Effect of chlorate on the sulfation of lipoprotein lipase and heparan sulfate proteoglycans. Sulfation of heparan sulfate proteoglycans affects lipoprotein lipase degradation.

In avian-cultured adipocytes 76% of the newly synthesized lipoprotein lipase is degraded before release into the medium (Cupp, M., Bensadoun, A., and Melford, K. (1987) J. Biol. Chem. 262, 6383-6388). The same group (Cisar, L. A., Hoogewerf, A. J., Cupp, M., Rapport, C. A., and Bensadoun, A. (1989) J. Biol. Chem. 264, 1767-1774) has proposed that the interaction of lipoprotein lipase with a class of cell surface heparan sulfate proteoglycans is necessary for degradation to occur. To test further this hypothesis, the binding capacity of the plasma membrane for the lipase was decreased by inhibiting the sulfation of glycosaminoglycans with sodium chlorate, an inhibitor of sulfate adenyltransferase. Chlorate decreased sulfate incorporation into trypsin-releasable heparan sulfate proteoglycans to 20% of control levels. The amount of uronic acid in the trypsin-releasable heparan sulfate proteoglycans remained constant. Therefore, chlorate decreased sulfation density on heparan sulfate chains by approximately 5-fold. In the same fractions, chlorate increased the median heparan sulfate Mr measured on Sephacryl S-300. Chlorate decreased the maximum binding of 125I-lipoprotein lipase to adipocytes by 4-fold, but no significant effects on the affinity constants were observed. Chlorate increased lipoprotein lipase secretion in a dose-dependent relationship up to 30 mM. Utilizing a pulse-chase protocol, it was shown that lipase synthesis in control and chlorate-treated cells was not significantly different and that the increased secretion could be accounted for by a decreased lipoprotein lipase degradation rate. In control cells 77 +/- 11% of the synthesized enzyme was degraded whereas in chlorate-treated cells degradation was reduced to 42 +/- 9% of the synthesized amount. The present study shows that decreased sulfation of heparan sulfate proteoglycans decreases the maximum binding of the lipase for the adipocyte cell surface. Consistent with the model that binding of lipoprotein lipase to cell surface heparan sulfate is required for lipase degradation, degradation is reduced in chlorate-treated cultures. In this report it is also shown that chlorate inhibits lipoprotein lipase sulfation and that desulfation of the enzyme has no effect on its catalytic efficiency or on its binding to cultured adipocytes.

ate increased the median heparan sulfate M, measured on Sephacryl S-300. Chlorate decreased the maximum binding of '2SI-lipoprotein lipase to adipocytes by 4fold, but no significant effects on the affinity constants were observed. Chlorate increased lipoprotein lipase secretion in a dose-dependent relationship up to 30 mM. Utilizing a pulse-chase protocol, it was shown that lipase synthesis in control and chlorate-treated cells was not significantly different and that the increased secretion could be accounted for by a decreased lipoprotein lipase degradation rate. In control cells 77 f 11 % of the synthesized enzyme was degraded whereas in chlorate-treated cells degradation was reduced to 42 2 9% of the synthesized amount. The present study shows that decreased sulfation of heparan sulfate proteoglycans decreases the maximum binding of the lipase for the adipocyte cell surface. Consistent with the model that binding of lipoprotein lipase to cell surface heparan sulfate is required for lipase degradation, degradation is reduced in chlorate-treated cultures. In this report it is also shown that chlorate inhibits lipoprotein lipase sulfation and that desulfation of the enzyme has no effect on its catalytic efficiency or on its binding to cultured adipocytes. * This work was supported by Grants HL-14990 and HL-24873 from the National Institutes of Health. 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.
