Cell Tissue Res (1987) 249:221-226 Cell and T'tssue Resealr �9 Springer-Verlag 1987

In vivo binding and uptake of low-density lipoprotein-gold- and albumin-gold conjugates by parenchymal and sinusoidal cells of the fetal rat liver H. Franke, U. Diirer, B. Schlag, and R. Dargel Institute of Pathological Biochemistry, Friedrich Schiller University, Jena, German Democratic Republic

Summary. To elucidate the participation of fetal rat liver cells in the receptor-mediated internalization of low-density lipoproteins (LDL), rat fetuses were injected with either LDL-gold or albumin-gold conjugates. The degree of bind- ing and uptake of LDL-gold and albumin-gold by paren- chymal and sinusoidal cells of the fetal rat liver differs markedly. Endothelial cells exhibit low LDL-gold uptake. In contrast, parenchymal cells internalize LDL-gold more actively (45+ 8 LDL conjugates/100 g m 2 cytoplasm within 60 min). Kupffer cells exceed this value by a factor of 20. The uptake of albumin-gold by endothelial and Kupffer cells is high, whereas it is extremely low in parenchymal cells. Estradiol pretreatment causes a significant doubling (p<0.05) of the LDL-gold particle density/100 gm 2 cyto- plasm both in parenchymal and Kupffer cells, whereas es- tradiol has no effect on the albumin uptake. The results strongly indicate that LDL uptake by parenchymal and Kupffer cells in the fetal rat liver is mediated by estrogen- inducible receptors, which may correspond to B, E recep- tors in the adult liver.

Key words: Liver - LDL-gold conjugates - Albumin-gold conjugates- Endocytosis B, E receptor - Rat

The liver plays a key role in the formation and degradation of lipoproteins. In various mammalian species more than 50% of the circulating low-density lipoproteins (LDL) in the blood of adults is cleared by the liver (for review, see Mahley and Innerarity 1983; Newton 1985). Biochemical and morphological studies reveal that liver parenchymal and non-parenchymal ceils contribute to the in vivo degra- dation of LDL. This process can be stimulated by the ad- ministration of estrogen, which is known to affect the ex- pression of the B, E receptor (van Berkel 1980; Brown and Goldstein 1979; Floren et al. 1981 ; Handley et al. 1981a, 1983; Kovanen et al. 1979; van Tol and van Berkel 1980; Windler et al. 1980).

However, there are no reports of morphological investi- gations concerning the contribution of fetal rat liver cells to the receptor-mediated catabolism of serum LDL. We have previously examined the occurrence and formation of lipoprotein particles in the developing rat liver from mid-

Send offprint requests to: Dr. H. Franke, Institute of Pathological Biochemistry, Friedrich Schiller University, DDR-6900 Jena, Ger- man Democratic Republic

gestation to term (Diirer et al. 1986; Franke et al. 1985). The aim of the present paper is to analyze the participation of parenchymal and sinusoidal lining cells of the fetal rat liver in the binding and uptake of serum LDL via high- affinity receptors. Specific apo B, E receptors have been demonstrated in the fetal liver of man, dog and swine (Mah- ley et al. 1981), but not in rats. In the fetal rat, the question of a receptor-mediated clearance of LDL from the plasma is of particular interest insofar as we have found in previous biochemical studies that LDL constitute the main lipopro- tein fraction in fetal rat serum at term, the level of LDL being 5 times as high as in adults (Schlag and Winkler 1978).

It was not possible to perform competitive binding ex- periments in vivo using an excess of unlabeled LDL, be- cause of the naturally occurring high LDL level in fetal rat serum. Therefore, fetuses were administered albumin- gold conjugates in order to compare the manner in which this ligand is taken up by fetal liver cells, as compared with the uptake of LDL-gold complexes. Additionally, some of the experimental groups were pretreated with estra- diol, which is known to stimulate the expression of B, E receptors.

Materials and methods

Animals'. Fetuses from Wistar rats, strain Uje: Wist, on gestational day 22 were used. The pregnant rats, approxi- mately 3 months old (180-220 g body weight), were housed under standardized conditions. The day of the onset of pregnancy was determined by vaginal smears.

Isolation of LDL. Fetal LDL were isolated from freshly pooled serum of 22-day-old rat fetuses by preparative ultra- centrifugation (d < 1.020-1.064 g/ml), as described in detail elsewhere (Winkler et al. 1981). The LDL protein content was determined by the method of Lowry et al. (1951).

