Selective Effects of Glucocerebroside (Gaucher's Storage Material) on Macrophage Cultures

Igal Gery, … , Roscoe O. Brady, John A. Barranger

J Clin Invest. 1981;68(5):1182-1189. https://doi.org/10.1172/JCI110363.

Research Article

Although the enzymatic lesion in Gaucher's disease is well established, little is known concerning the pathogenic mechanisms involved in the clinical manifestations of the disease. In order to obtain insight into this unexplored aspect of Gaucher's disease, we examined the effects of glucocerebroside (GL1) at the cellular level in monolayers of cultured murine macrophages. The addition of GL1 to these cultures stimulated the macrophages to release increased amounts of lymphocyte-activating factor (LAF) and lysosomal enzymes into the medium. These responses were proportional to the amount of GL1 added to the culture. At higher levels of GL1 (≥20 μg/ml), lactic dehydrogenase, a cytoplasmic enzyme was also released indicating cellular damage at these doses. Intracellular LAF also increased in macrophages incubated with the high doses of GL1, demonstrating an increase in total LAF production by these cells. Lipopolysaccharide acted synergistically with GL1 and stimulated the release of exceedingly high levels of LAF which had a molecular weight profile similar to that of LAF released by exposure to lipopolysaccharide alone. Unlike GL1, galactocerebroside, , and ceramidetrihexoside, exerted little or no effect on the release of macrophage products. The effect of GL1 was selective for macrophages since addition of this material to mouse lens epithelial cells had no detectable cytotoxic effect and it was only slightly toxic to lymphocytes or P815 cells in concentrations […]

Find the latest version: https://jci.me/110363/pdf Selective Effects of Glucocerebroside (Gaucher's Storage Material) on Macrophage Cultures

IGAL GERY, J. SAMUEL ZIGLER, JR., ROSCOE 0. BRADY, and JOHN A. BARRANGER, Laboratory of Vision Research, National Eye Institute and Development and Metabolic Neurology Branch, National Institute of Neurological and Communicative Disorders and Stroke, National Institutes of Health, Bethesda, Maryland 20205

A B S T R A C T Although the enzymatic lesion in lated large amounts of GL1 but not sphingomyelin, Gaucher's disease is well established, little is known whereas the other cells examined in this investigation concerning the pathogenic mechanisms involved in the did not accumulate either of these . Human clinical manifestations of the disease. In order to ob- monocytes, like murine macrophages, also release tain insight into this unexplored aspect of Gaucher's increased amounts of LAF when incubated with GL1. disease, we examined the effects of glucocerebroside The effect of GL, was dose-responsive and synergy (GL1) at the cellular level in monolayers of cultured was found with lipopolysaccharide. The relevance of murine macrophages. The addition of GL1 to these cul- these findings to the pathogenesis of Gaucher's tures stimulated the macrophages to release increased disease is considered. amounts of lymphocyte-activating factor (LAF) and lysosomal enzymes into the medium. These responses INTRODUCTION were proportional to the amount of GL1 added to the culture. At higher levels of GL1 (.20 ,ug/ml), lactic Gaucher's disease, an autosomal recessive dehydrogenase, a cytoplasmic enzyme was also re- storage disorder, is characterized by the deficiency leased indicating cellular damage at these doses. of and the accumulation of its sub- Intracellular LAF also increased in macrophages strate, glucocerebroside (GL1)1 in phagocytic cells incubated with the high doses of GL1, demonstrating ("Gaucher cells") throughout the reticuloendothelial an increase in total LAF production by these cells. system. The disease is multifaceted and its pathogene- Lipopolysaccharide acted synergistically with GL1 and sis is not completely understood (1, 2). stimulated the release of exceedingly high levels of Very little is known about the effects of GL1 on LAF which had a molecular weight profile similar to macrophages (MO) in culture and to the best of our that of LAF released by exposure to lipopolysac- knowledge, no study on this topic has been reported charide alone. Unlike GL1, galactocerebroside, sphin- previously. The present study was aimed at examining gomyelin, and ceramidetrihexoside, exerted little or no the effects of GL1 on MO cultures. The parameters effect on the release of macrophage products. The tested included the release of lysosomal and cyto- effect of GL, was selective for macrophages since plasmic enzymes, as well as of a monokine, lympho- addition of this material to mouse lens epithelial cells cyte-activating factor (LAF (3-6), also designated had no detectable cytotoxic effect and it was only Interleukin 1 (7)). The specificity of the responses of slightly toxic to lymphocytes or P815 cells in con- MO to GL1 was established by examining the effects centrations at which macrophages were clearly af- of other clinically relevant , namely, fected. A direct relationship was observed between 1Abbreviations used in this paper: CTH, tri- the cytotoxicity of the sphingolipids and their ac- hexoside; FCS, fetal calf serum; Gal-cer, galactocerebroside; cumulation in various cells. Macrophages accumu- GL1, glucocerebroside; LAF, lymphocyte-activating factor; LDH, lactate dehydrogenase; LPS, lipopolysaccharide; MO, macrophages; NIH, National Institutes of Health; PBS, Received for publication 3 July 1980 and in revised form Dulbecco's phosphate-buffered saline; ROS, rod outer seg- 26 June 1981. ments; SPM, sphingomyelin.

