Copyright @ 1981 by Academic Press. Inc. All rights of reproduction in any form reserved 0014.4827/811020251-16$02.00/O
Cell Research 131 (1981) 251-266
ERIK HARMS,’ JURGEN KARTENBECK:
OF HUMAN LYMPHOBLASTS
DARAB and JERRY SCHNEIDER4
‘Universitiits-Kinderklinik; 2Abteilung Membranbiologie und Biochemie, Institut fiir Zell- und Tumorbiologie, Deutsches Krebsforschungszentrum; %stitutfiir Medizinische Virologie der Universitiit, D-6900 Heidelberg, Germany and 4Department of Pediatrics, University of California, San Diego, La Jolla, CA 92023, USA
SUMMARY A method was developed for the isolation of unmodified lysosomes of human origin using cultured EB-virus transformed lymphoblasts. The cells were lysed carefully by repeated resuspension in buffered isotonic sucrose. A crude granular fraction derived from this lysate was further purified by isopyknic centrifugation in an isotonic colloidal silica gel gradient and by free-flow electrophoresis. The following relative specific activities (mean + S.D.) of lysosomal marker enzymes were measured in a pooled lysosomal fraction obtained from the final electrophoresis step (representing less than 0.1% of the initial protein): /3-N-acetylglucosaminidase 85.6+ 15.5; /?-galactosidase 87.6k13.4; acid )V-glycerophosphatase 41.7f3.5; &ducuronidase 36.6k6.1. Wtth respect to the fmal two enzymes the recovery within this pooled fraction was 5d% of the initial lysate. The great differences in relative specific activities achievable may be due mainly to different extralysosomal portions of the lysosomal marker enzymes, as was found for acid &lycerophosphatase which was largely distributed within non-lysosomal structures in lymphoblasts when studied by histochemical staining. The final fraction consisted almost exclusively of lysosomes when examined by electron microscopy. Most lysosomes appeared club-shaped immediately after cell lysis and throughout the preparation procedure. Examination by electron microscopy and measurement of the latency of lysosomal enzyme activity revealed an exceptional integrity of the lysosomal membrane. This method provides the opportunity to study highly purified lysosomes from patients with lysosomal disorders.
Lysosomes are cell organelles which degrade macromolecules. A number of inherited metabolic diseases in man are characterized by the lack of specific lysosomal hydrolases and abnormal storage of undigested material within lysosomes [l]. However, other lysosomal disorders appear more complex and may involve abnormalities of transport across the lysosomal membrane  or of incorporation of hydrolases into the lysosome . Progress in studying these conditions has been hampered by the lack of methods to obtain pure human lysosomal preparations. 17-801810
In general, the defects in lysosomal storage diseases can be demonstrated in cultured cells derived from affected individuals. A variety of methods have been described for the isolation of lysosomes . More recently, the use of density gradients of metrizamide  and polyvinyl-pyrrolidon-coated silica gel (Percolle) has been described [6-8]. However, most methods have used rat liver as starting material. Relatively few attempts have been reported on the isolation of lysosomes from tissue culture cells [7-161. All these attempts have used known separation methods, but have Exp Cd Res /3/ (1981)
Harms et al. Lymphoblasts grown in RPNI 1640/20: fetal bovine
1,000 x g 10 mill GSA rotor I
I cell pellet resuspended isolation buffer without
1 cells lysed by three times resuspension in isolation buffer (0.25 I4 sucr~~se, 10 mM tsiethanolamine, 10 mM acetic acid, 1 mM EDTA, pH 7.40)
supesnatancs thrygh filter
(FRACTION B) paper
vol (FRAC+ION C)
mixed with Percoll 10 mM triethanolamik, 10 mI4 acetic' acid. 1 rH EDTA, pH 7.40 to final density of 1.078 gjml
bottom layer between densities and 1.073 g/ml (FRACTION D/l) 1 : 5 with isolation buffer
times in isolarion buffer 40,000 x g 10 min ss-34 rotor
I resuspended in two to four isolati.on uffer 'r
supernerant (FRACTION Dj2) separacod by free-flow electrophoresia in isolation buffer in a VaP 5 apparatus: buffer temp. 6'C, buffer flow 2 ml/fraction/hour, 112 - 122 V/cm. 185 - 190 mA
Fig. I. Scheme of isolation procedure of lysosomes from human EB-virus transformed lymphoblasts.
Res 131 (1981)
yielded incomplete separation of lysosomes from non-lysosomal cell components. Accordingly, the resulting fractions have not been sufficiently characterized. The aim of the present study was to develop methods for thorough purification of lysosomes from cultured human cells and to characterize the fractions obtained both enzymatically and morphologically. This has been achieved by a combination of differential centrifugation, isopyknic centrifugation and free-flow electrophoresis, three techniques based upon different physical properties of the organelles. MATERIALS
Virus EBV producer cell line B95-8 [17, 181 was kindly provided by Dr F. Deinhardt, Munich, as a cell suspension and grown using RPMI-1640 supplemented with 10% fetal bovine serum at 37°C. Preparation of virus stocks was made by freezing and thawing 500 ml of B95-8 cell suspensions three times at a density of 2.5X106 cells/ml. These cells had been grown at 37°C and subsequently incubated for 1 week at 33°C. Cell debris was removed by centrifugation at 800 g for 15 min at 4°C. Supematants were filtered through a 0.45 pm filter (Millex, Millipore Co.) and frozen at -70°C until use.
