space for expansion of the hematopoietic elements. In addition, adipocytes may be involved in regulating the hematopoietic microenvironment of the bone marrow. I n vitro hematopoietic cultures appear to require marrow stromal elements, including adipocytes, for the normal maintenance, proliferation, differentiation, and maturation of bone marrow stem cells (38). Further studies are required to distinguish among these possibilities.

In rats, guinea pigs, and dogs, the spleen was the organ

with the second highest uptake of radiolabeled chylomicrons (after the liver). However, the spleen only accounted for -1.5% of the injected dose of chylomicrons in dogs and 5- 13% of the injected dose in rats and guinea pigs, calculated per organ (Table I). Previously, the uptake of chylomicrons has been shown to occur in the rat spleen (27,28). To establish that the uptake by rat and guinea pig spleen was not due to the use of canine chylomicrons or to chylomicron alteration during isolation, these animals were fed ['4C]cholesterol plus fat to label endogenous intestinal particles. Significant amounts of radiolabeled lipoproteins were detected in the spleens of these animals (Table 111). In fact, the liver and the spleen in rats and guinea pigs accumulated similar amounts of radiolabel when calculated per g of tissue. This demon- strates that the spleen does accumulate native endogenous intestinal lipoproteins. In addition, it was noted that a similar amount of ['4C]cholesterol was also taken up by the rabbit spleen (liver/spleen ratio; Table 111). However, because the spleens of rats and guinea pigs are much larger than those of rabbits, there is a greater total uptake (per organ) in the spleens of rats and guinea pigs. I t is intriguing to consider

17938 Chylomicron Uptake by the Bone Marrow

that the relative differences in the uptake of chylomicrons by the bone marrow versus the spleen (greater bone marrow uptake in rabbits and marmosets uers’sus greater spleen uptake in rats and guinea pigs) might reflect different metabolic roles for these organs in the different species. Furthermore, the difference between bone marrow and spleen uptake of chylo- microns may provide insights into the role of chylomicron catabolism in these organs, both of which function in hema- topoiesis. A correlation between hematopoiesis in the bone marrow and the spleen and uptake of chylomicrons could be of significant physiological importance.

Acknowledgments-We thank Dennis Miranda and Linda Ander- son for excellent technical assistance. In addition, we thank Sylvia Richmond for manuscript preparation, AI Averbach and Sally Gullatt Seehafer for editorial assistance, and Charles Benedict and Tom Rolain for graphics.

1. 2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

REFERENCES Redgrave, T. G. (1983) Int. Reu. Physiol. 28, 103-130 Sherrill, B. C., Innerarity, T. L., and Mahley, R. W. (1980) J.

Windler, E., Chao, Y., and Havel, R. J. (1980) J. Biol. Chem.

Shelburne, F., Hanks, J., Meyers, W., and Quarfordt, S. (1980)

Borensztajn, J., and Kotlar, T. J. (1984) Proc. Natl. Acad. Sci. U.

Hui, D. Y., Innerarity, T. L., Milne, R. W., Marcel, Y. L., and

Borensztajn, J., Getz, G. S., and Kotlar, T. J. (1988) J. Lipid Res.

Hussain, M. M., Mahley, R. W., Boyles, J. K., Fainaru, M., Brecht, W. J., and Lindquist, P. (1989) J. Biol. Chem. 2 6 4 , 9571-9582

Williams, K. J., Vallabhajosula, S., Rahman, I. U., Donnelly, T. M., Parker, T. S., Weinrauch, M., and Goldsmith, S. J. (1988) Proc. Natl. Acad. Sci. U. S. A. 8 5 , 242-246

Wiklund, O., Witztum, J. L, Carew, T. E., Pittman, R. C., Elam, R. L., and Steinberg, D. (1987) J . Lipid Res. 28, 1098-1109