Preparation of LDL-gold and albumin-gold conjugates. Col- loidal gold (cAu) was prepared according to Frens (1973) having particles with a diameter of 19 nm. 100 ml cAu were conjugated to 2 mg LDL protein according to Handley et al. (1981 b). Albumin-gold complexes were made by con- jugating 1.5 mg bovine serum albumin (Serva, Heidelberg) to 10 ml cAu (Warchol eta . 1982). Samples of the freshly prepared and extensively dialysed LDL, and also gold con- jugates of LDL and albumin, were negatively stained with

222

Fig. I a, b. Fetal liver, 5 min after administration of LDL-gold conjugates. Perisinusoidal region with endothelial (EN) and parenchymal cells (PC). In the space of Disse (SD), clusters of LDL-gold particles are seen; some are at the stage of internalization via coated pit regions ( ~ ), whereas others have already been endocytosed (e) by parenchymal cells. The endothelial projections contain no conjugates. gl Glycogen, S sinusoidal space, M mitochondria, x 55000. Inset: Freshly prepared LDL-gold conjugates negatively stained with 1% sodium phosphotungstate, x 140000

Fig. 2. Fetal rat liver, 60 min after administration of LDL-gold conjugates. Golgi region (GA) of a sinusoid-distant parenchymal cell with some electron-dense endosomes (e) that contain a variable number of LDL-gold conjugates in a diffuse arrangement, x 31000. Inset: One of the dense endosomes (e) at higher magnification, x 60000

Fig. 3. Fetal rat liver, 60 min after administration of albumin-gold conjugates. Perisinusoidal region showing an endothelial cell projection (EN). Clusters of gold particles are bound to coated pits ( / ' ) at the adluminal and abluminal side of the cell surface, lntracellularly, albumin-gold-containing light and dark endosomes (e) are to be seen; the former are occasionally of a tubular-shaped appearance (re). S Sinusoidal lumen, E erythrocyte, M mitochondria, x 22000

223

Fig. 4. Fetal rat liver, 5 min after administration of LDL-gold conjugates showing an area of a Kupffer cell (KC), the surface of which is densely covered with LDL-gold clusters. A large number of LDL-gold particles are endocytosed via coated and uncoated vesicles (see inset b). ICS Intercellular space, Fe ferritin-containing endosome, x 43000. Inset a." One of the Kupffer cell projections (KC) traversing a sinusoid (S). Its surface is densely covered by numerous LDL-gold particles, x 33000. Inset b: An assembly of LDL-gold particle-containing coated vesicles (cv) at the Kupffer cell periphery, x 31000

Fig. 5. Fetal rat liver, 60 min after administration of LDL-gold conjugates demonstrating an area of a Kupffer cell. At this time LDL-gold particles are preferentially enclosed in electron-dense endosomes (e). In contrast to the 5-min findings, their number has increased in comparison with the surface-bound conjugates, which have decreased. 1CS Intercellular space, x 55000

1% sodium phosphotungsta te for electron microscopic ex- aminat ion (Fig. 1 a, inset).

Experimental conditions. 16 pregnant rats were divided into two equal groups. Rats of group 1 were injected i.m. with 0.1 ml estradiolvaler ianate (31 gg/100 g body weight, VEB Jenapharm), dissolved in oleum arachidis, on gestational day 20 and 21. Rats of group 2 were sham-injected with oleum arachidis. In one half of the rats of group I (estra- diol-treated), l apa ro tomy was performed and, after opening either of the uterine horns, 3 4 fetuses from each rat were exposed. The fetal membranes were retracted and 25 gl LDL-gold (1 x 1014 conjugates/ml) were injected into the umbilical vein. The other half of the rats of group 1 received an equivalent dose of a lbumin-gold conjugates. The concen- trat ion of gold conjugates was determined photometr ical ly

by measuring the absorbance at 525 nm (Horrisberger 1980).

Immediate ly after the injection, the umbilical cord was ligated, cut through and each fetus transferred into a humi- dified micro incubator at 37 ~ C. 5 min and 60 min following the injection of the gold conjugates (LDL- or a lbumin- gold), 4 liver tissue samples were randomly taken from each fetus for electron-microscopic analysis. F ixat ion and em- bedding procedures have been described elsewhere (Franke et al. 1985).

Morphometry. To obtain a semiquanti tat ive estimate of the extent to which liver parenchymal cells and Kupffer cells part ic ipated in the uptake of LDL-go ld and albumin-gold 2 liver samples from each fetus were randomly chosen for sectioning 60 min after the injection of the conjugate. In

224

each case, cytoplasmic areas of 10 nucleated parenchymal cells and 5 nucleated Kupffer cells were randomly photo- graphed at x 7500 and enlarged to x 30000 for scoring the number of gold particles/100 gm 2 cytoplasm, 60 min after administration. Since the sinusoidal endothelial cells from controls and estradiol-treated fetuses did not contain appreciable amounts of LDL-gold, they were excluded from the morphometric analysis, which was performed according to Weibel (1979). All morphometric data were subjected to statistical analysis by Student's t-test. Values are given as mean_+ SEM. Comparable values that were significantly different at p < 0.05 were noted.