Cells Human leukocytes were derived from a 2-year-old healthy boy, a 2-year-old boy with infantile nephropathic cystinosis and a 7-year-old girl with adolescent cystinosis. The separation and purification of the leukocytes were carried out using Lymphodex LSM 1077 (Fresenius, Bad Homburg) according to a previously described method [ 191.
In order to transform isolated lymphocytes with the virus stock, [email protected]
cells were inoculated with 10 ml of virus stock for 2 h at 37°C. Thereafter the cultures were washed five times with Hanks’ solution and refed with RPMI-1640 sunnlemented with 20% fetal bovine serum and incubzed at 37°C. The human lymphocytes were transformed efftciently 9-14 days post-infection with formation of floating cell clumps. The transformed lymphocytes were EBNA-positive using the standard method .
Tissue culture EBV-transformed lymphoblasts were grown in RPMI1640 supplemented with 20% fetal bovine serum,
of lysosomes from human lymphoblasts
penicillin G 100 IU/ml and streptomycin sulfate 0.1 mgfml. The cells were grown in 2- or S-liter glass bottles and were mixed once or twice each day by gentle shaking and placed in an incubation room at 3PC. Cells were grown to a final density of between 5~ 105and 7~ lo5 cells/ml. Viable cells were counted after staining in 0.05% Trypan blue. For each lysosome preparation between 5~10~ and lOlo cells were used. The cells were free of mycoplasms as judged from the measurement of uracil incorporation .
Tissue fractionation A scheme of the isolation procedure is given in fig. 1. The cells were allowed to settle for 24 h before preparation. After 4 liters of the culture medium of 5-liter flasks had been decanted, the remaining medium containing the bulk of cells was centrifuged for 10 min at 1000 g (GSA rotor) at 15°C. All centrifugations were done in a RC-5 centrifuge (DuPont-Sorvall, Newton, Conn.). The following preparation steps were carried out between 0 and 4°C: Collected cell pellets were washed once by gentle resuspension in 220 ml 0.25 M sucrose, 10 mM&iethanola&ne, 10 mM acetic acid, pH 7.40, and centrifuged for 10 min at 3000 g (GSA rotor). Cells were lysed in an isolation buffer (0.25 M sucrose, 10 mM triethanolamine. 10 mM acetic acid, 1 mM EDTA, pH 7.40) by being drawn into and exnelled from a 10 ml alass Dinette (internal diameter of tip 1.45 mm) 15 tymes a( a density of 5X 10’ cells/ml. This buffer is a slight modification of the buffer used by Stahn et al. . Unbroken cells and nuclei were collected by a 10 min centrifugation at 750 g (SS-34 rotor). This procedure was repeated three times and the supematants were combined and filtered through one layer of prewashed filter paper (Schleicher & Schtill, No. 589/3, Dassel, West Germany). The filtrate was centrifuged for 10 min at 30000 g (SS-34 rotor) and the pellet resuspended in the isolation buffer at approximately the volume used for each resuspending step during lysis. If aggregations were visible by phase microscopy, they were removed by a 10 min centrifugation at 750 g (SS-34 rotor). To obtain the silica gel gradient, 22 ml of the supernatant was mixed with 15 ml of a mixture of 9 parts Percoll and 1 part IO-fold concentrated isolation buffer adjusted to pH 7.40 with HCI. The final density of the well-mixed suspension was 1.078 g/ml. After centrifueation for 90 min at 40000 a (SS-34 rotor. brake turned off) a 12 ml portion of-the bottom be: tween densities 1.132 and 1.073 g/ml was removed by aspiration into a syringe and diluted 1 : 5 with isolation buffer. After centrifugation for 20 min at 40000 a (SS-34 rotor) the clear sipematant was removed by aspiration and discarded. The remaining suspension was diluted with 3 vol of isolation buffer, followed by centrifugation at 40000 g (SS-34 rotor) for 10 min. This wash step was repeated 4-5 times. After three or four resuspension steps the organelles sedimented as a pellet, indicating that most of the colloidal silica was washed out. The final pellet was resuspended in 2-4 ml of isolation buffer and centrifuged again for 10 min at 480 g (SS-34 rotor). The particles in the supematant were separated in isolation buffer by freeflow electrophoresis in a VaP5 apparatus (Bender & Exp Cell Res 131 (1981)
Harms et al.
2 3 RESUSPENSION
Fig. 2. Enzyme recovery (% of total activity) of pN-acetylglimosaminidase (Cl) and succinate-INT reductase @) during cell lysis. Each pair of columns represents the enzyme activity found in the 750 g supematant after successive resuspensions of the same cell pellet (n=6, S.D. indicated by range bars). See Methods for experimental details. Hobein GmbH, Munich. BRD) usina the followina conditions :’ buffer temperature 6”e, 185-190 mA; 112-122 V/cm, buffer flow 2 ml/fraction/h. The time required for the entire procedure was approx. 10 h.
dylate (pH 7.2) containing 50 mM KC1 and 2.5 mM MgCI, for 30 min. After repeated washes with cacodylate buffer (as above) the samples were post-fixed with 2% osmium tetroxide for 120 mitt, washed with water, and were then soaked overnight in 0.5 % aqueous uranylacetate. After dehydration in graded ethanol solutions the samples were transferred into propylene oxide and embedded in Epon 812 . Ultrathin sections were double stained with uranylacetate (5% methanolic) and lead citrate. Micrographs were obtained with a Siemens Elmiskop 101 instrument. Cytochemical staining for acid phosphatase was performed as described by Leuenberger & Novikoff . The cytochemical assay of cells was carried out in suspension in a shaker at 37°C. Fixation, post-fixation and staining were the same as described above but without uranylacetate.