$ T o whom correspondence should be addressed Div. of Nutritional Sciences, Cornel1 University, 321 Savage Hall, Ithaca, NY 14853. In plasma, the major enzyme responsible for the hydrolysis of VLDL' and chylomicron triglycerides to free fatty acids is lipoprotein lipase (LPL). In recent years it has become evident that posttranslational events may be important in the regulation of LPL. In studies utilizing 3T3-Ll adipocytes (l), it was concluded that regulation of LPL by insulin occurred mainly at posttranscriptional and posttranslational levels. Others (2) similarly report that in response to fasting and feeding, the regulation of LPL occurs mainly at the posttranslational level.
One potential posttranslational mechanism in the regulation of LPL is a modulation of the degradation rate of LPL with little or no change in the synthetic rate. Three laboratories have shown, independently, that in isolated cultured adipocytes, 70-80% of the newly synthesized enzyme is degraded before appearing in the culture medium, and that addition of heparin to the medium reduces the amount of LPL degraded (3)(4)(5). Furthermore,  have shown that degradation of LPL is dramatically reduced when the adipocyte cell surface heparan sulfate proteoglycans (HSPG) are removed by endoglycosidases. Therefore, modulation of the interaction of LPL with HSPG on the adipocyte plasma membrane may have dramatic effects on the degradation of LPL and may determine the efflux of LPL from cells. To test this possibility, we have decreased the LPL binding capacity of the plasma membrane by decreasing the sulfation density of glycosaminoglycans. Sodium chlorate, a potent inhibitor of sulfate adenylyltransferase, reduces protein and carbohydrate sulfation (7).  concluded that treatment of human fibroblasts with chloratereduced proteoglycan sulfation but did not affect proteoglycan chain polymerization. This paper reports the effects of chlorate on the sulfation of HSPG and LPL, and the effects of altered HSPG sulfation on LPL turnover in adipocytes. EXPERIMENTAL  rather than to decrease the mass. The effect of chlorate on the incorporation of 35s04 into glycosaminoglycans was more pronounced for chondroitin sulfate/dermatan sulfate than for heparan sulfate. To determine if chlorate altered the average length of heparan sulfate chains in control and chloratetreated cells, the trypsin-releasable fraction was digested with Pronase, and the glycosaminoglycans were precipitated. Heparan sulfate was obtained by digestion of glycosaminoglycans with chondroitinase ABC followed by incubation with alkaline borohydride. The chains were sized on a calibrated Sephacryl S-300 column. Fig. 1 shows that chlorate treatment increased the median size of heparan sulfate chains from 61,790 f 1,710 in control cells to 90,100 f 260 in the chlorate-treated cells.
Effect of Chlorate Treatment of Adipocytes on the Binding of lz5I-LPL to the Cell Surface-Since chlorate treatment reduced the incorporation of 35S04 into HSPG, which are known to bind to LPL (6), it was important to assess whether this change affected the binding of LPL to the adipocyte cell surface. Adipocytes were cultured in the presence and absence of 10 or 20 mM chlorate for 15 h. Following this pretreatment, equilibrium binding experiments were performed with purified lZ5I-LPL (Fig. 2). Incubations were conducted for 2 h at 4 "C. Previous experiments (6) established that equilibrium was reached by 1 h, and that at 4 "C, 95% of the cell surfaceassociated LPL was released by 2 X 1 ml of heparin washes. Additionally, it was shown that iodination of LPL did not affect binding behavior (6). By assuming that total binding is the sum of a high affinity component and a nonspecific component, the binding function can be described by where u represents the pg of LPL bound/dish; [SI, the concentration of free enzyme at equilibrium in pg/ml; K,, the association constant; nl, the maximum amount of enzyme specifically bound/dish; and a, the slope of the nonspecific binding function. The results of four separate binding experiments were fitted by the least squares technique of Marquardt (23), utilizing the computer program of Dell et al. (24). A maximum specific binding of 1.57 f 0.25 pg of LPL/GO-mm dish was obtained for control cultures. Chlorate-treated cultures displayed a reduced maximum specific binding of 0.48

TABLE I1
Effect of chlorate on the degree of sulfation o/glycosaminoglycans in chicken adipocytes Adipocytes were incubated with Na2["'SO4] and 0 or 20 mM sodium chlorate for 48 h after an overnight pretreatment with or without sodium chlorate. The intracellular (IC), trypsin-releasable (TR), and media (M) pools were collected. The TR pool was successively dialyzed against 10 mM Na2SO4 and water, was digested with thermolysin or Pronase, and the products of digestion were separated by gel permeation chromatography. The percentage of heparan sulfate (HS) or chondroitin sulfate and dermatan sulfate (CS/DS) labeled with 3sS0, was determined by digesting the glycosaminoglycans with heparitinase or chondroitinase ABC, respectively. After digestion, the digestion products were separated by gel permeation chromatography. The material resistant to digestion by chondroitinase ABC was used to measure the mass and radioactivity of HS, while the low molecular weight products of the chondroitinase ABC digestion were used for the analysis of CS/DS. Uronic acid was measured, using glucuronic acid as a standard. Values are means f S.D. for three pools of 5-60-mm dishes or from one DOOI for the determination of the Dercent of HS in the total alvcosaminoalvcans.