1182 The Journal of Clinical Investigation Volume 68 November 1981 *1182 -1189 galactocerebroside (Gal-cer), sphingomyelin (SPM) Chemical Company, Hamden, Conn.) and used immediately and ceramidetrihexoside (CTH) which accumulate in after removal of the nonadherent cells. Agents to be tested were added in 1 ml of M-199 medium (for Mqb) or RPMI 1640 , Niemann-Pick disease, and Fabry medium (for human monocytes), with 5% FCS, and the cul- disease, respectively. tures were further incubated for 20 h. After incubation, the supemates were collected and centrifuged at 900 g for 15 min' Monolayers assayed for intracellular LAF activity were METHODS washed three times with Hanks' balanced salt solution (NIH Reagents. GL1, SPM, and CTH, were purchased from media unit), covered with 1 ml of M-199 or RPMI 1640 medium Supelco, Inc. Bellefonte, Pa. Gal-cer, purified from sheep (for MO or monocytes, respectively) with 5% FCS, and frozen brain, was a gift from Dr. A. Gal (National Institutes of at -20°C. After thawing, the cells were scraped off by Health, Bethesda, Md.). The sphingolipids were suspended rubber policeman, sonicated and filtered (0.45 ,um). Mono- in water by sonication and sterilized by irradiation with layers tested for intracellular enzyme activities were washed 10,000 rad. GL, was examined for purity by thin-layer chroma- three times with PBS, lysed with 1.0 ml of 0.1% Triton X-100 tography; the preparation used here gave a single spot and removed by scraping. on silica gel G plates in a solvent system of chloroform: Chromatography on sephadex columns. Supernates from methanol:water (60:35:4). In addition, high-performance Mq cultures were concentrated (x8) by Amicon UM-10 mem- liquid chromatography of the perbenzoylated branes (Amicon Corp. Scientific Sys. Div.) and fractionated by the method of Ullman and McCluer (8) gave a single on G-75 superfine Sephadex columns (2.6 x 35 cm), equi- peak on a Ultrasphere-Si column (Altex Scientific, Inc., librated with RPMI 1640 medium with HEPES (Gibco Berkeley, Calif.). Soybean intralipid (SB-IL) was sup- Laboratories). Elution, with the same medium, was carried plied by Cutter Medical (Berkeley, Calif.) in a concentra- out by gravity and the fractions, 3 ml each, were supple- tion of 10% and was further diluted in the tested medium. mented with FCS (5%) before being filtered (0.45 j,m). Quartz silica particles, <5 Am, a gift from Dr. P. Davies (Merck Bovine L- and y-lens crystallins were used as molecular Institute for Therapeutic Research, Rahway, N. J.), were weight markers. suspended in Dulbecco's phosphate-buffered saline (PBS, LAF assay. The original method (3) was modified for NIH Media Unit); lipopolysaccharide W from Salmonella microcultures as follows. Samples were diluted in RPMI typhimurium (LPS, Difco Laboratories, Detroit, Mich.), 1640 medium (Gibco Laboratories) with 5% FCS and added phytohemagglutinin P (Difco Laboratories), concanavalin A in triplicate to flat bottomed microplates (Linbro, 8 x 6 mm), and lysophosphatidyl choline (LPC, or lysolecithin, from in aliquots of 0.1 ml. Thymocytes were added in aliquots egg, Sigma Chemical Co., St. Louis, Mo.) were dissolved of 1.5 x 106 cells in 0.1 ml shortly after being mixed with PHA in PBS. at a final dilution of the commercial product of 0.5 Il/ml. Preparations of cell cultures. Cultures were incubated After an incubation for 66 h the cultures were pulsed for 6 h in a humidified atmosphere at 37°C with 5% CO2 in air. All with [3H]thymidine (New England Nuclear, Boston, Mass., media contained 100 U penicillin and 100 ,ug streptomycin 2 Ci/mM, 0.5 ,uCi/well) and harvested by a MASH II (Micro- per ml. (a) Peritoneal Mo were collected by lavage from biological Associates, Walkersville, Md.). The radioactivity Swiss outbred mice (6-8 wk of age, NIH breeding facilities), in the collecting filter discs was measured by an LKB 1216 washed once and resuspended in M-199 medium (Gibco Rackbeta scintillation counter. The levels of LAF activity in Laboratories, Grand Island Biological Co., Grand Island, the tested MO preparations are expressed in two ways: (a) the N. Y.) supplemented with 10% fetal calf serum (FCS, Gibco mean counts per minute values+standard error of the mean Laboratories). Cultures were set up with 3-4 x 106 cells/ml, of [3H]thymidine uptake in thymocyte cultures stimulated by in two types of dishes, according to the experimental design one dilution of the tested preparation and (b) "LAF units". (see below). After an incubation for 2 h, nonadherent cells 1 unit was defined as the LAF activity found in supernate were removed by thotough washing and attached Mo were of the control M4 culture to which no additions were made; used for testing. (b) Human monocytes were isolated from the unit values in tested preparations were calculated by the the mononuclear fraction of peripheral leukocytes of normal dilution needed to bring the LAF activity to the 1 unit level. donors. Mononuclear cells, fractionated on Isolymph gradi- The use of LAF units is introduced here in order to simplify ents (Gallard-Schlesinger Chemicides, Inc., Carle Place, the data presentation and to accurately illustrate the ex- N. Y.) were cultured at 5-7 x 106 cells/ml in RPMI 1640 ceptionally high LAF activities in certain preparations ob- medium (Gibco Laboratories), with 10% FCS. Following an tained in this study; the dose-response curves of LAF incubation for - 1 h, the nonadherent cells were removed by activity of these preparations gave a high plateau at a wide thorough washing and the remaining monocytes were tested range of dilutions and all dilutions (1:10-1:640) were taken as described below. (c) Murine lens epithelial cells were into account when LAF units were calculated. Control cul- kindly provided by Dr. Paul Russell (National Eye In- tures, incubated with the medium alone, incorporated 150- stitute). These cells were originally isolated (9) and used 400 cpm. The sphingolipids tested here had no apparent in this study after being periodically subcultured for -4 yr. effect on the thymocyte cultures. (d) Normal human skin fibroblasts were supplied by Human Enzyme assays. l3-N-Acetylglucosaminidase (hexosamini- Genetic Mutant Cell Repository (Camden, N. J.) and sub- dase) was determined by the method of Pugh et al. (10). The cultured 3-5 times before being used in this study. (e) assays were carried out in duplicate, with the means dif- Mouse thymus or spleen lymphocytes were obtained from fering from the actual values by <15%. Lactate dehydroge- CBA/N mice (6-9 and 8-12 wk of age, respectively, NIH nase (LDH) was assayed by the method of Amador et al (11). breeding facilities) by gentle teasing and used after washing The levels of enzymatic activities in the supernates are once. (f) Cells of the P815 cell line (murine mastocytoma) presented as percent of the total activity in the culture were kindly provided by Dr. G. M. Shearer (National Cancer (supemate + cell lysate). Institute) and subcultured in suspension once a week until Inhibition of nucleic acid synthesis. This assay was being used. carried out in flat bottomed microplates (6 x 8 mm, Linbro), Content and release of cell products. Monolayer cell with the cultures set up in triplicate, in a final volume of 0.2 cultures were set up in multiwell plates, 17 x 24 mm (Linbro ml of RPMI 1640 medium with 5% FCS. The cultures