R e s u l t s

Liver parenchymal cells"

Shortly after administration, varying amounts of LDL-gold conjugates can be seen diffusely distributed in the space of Disse and the adjoining intercellular regions (Fig.I a, b), whereas the more sinusoid-distant intercellular spaces are still free from conjugates. Some of the LDL-gold parti- cles are attached in clusters to indented coated areas of the plasma membrane of parenchymal cells, others are be- ing internalized (Fig. 1 a, b). Initially, most of the endocy- tosed LDL are located near the sinusoidal cell pole of the parenchymal cells. They are enclosed either in coated vesi- cles or in larger electron-lucent endosomes. In the latter case, they contain up to 30-40 LDL-gold conjugates, mostly arranged along the periphery of the endosomal ma- trix. 60 min after administration, in both sinusoid-near and sinusoid-distant parenchymal cells, larger amounts of LDL conjugates are accumulated in endosomes having a diame- ter of about 100-200 nm. At this stage, they are preferen- tially assembled around the Golgi complex (Fig. 2) and ar- ound the bile capillaries. Most of the endosomes that con- tain LDL-gold are of the dense type (endolysosomes). In contrast to LDL-gold, the binding of albumin-gold to liver parenchymal cells appears to be low at both observation times. In spite of the presence of larger numbers of albumin conjugates in the space of Disse and the close contact of some albumin-gold particles with the surface membrane of parenchymal cells, the incidence of endocytosed albumin conjugates is verly low. Their number per endosome varies between 2 and 10 and few endosomes have more than 10 particles, in contrast to the 30-50 LDL-gold conjugates per endosome. Morphometric analysis demonstrates that parenchymal cells internalize significantly more LDL-gold than albumin-gold conjugates (see controls, Fig. 6).

Endothelial and Kupffer cells

Stages of LDL-gold binding and uptake by endothelial cells are not seen as frequently as in parenchymal cells. On the other hand, the internalization of albumin-gold particles is high. This is indicated by numerous electron-lucent and electron-dense endosomes occurring in the endothelial cells and their long flattened projections, which contain large numbers of albumin-gold particles (Fig. 3).

Kupffer cells exhibit the greatest activity in binding and uptake both of LDL-gold and albumin-gold conjugates. 5 min following the injection of the gold conjugates, the total Kupffer cell surface, including the long cell projections and slender plasma membrane tentacles, are covered by small clusters of LDL-gold (Fig. 4 and inset a) or by albu-

%

~100--" c- o u

- o

o

~ 5 0 -

0

E

Z

T

r LDL Albumin Hepatocytes

Number of LDL-gold

T

1 T

T

I

C E C I E " LDL Albumin Kupffer Cells

-2000

-151)0

%

.1000

"E 0 0

" 0

0 �9 500 ~

E

Z

Fig. 6. and albumin-gold conjugates/ 100 gm 2 cytoplasm in fetal rat parenchymal and Kupffer cells from controls (C) and after estrogen pretreatment (E), 60 min following injection. The difference in the LDL-gold density/100 gm z between untreated and estrogen-pretreated parenchymal cells (o vs x), be- tween both groups of Kupffer cells (o vs + ) and between parenchy- mal and Kupffer cells (o vs o; x vs +) are significant at p<0.05

min-gold particles (not shown). At the same time, varying amounts of electron-lucent endosomes are present along the cell periphery. These endosomes are filled with LDL- gold (Fig. 4) or with albumin-gold particles (not shown). The internalization of the LDL-gold particles occurs via coated pits and vesicles (Fig. 4, inset b), and partly also via uncoated pits. 60 min after administration, the size and number of the endosomes, their electron density and the amount of LDL-gold or albumin-gold conjugates that they contain has markedly increased (Fig. 5). Morphometric analysis reveals that Kupffer cells have taken up about 30% more albumin-gold than LDL-gold particles, 60 min after administration (Fig. 6).