Molecular Weight
FIG. 1. Molecular weight distribution of 35SOa radioactivity associated with heparan sulfate chains from control (A) and chlorate-treated ( B ) cultures. Total glycosaminoglycans were isolated from the trypsin-releasable fraction of cells and digested with chondroitinase ABC and alkaline borohydride as described under "Experimental Procedures." The molecular weight distribution (0) and the integral distribution curve (0) were determined by gel chromatography on a calibrated column of Sephacryl S-300. Each value represents the mean & S.D. for three separate molecular weight determinations. Where error bars cannot be seen, the S.D. is smaller than the symbol. Effect of Chlorate on LPL in Cultured Adipocytes-Since chlorate altered the degree of sulfation of HSPG and reduced the binding of lZ5I-LPL to the cell surface, it was important to assess how these changes affected the levels of LPL in adipocytes. The secretion of LPL in adipocytes was measured by pretreating cells with increasing concentrations of chlorate and then incubating the cells with the indicated chlorate concentrations for 5 h. LPL was measured in the medium by ELISA. Fig. 3 shows data from one such experiment. Incubation of cells with graded levels of chlorate resulted in a dose-dependent increase in LPL secretion. At 20 mM chlorate concentration, a 3.9-fold increase in LPL secretion was observed. Total cellular protein did not change as a result of chlorate treatment at concentrations up to 50 mM chlorate, and increased LPL secretion was observed at concentrations up to 30 mM chlorate (data not shown). All further experiments utilized either 10 or 20 mM chlorate, since these concentrations consistently produced increased LPL secretion levels without any decrease in cellular protein or DNA levels. Table I11 presents the results of a n experiment where the effects of 20 mM chlorate and 10 unit/ml heparin on LPL levels in the intracellular and cell surface compartments were measured. Chlorate was compared to heparin, a glycosami- Concentration-dependent binding of LPL in adipocytes treated with chlorate. Adipocytes were pretreated with or without chlorate for 15 h. Adipocytes were then maintained in complete RPMI-1640 containing 0.2% bovine serum albumin for 2 h at 37 "C in the presence or absence of 10 mM chlorate. Cells were then incubated with increasing amounts of '"1-LPL (specific activity = 7110 cpm/pgprotein) for 2 h at 4 "C. Incubation media were collected for measurement of free lZ5I-LPL. Bound lZ5I-LPL was released in 2 X 1 ml of heparin washes. Binding curves are from control (upper panel) and chlorate-treated (lower paneel) cells. Total binding (0) and specific binding (A) are shown. Each point represents data from a single 60-mm dish. Shown is data representative of four experiments. noglycan which exerts its action on LPL secretion by decreasing the interaction of LPL with the cell surface, and thereby decreases LPL degradation. The effect of heparin was consistent with the earlier observation that heparin decreases the LPL degradation rate (3, 6). With heparin, LPL in the medium was increased 6-fold. LPL levels on the cell surface and in the cell were decreased; this decrease in the cell-associated pool size is consistent with a lower amount of LPL binding to cell surface HSPG and decreased LPL internalization.