Materials Percoll and Ficoll were purchased from Pharmacia Fine Chemicals AB (Uppsala). [U-r4C]UDP-D-galactose was obtained from The Radiochemical Centre (Amersham, Bucks, UK). Desialylated and degalactosylated fetuin was kindly supplied by Dr C. H. Bauer, Freiburg im Br., West Germany. All other reagents of analytic1 grade were obtained from commercial sources.
Enzyme assays Protein was determined calorimetrically [24, 251 by Bio-Rad protein assay (Bio-Rad, Richmond, Calif.) using bovine serum albumin as a standard. Marker enzymes were assayed according to the following references: acid phosphatase by the use of /3-glycerophosphate as substrate, EC 184.108.40.206 ; acid p-nitrophenylphosphate phosphatase ; PN-acetylglucosaminidase, EC 220.127.116.11 ; p-galactosidase, EC 18.104.22.168 ; P-glucuronidase, EC 22.214.171.124 ; cathepsin Bl, EC 126.96.36.199 ; NADPH-cytochromec reductase, EC 188.8.131.52 [30, 311; catalase, EC 184.108.40.206 [32, 331; succinate-INT reductase (INT, 3-(4-Iodophenyl) - 2 - (4- nitrophenyl) - 5 - phenyl- 2H- tetrazolium chloride), EC 220.127.116.11 ; UDP-galactose-glycoprotein galactosyltransferase, EC 2.4.1.- . Either sonication or addition of 0.1% Triton X-100 were utilized in an attempt to achieve maximal enzyme activity. Specific activities of enzymes were calculated as mU/mg protein. Relative specific activities are the ratio specific activity in the fraction/specific activitv in the -crude lysate.. The structure-linked latency of P-N-acetylglycosaminidase was tested according to Sellinger et al.  in the modification of Baccino et al.  using 8 mM p-nitrophenyl-fi-u-N-aetylglucosaminide as substrate. When the structure-linked latency of p-nitrophenylphosphate phosphatase was tested the enzyme assay was carried out in the presence of 0.25 M sucrose without addition of Triton X-100.
Electron microscopy Pellets of isolated material were fixed with 2.5% glutaraldehyde buffered with 50 mM sodium cacoExp Cd/ Res 131 (1981)
RESULTS Cell lysis
In preliminary studies, a substantial loss of lysosomal enzyme activity from the cell pellet was noted after the cells had been washed several times in isotonic isolation buffer (0.25 M sucrose, 10 mM triethanolamine, 10 mM acetic acid, 1 mM EDTA, pH 7.40). Systematic studies (fig. 2) showed a release of about 75% of the total mitochondrial (succinate-INT reductase) and lysosomal (P-N-acetylglycosaminidase) enzyme activity after five resuspensions in isolation buffer with a 10 ml glass pipette. Seventy-six percent of the released /3-Nacetylglucosaminidase activity was particlebound and recovered by 10 min centrifugation at 40000 g. Only 24% of the released enzyme activity was free (non-sedimentable under these centrifugation conditions) and this was found mainly in the first 750 g supernatant. Only slight aggregation was produced by
of lysosomes from human lymphoblasts
Table 1. Recovery of protein and marker enzyme activities following tion procedure
B, combined 750 g supematants after cell lysis. C, 30000 g pellet, crude granular fraction. D/l, lysosomal fraction, removed from Percoll gradient. D/2, lysosomal fraction, five times washed after Percoll gradient Preparation step . . .
Protein (n=5) /3-N-Acetylglucosaminidase (n =5) /3-Galactosidase (n =4) Acid #I-glycerophosphatase (n =4) P-Glucuronidase (n =4) Cathepsin B 1 (n=4) Succinate-INT reductase (n =4) Catalase (n=4) NADPHxytochrome-c reductase (n =4) UDP-galactose-glycoprotein (a =4) galactosyltransferase
44.8k6.3 72.7f13.3 78.4k9.1 59.9f 13.0 56.2k11.4 67.4f13.8 70.7+ 10.4 59.5k7.0 70.8f4.4
9.31k2.58 54.4k11.6 50.9k7.7 31.5f7.3 31.7k9.4 168.2f25.9 60.9+ 17.2 21.5k5.3 30.8f8.5 32.4f7.0
2k4k4.2 37.7k3.0 11.9f2.0 13.8f2.3 55.6f9.0 1.71+1.00 b 0.39f0.14
0.21+0.06 lO.OOf1.56 11.77f2.42 4.58kO.93 4.85+ 1.60 30.6f6.5 0.50f0.24 0.13+0.03 0.2OkO.04
Results are given as percentage values, mean f S.D. (percent of protein or enzyme found in the crude lysate). o No method available for protein measurement at this step. * Not done.
the first three resuspension steps as seen washed filter paper as described under by phase microscopy whereas additional Methods. This resulted in the loss of only lysis by continuing resuspension resulted in some 24% of the total activity of P-Nincreasing damage to nuclei and the appear- acetylglucosaminidase. ance of aggregates. Since the bulk of lysosomes were released during the first three Differential centrifugation resuspension steps, the lysis procedure was The amount of enzyme activity recovered confined to three resuspensions. The slight from the combined supernatants (fraction aggregation observed could be diminished B) within the crude granular fraction (fracby passing the suspension through a pre- tion C) varied depending on the enzyme Table 2. Relative
tion procedure B, combined 750 g supematants after cell lysis. C, 30000 g pellet, crude granular fraction. D/2, lysosomal fraction five times washed after Percoll gradient. E, pooled five fractions of highest lysosomal activity after free-flow electrophoresis Preparation step . . .
