Isolation and Characterization of Plasma Membranes from ......T355, and M21) contained the purified endoplasmic reticulum. The collected membrane fractions were diluted 2- to 3-fold

[CANCER RESEARCH 41. 4031-4038, October 1981]0008-5472/81/0041-OOOOS02.00

Isolation and Characterization of Plasma Membranes from TransplantableHuman Astrocytoma, Oat Cell Carcinoma, and Melanomas1

Aileen F. Knowles,2 Jose F. Leis, and Nathan O. Kaplan

Department of Chemistry and the Cancer Center, University of California, San Diego, La Jol/a, California 92093

ABSTRACT

Purified plasma membranes were obtained from five trans-

plantable human tumors, a Grade IV astrocytoma, an oat cellcarcinoma, and three melanomas. Plasma membrane fractionswere isolated from tumor homogenates by differential anddiscontinuous sucrose gradient centrifugation. Determinationof enzyme activities indicated that the plasma membranes wereenriched 10- to 20-fold with respect to 5'-nucleotidase, nico-tinamide adenine dinucleotide glycohydrolase, Mg2+-activated

nucleoside triphosphatase, and sialic acid. Specific activitiesof nearly all the enzymes varied with the individual tumors,even among tumors of the same type, i.e., the melanomas.Electron micrographs of the plasma membrane fractionsshowed smooth single-membrane vesicles with slight contam

ination by lysosomes. Therefore, these membranes are suitablefor comparative biochemical studies and for the preparation oftumor-specific monoclonal antibodies.

Plasma membranes from all five tumors contained very highMg2+-adenosine triphosphatase (ATPase) activities. The Na+-K+-ATPase was a minor component of the total ATPase of

these membranes (

A. F. Knowles et al.

to 10 g/mouse was 3 weeks for astrocytoma, 6 weeks for T355, and2 to 3 months for M21, T242, and T293.

Isolation of Purified Plasma Membranes and Endoplasmic Retic-

ulum. Usually 30 to 50 g of tumor tissue were used for each preparation. Crude microsomal fractions were obtained by centrifuging thepostmitochondrial supernatant at 60,000 x g for 90 min as describedpreviously (20). The further separation of the plasma membrane andendoplasmic reticulum on a discontinuous sucrose gradient was carriedout by a modification of the procedure described by Aronson andTouster (1). The microsomal pellet was resuspended in 56% sucrose(w/w) containing 5 mM Tris-CI, pH 8.0, so that the final sucrose

concentration was 43% (w/w) at 1 to 2 ml/g tissue. For T293, 15 mlof the suspension were placed in 25.4- x 89-mm cellulose nitrate tubes.

The suspension was overlaid by 10 ml each of 31 and 7% sucrose (w/w) containing 5 mM Tris-CI, pH 8.0. For T24, T242, T355, and M21,

the suspension (10 ml) was overlaid by 6.5 ml each of 34, 31, 27, and7% sucrose (w/w) containing 5 mM Tris-CI, pH 8.O. The tubes were

centrifuged overnight in a Beckman SW 27 rotor at 26,000 rpm. Aftercentrifugation, the protein bands at the interfaces were collected fromthe top by Pasteur pipets with tips bent at 90Â°. The band collected at

the interface of 7% and the next layer of sucrose contained purifiedplasma membranes. The band collected at the 43 and 31% sucroseinterface (for T293) or at the 43 and 34% interface (for T24, T242,T355, and M21) contained the purified endoplasmic reticulum. Thecollected membrane fractions were diluted 2- to 3-fold with ice-cold 5mM Tris-CI, pH 8.0, and concentrated by centrifugation at 60,000 x gfor 90 min. The pellet was resuspended in 0.25 M sucrose: 10 mw Tris-

CI, pH 8.0, at approximately 10 mg protein per ml and stored in smallaliquots at -20Â°.

Enzyme Assays. All enzyme reactions were carried out at 37Â°.Mg2*-ATPase was assayed in a reaction mixture containing 50 mM

Tris-CI (pH 7.5), 5 mM MgCI2, and 5 mM sodium ATP for 10 min.

Protein was denatured by 5% trichloroacetic acid (final concentration).Aliquots of the deproteinized supernatant solution were taken fordetermination of P, using the procedure of Lohmann and Jendrassik

(23).5'-Nucleotidase was assayed in a reaction mixture containing 0.1

M glycine (pH 9.1), 5 mw AMP, and 10 mM MgCI2 (1). Reaction wascarried out for 10 min.