Effects of Glucocerebroside on Macrophage Cultures 1183 consisted of (a) 4 x 104 P815 cells, (b) murine lens epithelial 24 p cells, cultured 20 h before being used, at 2.5 x 104 cells/ 60 well, or (c) 3 x 105 mouse spleen cells, collected shortly 20 before use and stimulated with concanavalin A (2 ,ug/ml). x / After addition of the tested compounds, the cultures were E 16 incubated for 24 or 48 h and pulsed with [3H]uridine (New 40 0 0 England Nuclear, 8 Ci/mmol) or [3H]thymidine, respectively, 0 12 0.5 ,uCi/well, for the last 6 h of incubation. All cultures were harvested by MASH II, with the lens epithelial cells being 8 20 treated with trypsin beforehand (12). - 2 0 Accumulation of labeled particles.1 Monolayer cultures 0 4 0 0./I-I 0 X were set up in duplicate in plastic petri dishes, 35 mm in U I,, m Diam (Falcon Labware, Div. of Becton, Dickinson & Co., 0" 7.5 15 30 60 20 £c Oxnard, Calif.) and consisted of freshly prepared peritoneal 4- MO and confluent monolayers of lens epithelial cells and skin SL 48 - fibroblasts cultured for 48 h in advance. After removal of the original culture media, the labeled particles were added, sus- I pended in 2-ml aliquots of RPMI 1640 medium with HEPES, 40H supplemented with 5% FCS. The labeled particles included 60 1 32H [3H]GL1 (New England Nuclear, 8 Ci/mmol), [14C]SPM I w I (New England Nuclear, 45 mCi/mmol) or 3H-labeled bovine 0 / z retinal rod outer segments (ROS). [3H]GL, was prepared 24 40 from the same unlabeled GL, used in this study and co- _C, /~~~~~~~~~~~~I,, chromatographed with it on thin layer silica gel G. Labeled 16 a" . ROS were prepared as described (13) and kindly provided by , ---~-- - x ' t Dr. Paul O'Brien (National Eye Institute). Unlabeled >1 20 sphingolipids, at 30 .g/ml, were added to the suspensions 8 .21 of the labeled GL1 and SPM. After incubation for various intervals, at 37°C or 4°C, the monolayers were washed 2;6! -- I 1 thoroughly with PBS and lysed in 0.5 ml NaOH, 0.2 N. The 0 6 12 24 48 96 20 radioactivity was measured by scintillation counting in GL1 LPS aliquots of 0.1 ml, treated with 0.5 ml of NCS tissue solu- added agents (yg/ml) bilizer (Amersham Corp., Arlington Heights, Ill) and mixed with 10 ml of Betafluor (National Diagnostics Inc., FIGURE 1 Effects of GL1 at different concentrations on Advanced Applications Institute Inc., Parsippany, N. J.) release of LAF, hexosaminidase and LDH by murine The protein content of the monolayer lysates was deter- peritoneal MO cultures. The levels of LAF activity presented mined by the method of Lowry et al. (14). The results are here were obtained with the supernates diluted to 1:40. presented as both the specific uptake, (mean disintegrations per minute per milligram protein) and as the percent of the total added isotope. The levels of specific uptake in the moderately increased the release of LAF and hexos- individual dishes differed by <10% from the mean values. aminidase, but had no effect on LDH release. GL1 also increased the release of two other lysosomal enzymes, RESULTS ,B-galactosidase and glucuronidase (data not shown), Effects of GL1 on the release of LAF and enzymes by but the changes were less consistent than those M+O. The results obtained in two experiments in observed with hexosaminidase. Silica particles, a which the release of various cell products was meas- known cytotoxic agent for MO (17), were used as control ured after incubation of MO monolayers with various in other experiments to examine the effect of damage on concentrations of GL1 are illustrated in Fig. 1. The the MO product release. The effects of silica resembled addition of GL1 caused an increase in the release of the those of GL1, namely, at subtoxic concentrations (<25 monokine, LAF, the lysosomal enzyme, hexosamini- Aglml), only LAF release was increased, whereas at dase, and the cytoplasmic enzyme, LDH, in a dose- toxic higher doses, larger amounts of LAF were dependent fashion. The increase in release of LDH released, along with an increased release of LDH and indicatps cell damage (15, 16). It is noteworthy that the hexosaminidase. release of LAF and hexosaminidase was particularly Synergy between GL1 and LPS. MO monolayers elevated at toxic doses of GL1, although an apparent incubated with combinations of GL1 and LPS released increase in LAF release was observed in cultures with LAF in amounts greatly exceeding those secreted by no detectable elevation of LDH release (with GL1 at cultures stimulated by the individual agents (Table I). -20 ,ug/ml or less). A known M4 stimulant, LPS, It is noteworthy that the release of substantial amounts of LAF in the presence of LPS and GL1 at 10 ,ug/ml was not 2 The term "accumulation" was used in the present study accompanied by an increase in the release of LDH; to describe the process in which labeled particles become the combination of LPS and GL1 at 30 ,ug/ml produced associated with cell monolayers. This process may include very high levels of LAF, along with elevated LDH both binding and actual phagocytosis or pinocytosis. release. Hexosaminidase secretion in these experi-