Rajpal, S. G., and Kirkpatrick, J. R. (1972) J. Surg. Res. 13,260- 261

Biol. Chem. 255,1804-1807

255,5475-5480

J . Clin. Invest. 6 5 , 652-658

S. A. 8 1 , 5863-5866

Mahley, R. W. (1984) J. Biol. Chem. 259 , 15060-15068

29, 1087-1096

12. Melchior, G. W., Mahley, R. W., and Buckhold, D. K. (1981) J. Lipid Res. 22, 598-609

13. Tavassoli, M., and Yoffey, J. M. (1983) Bone Marrow: Structure and Function, pp. 105-123, Alan R. Liss, Inc., New York

14. Nye, R. N. (1931) Proc. SOC. Exp. Biol. Med. 29, 34-37 15. Dietz, A. A. (1944) Proc. SOC. Exp. Biol. Med. 57,60-62 16. Fairman, E., and Whipple, G. H. (1933) Am. J. Physiol. 104,

17. Fairman, E., and Corner, G. W. (1934) Anat. Rec. 6 0 , 1-4 18. Hudson, G. (1958) J. Anat. 92, 150-161 19. Goodman, D. S., Huang, H. S., and Shiratori, T. (1965) J. Lipid

20. Redgrave, T. G. (1973) Biochem. J. 136, 109-113 21. Bergman, E. N., Havel, R. J., Wolfe, B. M., and Bbhmer, T.

(1971) J. Clin. Inuest. 5 0 , 1831-1839 22. Ross, A. C., and Zilversmit, D. B. (1977) J. Lipid Res. 18 , 169-

181 23. Redgrave, T. G., Dunne, K. B., Roberts, D. C. K., and West, C.

E. (1976) Atherosclerosis 24, 501-508 24. Kita, T., Goldstein, J. L., Brown, M. S., Watanabe, Y., Hornick,

C. A.. and Havel. R. J. (1982) Proc. Natl. Acad. Sci. U. S. A.

352-357

Res. 6,390-396

79; 362313627 ’ ~I

25. Blomhoff. R.. Heleerud. P.. Rasmussen. M.. Bera. T.. and Norum. K. R. (1982) Przc. Natl. Acad. Sci. U . S. A. 79, 7326-7330

26. Quarfordt, S. H., and Goodman, D. S. (1967) J. Lipid Res. 8 ,

27. Lossow, W. J., Brot, N., and Chaikoff, I. L. (1962) J. Lipid Res.

28. Goodman, D. S. (1962) J. Clin. Znuest. 41 , 1886-1896 29. Nestel. P. J.. Havel. R. J.. and Bezman. A. (1963) J. Clin. Znuest.

264-273

3,207-215

42,1313-1321 ’

30. Fredrickson. D. S., Goldstein, J. L., and Brown, M. S. (1978) in The Metabolic Basis of Inherited Disease (Stanbury, J. B.,

604-655, McGraw-Hill Book Co., New York Wyngaarden, J. B., and Fredrickson, D. S., eds) 4th Ed., pp.

31. Mjbs, 0. D., Faergeman, O., Hamilton, R. L., and Havel, R. J.

32. Redgrave. T. G. (1970) J. Clin. Invest. 49, 465-471 (1975) J. Clin. Znuest. 56,603-615

33, Kayden, H. J., and Silber, R. (1965) Trans. Assoc. Am. Physiol.

34. Lie, S. O., Wathne, K.-O., Petersen, L. B., Slbrdahl, S. H., and

35. Wathne, K.-O., Norum, K. R., Smeland, E., and Blomhoff, R.

36. Trubowitz, S., and Bathija, A. (1977) Blood 4 9 , 599-605 37. Tavassoli, M. (1984) Exp. Hematol. (Copenh.) 12, 139-146 38. Dexter, T. M., Allen, T. D., and Lajtha, L. G. (1977) J. Cell.

78,334-342

Norum, K. R. (1988) Eur. J. Haematol. 40,460-465

(1988) J . Biol. Chem. 263,8691-8695

Physiol. 91,335-344