Rotenone-insensitive NADH-cytochrome c reducÃ-ase was deter

mined in a reaction mixture containing 50 mM potassium phosphate(pH 7.7), 0.1 mM EDTA, 36 mM cytochrome c, 2 mM sodium azide, and10 ng rotenone per ml. Reduction of cytochrome c was followed at550 nm. A molar extinction coefficient of 18.5 x 103 M~1 cm~' was

used in the calculation of the activity.NAD* glycohydrolase was determined by the disappearance of

NAD* which was inhibitable by nicotinamide (7). Residual NAD* was

monitored by measuring the NAD-cyanide adduci at 325 nm (5). Inmeasuring total NAD* degradation, residual NAD* was determined

coupled to alcohol dehydrogenase (5).Acid phosphatase was assayed in 1 ml of reaction mixture containing

0.1 M sodium acetate (pH 5.0), 10 mM p-nitrophenyl phosphate, and0.2% Triton X-100. After 5 min at 37Â°, the reaction was stopped by

the addition of 2 ml of 0.1 N NaOH, and absorbance at 410 nm wasdetermined. A molar extinction coefficient of 18.3 x 103 M~' cm~' was

used in calculation of the data.Determination of Sialic Acid. Sialic acid content of subcellular

fractions was determined according to Warren (34). In order to eliminate a strong interference by sucrose with the thiobarbituric acid assay,the cellular fractions (in 0.25 M sucrose:5 mM Tris-HCI) were diluted 5-to 10-fold with water and sedimented at 100,000 x g for 60 min. The

pellet was washed 3 times with water by resuspension and centrifugation. Sucrose could also be removed by dialysis against 5 mw Tris-

HCI, pH 8.0. The cellular fractions were then lyophilized and delipidatedby extraction with chloroform:methanol (2:1 ). Sialic acid content of thedelipidated membranes was then determined. Data were corrected forthe presence of interfering deoxyribose as described by Warren (34)

except that the following molar extinction coefficients were used: 6.6x 104 and 3.2 x 104 M~' cm"' for sialic acid at 549 and 532 nm,respectively; and 3.8 x 10'andi.37 x 105 M~' cm "' for deoxyribose

at 549 and 532 nm, respectively.['H]Ouabain Binding to Plasma Membrane. [3H]Ouabain binding

assays were carried out according to the method of Matsui andSchwarz (24). To a reaction mixture containing 25 mM Tris-CI (pH 7.5),

120 mM NaCI, 3 mM MgCI2, 3 mw ATP, 0.2 mg membrane proteins,and 0.1 /iCi carrier-free [3H]ouabain was added either 1 JIM or 1 mMnonradioactive ouabain. Binding was complete in 30 min at 37Â°, and

the membranes were centrifuged at 100,000 x g for 30 min. Thesupernatant solution was decanted, and residual solution on the wall ofthe tubes was removed carefully by Kimwipes. The pellet was thendissolved in 0.2 ml 10% sodium dodecyl sulfate. An aliquot (0.1 ml)was counted with ACS counting solution (Amersham/Searle Corp.,Arlington Heights, III.).

Electron Microscopy. Membrane fractions for electron microscopywere prepared according to the method of Fleischer and Kervina (11).To 0.5 ml of a suspension containing 0.1 to 0.3 mg membrane proteinwas added 0.5 ml 5% glutaraldehyde in 0.1 M sodium cacodylate, pH7.4. After overnight fixation at 4Â°,the membranes were pelleted in an

SW 50.1 rotor for 15 min at 13,000 rpm. The pellet was washed andthen fixed in 1% osmium tetroxide. After removal of osmium tetroxide,the samples were dehydrated in a graded ethanol:water series followedby a graded ethanol:propylene oxide series and embedded in Epon.Thin sections were cut on a Sorvall MT-2 microtome and postfixed in

uranyl acetate and lead citrate. Electron micrographs were taken on aZeiss 10 electron microscope.

Materials. Na-AMP, Tris-ATP, horse heart cytochrome c (type c),ouabain, concanavalin A, p-nitrophenyl phosphate, sialic acid, rote

none, glutaraldehyde, and yeast alcohol dehydrogenase were obtainedfrom Sigma Chemical Co., St. Louis, Mo. Disodium ATP was fromBoehringer Mannheim Biochemicals, Indianapolis, Ind. NAD* was purchased from P-L Biochemicals, Inc., Milwaukee, Wis. [3H]Ouabain (49

Ci/mmol) was obtained from Amersham/Searle Corp.

RESULTS

Enzyme Profiles of the Purified Tumor Plasma Membranes.The purified plasma membranes obtained from discontinuoussucrose gradient were assayed for an array of enzyme activitiesincluding marker enzymes for plasma membranes, i.e., 5'-nucleotidase and Mg2*-ATPase; marker enzymes for endo

plasmic reticulum, i.e., rotenone-insensitive NADH-cytochromec reducÃ-aseand NAD* glycohydrolase; and marker enzyme for

lysosomes, i.e., acid phosphatase. The activities of these enzymes of the various tumor plasma membranes are listed inTable 1 together with the values obtained for homogenate andpurified endoplasmic reticulum for comparison. Table 1 showsthat the specific activities of 5'-nucleotidase and Mg2+-ATPase

of the plasma of all 5 tumors were 4 to 20 times higher than inthe homogenates and were 2.5 to 8 times higher than thatfound in the endoplasmic reticulum fractions.