P-N-Acetylglucosaminidase (n =5) P-Galactosidase (n =4) Acid P-glycerophosphatase (n =4) P-Glucuronidase (n =4) Succinate-INT reductase (n =4) Catalase (n =4) NADPHxytochrome-c reductase (n =4) UDP-galactose-glycoprotein (n =4) galactosyltransferase
1.63kO.30 1.79kO.31 1.28kO.10 1.27kO.24 1.57kO.35 1.37f0.27 1.52f0.32
6.00+1.19 5.31f0.93 3.22f0.42 3.2OkO.45 6&t+ 1.39 2.2220.48 3MVO.83
50.4t15.0 56.5f11.6 23.4f6.0 22.9k6.2 2.17kO.59 0.65kO.26 0.72f0.09
85.6f 15.5 87.6f 13.4 41.7f3.5 36.6f6.1 0.63f0.26 n.d. n.d.
Results are given in relative specific activities, mean f S.D. (spec. act. in the fraction/spec. act. in crude lysate). n.d., Not detectable by the enzyme assay used. Exp Cell Res 131 (1981)
Harms et al.
Isopyknic centrifugation A crude granular fraction (fig. 1, fraction C) was separated by an isotonic, continuous, self-generating gradient of colloidal silica as described by Wolff & Pertoft . As shown in fig. 3, all non-lysosomal enzymes were banding at lower densities between 1.04 and 1.06 g/ml (corresponding to fig. 3, fraction 3-6). Among the lysosomal enzymes measured two different manners of distribution were noted. Acid P-glycerophosphatase and /3-glucuronidase showed a clear bimodal distribution with one peak concomitant with the non-lysosomal enzymes in the lower density region and the second peak with the other lysosomal enzymes. These latter enzymes (P-N-acetylglucosaminidase, P-galactosidase and cathepsin Bl) banded predominantly at higher densities between 1.07 and 1.10 g/ml with Fig. 3. Distribution of marker enzyme activities in maximum activity at 1.085 g/ml. But even fractions obtained after centrifugation of a crude gran- for these enzymes there was some indicaular fraction of EB-virus transformed lymphoblasts tion for bimodal distribution to a minor exin an isotonic colloidal silica gel gradient. Enzyme activity is expressed as percentage of total activity tent as seen from the hump of activity in recovered from gradient. For experimental details, see the lower density region (fig. 3). However, Methods. it was noted that whenever aggregates became visible these were distributed within the low density area and therefore could (table 1). The sedimented activity consist- also account for the observed distribution. ed of about 40% of the non-lysosomal The separation in colloidal silica gel gradienzymes catalase, NADPH-cytochrome-c ents is in contrast to linear gradients of reductase and galactosyltransferase, but sucrose in 10 mM triethanolamine, 10 mM 65-75% of the lysosomal enzymes P-N- acetic acid, 1 mM EDTA pH 7.40 or Ficoll acetylglucosaminidase and P-galactosidase. in isolation buffer where lysosomal P-NA more intermediate recovery of about acetylglucosaminidase and mitochondrial 55 % was noted for acid P-glycerophospha- succinate-INT reductase did not separate tase and P-glucuronidase. Accordingly, a (data not shown). 2.5- to 3.7-fold increase in specific activities of lysosomal marker enzymes resulted from Free-flow electrophoresis Further purification of lysosomes was this preparation step (table 2). Although slightly higher specific activities could be achieved by additional separation of the achieved by varying the centrifugation con- granular fraction (fig. 1, D/2) obtained from ditions this caused too great a decrease in differential and isopyknic centrifugation in total activity to be of use. a high voltage field on a free-flow electroE.rp Cell Res 131 (1981)
of lysosomes from human lymphoblasts
a Fig. 4. (a) Distribution of enzyme Cothode
activities of /3-N-acetylglucosaminidase (0) and succinate-INT reductase @I) in fractions obtained after separation by electrophoresis. Enzvme activitv is exnressed as (mU/ml in eachfraction/mU/ml in the crude Ivsate)x 100. (b) Ratio of lysosomal to mitochondrial enzyme activities; crude lysate (A)= 1, fraction injected into the separation chamber (D/2)=32.6.
phoresis apparatus [22,40] (fig. 4). The two enzyme activities succinate-INT reductase, within mitochondria, and /3-N-acetylglucosaminidase, within lysosomes, were deflected towards the anode, indicating the negative surface charge of both organelles. Lysosomes were deflected more than mitochondria in the equipment used and several fractions were separated towards the anode containing substantially higher specific activities of lysosomal enzymes compared with the granular fraction injected into the separation chamber. This can be shown by the large increase, up to 4- to 5-fold, of the ratio of lysosomal to mitochondrial enzyme activity in these fractions after electrophoresis (P-N-acetylglucosaminidaselsuccinate-INT reductase in fig. 4).