1184 I. Gery,J. S. Zigler, Jr., R. 0. Brady, andJ. A. Barranger TABLE I Synergy between GL1 and LPS*

MO products in supernates

_s _ LAF activity _ ._~ Enzymesf Agent added to [3H]Thymidinet LAF§ MO cultures uptake units Hexosaminidase LDH 10-s /g/ml -0' None 394+81 1 25.3 20.1 GL,, 10 425±+106 1 29.7 22.4 GLj, 30 426±37 2 37.5 36.5 LPS, 20 2,670+338 5 20.0 18.2 LPS + GLj, 10 5,171+447 10 26.2 18.4 LPS + GLb, 30 60 47.6 46.7 E0-4 F_ \LPS+GL1I 33,827+1,837 X 10t (1:10) * The data recorded here are of one experiment, the findings were reproduced in three other experiments performed on different dates. t Data represent counts per minute±standard error. § LAF activity is presented as mean cpm of [3H]thymidine incorporated by thymocyte cultures, in triplicate, incubated with the tested supernates at a final dilution of 1:160. The definition of "LAF units" is detailed in Methods. 10123 ¶ Levels of enzymatic activities in supernates, as percent of total activities in culture (supernate + cell lysate). 02 I \ / ments paralleled the release of LDH and reached 10~ 15 2 5 30 3 0 4 particularly high levels in cultures with LPS and GL1 at 30 ,ug/ml. Chromatographic analysis of LAF released by GL1 + LPS. The possibility that the high levels of LAF activity released by the combination of GL1 plus FRACTION NUMBER LPS are due to a mediator other than LAF was FIGURE 2 Chromatographic profiles of LAF activities re- examined by testing the chromatographic profiles ofthe leased by murine M45 cultured with LPS alone (at 20 gtg/ml) mediator activity released by these cultures; the or with LPS + GL1 (at 20 and 60 ,ug/ml, respectively). The LAF fractionation was carried out on Sephadex G-75 superfine chromatographic properties of have been well columns and the fractions were tested at the recorded defined in previous studies (5, 18). Fig. 2 shows the dilutions. chromatographic comparison between the activities released by MO incubated with LPS alone or with the combination of LPS plus GL1. The results demonstrate LAF and moderately elevated the level of intracellular a close similarity between the elution profiles of the LAF activity. LPS differed from GL1 by remarkably LAF activity from the two types of MO cultures, despite increasing the intracellular LAF activity, while having the substantial difference in their levels of activity. It only a small effect on LAF release. The combination of seems, therefore, that the synergistic effect of GL1 plus GL1 and LPS again showed synergy in LAF release, LPS brings about the release of extremely large along with the stimulation of high intracellular activity amounts of the same mediator secreted by cultures (which surpassed the elevated activity found in stimulated with LPS alone. cultures with LPS alone). It is apparent, therefore, that Effects of GL1 on intracellular LAF. In view of the the increase in LAF release by MO cultured with GL1 is cytotoxic effect of high doses of GL1 on MO, intra- not simply the result of depletion of the intracellular cellular LAF activity was also measured, to test the LAF pool. Rather, incubation with GL1 brings about an possibility that the increased release of LAF by GL1 increase of both intracellular and extracellular ac- was due only to leakage of the intracellular pool of the tivities. mediator. The data of a typical experiment are Effects of other sphingolipids on Mo. Three other presented in Table II; similar results were repeatedly sphingolipids involved in storage lipid diseases were obtained in other experiments. GL, at 80 ,ug/ml (a toxic tested, CTH, SPM and Gal-cer. As shown in Table III, dose, see Fig. 1) substantially increased the release of these lipids caused minimal or no increase in LAF