Total amounts of LPL recovered in all three compartments were increased 2.4-fold. The effect of chlorate treatment on LPL levels was similar to heparin, although the effect on LPL secretion was not as dramatic. LPL secretion was increased 3.3-fold, and intracellular and cell surface-associated LPL levels were decreased t o 85 and 62% of control levels, respectively. Total amounts of LPL recovered in all three compart-  ments were increased 1.6-fold. Chlorate did not have an additive effect on LPL secretion when added with heparin. The Effect of Chlorate on LPL Synthesis-To evaluate the effect of chlorate on enzyme synthesis, cultures were pretreated in the presence and absence of 10 mM chlorate for 5 or 17 h at 37 "C, and then pulsed for 15 min with Tran35S-Label". Whether cells were pretreated for 5 or 17 h with chlorate, there was no statistical difference in the rate of incorporation of the label into control or chlorate-treated cells (Table IV).

Effect of Chlorate on LPL Turnover in Cultured
Adipocytes-Since chlorate did not significantly change LPL synthesis, the increase in enzyme secretion can be explained only by a decreased rate of enzyme degradation. Enzyme degradation was measured in three separate experiments by a pulsechase protocol (25). Cells were pretreated with and without 20 mM chlorate for 15 h and then pulsed with Tran3'S-Label@ for 1 h to label the intracellular pool. Cells were then chased for up to 1 h with medium containing excess unlabeled methionine.
[35S]LPL was measured following immunoadsorption, and LPL protein was measured by ELISA. Values for intracellular, cell surface-associated, and medium LPL radioactivity for one experiment are presented in Fig. 4. Loss of LPL radioactivity in the top panel is due to both degradation and appearance of labeled enzyme on the cell surface and medium. For the three experiments, chlorate treatment re-Time (minutes)

FIG. 4. Secretion of labeled LPL during a 60-min chase in the presence of control medium (closed circles) or medium containing 20 mm of chlorate (open circles).
Adipocytes were radiolabeled with Tran%-label@ (50 pCi/ml; 2 p~ L-methionine) for 60 min at 37 "C and then chased for the indicated time in label-free medium containing 300 p~ L-methionine. Radiolabeled LPL was isolated by immunoadsorption followed by electrophoresis and fluorography. The data shown are: top, intracellular; center, cell surfaceassociated; and bottom, medium 35S-labeled LPL. Each value was derived from a single pool of 4-100-mm dishes. Shown are data representative of three separate experiments. duced the radiolabeled cell surface associated LPL to 17-46% of control levels at the time points measured (Fig. 4, center  panel). After 60 min of chase, radiolabeled LPL in the medium (Fig. 4, bottom panel) was 2.1-fold higher in chlorate-treated cells than in control cells. Cell surface-associated LPL levels decreased 4.5-fold with chlorate treatment. Values for one experiment were 7.3 f 1.0 ng/lOO-mm dish for control cells, and 1.6 f 0.7 ng/lOO-mm dish for chlorate-treated cells ( p < 0.0001). In the other experiments, cell surface-associated LPL levels were undetectable in the chlorate-treated cells.
Intracellular LPL radioactivities in both the control and chlorate-treated cultures decreased by an apparent first order process (Fig. 4, toppanel). The first order rate constants were 0.0213 f 0.005 min" ( r = 0.98 f 0.02) and 0.0279 f 0.003 min-' ( r = 0.99 f 0.01), respectively, for the control and chlorate-treated cultures. These first order rate constants were not statistically different. Intracellular LPL remained relatively constant during the chase. The coefficient of variation for intracellular LPL measured by ELISA varied between experiments from 5 to 8.6% and from 5 and 9.9% for control and chlorate treated cultures, respectively. Synthesis rates were calculated as described previously by Cupp et al.