Rotenone-insensitive NADH-cytochrome c reducÃ-aseactivity

was measurable in the plasma membrane fraction although thespecific activities were usually lower than that of the endoplasmic reticulum fraction. Presence of acid phosphatase indicatedthat the plasma membrane fractions were contaminated withsome lysosomes, which were borne out by electron microscopic observations. However, lysosomal contamination couldnot account for the extremely high activity seen with the plasmamembranes of astrocytoma and melanoma (T355).

To our surprise, we found that NAD* glycohydrolase was a

better marker enzyme for the plasma membranes than endo-

4032 CANCER RESEARCH VOL. 41

on June 20, 2021. © 1981 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

Plasma Membranes of Human Tumors

Table 1Enzymatic activities and sialic acid content of cellular fractions from 5 transplantable human tumorsEnzyme activities were determined as described under "Materials and Methods." In the assay for NAD'

glycohydrolase, residual NAD* was determined by measuring NAD "-cyanide adduct.

ijmol/mm/mgproteinTumorsAstrocytoma

(T24)Oat

cell carcinoma(T293)Melanoma

(T242)Melanoma

(T355)Melanoma

(M21)Cellular

fractionH"I'M

ERHI'M

ERHI'M

ERH

PMERH

PMER5'-Nu-

cleoti-

dase0.025

0.7500.1080.013

0.0540.0160.023

0.2610.0800.012

0.1450.0180.068

1.1510.324Mg2'-

ATPase0.047

0.7830.1570.105

1.5260.6350.050

0.5080.1410.102

0.8610.2550.032

0.7900.164NAD4

glycohydrolase0.002

0.0330.0080.001

0.0220.0080.002

0.0130.0050.003

0.0420.0140.003

0.0440.007Rotenone-in-

sensitiveNADH-cyto-chrome c re

ducÃ-ase0.017

0.0580.1120.018

0.1250.2020.025

0.1250.1620.019

0.0450.2130.004

0.0700.094Acid

phos-

phatase0.077

0.8830.2290.021

0.1130.0940.023

0.1800.1190.063

0.4670.1760.041

0.1610.077Sialic

acid(M/mg)2.8

15.44.02.112.8

6.02.5

25.410.02.514.2

4.31.916.9

4.3a H, homogenate; PM. plasma membranes; ER, endoplasmic reticulum.

plasmic reticulum. The increase in specific activity of thisenzyme in the plasma membrane relative to the homogenatewas comparable to that of 5'-nucleotidase.

In comparing the results obtained with the 2 methods ofmeasuring NAD* degradation, we found the plasma membraneof melanoma T242 contained large amounts of a second NAD+-

degrading enzyme, the activity of which was not inhibited bynicotinamide. Table 2 shows that the NAD*-degradative activity

of the plasma membranes of T24, T293, T355, and M21 couldbe inhibited more than 70% by 30 rriM nicotinamide whereasless than 20% inhibition was obtained with the plasma membranes of T242 when the disappearance of NAD was monitoredby alcohol dehydrogenase activity. This indicated largeamounts of nucleotide pyrophosphatase activity were presentin the T242 plasma membrane. The product of the reactioncatalyzed by nucleotide pyrophosphatase, nicotinamide mon-onucleotide, would not be reduced by alcohol dehydrogenasebut would form a cyanide adduct with the same absorptioncharacteristics as the cyanide adduct of NAD.

Sialic Acid Content of Plasma Membranes. In the determination of sialic acid in the cellular fractions, we found stronginterference with the thiobarbituric acid assay caused by 2-deoxyribose in the homogenates of 2 of the 5 tumors. Astrocytoma and oat cell carcinoma have unusually high DNA:pro-tein ratios (data not shown). As can be seen in Chart 1, thechromophore obtained from the homogenates of oat cell carcinoma has an absorption maximum at 532 nm, which corresponds to the absorption maximum due to 2-deoxyribose. Thechromophore obtained from the plasma membrane shows anabsorption maximum at 549 nm which corresponds to that ofsialic acid. We believe the shift in absorption maxima is corroborating evidence of the purity of the plasma membrane fractions. After correction for 2-deoxyribose interference, it can beseen that sialic acid was enriched at least 6-fold in the plasmamembrane fractions with respect to homogenates (Table 1).