Effects of Glucocerebroside on Macrophage Cultures 1185 TABLE II Cytotoxicity of GL1 to cells other than Mo. The Effects of GL, on Intracellular and Extracellular Levels possibility that GL, affects Mo simply as a nonspecific of LAF Activity of M+* cytotoxic agent was examined by testing its effects Cell lysates on cells other than Mo. The cytotoxic effects of GL, Supernates and other lipids were measured by their capacity to Agent added [3H]Thymidine LAF [3H]Thymidine LAF inhibit nucleic acid synthesis; this effect has been to M44 uptake§ units uptake§ units shown to be directly related to cytotoxicity (19). The None 381+94 1 1,245±204 3 data of two experiments are shown in Fig. 3. Lens GL, 7,001+950 16 5,703+665 10 epithelial cells were found refractory to GL1, whereas LPS 1,662+355 3 30,169±525 60 spleen and P815 cells were moderately inhibited by GLI + LPS 40,574+548 100 51,467±5,679 100 high concentrations of this . SPM and CTH affected cells other than Mo similarly to GL, (data * The data recorded here are of one experiment; the findings with CTH are not shown). All tested cells were highly were reproduced in three other experiments. inhibited by the surface active agent, lysophosphatidyl I GL, was added to the concentration of 80 ,ug/ml and LPS, to 20 ,ug/ml. choline (LPC) and moderately affected by soybean § The values of [3H]thymidine uptake recorded here were intralipid (SB-IL). obtained with the supernates and lysates added at a final Accumulation of GL, by Mo. The relationship be- dilution of 1:40, and are expressed as mean counts per tween the effect of the lipids and their interaction with minute±standard error. the cells was further analysed by measuring the ac- cumulation of [3H]GL,, ['4C]SPM and 3H-labeled ROS by monolayers of Mo and two other cell types, human secretion and, unlike GL,, were not synergistic with skin fibroblasts and murine lens epithelium. The LPS. At the doses tested, none of the lipids other than sphingolipids were added to the cultures in the GL, had a detectable effect on the release of hexos- presence of 30 ug/ml of the nonradioactive lipid. aminidase or LDH (data not shown). The data recorded in Table IV show a striking dif- ference between the M4 and the other tested cells TABLE III in their capacity to accumulate [3H]GL,. Mo cultures Release ofLAF by MO Cultured with Different Sphingolipids accumulated -10% of the added tracer during an in- cubation of 20 h, while the accumulation of radio- LAF activity < LPS activity by the other cells was 1%. On the other hand, Tested Concentra- added, [3H]Thymidine LAF Mo, like the other cells, did not accumulate a sig- lipid tion 20 lAg/ml uptake* units nificant amount of ['4C]SPM. The selective lack of JAg/mi None No 383± 103 1 A B Yes 4,525±528 5 GL, 40 No 5,654±553 5 100 - Lens P815 120 No 17,332±524 20 - epithelium 40 Yes 29,739±986 100 60 CTH 40 No 724±76 1 zI20 120 No 1,487+318 2 0 DM_ 40 Yes 4,068±134 6 ED

. r.0% SPM 40 No 651±144 1 z 10RJf1W Spleen Spleen 120 No 698±218 1 * 40 Yes 3,673±505 5 Gal-cert 60 No 467±50 1 120 No 515±98 1 60 Yes 2,346±524 4 US §8tS 8 a 9 !i B!R8 * The values of [3H]thymidine uptake recorded here were GL1 SPM LPC GL1 SB-IL LPC obtained with the supernates added at a final dilution of Tested agents qug/mll 1:40, and expressed as mean counts per minute±standard FIGURE 3 Inhibitory activity of GL, and other lipids on the error. nucleic acid metabolism ofvarious cell cultures. Incorporation t Gal-cer was tested in a separate experiment. Supernate of of [3H]thymidine was used in experiment A and of [3H]- LPS alone (at 1:40 dilution) stimulated in this experiment uridine in experiment B. % inhibition = 100 - mean cpm in an uptake of 1,760±230, or 3 LAF units. cultures with lipid/mean cpm in control cultures x 100.