(3) as the product of the average intracellular pool size over 60 min, and the first order rate constant of the decrease in intracellular radioactivity (Fig. 4, top panel). Table V shows that the synthesis rate in control cells, 9.0 f 1.3 ng/h, is similar to the rate in chlorate-treated cells, 10.0 f 2.1 ng/h. These results confirm the conclusions drawn from the synthesis experiments shown in Table IV.
The degradation rate constant during the chase can be estimated as follows by the method described by Cisar et al. (25).
where A. and Aso are the total radioactive LPL at t = 0 and t

Effects of
Chlorate on Sulfation in Adipocytes in cultured adipocytes LPL synthesis rate was calculated as the product of the first order rate constant of the decrease in intracellular LPL radioactivity (Fig. 4,top) and the average intracellular LPL concentration. Degradation rates were obtained by multiplying the measured degradation rate constant by the average intracellular pool of LPL for a given set of dishes. The rates of LPL release to the medium were calculated by subtracting degradation rates from the synthesis rates. = 60 min, respectively, and C is the function describing the intracellular LPL radioactivity. The derivation assumed 1) a three-pool model consisting of intracellular, cell surface, and medium pools; 2) first-order processes directing the movement of LPL between these compartments; 3) irreversible degradation in the intracellular pool. In this equation, the integral in the denominator can be calculated by assuming that the loss of radioactivity from the cells is a first-order process or can be measured graphically. The degradation rate constants were 0.01651 f 0.006 and 0.01192 f 0.004 min", for control and chlorate-treated cultures, respectively. These two rate constants are not significantly different. The degradation rates were calculated as the product of the rate constant and the intracellular pool size, as measured by ELISA. Rates of release of LPL into the medium were estimated by difference. Table V, the amount of LPL degraded in 1 h is 42 f 9% of the amount synthesized in 1 h in chlorate-treated cells, compared to 77 f 11% in control cells ( p = 0.026).

Effect of Chlorate on the Incorporation of 35S04 into LPL-
In a recent report, we have shown that LPL is a sulfated glycoprotein (26). Since chlorate inhibits the formation of 3'phosphoadenylylphosphosulfate (PAPS), the ubiquitous activated sulfate donor, it was possible that chlorate reduced the sulfation of LPL. Indeed in cells pulsed with Na2["S04] for 12 h in sulfate-free media in the presence of chlorate, a complete inhibition of LPL sulfation is observed. Treatment of cultures with 10 mM chlorate inhibited the incorporation of 35S04 into cellular and medium LPL by 99 and loo%, respectively (Table VI). This decrease in sulfation was not accounted for by decreased LPL mass, since LPL protein decreased by only 18% in the cells, and increased 2.5-fold in the media. Effect of Sulfation State of LPL on Binding to the Adipocyte Cell Surface-The interaction of LPL with HSPG is considered to be ionic (27). Since chlorate affects the sulfation of both HSPG and LPL binding experiments were designed which would dissociate the individual roles of the sulfation of LPL and HSPG in their interaction. Adipocytes were pretreated with and without chlorate for 15 h, and then pulsed with Tran35S-Label@ for 90 min to obtain radiolabeled LPL in unsulfated and sulfated states, respectively. The cell extracts were prepared as described under "Experimental Procedures" and used in 4 "C equilibrium binding experiments.
Labeled cell extracts or purified LPL were added in increasing into lipoprotein lipase Adipocytes were pretreated with control media (30 dishes) or media containing 10 mM chlorate (30 dishes) for 15 h. Cells were then incubated with 0.15 mCi/ml Na2["SOo,] sulfate-free media, with and without chlorate, for 12 h. The radiolabeled LPL from cells and media was immunoadsorbed and separated by SDS-PAGE. After fluorography, bands representing LPL were excised from the gel, digested, and the radioactivity determined by LSC. LPL protein was determined bv ELISA in aliauots of cell extracts and media. Concentration-dependent binding studies of sulfated and unsulfated LPL were performed using cell extracts which had been metabolically labeled with Tran35S-Label" in the absence (sulfated) and presence (unsulfated) of 20 mM sodium chlorate. Control or chlorate-treated cells were incubated with the labeled cell extracts for 2 h at 4 "C. The media were collected, and heparin washes were performed to release the bound LPL. The amount of free and bound LPL was determined after isolation of LPL by immunoadsorption, SDS-PAGE, and fluorography (see "Experimental Procedures"). Shown are binding studies performed with total cell extracts (A), and partially purified cell extracts ( B ) .