Electron Microscopy of Plasma Membranes. Fig. 1 shows

Table 2NAD ' -degradative activities of tumor plasma membranes

NAD* degradation was determined in the absence or presence of 30 rnwnicotinamide. Residual NAD* was determined after being converted to NADH by

ethanol and alcohol dehydrogenase.NAD* degradation (nmol

min/mg)

Plasma membranesfromAstrocytoma

(T24)Oatcellcarcinoma(T293)Melanoma

(T242)Melanoma(T355)Melanoma

(M21)-Nicotin

amide39.016.971.443.841.9+Nicotinamide6.93.561.48.111.9%of inhibition84.979.314.081.571.6

that the best plasma membrane preparations were obtainedfrom astrocytoma (Fig. 1a) and oat cell carcinoma (Fig. 10) interms of morphology. The only discernible contaminating or-

ganelles were lysosomes. Plasma membrane fractions from themelanomas usually had a more heterogeneous composition(Fig. 1, c to e). In addition to lysosomes, autophagic vesicleswere often found in these preparations. However, all the plasmamembrane fractions were virtually free of vesicles arising fromendoplasmic reticulum. Fig. 1f is an electron micrograph of theendoplasmic reticulum fraction from melanoma (T242). It ischaracterized by vesicles studded with ribosome particles.Endoplasmic reticulum from the 4 tumors have very similarmorphology (not shown).

We have also carried out sodium dodecyl sulfate-polyacryl-

amide gel electrophoresis on the 5 plasma membrane preparations. The peptide composition of the membranes was verycomplex with more than 40 discernible bands. Distinct differences between the different tumor plasma membranes couldbe observed. This was especially pronounced when comparingpeptides of lower molecular weights (M.W.


Ouabain Binding to the Plasma Membranes. Anothermarker enzyme for mammalian plasma membranes Â¡sthe Na+-K+-dependent ATPase which is inhibited by ouabain. We found

that the ouabain inhibition of the tumor plasma membraneATPase was quite small as shown in Table 3. However, all themembranes bound [3H]ouabain in the range of 20 to 60 pmol/

mg protein, which was comparable to values obtained with aplasma membrane fraction from mouse brain, a tissue knownto have large amounts of Na+-K+-dependent ATPase.

Mg1'*-dependent Nucleoside Triphosphatase Activities of

the Tumor Plasma Membranes. Using ATP as the substrate,we found that all the tumor plasma membranes possessed avery active Mg2+-activated ATPase. Results in Table 4 showed

that other nucleoside triphosphates were also effective assubstrates. Activities obtained with AMP were probably derivedfrom 5-nucleotidase, although values presented in Table 4were obtained at pH 7.5 whereas 5'-nucleotidase activities

(Table 1) were usually measured at pH 9.1. The plasma membranes exhibited little hydrolytic activity toward p-nitrophenylphosphate and PP, at neutral pH (data not shown.)

Chart 2 shows that Ca2* could substitute for Mg2+ in activating the ATPase activity. Activities obtained with Mn2+ were

much lower. Although only the data obtained with the plasmamembrane ATPase of oat cell carcinoma are presented here,this was true of all other plasma membranes used in this study.

550Wavelength Irani

Chart 1. Absorption spectra of chromophores obtained in the thiobarbituricacid assays. Thiobarbituric assays were carried out according to the method ofWarren (34) with: 15 /Â¿gsialic acid (T); 2.68 jjg 2-deoxyribose (2); 0.63-mghomogenates from oat cell carcinoma (3); and 0.42-mg plasma membranes fromoat cell carcinoma (4).

Table 3Ouabain inhibition of the tumor plasma membrane A TPase and [3HÂ¡ouabain

binding to the plasma membranesATPase was assayed in a reaction mixture containing 30 HIM histidine-Tris

(pH 7.5), 100 mM NaCI, 10 mM KCI, 5 mM MgCI2, and 5 mM Tris-ATP with orwithout 1 mM ouabain. Reactions were carried out for 10 min at 37Â°with 50 to150 jig membrane protein. [3H]Ouabain binding was carried out as described in'Materials and Methods."

ATPase activity (Â¿imol/min/mg)Plasma

membranesfromAstrocytoma

Oat cell carcinomaMelanoma (T242)Melanoma (T355)Melanoma (M21)Mouse brain-Ouabain0.77

1.860.410.690.410.35+

Ouabain0.69

1.580.280.560.340.09%

of inhibition byouabain10.4

15.131.718.817.174.3[3H]Oua-

bain binding (pmol/

mg)19.1

51.733.139.124.260.6

Table 4Substrate specificity of the Mg**-dependent nucleoside triphosphatases of

tumor plasma membranesReactions were carried out in a reaction mixture containing 50 mM Tris-CI (pH

7.5), 2 mM MgCI.,, and 2 mM nucleoside phosphate esters.