1186 1. Gery, J. S. Zigler, Jr., R. 0. Brady, and J. A. Barranger TABLE IV Uptake of [3H]GL1 and Other Particles by MO, Fibroblasts, and Epithelial Cells Cell monolayers Particles Incubation Skin Lens Exp. tested time Temp. MX fibroblasts epithelium h °C I. 3H-GL, 20 37 700,402 (9.3)* 9,742 (0.1) 4,569 (0.15) 4 5,246 (0.1) 6,462 (0.03) 26,987 (0.8) [14C]SPM 20 37 21,673 (0.4) 6,539 (0.1) 5,194 (0.3) 4 4,796 (0.1) 8,449 (0.05) 2,678 (0.1) [3H]ROS 20 37 30,580 (7.6) 12,149 (1.4) 24,353 (9.2) II. [3H]GLI 3 37 201,196 (4.3) 2,861 (0.02) 4,472 (0.1) 4 3,947 (0.1) 1,046 (0.01) 7,255 (0.2) 20 37 569,390 (10.4) 7,360 (0.1) 5,054 (0.1) 4 4,402 (0.1) 7,266 (0.04) 19,714 (0.3) [3H]ROS 3 37 26,478 (5.8) 8,677 (0.7) 33,675(4.2) 4 2,542 (0.6) 307 (0.02) 2,216 (0.2) * Mean disintegrations per minute of labeled particles per milligram protein of cell lysate; in parentheses, the uptake calculated as percent of total activity added per dish. The labeled particles were added in aliquots containing (per dish): [3H]GL,, Exp. I, 7.5 x 105 dpm and Exp. II, 8.6 x 105; [14C]SPM, 5 x 105 dpm; [3H]ROS, Exp. I, 6.6 x 104 dpm and Exp. II, 8.2 x 104 dpm. accumulation of [3H]GL1 by the cells other than MO by GL1 similarly to the murine Mo. Although some contrasted sharply with their capacity to phagocytose donors did not yield GL1-responsive monocytes, those 3H-labeled ROS, which consist of discs of cell mem- who did, retained their responsiveness in repeated branes (13, 20). Lens epithelial cells phagocytosed tests for >2 yr. Results obtained with cells from a 3H-labeled ROS similarly to Mo. Skin fibroblasts, al- typical responsive donor are shown in Table V. GL1 though inferior to the other cells in their capacity to increased the release of LAF by the monocyte cultures, phagocytose ROS, nonetheless took up considerably with an apparent dose responsiveness, and synergized more ROS than the sphingolipids. with LPS to induce the release of very high levels The molecular form of the label accumulating in Mo of the mediator. SPM had no detectable effect on the incubated with [3H]CGL was examined and found to be release of LAF by these cultures. similar to that of the original [3H]GL1 added to the cultures. Chloroform/methanol extracts of the cell DISCUSSION lysates resulted in >90% recovery of activity. Further, >90% of the radioactivity in these extracts cochro- M4A cultures were found in this study to be useful in matographed with GL1 on thin layer silica gel G studying pathophysiologic activities of GL, at the plates. In addition, no change in molecular form was cellular level. Mo monolayers incubated with GL1 re- found in the label present in the media of cultures leased increased amounts of a specific mediator, LAF, of Mo or other tested cells. These data indicate, as well as the lysosomal enzyme, hexosaminidase. therefore, that no significant degradation of GL1 took In addition, GL, increased the release of a cyto- place under the conditions employed. plasmic enzyme, LDH, which is indicative of cell The accumulation of the particles by the cell mono- damage (15, 16). It is assumed, therefore, that the layers was reduced in most systems when carried out increased secretion of LAF and lysosomal enzymes at 4°C, thus indicating that active phagocytosis was is due at least in part to cellular injury. It should involved. An exception to this pattern was found in be pointed out, however, that increased LAF secretion cultures of lens epithelium incubated with [3H]GL,; was observed in Mo cultures incubated with GL1 in all repeated experiments the accumulation of this at doses that had no detectable effect on LDH release. lipid was markedly higher at 40C than at 370C. Moreover, substantial amounts of LAF were released Effects ofGL1 on human monocyte cultures. Mono- by the combination of low doses of GL1 and LPS, with layers of human blood monocytes were affected no elevation of LDH (Table I). It is not known at this