tivities of non-lysosomal marker enzymes resulted from the colloidal silica gradient. Apart from lysosomal enzymes, only succinate-INT reductase was found in significant amount after this preparation step. The pellet following differential centrifugation contained more cathepsin Bl activity than the original crude homogenate (see table 1). Mixing experiments using crude homogenate and fractions from various purification steps always showed the cathepsin Bl activity to be the expected sum of the added activities. Thus, there was no evidence of an enzyme inhibitor in the crude homogenate. Although the distribution of this enzyme in the colloidal silica gel gradient (fig. 3) and the increase of specific activity following the preparation procedure was concomitant to other lysosomal enzymes it was not used as a purification marker for Measurement of marker enzymes lysosomes from lymphoblasts. The relative specific activities of lysoThe recovery of protein and several marker somal and non-lysosomal marker enzymes enzymes following a lysosome preparation until the electrophoresis step is listed in during a preparation are given in table 2. table 1. The most relevant decrease of ac- It should be noted that the data given for Exr, Cd
Harms et al.
the fraction after electrophoresis (E) were pyknic centrifugation and free-flow electroaverage values obtained from five pooled phoresis (data not shown for free-flow elecfractions of highest lysosomal activity trophoresis) differed from lysosomal and (highest ratio P-N-acetylglucosaminidase/ mitochondrial marker enzymes. The parsuccinate-INT reductase in fig. 4). Higher ticle-bound p-nitrophenylphosphatase was values would be obtainable by testing only deflected more towards the anode than the fraction of highest ratio after electro- were lysosomal and mitochondrial enphoresis, but this single fraction contained zymes. The banding separate from lysoinsufficient material for all the measure- somal enzyme activity in a gradient of colloidal silica is shown in fig. 3. Electron ments done. The gradient of colloidal silica resulted in microscopic examinations of the fractions a striking increase in specific activities of high in p-nitrophenylphosphatase showed all lysosomal enzymes tested. After wash- only empty membrane vesicles (data not ing out the colloidal silica the granular frac- shown). In the presence of low4 M amtion before electrophoresis (D/2) contained monium molybdate, the enzyme activity high specific activities of lysosomal en- diminished by about 80% as described for zymes, whereas only traces of catalase phosphoprotein phosphatase (EC 18.104.22.168) (indicating peroxisomes), NADPHxytoby Paigen . The enzyme activity inchrome-c reductase (indicating endoplas- creased 2- to 5-fold (0.25 M sucrose presmic reticulum) and UDP-galactose-glyent) with the addition of 0.2% Triton Xcoprotein galactosyltransferase (indicating 100. Yet, the activity was unchanged when Golgi vesicles) were found. Most of the re- preincubated for 1 h in either hypotonic maining impurity consisted of mitochon- (0.025 M) or isotonic (0.25 M) sucrose . dria (indicated by succinate-INT reductase) The enzyme was relatively unstable with which was markedly decreased in the the loss of activity up to 40% in 24 h when pooled final fraction after electrophoresis stored at 4°C. (E). Catalase and NADPH
5. Survey electron micrograph showing high purity of final lysosomal fraction as revealed after free-flow electrophoresis step. Bar, 0.5 pm. x22000.
Harms et al.
Fig. 6. Survey micrograph of fraction D/2 (see isolation scheme, fig. 1) as it appears after the final centrifugation step. Numerous primary and secondary lysosomes are identified as well as some mitochondria (M), multivesicular bodies (ME), some smooth sur-
faced vesicles (arrows), and remnants of broken organelles (arrowheads). Inset: The obvious club-like appearance of densely stained primary lysosomes is observed after the very first isolation step. Bar, 1 pm; inset: 0.5 pm. x20000 (inset) x32000.
to be mostly unaffected by the isolation procedure.
step when the cells are lysed in the isolation buffer. The electron-dense content of the lysosomes is surrounded by an intact membrane. The same club-like shape of the lysosomes was found whether the cells were lysed in isotonic sucrose buffered with either 10 mM triethanolamine-acetic acid, 10 mM Tris-acetic acid, 10 mM sodium phosphate, or without buffer salts. Addition of EDTA to these isotonic sucrose solutions did not result in notable changes in morphology of the lysosomes in the first
Electron microscopic observations The final fraction (fraction E) obtained after free-flow electrophoresis consisted nearly exclusively of primary and secondary lysosomes (fig. 5). Most of the lysosomes formed elongated organelles with a clublike shape. This obvious morphological alteration of their in vivo structure was observed after the very first fractionation E.vp Cd
Fig. 7. Electron micrographs demonstrating separation
effectiveness of the free-flow electrophoresis. Material deflected in direction of anode consisted mainly of secondary lysosomes, broken mitochondria, and some multivesicular bodies (a), while smooth-surfaced vesi-
lysate. In phosphate-buffered saline (without calcium and magnesium ions) or isotonic NaCl or KC1 the cells did not lyse spontaneously. When lymphoblasts suspended in these solutions were broken by 80 strokes of a tight-fitting glass-glass homogenizer (Dounce) the lysosomes in the homogenate were found to be round shaped, not distinguishable from their in vivo structure.
of lysosomes from human lymphoblasts
cles and membrane sheets were deflected toward the cathode (c). Final fraction of highly purified lysosomes (b) could be collected in between these fractions. Bar, 0.5 pm. X20000.
The fraction obtained after the final centrifugation step (Fraction D/2) is demonstrated in fig. 6. Mitochondria, smooth surfaced vesicles and membrane sheets, and remnants of broken organelles are impurities which could not be excluded completely during the various centrifugation steps. After free-flow electrophoresis these contaminants were separated into different fractions (fig. 7). No morphological changes Exp Cell Rrs 131 (1981)
Harms et al.