Enzyme activity (fimol P./min/mg) of

Oat cellAstrocy- carcinoma Melanoma Melanoma Melanoma

Substrate toma (T24) (T293) (T242) (T355) (M21)

ATPADPAMPGTPITPCTPUTP0.880.540.650.951.070.960.981.240.280.201.201.261.181.280.560.340.260.420.490.440.481.130.460.300.801.110.950.990.630.330.770.470.590.480.49

1.5

Ei

1.0

0.5

2468

Divalent ions I mM )10

Chart 2. Effect of divalent ions in activating the ATPase activity of oat cellcarcinoma plasma membranes. Plasma membranes (70 fig) of oat cell carcinomawere preincubated with the indicated concentrations of divalent ions in 1 ml of50 mM Tris-CI, pH 7.5. for 5 min at 23Â°.Reaction was initiated by the addition of5 /Â¿molof ATP, and incubation was carried out for 10 min at 37Â°.

Effect of Concanavalin A on the Mg11' -dependent ATPaseand S'-Nucleotidase of Tumor Plasma Membranes. It has

been reported that concanavalin A stimulated the plasma membrane ATPase of several types of cells (16, 27, 29). Chart 3shows that concanavalin A stimulated the plasma membraneATPases of astrocytoma and oat cell carcinoma considerably,whereas it had little or no effect on the ATPases of the melanoma cell membranes. The concentration of concanavalin Arequired for stimulation of the ATPase of oat cell carcinomaand astrocytoma plasma membranes was one order of magnitude lower than the ones used for stimulation of the ATPase ofliver plasma membranes (29). Concanavalin A is also a knowninhibitor of 5'-nucleotidase (3). Chart 4 shows that 5'-nucleo-

tidase activity of all 5 tumor plasma membranes was inhibitedbetween 70 and 80% by concanavalin A at 200 fig/ml.

DISCUSSION

The 5 transplantable human tumors used in this study weremaintained in the athymic mice. These tumors, in spite of manypassages in the athymic mice, retained their histological characteristics and responded to drugs similar to those in human





2.8

2"

3 1-6

- 1 2S "

Â§0.8Â£5

0.4

20 30 40 50Concanavalin A Uug/ml)

60

Chart 3. Effect of concanavalin A on the Mgz*-ATPase of tumor plasma

membranes. Plasma membranes (60 to 140 fig) from astrocytoma (O). oat cellcarcinoma (A), melanoma (T242) (O). melanoma (T355) (â€¢).and melanoma(M21) (â€¢)were preincubated with various amounts of concanavalin A at 23Â°for

10 min in 1 ml solution containing 50 mM Tris-CI (pH 7.5) and 5 mM MqCI.Reaction was initiated by the addition of 5 /imol ATP.

20 30 40 50 60^100 200

Concanavalin A lug/milChart 4. Effect of concanavalin A on the 5'-nucleotidase of tumor plasma

membranes. Plasma membranes (100 to 500 /ig) from astrocytoma (O), oat cellcarcinoma (A), melanoma (T242) O. melanoma (T355) (â€¢),and melanoma(M21) (â€¢)were preincubated with various amounts of concanavalin A at 23Â°for

10 min. Reaction was then initiated by the addition of 1 ml reaction mixturecontaining 0.1 M glycine (pH 9.1), 10 mM MgCI2, and 5 mM AMP.

patients (12). Although we cannot ascertain if the passage inthe mice has altered the biochemical composition of the tumorcells, we have found remarkable reproducibility in the mito-

chondrial functions and enzyme activities of the plasma membranes from these tumors over an extended period of time (2years). We therefore feel it is a valid system for the comparativestudy of the biochemistry of human tumors.

In the evaluation of specific alterations in transformed cells,it is necessary to compare tumor tissue and the normal tissuefrom which it originates. Thus, animal hepatoma has becomethe best documented system. In the study of human tumors,

the availability of normal human tissues poses a problem,especially in amounts necessary for any isolation work and forstudy of labile function (e.g., energy-linked functions in mitochondria). However, meaningful comparison can still be madefrom information obtained with normal and malignant tissuefrom other species.

This paper deals mainly with the methodology involved in thepreparation and characterization of plasma membranes andthe best way to determine purity of these membranes. We findit to be easily applicable to other types of tumors.

The preparative procedure described herein was not developed solely for the isolation of plasma membranes. We havealso studied mitochondria obtained from these tumors andhave avoided using substances, e.g., Zn2+ or fluorescein mer

curic acetate, which would damage mitochondria. The procedure of Aronson and Touster (1 ) was followed with 2 modifications, (a) Only the crude microsomal fraction was used asthe source of plasma membranes and (Ã³)2 additional layers ofsucrose solution were used in the preparation of plasma membranes from astrocytoma and the 3 melanomas at the step ofdensity gradient centrifugation. In spite of the fact that the 5tumors were histologically different, this procedure consistentlyyielded pure plasma membranes.