Effects of Glucocerebroside on Macrophage Cultures 1187 TABLE V cessive amounts of LAF by murine or human Mo. The Release of LAF by Human Monocytes Cultured combined stimulation provides, therefore, a new with GL, and Other Agents* method for obtaining high levels of LAF, a mediator LAF Activity in Supernate of various biological reactions (5, 6, 24, 25). The selectivity of the effect of GL1 on Mo was Agentadded [3H]Thymidine uptaket LAF units underlined by the finding that three other sphingo- AgIml lipids, Gal-cer, SPM and CTH had very little effect on these cells in terms of LAF release (Table III). This None 791 ± 178 1 selectivity was further demonstrated by studying the GL1, 13 2,554±545 4 cytotoxic activities of GL, on a variety of cells. Lens GL1, 40 5,056+533 10 epithelial cells were unaffected by GL1 at all tested GL1, 120 5,838+520 10 doses, whereas lymphocytes or P815 cells, which are more susceptible to cytotoxic agents, were minimally LPS§ 1,876+206 3 LPS + GL1, 13 11,950±1,855 20 affected by GL1 at concentrations that increased LPS + GL1, 40 20,604+1,958 40 markedly the release of LDH from Mo, i.e., 20-100 ,ug/ml. (The release of LDH is considered a reliable SPM, 120 494±183 <1 method for determining Mo injury (15, 16)). The * selectivity of GL1 for Mo may derive from the unique The data recorded here were obtained with monocytes of capacity of these cells to bind or take A one donor; similar results were obtained with cultures from up GL1. direct six other donors. relationship was found between the levels of accumu- t The [3H]thymidine values recorded here were obtained with lation and of cytotoxic activity of the sphingolipids as the supernates at a final dilution of 1:160, and expressed as shown by the following results: (a) lens epithelial mean counts per minute±standard error. cells accumulated negligible amounts of GL1 and were § LPS was added to the concentration of 25 ,ug/ml. refractory to the damaging effects of this lipid, whereas Mo accumulated large quantities of GL1 and showed time whether the toxicity of GL1 in culture has any high susceptibility to its effects. (b) SPM was not bearing on the pathogenesis of Gaucher's disease in accumulated by M4 and had minimal or no effect on vivo. Plasma of Gaucher's patients contains -10-15 these cells. The relationship between accumulation ,ug GLJ/ml. This is in the same range that produces and damaging capacity of GL1 was further confirmed toxicity in our cultures. Moreover, storage MO in experiments in which untreated ("resident") (Gaucher cells) may contain several hundred micro- Mo were compared to MO activated in vivo by BCG grams of GL1 per 106 cells. Biopsies of liver or bone bacteria. The activated Mo accumulated 3-10 times marrow from these patients often show evidence of cell less [3H]GL1 than the resident MO and were refractory damage (21). Whether this results from toxic products to its damaging effects.3 released from M4 (22) remains to be tested. Our data The selectivity of GL1 for Mo in vitro (Table IV) and the observation that Gaucher tissues contain ex- may be related to the exclusive accumulation of this cess lysosomal enzymes (23) support this possibility. sphingolipid in the reticuloendothelial system of The relationship between MO damage and produc- patients with Gaucher's disease (1, 2). The mechanism tion of LAF was further examined by measuring the that regulates the selective uptake of GL1 by Mo is not levels of intracellular LAF activity in the GL1-affected known. The possibility that cells other than M4 did not cultures. Lysates of these MO contained more LAF accumulate GL, because of high levels of the specific activity than lysates of untreated controls (Table II). enzyme, glucocerebrosidase (GL1ase) is not consistent It seems, therefore, that the increased release of LAF with the finding that the levels of this enzyme were by toxic doses of GL, is not due merely to a leakage similar in Mof and lens epithelial cells (114.7 and of the cellular content of this mediator. Rather, the 128.7 U/mg protein, respectively). It is known that results suggest that injured MO increase their produc- moieties have some affinity for the mannose tion of LAF and indeed, similar observations were receptor of M4 (26). The possibility that this or another made in experiments not included here, in which M4 receptor is involved in binding and uptake of glyco- cultures were incubated with silica, a selective cyto- lipids is currently under investigation. toxic agent for MO (17). It is also noteworthy that The present study thus shows that Mo cultures may M4b "activation" is often accompanied by cell damage be useful for investigating the interactions between and increased LDH release, as has been shown in cul- GL1 and the reticuloendothelial system. Although the tures incubated with high concentrations of antigen- antibody complexes or zymosan (16). 3 Gery, I., R. Prendergast, and J. A. Barranger. Production Of a particular interest is the finding that GL1 and release of lymphocyte-activating factor by macrophages. synergizes with LPS to induce the secretion of ex- Effects of macrophage activation in vivo. In preparation.