amounts. The [35S]LPL bound to the cell surface or in the media at equilibrium was measured following immunoadsorption. The maximum concentration of free LPL in the medium was 20 ng/ml. Previous binding experiments with iodinated LPL showed that the nonspecific binding component in this range is negligible. Fig. 5 shows the results of experiments performed with total cell extracts ( A ) , or partially purified

Effects of
Chlorate on Sulfation in Adipocytes 16569 LPL ( B ) . There was no difference in the binding of sulfated uersus unsulfated LPL to the adipocyte cell surface. However, the binding of both sulfated and unsulfated 35S-labeled LPL to chlorate-treated cells was less than the binding to control cells. This is consistent with the results of the binding studies performed with '9-LPL. Effect of Chlorate on the Lipolytic Activity of LPL in Cultured Adipocytes-Since chlorate treatment of cultured adipocytes provides a convenient method of producing unsulfated LPL it was of interest to determine if the sulfate moiety of LPL is important for catalytic activity. The specific activity of lipoprotein lipase in control and chlorate-treated cells was calculated by dividing the triacylglycerol hydrolase activity by the enzyme mass for four pools of four dishes. The specific activity of lipoprotein lipase was 9.77 f 1.18 and 12.09 f 1.54 nmol of free fatty acid released/ng LPL/h for control and chlorate-treated cells, respectively. These differences were not considered statistically significant ( p = 0.062).

DISCUSSION
The results of this study show that the sulfation of the cell surface HSPG plays an important role in the binding, secretion, and degradation of LPL in cultured adipocytes. Chlorate treatment of adipocytes was used to inhibit the formation of PAPS, the activated form of sulfate. Treatment of adipocytes with chlorate resulted in reduced incorporation of 35S04 into HSPG and LPL, reduced binding of lZ5I-LPL to the adipocyte cell surface, increased LPL secretion, and reduced LPL degradation.
These data support the following hypothesis proposed by . In control cells, LPL is transported to the cell surface where it binds to high affinity sites on HSPG. If the enzyme is not released from the cell surface HSPG, it is internalized as a LPL-HSPG complex and is then either degraded or recycled to the surface. The net release of LPL from the surface would be determined by the association constant of LPL for the HSPG-binding site, the number of HSPG-binding sites, the residence time of HSPG. LPL complex on the surface, and the presence of soluble molecules which bind LPL and compete with the binding of LPL to HSPG. In the absence of competing molecules, 76% of the synthesized enzyme is degraded; the addition of heparin reduced the degradation rate to 21% of the synthetic rate (3). Treatment with heparinase and heparitinase, endoglycosidases which specifically cleave HSPG and therefore reduce the number of HSPG-binding sites, resulted in reduced LPL binding and degradation (6). Alternatively, it remains possible that sulfation of a cellular protein other than LPL and HSPG is inhibited by chlorate and that this factor contributes to the reduced LPL degradation observed in chlorate-treated adipocytes.