The purity of these membranes was evaluated by both morphology and the enrichment of the plasma membrane markerenzymes. 5'-Nucleotidase has been found to be exclusively

localized in the plasma membranes of HeLa cells (3) and murineependymoblastoma cells (22) by cytochemical methods. Thisenzyme was enriched 10-fold in the membrane preparations

described above. Similar enrichment was also obtained withrespect to Mg2+-ATPase and sialic acid.

We have found that the use of marker enzymes for quanti-

tating the amount of other organelles in the plasma membranefractions yielded misleading results. This was demonstrated in3 cases, (a) Cross-contamination of the plasma membrane

fraction by endoplasmic reticulum was determined from measurements of NAD* glycohydrolase, a putative marker enzyme

for endoplasmic reticulum (15), and the rotenone-insensitiveNADH-cytochrome c reducÃ-ase, a standard marker enzyme forendoplasmic reticulum. Our results indicated that NAD* gly

cohydrolase was actually a marker enzyme for plasma membranes, in agreement with the finding of Bock ef al. (2), whohave shown the presence of this activity in liver plasma membranes, (b) Determination of the rotenone-insensitive NADH-

cytochrome c reducÃ-ase indicated that there might be considerable contamination of the plasma membrane fraction withendoplasmic reticulum since the specific activities of this enzyme by the 2 membranes were quite similar. This was notsupported by electron-microscopic observation. In considera

tion of recent reports that electron transport activity has beenfound in the plasma membranes of liver (13) and Ehrlich ascitestumor cells (19), an inherent electron transport activity couldnot be excluded from human tumor plasma membranes, (c)Electron micrographs of these membrane fractions showed thepresence of a few lysosomes. Determination of acid phospha-

tase, a marker enzyme for lysosomes, was low in the membranes of 3 tumors (T293, T242, and M21) and high in theplasma membranes of astrocytoma (T24) and melanoma(T355). In the latter instances, the specific activities weresimilar to or higher than that obtained with purified mouse liverlysosomes. Such high acid phosphatase activities certainly

OCTOBER 1981 4035




could not be accounted for by lysosomal contamination, aselectron micrographs of these membranes showed that lysosomal contamination was very slight. We have also observed ahigh phosphatase activity at acid pH's in intact cells establishedin culture from astrocytoma.3 These preliminary results sug

gested that, at least in the astrocytoma, there was an activeacid phosphatase on the outer surface of the cells.

Our experience with the use of marker enzymes in evaluatingthe purity of membrane fractions has shown that some of thestandard membrane marker enzymes had different distributionpatterns in different cells. Thus, extreme caution needs to beexercised in interpreting results obtained from marker enzymemeasurements alone.

Among the enzymes we examined, the Mg2+-ATPase

showed much higher specific activity than all the other enzymes. This activity was apparently the main contributor to thetotal cellular ATPase activity, accounting for the low inhibitionof the ATPase in the tumor homogenates by oligomycin, anmitochondrial ATPase inhibitor (20). The abundance of thisenzyme in tumor cells was a point of great interest. From itslack of substance selectivity and its ability to utilize Ca2+ aswell as Mg2+ as the divalent ion activator, this enzyme was

similar to the cell surface ATPase (17, 30, 31) or CTPase (26)found on the surface of many different cells. In the 5 tumorsstudied, this activity was so strong as to render the evaluationof the Na+-K*-ATPase difficult, the presence of which was

nevertheless established definitively by the capacity of themembranes to bind [3H]ouabain (Table 3). The fact that theNa+-K+-ATPase was overwhelmed by Mg2+-ATPase in the

astrocytoma was striking. Astrocytoma was of glial origin, andthe Na+, K+-ATPase usually accounted for 50% of the total

ATPase activity. We found that ouabain inhibition of the plasmamembrane ATPase prepared from a rat glioma line (C6) wasalso 40 to 50% (data not shown) whereas the inhibition of theplasma membrane ATPase of astrocytoma by ouabain was lessthan 20%. In addition to the reduced sensitivity to ouabain,astrocytoma (T24) also showed a decreased response to thecatecholamines when compared to the C6 cell line." Thus, theexcessive Mg2+-ATPase activity of the astrocytoma might be

of significance in either dedifferentiation or neoplastic transformation.

The role of the cell surface ATPase in the physiology of thecell has not been established. However, Karasaki and Okigaki(17) and Karasaki ef al. (18) have correlated the presence ofa cell surface ATPase with high Km and Vmaxon liver epithelialcells with oncogenicity. The possibility that the high ATPaseactivities in the plasma membranes of the human tumors wasalso closely correlated with tumorigenicity should be seriouslyconsidered.