Estradiol experiments

In comparison with the controls, those fetuses whose mother was hormone-pretreated, exhibit an increased bind- ing and uptake of LDL-gold particles both in hepatocytes and Kupffer cells, but not in endothelial cells. A measurable estradiol effect on the endocytosis of albumin-gold was not observed. Since the differences in the uptake of LDL-gold conjugates between controls and estradiol-treated animals are most evident 60 min after administration, we have ex- amined this period morphometrically (Fig. 6). Control he- patocytes take up 45 _+ 8 LDL-gold conjugates/100 gm 2 cy- toplasmic area, whereas control Kupffer cells internalize

225

about 20 times more LDL-go ld conjugates (880+26 gold particles/100 gm2). Estradiol t reatment causes an approxi- mate twofold increase in the density of LDL-gold conju- gates/100 t.tm 2 both in parenchymal and in Kupffer cells. The uptake of a lbumin-gold by parenchymal cells in con- trols and after hormone treatment is equal, with a value of 1 2 + 3 albumin-gold conjugates/100 ldm2; however, in compar ison with the LDL-gold uptake, the endocytosed quantit ies of a lbumin-gold particles are 3 times and 7 times lower in controls and estradiol- t reated livers, respectively. Unst imula ted Kupffer cells take up about 30% more albu- min-gold than LDL-go ld conjugates. The difference in the endocytosis of a lbumin-gold conjugates between controls and hormone-s t imula ted Kupffer cells is not statistically significant.

Discussion

Our quali tative and quanti tat ive findings demonstra te that parenchymal and sinusoid lining cells contr ibute differently to the L D L clearance from plasma. Al though liver endothe- lial cells possess a high endocytot ic capacity (Bankston and de Bruyn 1974; de Bruyn et al. 1985; Handley et al. 1983), the internal izat ion rate of LDL-go ld by this sinusoidal cell type is low in the fetal rat liver near term. This is in contrast to the good LDL-go ld binding and uptake capacity of par- enchymal and Kupffer cells. The lat ter show the highest endocytot ic activity with respect to both LDL-go ld and albumin-gold conjugates. This indicates that Kupffer cells have a " scavenger" function in fetal rat liver; this is appar- ently not restricted to specific ligands. Even though Kupffer cells exhibit a 20-fold higher uptake of LDL-go ld particles than parenchymal cells, the lat ter p robab ly contr ibute more effectively to the total catabolism of serum L D L in rat, because at term the volume fraction of fetal parenchymal cells is about 85%, whereas that of Kupffer cells is only about 4% (Greengard et al. 1972)

Our morphometr ic da ta reveal that the uptake of LDL- gold is significantly increased in parenychmal and Kupffer cells following estradiol pre t reatment of pregnant rats. This finding supports our view that B, E receptors are involved in the internalizat ion of LDL-gold conjugates in a similar way to that repor ted for parenchymal cells in adul t rat liver (Chao e ta l . 1979; Handley e ta l . 1981a; Kovanen et al. 1979; Windler et al. 1980). Our suggestion that L D L uptake is mediated by the B, E receptor is also consistent with da ta showing that native serum LDL of rat fetus con- tains only apol ipoprote in B, other protein consti tuents amount ing to less than 0.5% of total L D L protein (unpub- lished data). Fur ther evidence for the existence of B, E receptors in fetal liver parenchymal cells in the rat at term is provided by results showing the lack of a s t imulatory action of estrogen on binding and uptake of albumin-gold, and by incubat ion experiments using isolated fetal hepato- cytes (gestational day 22), which reveal a statistically signifi- cant inhibit ion of binding and uptake of gold-labeled L D L in the presence of a 15-fold excess of unlabeled L D L (Franke et al. 1985a).

In comparison with liver parenchymal cells in adult rat (Handley et al. 1981 a, 1983), the s t imulatory effect of estro- gen is, however, lower in the rat fetus. This difference in the response to estrogen can be explained by a dosage effect. In order to prevent developmental disturbances in the rat fetus, we had to choose an estradiol dose much lower than

that appl ied by Handley et al. (1981a, 1983). In addit ion, the competi t ive action of unlabeled serum L D L of the fetus must be taken into considerat ion with respect to our in vivo experiments.

Acknowledgements. We wish to thank Mrs. H. Guder and Mr. F. Reim for their excellent technical assistance.