According to the current model for HSPG biosynthesis, unsulfated glycosaminoglycans are first elongated on the protein core and are then subsequently modified and sulfated (28). Chlorate has been used to inhibit the sulfation of proteoglycans by several groups. In bovine endothelial proteoglycans (29), 30 mM chlorate treatment reduced the sulfation to 35% of control levels, as measured by ratios of "SOo4 to [3H] glucosamine incorporation into HSPG. Another study utilizing human skin fibroblasts (8) showed that chlorate reduced the sulfation of proteoglycans but did not affect the degree of chain polymerization. This was measured by the effect of chlorate on [35S04]/[3H]gl~~osamine ratios, anion-exchange chromatography, and measurement of the molecular sizes of ['35S]proteoglycans. These studies show that the effect of chlorate on HSPG biosynthesis is not to inhibit chain elon-gation, but to specifically inhibit sulfation. Consistent with the above observation, our data with cultured adipocytes show also that chlorate has no effect on the uronic acid content of the trypsin-releasable HSPG. However, in our study where the distribution of heparan sulfate chains was measured, chlorate treatment increased the median size of heparan sulfate chains from 61,790 f 1,710 in control cells to 90,100 f 260 daltons in the chlorate-treated cells.
Chlorate treatment decreased the binding capacity of adipocytes for lZ5I-LPL. However, the affinity constant was not significantly different. Considering that the size of heparan sulfate chains and their content of uronic acid were not reduced, it is conceivable that for a given binding domain on a heparan sulfate chain, inhibition of sulfation at a specific site would either have no significant effect, or would inhibit LPL binding quantitatively. An example of the necessity of the sulfation of a specific saccharide sequence for the binding of a protein to a proteoglycan is known. The heparin-binding region of antithrombin is a distinct pentasaccharide sequence with a 3-O-sulfate residue on the internal glucosaminyl unit. Heparin fractions which lacked the 3-O-glucosaminyl groups had low affinity for antithrombin (30). A similar model that is consistent with the present data is that LPL binds to a specific sequence of sulfated saccharides of HSPG, and the abolition of specific sulfation sites within this sequence renders it unsuitable as a high affinity binding site for LPL.
It is interesting to note that the sulfate moiety of LPL did not contribute to the binding of LPL to the adipocyte cell surface. The most proximal N-acetylglucosamine (GlcNAc) residue of the complex oligosaccharide conjugated at Asn-45 contains the sulfate moiety of LPL (26). Consensus sequences for heparin-binding domains have been postulated by comparisons of known heparin-binding domains of apolipoproteins B-100 and E (31). The consensus sequences for these domains are B1-B2-B3-X-X-B4 and B1-B2-X-B3, where B represents a basic and X a hydropathic amino acid. In chicken LPL, these sequences occur at amino acids 148-153, and 281-284 (32). It was of interest to consider the effect of sulfation of LPL on its binding to cell surface HSPG since in the native LPL the sulfated GlcNAc conjugated to Asn-45 may be in close proximity to the postulated HSPG-binding regions.
Chlorate is an effective inhibitor of sulfation and is a useful tool to investigate the role of sulfate moieties in macromolecules. However, chlorate treatment of cells has no physiological relevance. Since sulfation may play a significant role in the binding properties of heparan sulfate chains, it will be important to identify physiological modulators of sulfation. There is very limited information on the effect of hormones on glycosaminoglycan sulfation. Shishiba and co-workers (33) reported that thyroid-stimulating hormone (10 milliunits/ml) increased the incorporation of [3H]glucosamine and "SO4 into HSPG of the cell layer of rat thyroid cells by 7 -and 3-fold, respectively. This would result in an alteration of the sulfation state of HSPG. The same report also noted that thyroidstimulating hormone enhanced the degradation of cell-associated proteoglycans. Perhaps other hormones which are relevant to adipocytes will be found to influence the sulfation of HSPG.
The present study has shown that chlorate treatment decreases the sulfation of both LPL and the adipocyte HSPG. The decreased sulfation of HSPG reduces the number of binding sites for LPL on the adipocyte cell surface. Consistent with the model that binding of LPL to cell surface HSPG is necessary for degradation of LPL to occur, LPL degradation is reduced in chlorate-treated cultures. The decreased sulfation of LPL does not affect the affinity of LPL for cell surface

Effects of Chlorate on Sulfation in Adlpocytes
HSPG, nor does the decreased sulfation affect the specific 16. Blumenkrantz, N., and Asboe-Hansen, G.