1188 I. Gery, J. S. Zigler, Jr., R. 0. Brady, and J. A. Barranger relationship between the phenomena observed in vitro Serum lactic dehydrogenase activity: an analytical and those occurring in vivo is not clear, it is tempting assessment of current assays. Clin. Chem. 9: 391-399. to are re- 12. BenEzra, D. 1977. A microculture technique for the speculate that the findings reported here evaluation of corneal cell metabolism in vitro. Invest. lated to certain clinical features of Gaucher's disease. Opthalmol. Visual Sci. 16: 893-898. It is conceivable that an increased release of products 13. Goldman, A. I., P. J. O'Brien, E. Masterson, P. Israel, by GL,-affected M4 could participate in bringing about P. Teirstein, and G. Chader. 1979. A quantitative system damage to liver and bone marrow (21), increased levels for studying phagocytosis in pigment epithelium tissue culture. Exp. Eye Res. 28: 455-467. of lysosomal enzymes (23), or the abnormally high 14. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. frequency of gammopathies (27) in Gaucher's disease. Randall. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193: 265-275. ACKNOWLEDGMENTS 15. Davies, P., R. C. Page, and A. C. Allison. 1974. Changes in cellular enzyme levels and extracellular release of We thank Drs. J. J. Oppenheim and F. Chu for stimulating lysosomal acid hydrolases in macrophages exposed to this study, Ms. J. Derr and Ms. N. Krett for excellent tech- group A streptococcal cell wall substance. J. Exp. Med. nical assistance, Dr. F. S. Furbish for preparing the glyco- 139: 1262-1282. lipids, Dr. P. O'Brien for the labeled ROS, Dr. P. Davies 16. Dean, R. T., W. Hylton, and A. C. Allison. 1979. Lysosomal for the silica particles, and Mrs. N. Beavers, Ms. J. McIntyre enzyme secretion by macrophages during intracellular and Mrs. A. Brown for typing the manuscript. storage of particles. Biochim. Biophys. Acta. 584: The generous support of National Lipid Disease Founda- 57-65. tion and the Mitchell Fund is also acknowledged. 17. Allison, A. C., J. S. Harington, and M. Birbeck. 1966. An examination of the cytotoxic effects of silica on macro- REFERENCES phages. J. Exp. Med. 124: 141-153. 18. Gery, I., and R. E. Handschumacher. 1974. Potentiation 1. Brady, R. 0. 1978. Glucosyl ceramide lipidosis: Gaucher's of the T-lymphocyte response to mitogens. III. Properties disease. In The Metabolic Basis of Inherited Disease. of the mediator(s) from adherent cells. Cell. Immunol. J. B. Stanbury, J. B. Wyngaarden, and D. S. Fredrickson, 11: 162-169. editors. McGraw-Hill, Book Co., Inc., New York. 4th 19. Eidinger, D., E. Bello, and A. Mates. 1977. The hetero- edition. 1: 731-746. cytotoxicity of human serum. I. Activation of the alterna- 2. Peters, S. P., R. E. Lee, and R. H. Glew. 1977. Gaucher's tive complement pathway by heterologous target cells. disease, a review. Medicine (Baltimore). 56: 425-442. Cell. Immunol. 29: 174-186. 3. Gery, I., R. K. Gershon, and B. H. Waksman. 1972. 20. Daeman, F. J. M. 1973. Vertebrate rod outer segment Potentiation of the T-lymphocyte response to mitogens. membranes. Biochim. Biophys. Acta. 300: 255-288. I. The responding cell. J. Exp. Med. 136: 128-142. 21. James, S., P. Stromeyer, F. W. Chang, and J. A. Barranger. 4. Gery, I., and B. H. Waksman. 1972. Potentiation of the 1981. Liver abnormalities in patients with Gaucher's T-lymphocyte response to mitogens. II. The cellular disease. Gastroenterology. 80: 126-133. source of potentiating mediator(s). J. Exp. Med. 136: 22. Ferluga, J., and A. C. Allison. 1978. Role of mononuclear 143-155. infiltrating cells in pathogenesis of hepatitis. Lancet II: 5. Unanue, E. R. 1978. The regulation of lymphocyte 510-611. functions by macrophages. Immunol. Rev. 40: 227-255. 23. Moffitt, K. D., J. P. Chambers, W. F. Diven, R. H. Glew, 6. Oppenheim, J. J., S. B. Mizel, and M. S. Meltzer. 1979. D. A. Wenger and D. F. Farrell. 1978. Characterization Biological effects of lymphocyte and macrophage- of lysosomal hydrolases that are elevated in Gaucher's derived mitogenic "amplification" factors. In Biology of disease. Arch. Biochem. Biophys. 190: 247-260 the Lymphokines. S. Cohen, E. Pick, and J. J. Oppen- 24. Smith, K. A., K. J. Gilbride, and M. F. Favata. 1980. heim, editors. Academic Press, Inc., New York. 291-323. Lymphocyte activating factor promotes T-cell growth 7. Aarden, L. A. et al. 1979. Revised nomenclature for factor production by cloned murine lymphoma cells. antigen-nonspecific T-cell proliferation and helper Nature (Lond.). 287: 853-855. factors. J. Immunol. 123: 2928-2929. 25. Rosenwasser, L., C. A. Dinarello, and A. S. Rosenthal. 8. Ullman, M. D., and R. H. McCluer. 1978. Quantitable 1979. Adherent cell function in murine T-lymphocyte microanalysis of perbenzoylated neutral glycosphingo- antigen recognition. IV. Enhancement of murine T-cell lipids by high-performance liquid chromatography with antigen recognition by human leukocytic pyrogen. deletion at 230 nm. J. Lipid Res. 19: 910-913. J. Exp. Med. 150: 709-714. 9. Russell, P., H. N. Fukui, Y. Tsunematsu, F. L. Huang, 26. Stahl, P., P. H. Schlesinger, E. Sigardson, J. S. Rodman, and J. H. Kinoshita. 1977. Tissue culture of lens epithelial and Y. C. Lee. 1980. Receptor-mediated pinocytosis of cells from normal and Nakano mice. Invest. Ophthalmol. mannose by macrophages: characteri- Visual Sci. 16: 243-246. zation and evidence for receptor recycling. Cell. 19: 10. Pugh, D., D. H. Leaback, and P. G. Walker. 1957. Studies 207-215. on glucosaminidase. N-acetyl-,3-glucosaminidase in rat 27. Pratt, P. W., S. Estren, and S. Kochwa. 1968. Immuno- kidney. Biochem. J. 65: 464-469. globulin abnormalities in Gaucher's disease. Report of 16 11. Amador, E., L. E. Dorfman, and W. E. C. Wacker. 1963. cases. Blood. 31: 633-640.

Effects of Glucocerebroside on Macrophage Cultures 1189