Is the same enzyme molecule responsible for the ATPaseactivity in the plasma membranes of all 5 tumors? This questioncannot be answered with certainty. We observed that theplasma membrane ATPase of astrocytoma and oat cell carcinoma was stimulated considerably by concanavalin A whereasthe ATPase of the 3 melanomas showed no response (Chart3). This could not be due to a lack of interaction of the lectinwith the membranes since the 5'-nucleotidase of all 5 mem

branes was inhibited by concanavalin A (Chart 4). These results

3 A. F. Knowles, J. Abare, and N. O. Kaplan, unpublished observations.4 G. Beattie and N. O. Kaplan, unpublished observations.

would suggest that either the ATPase molecules themselvesdiffer or that the membrane environment of the ATPase wassignificantly different in the various tumors.

In conclusion, the present study has pointed out the diversityof the plasma membranes from different tumors and betweentumors of the same type. These were exemplified by (a) thevery low 5'-nucleotidase activity of the oat cell carcinomaplasma membrane, (b) large amounts of NAD+ pyrophospha-

tase in the plasma membranes of melanomas (T242), (c) largeamounts of acid phosphatase in the plasma membranes ofastrocytoma and melanoma (T355), and (oOvariable responseof the Mg+-ATPase to concanavalin A. With the purified plasma

membranes from these human tumors available to us, we shallbe able to examine some of the enzymes of interest in greaterdetail.

ACKNOWLEDGMENTS

We thank the staff of the Athymic Mice Facility for support. Beverly Kelly forperforming electron microscopy, Neil Talbert for transplanting the tumors, andJon Dickinson for help with the subcellular membrane preparations. Dr. Knowleswishes to thank Dr. George Fortes for suggesting the [3H]ouabain-binding exper

iments.

REFERENCES

1. Aronson. N. N.. Jr., and Touster, O. Isolation of rat liver plasma membranefragments in isotonic sucrose. Methods Enzymol., 37. 90-102, 1974.

2. Bock, K. W., Siekevitz, P., and Palade, G. E. Localization and turnoverstudies of membrane nicotinamide adenine dinucleotide glycohydrolase inrat liver. J. Biol. Chem., 246. 188-195. 1979.

3. Brake. E. T., Will, P. C.. and Cook, J. Characterization of HeLa 5'-Nucleo-tidase: a stable plasma membrane marker. Membr. Biochem., 2. 17-46,1978.

4. Chatterjee, S. K., Kim, U., and Bielat. K. Plasma membrane associatedenzymes of mammary tumors as the biochemical indicators of metastasizingcapacity. Analyses of enriched plasma membrane preparations. Br. J. Cancer, 33. 15-26, 1976.

5. Ciotti, M. M., and Kaplan, N. O. Procedures for determination of pyridinenucleotides. Methods Enzymol., 3. 890-892. 1957.

6. Collett. M.. and Erikson, R. Protein kinase activity associated with the aviansarcoma virus src gene product. Proc. Nati. Acad. Sei. U. S. A., 75: 2021-2025, 1978.

7. Colowick, S. P., Kaplan, N. O., and Ciotti, M. M. The reaction of pyridinenucleotide with cyanide and its anaytical use. J. Biol. Chem., 19Ã•.447-459,1951.

8. Emmelot, P., and Bos, C. J. Studies on plasma membranes. IX. A survey ofenzyme activities displayed by plasma membranes isolated from normal andpreneoplastic livers and primary and transplanted hepatomas of the rat. Int.J. Cancer, 4. 705-722, 1969.

9. Emmelot, P., and Bos, C. J. Studies on plasma membranes. X. A survey ofenzyme activities displayed by plasma membranes isolated from mouse liverand three mouse hepatoma strains. Int. J. Cancer, 4. 723-734, 1969.

10. Emmelot, P.. and Bos. C. J. Studies on plasma membranes. XVII. On thechemical composition of plasma membranes prepared from rat and mouseliver and hepatomas. J. Membr. Biol., 9. 83-104, 1972.

11. Fleischer, S., and Kervina, M. Long-term preservation of liver for subcellularfractionations. Methods Enzymol., 31: 3-41, 1974.

12. Giuliani, F. C.. Zirvi, K. A., and Kaplan, N. O. Therapeutic response of humantumor xenografts in athymic mice to doxorubicin. Cancer Res.. 4). 4682-4687, 1980.

13. Goldenburg, H., Crane, F. L., and MorrÃ©.D. J. NADH oxidoreductase ofmouse liver plasma membranes. J. Biol. Chem., 254. 2491-2498, 1979.

14. Hynes, R. O. Cell surface proteins and malignant transformation. Biochim.Biophys. Acta, 458. 73-107, 1976.

15. Jacobson, K. B., and Kaplan, N. O. Distribution of enzymes cleaving pyridinenucleotides in animal tissue. J. Biophys. Biochem. Cytol., 3. 31-43, 1957.