Fig. 8. Cytochemical staining for acid phosphatase.
Intact human EB-virus-transformed lymphoblast; N, nucleus; M, mitochondria; arrow, lysosome; arrowhead, Golgi region. Bar, 1 pm. X 13500.
of the separated organelles were observed during the electrophoresis separation procedure. Histochemical staining for acid phosphatase of intact lymphoblasts (fig. 8) revealed a considerable extralysosomal distribution of this enzyme in endoplasmic reticulum and Golgi vesicles. DISCUSSION The method described yields highly purified fractions of isolated lysosomes from human EXP Cell Res 131 (1981)
tissue culture cells. In contrast to most other methods described for rat liver , this technique does not require prior moditication of the lysosomes. Yet, the purity of the fractions obtained compares well with preparations from rat liver with respect to the lysosomal enzyme activities and the morphological findings. The results presented in the figures and tables were obtained from one cell line derived from a normal individual. However, almost identical results were obtained from cell lines from patients with cystinosis, an inherited disorder characterized by lysosomal cystine storage . This suggests that this method may also be suitable for the study of other inherited lysosomal disorders. The disintegration of cells at the beginning of a preparation is a critical step. We were unable to find any condition where mechanical forces would selectively disrupt the plasma membranes of tissue culture cells. All such attempts resulted in aggregation of cell organelles, especially if carried out in hypotonic solution. Our impression is that damage of the nuclear membranes leads to aggregation of cellular organelles . Cohn & Hirsch  first described the spontaneous cell lysis of rabbit polymorphonuclear leukocytes which occurred in an almost isotonic salt-free sucrose solution. The method for lysing human lymphoblasts in an isotonic solution presented in this study is similar and approaches an ideal type of homogenization. Resuspension by pipetting appears to be a more gentle method than homogenization in Teflon-glass [lo, 131or glass-glass [7, 11, 14, 151homogenizers. The mechanism of action of the isotonic isolation buffer described is not completely understood. The low ionic strength of this isotonic buffer may be the primary reason it dissociates plasma membranes. Although the important role of cal-
cium ions in stabilizing membrane structures is well known [44-46] this cannot be the only explanation for the lysis effect since it occurs in isotonic sucrose without EDTA and not with EDTA alone. Isopyknic centrifugation in an isotonic gradient of colloidal silica is an excellent method for separation of lysosomes derived from human lymphoblastoid cells. Many methods for separation of lysosomes have been described using density gradient centrifugation. Such attempts have been most successful for the isolation of modified lysosomes from pretreated animals (i.e., Triton WR 1339 and Dextran 500 , colloidal gold , colloidal iron ). In the case of tissue culture cells isopyknic centrifugation of crude granular fractions have resulted in lysosome-rich preparations, but with contamination by other cell organelles, especially mitochondria and peroxisomes [7, 11, 13-151. Wolff & Pertoft  first described the use of isotonic gradients of colloidal silica for the isolation of lysosomes from HeLa cells. Their findings were quite different from the results obtained for lymphoblasts in this study. They found the lysosomal enzyme @-glucuronidase distributed at lower densities compared with succinate-INT reductase and catalase. Moreover, the mitochondrial succinate-INT reductase overlapped both lysosomal and peroxisomal enzyme activities. In contrast our results with lymphoblasts, using a very similar gradient, show a much sharper distribution of the succinate-INT reductase, with the main portion of lysosomal enzyme activity at higher densities mostly free from mitochondrial contamination. In addition, almost no measurable catalase activity, indicating peroxisomes, was found in the lysosomal fraction from this gradient. More recently Pertoft et al. used a colloidal silica gradient for the isolation of rat liver lyso-
of lysosomes from human lymphoblasts
somes from untreated animals . In these experiments the distribution of the lysosomal arylsulfatase is similar to the distribution of /3-N-acetylglucosaminidase as shown for lymphoblasts in this study. Rome et al.  described the use of colloidal silica gradients in an attempt to isolate lysosomes from human skin fibroblasts. Unfortunately, their data are difficult to compare because the isolation was not complete and therefore fractions which were obtained. were not sufficiently characterized. Interestingly they reported a bimodal distribution of lysosomal enzymes in such gradients. They interpreted this as an indication of the existence of two different forms of lysosomal organelles in these cells. Rather than interpret the bimodal distribution found in lymphoblasts the same way, we think it is more likely that there is some extralysosomal distribution of lysosomal enzymes (see below). Moreover, the distribution of aggregates within the more buoyant fraction from lymphoblasts makes the interpretation of a bimodal distribution difficult. No lysosomal organelles were found in micrographs obtained from these lighter fractions. Carrier-free continuous electrophoresis (free-flow electrophoresis) was first introduced for the preparation of rat liver lysosomes of untreated animals by Stalm et al. . This method requires differential centrifugation and extensive washing of the crude granular fraction before a separation of lysosomes from other cell organelles can be achieved. However, the application of the same preparation steps to lymphoblasts did not result in separation of lysosomes. The additional purification by isopyknic centrifugation in colloidal silica combined well with the electrophoresis step. In this procedure, the same buffer system of constant osmolarity is used throughout the Exp Cell Res 131 (1981)