References

Bankston PW, de Bruyn PPN (1974) The permeability to carbon of the sinusoidal lining cells of the embryonic rat liver and rat bone marrow. Am J Anat 141:281 290

Berkel TJC van (1980) The role of non-parenchymal cells in liver metabolism. Trends Biochem Sci 4:202 205

Brown MS, Goldstein JL (1979) Receptor-mediated endocytosis: Insights from the lipoprotein receptor system. Proc Natl Acad Sci USA 76:3330-3335

Bruyn de PPH, Michelson S, Bankston PW (1985) In-vivo endocy- tosis by bristle-coated pits and intracellular transport of endo- genous albumin in the endothelium of the sinuses of liver and bone marrow. Cell Tissue Res 240 : 1-7

Chao Y, Windler EE, Chen GC, Havel RJ (1979) Hepatic catabol- ism of rat and human lipoproteins in rats treated with 17~- ethinyl estradiol. J Biol Chem 254:11360-11366

Dfirer U, Franke H, Dargel R, Ude J (1986) A post-embedding immunoelectron-microscopic demonstration of apolipoprotein- B-containing lipoprotein particles in hepatocytes of fetal rats. Histochemistry 84:263-270

Floren C-H, Kushwaha RS, Hazzard WR, Albers JJ (1981) Estro- gen-induced increase in uptake of cholesterol-rich very low den- sity lipoproteins in perfused rat liver. Metabolism 30:367-375

Franke H, Dargel R (1985) Morphometric studies on lipoprotein particles in developing rat liver and their corticosteroid-induced changes during the late gestational period. Cell Tissue Res 242 : 661-667

Franke H, Dfirer U, Dargel R (1985) Elektronenmikroskopisch- morphometrische Untersuchungen fiber den Lipoproteinmeta- bolismus in der fetalen Rattenleber. 2. Hepatolog Symp Bio- chem Ges DDR and Ges Klin Chemic Lab Diagn 1985. Wiss Beitr Fr Schiller Univ Jena (in press)

Frens G (1973) Controlled nucleation for the regulation of the particle size in monodisperse gold solution. Nature Phys Sci 241:20~22

Greengard O, Federman M, Knox WE (1972) Cytomorphometry of developing rat liver and its application to enzymic differenti- ation. J Cell Biol 52:261 272

Handley DA, Arbeeny CM, Eder HA, Chien S (1981a) Hepatic binding and internalization of low density lipoprotein-gold con- jugates in rats treated with 17~-ethinyl-estradiol. J Cell Biol 90:778-787

Handley DA, Arbeeny CM, Witte LD, Chien S (1981 b) Colloidal gold-low density lipoprotein conjugates as membrane receptor probes. Proc Natl Acad Sci USA 78:368 371

Handley DA, Arbeeny CM, Chien S (1983) Sinusoidal endocytosis of low density lipoprotein-gold conjugates in perfused livers of ethinyl-estradiol treated rats. Eur J Cell Biol 30:266-271

Horrisberger M (1980) Evaluation of colloidal gold as a cytochemi- cal marker for transmission and scanning electron microscopy. Biol Cell 36:253-258

Kovanen PT, Brown MS, Goldstein JL (1979) Increased binding of low density lipoprotein to liver membranes from rats treated with 17c~-ethinyl estradiol. J Biol Chem 254 : 11367-11373

Lowry OH, Rosebrough NJ, Farr AL, Randall RJ (1951) Protein measurement with the Folin phenol reagent. J Biol Chem 193 : 265-275

Mahley RW, lnnerarity TL (1983) Lipoprotein receptors and cho- lesterol homeostasis. Biochern Biophys Acta 737:197-222

Mahley RW, Hui DY, Innerarity TL (1981) Two independent lipo-

226

protein receptors on hepatic membranes of dog, swine and man. J Clin Invest 68 : 1197-1206

Newton RS (1985) Modulation of hepatic and extrahepatic LDL receptors: Involvement in the progression of atherosclerosis. Drug Dev Res 6:141-154

Schlag B, Winkler L (1978) Konzentration und Zusammensetzung der Serumlipoproteinklassen der fetalen Ratte. Acta Biol Med Germ 37:233-237

Tol A van, Berkel TJC van (1980) Uptake and degradation of rat and human very low-density (remnant) apolipoprotein by parenchymal and non-parenchymal rat liver cells. Biochem Bio- phys Acta 619 : 156-163

Warchol J B, Brelihska R, Herbert DC (1982) Analysis of colloidal gold methods for labelling proteins. Histochemistry 76 : 567-575

Weibel ER (1979) Stereological methods. Vol I, Practical methods for biological morphometry. Academic Press, London

Windler EET, Kovanen PT, Chao YC, Brown MS, Havel R J, Goldstein JL (1980) The estradiol-stimulated lipoprotein recep- tor of rat liver. J Biol Chem 255:10463-10471

Winkler L, Schlag B, Leutert F, Dargel R (1981) Quantitative Lipoproteinanalytik mit der Ultrazentrifuge VAC 601 unter Verwendung eines Winkelrotors und eines Fraktionsschneide- ger/ites. Z Med Labor-Diagn 22:210 214

Accepted December 2, 1986