preparation procedure. The only disadvantage of the polyvinylpyrollidon-coated colloidal silica ([email protected]
) is that it disturbs the separation of lysosomes and mitochondria in the electric field and must be washed out for the electrophoresis step to be effective. Some loss of lysosomal enzyme activities resulted from this washing procedure. The further separation by electrophoresis is essential to obtain the most highly purified lysosome preparations. The degree of purification (and of lysosome recovery) is determined by the fractions which are used (fig. 4). The increase of the specific activities of lysosomal enzymes during preparation differs depending on the enzyme measured. This could be explained either by a selection of a single lysosomal species from a fairly heterogeneous population differing in content of hydrolases  or by the markers being not completely selective. The bimodal distribution in the colloidal silica gel gradient of acid /3-glycerophosphatase and &glucuronidase as well as the largely nonlysosomal distribution of acid phosphatase, as revealed by the cytochemical studies, suggest the latter possibility is likely. This finding compares well to the distribution of acid phosphatase within the GERL as extensively studied and reviewed by A. B. Novikoff and co-workers (for review see [SO]). Extralysosomal distribution of p-glucuronidase was described by deDuve et al.  and Fishman et al. . From this it is understandable that these two enzymes are distributed in the colloidal silica gel gradient to a greater extent within the endoplasmic reticulum and Golgi vesicles and therefore the relative specific activities of these enzymes are lower in the final lysosomal fraction than are other enzymes selectively found within lysosomes. The lower recovery by differential centrifugation of the Exp Cell Res 131 (1981)
non-lysosomal enzymes catalase, NADPHcytochrome-c reductase and galactosyltransferase may be explained by damage to the corresponding organelles leading either to release of soluble enzymes or the formation of small non-sedimentable remnants. Numerous small remnants and vesicles (0 approx. 0.1 pm) were seen in micrographs from fraction B (data not shown). The activity of acid p-nitrophenylphosphatase was clearly distinct from other lysosomal enzyme activities by all separation methods used. The non-lysosomal localization of acid p-nitrophenylphosphatase activity has been reviewed by Beaufay . The distribution and corresponding morphological appearance we observed are analogous to the finding of Baggiolini et al. [53, 541 in rabbit heterophil leukocytes. Although we found the enzyme activity increased by the addition of the detergent Triton X-100, osmotic activation in hypotonic sucrose (0.025 M) was not observed. Since the latter procedure is known to eliminate the structure-linked latency of lysosomal acid phosphatase , the most likely interpretation for the effect of Triton X-100 is the activation of a membranebound enzyme. Two types of lysosomal morphology were observed, the elongated organelles resembling primary lysosomes and other structures containing material under degradation and therefore assumed to be secondary lysosomes. These types are not separated by the colloidal silica gel gradient and at best have only a tendency to separate in the electric field. The shape of the elongated lysosomes differs from their round in vivo shape observed in unbroken cells. The results obtained from the disintegration experiments in different buffers indicate that the low ionic strength of the isotonic buffer is the primary reason for the variation
of lysosomes from human lymphoblasts
of the lysosomal shape. But the clublike shape which is found immediately after cell lysis and at all steps of preparation is not due to damage of the organelles as judged by the high structure-linked latency of /3-N-acetylglucosaminidase. The morphology of the more dense lysosomes from libroblasts reported by Rome et al. [g] is to some extent similar to the club-like shape observed from lymphoblasts although the fibroblast lysosomes are obviously altered as indicated by the very low structurelinked latency reported. In conclusion, the combination of careful cell lysis, differential centrifugation, isopyknic centrifugation and free-flow electrophoresis has permitted the isolation of the most highly purified lysosomes from human cultured cells yet reported. Since large quantities of human lymphoblasts can be grown in vitro, this method may provide investigators the opportunity to study a large number of lysosome-related questions in human cells. This method can serve as a tool for investigations in the wide field of human inherited lysosomal disorders. The excellent technical assistance of Hildegard Kern, Ocean Pellett, Christine Grund and Marilyn Woodward is gratefully acknowledged. The authors wish to thank Dr Hans-G. Heidrich, Munich, for his valuable suggestions for the application of the free-flow electrophoresis technique, and Dr Stanley A. Mendoza and Dr Robert G. Oshima. La Jolla. for their helo in preparing the manuscript.’ This work was supported by Deutsche Forschungsgemeinschaft, Bonn-Bad Godesberg, grants Ha 756/2 to Ha 756/6; NIH grants GM 24648, AM 18434 and GM 17702;The National Foundation-March of Dimes, no. NFl-373; and the Medical Research Service of the Veterans Administration. A preliminary report of this work was presented at the Annual Meeting of the Western Society for Pediatric Research in Carmel, California, February 7-9, 1979 [551.
REFERENCES 1. Hers, H G & van Hoof, F, Lysosomes and storage diseases. Academic Press, New York (1973). 2. Schneider, J A, Schulman, J D & Seegmiller, J E,
The metabolic basis of inherited disease (ed J D Stanbury, J B Wyngaarden & D S Fredrickson) 4th edn, p. 1660. McGraw-Hill, New York (1978). 3. McKusick, V A, Neufeld, E F & Kelly, T E, The metabolic basis of inherited disease (ed J D Stanbury, J B Wyngaarden & D S Fredrickson) 4th edn, p. 1282. McGraw-Hill, New York (1978). 4. Beaufay, H, Lysosomes. A laboratory handbook (ed J T Dingle) p. 1. North-Holland Publ, Amsterdam (1972). 5. Wattiaux, R, Wattiaux-DeConinck, S, Ronveaux-Duoal, M F & Dubois, F, J cell biol 78 (1978) 349. 6. Pertoft, H, Warnregard, B & Hook, M, Biochem j 174 (1978) 309. 7. Wolff, DA & Pertoft, H, Biochim biophys acta (1972) 197. 8. Rome, L H, Garvin, A J, Allietta, M M & Neufeld, E F, Cell 17 (1979) 143. ^ Wattiaux, R, Arch int physiol biochim 70 (1962) _ ‘. 765. 10. Gordis, L & Nitowsky, H M, Exp cell res 38 (1965) 11 556. Thacore, H & Wolff, D A, Exp cell res 49 (1968) ’ 266. 12. Tulkens, P & Trouet, A, Arch int physiol biochim 80 (1972) 623. 13. Milsom, J P & Wynn, C H, Biochem j 132 (1973) t4 493. Tulkens. P. Beaufav. H & Trouet. A. J cell bio163 . ’ (1974) 383.’ 15. Sun. A & Poole. B. Anal biochem 68 (1975) 260. 16. Noel, G, Peterson’, C, Trouet, A &‘Tulkens, P, Eur j cancer 14 (1978) 363. 17. Miller, G, Shope, T, Lisco, H, Stitt, D & Lipman, M, Proc natl acad sci US 69 (1972) 383. 18. Miller, G & Lipman, H, Proc natl acad sci US 70 (1973) 190. 19. Harris, R & Ukaejiofo, E 0, Lancet ii (1%9) 327. 20. Reedman, B M & Klein, G, Int j cancer 11 (1973) 499. 21. Kenny, G E, Microbiology 1975 (ed D Schlessinger) p. 32. American Society for Microbiology, Washington (1975). 22. Stahn, R, Maier, K P & Hannig, K, J cell biol 46 (1970) 576. 23. Henning, R & Heidrich, H G, Biochim biophys acta 345 (1974) 326. 2$ Bradford; M M, Anal biochem 72 (1976) 248. . Sedmak, J J & Grossbera. S E, Anal biochem 79 (1977) 544. 26. Gianetto, R & deDuve, C, Biochemj 59 (1955)433. 27. Walter, K & Schiitt, C, Methods of enzymatic analysis (ed H U Bergmeyer) 2nd edn, vol. 2, p. 856. Verlag Chemie, Weinheim, and Academic Press, New York (1974). 28. Koldovsky, 0 & Palmieri, M, Biochem j 125(1971) 697. 29. Barrett, A J, Anal biochem 47 (1972) 280. 30. Sottocasa, G L, Kuylenstiema, B, Emster, L & Bergstrand, A, J cell bio132 (1%7) 415. 31. Van Gelder, B F & Slater, E C, Biochim biophys acta 58 (1%2) 593. 32. Beers, R F & Sizer, I W, J biol them 195 (1952) 133. 33. Luck, H, Methods of enzymatic analysis (ed H U Exp Cell RPS 131 (1981)
34. 35. 36. 37.
Harms et al.
Bergmeyer) 1st edn, 2nd printing, p. 885. Verlag Chemie, Weinheim, and Academic Press, New York (1%5). Pennington, R J, Biochem j 80 (1961) 649. Bauer, C H, Hassels, B F & Reutter, W G, Biochem j 154(1976) 141. Sellinger, 0 Z, Beaufay, H, Jacques, P, Doyen, A & deDuve, C, Biochem j 74 (1960) 450. Baccino, F M & Zuretti, M F, Biochem j 146(1975)
146. 38. Franke, W W, Luder, M R, Kartenbeck, J, Zer-
ban, H & Keenan, T W, J cell bio169 (1976) 173.
45. Manery, J F, Fed proc 25 (1966) 1804. 46. Wolman, M & Wiener, H, Nature 200 (1963) 886. 47. Henning, R & Plattner, H, Biochim biophys acta 354 (1974) 114. 48. Glaumann, H, Jansson, H, Arborgh, B &
Ericsson, J L E, J cell biol67 (1975) 887. 49. Davies, M, Lysosomes in biology and pathology 50. 51.
39. Leuenberger, P M & Novikoff, A B, J cell bio165
(1975) 324. K & Heidrich, H G, Methods in enzymology. Biomembranes, part A (ed A S Fleischer & L Packer) vol. 31, p. 746. Academic Press, New York (1974). 41. Paigen, K, J biol them 233 (1958) 388. 42. Appelmans, F & deDuve, C, Biochem j 59 (1955)
426. 43. Cohn, Z A & Hirsch, J G, J exp med 112 (1960) 983. 44. Benedetti, E L & Emmelot, P, Ultrastructure . in
biological systems, vol. 4: The membranes (ed A G Dalton & F Hagnenau) p. 33. Academic Press, New York (1968).
(ed J T Dingle & R T Dean) vol. 4, p. 305. NorthHolland Pub], Amsterdam (1975). Novikoff, A B, Proc natl acad sci US 73 (1976) 2781. deDuve, C, Pressman, B C, Gianetto, R, Wattiaux, R & Appelmans, F, Biochem j 60 (1955) 604. Fishman, W H, Goldman, S S & delellis, R, Nature 213 (1%7) 457. Baggiolini, M, Hirsch, J G & deDuve, C, J cell biol 40 (1%9) 529. - Ibid 45 (1970) 586. Harms, E, Kartenbeck, J & Darai, G, Clin res 27 (1979) 103A.
Received May 30, 1980 Revised version received September 16, 1980 Accepted September 19, 1980