16. Jarret, L., and Smith, R. M. The stimulation of adipocyte plasma membranemagnesium ion-stimulated adenosine triphosphatase by insulin and concanavalin A. J. Biol. Chem., 249. 5195-5199, 1974.

17. Karasaki, S., and Okigaki. T. Surface membrane nucleoside triphosphataseactivity and tumorigenicity of cultured liver epithelial cells. Cancer Res., 36:4491-4499, 1976.

18. Karasaki, S., Suh, M. H., Salas, M., and Raymond, J. Cell surface adenosine5'-triphosphatase as an in vitro marker of the lineage and cytodifferentiation





of oncogenic epithelial cells from rat liver parenchyma. Cancer Res., 40:1318-1328, 1980.

19. Kilberg, M. S., and Christensen, H. N. Electron transferring enzymes in theplasma membrane of the Ehrlich ascites tumor cell. Biochemistry, 18:1525-

1530, 1979.20. Knowles, A. F., and Kaplan, N. O. Oxidative phosphorylation and ATPase

activities of human tumor mitochondria. Biochim. Biophys. Acta, 590: 170-

181, 1980.21. Krueger, J. G., Wang, E., and Goldberg, A. R. Evidence that the src gene

product of Rous sarcoma virus is membrane associated. Virology. 101: 25-

40, 1980.22. Laurent, G., Doriaux. M., and Hildebrand, J. Isolation of plasma membranes

from murine ependymoblastoma and subcellular distribution of amphotericinB in this tumor. Biochim. Biophys. Acta, 466. 123-135, 1977.

23. Lohmann, K., and Jendrassik, L. Kolonmetrische PhosphorsÃ¤urebestimmu-gen in Muskelextrakt. Biochem. Z., J78: 419-426, 1926.

24. Matsui, H., and Schwartz, A. Mechanism of cardiac glycoside inhibition ofthe (Na*-K*)-dependent ATPase from cardiac tissue. Biochim. Biophys.

Acta, Ã•5Ã•.655-663, 1968.25. Nicolson, G. L. Trans-membrane control of the receptors on normal and

tumor cells. II. Surface changes associated with transformation and malignancy. Biochim. Biophys. Acta, 458. 1-72, 1976.

26. Perdue, J. I., and Sneider, J. The isolation and characterization of theplasma membrane from chick embryo fibroblasts. Biochim. Biophys. Acta,Ã•96: 125-140, 1970.

27. Pommier, G., Ripert, G., Azoulay. E., and Depieds, R. Effect of concanavalinA on membrane-bound enzymes from mouse lymphocytes. Biochim. Biophys. Acta, 389. 483-494, 1975.

28. Reid, L. M.. Holland, J., Jones, C., Wolf, B., Niwayama, C. T., Williams, R..Kaplan, N. 0., and Sato. G. Some of the variables affecting the success oftransplantation of human tumors into the athymic nude mouse. In: 0. P.Houchens and A. A. Ovejera (eds.), Proceedings of the Symposium on theUse of Athymic (Nude) Mice in Cancer Research, pp. 107-122. New York:

Gustav Fisher, 1978.29. Riordan, F. R., Slavik, M., and Kartner, V. Nature of the lectin-induced

activation of plasma membrane Mg2* ATPase. J. Biol. Chem., 252: 5449-

5455, 1977.30. Ronquist, G., and Agren. G. K. A Mg2* and Ca2*-stimulated adenosine

triphosphatase at the outer surface of Ehrlich ascites tumor cells. CancerRes.. 35: 1402-1406, 1975.

31. Trams, E. G., and Leuter, C. J. On the sidedness of plasma membraneenzymes. Biochim. Biophys. Acta, 345: 180-197, 1974.

32. Vaheri, A. Surface proteins of virus-transformed cells. In: C. Nicolau (ed.).Virus-transformed Cell Membranes, pp. 1-87. New York: Academic Press,Inc., 1978.

33. Van Hoeven, R. P., and Emmelot, P. Studies on plasma membranes. XVIII.Lipid class composition of plasma membranes isolated from rat mouse liverand hepatomas. J. Membr. Biol., 9: 105-126, 1972.

34. Warren, L. The thiobarbituric acid assay of sialic acids. J. Biol. Chem., 234.1971-1976, 1959.

OCTOBER 1981 4037




N! 'V ; 4 i\ Â¿' ''r v- :

rcSÃ¤ierra-.eiSafvÂ«f *

W- 4T>Â»i

Jity, ' .

' -*:Fig. 1. Electron micrographs of tumor plasma membrane and endoplasmic reticulum fractions. Plasma membranes were from: astrocytoma (a): oat cell carcinoma

0; melanoma (T242) (c); melanoma (T355) (d); melanoma (M21 ) (e) and endoplasmic reticulum from melanoma (T242) ( f). X 25,700.

/ r>




1981;41:4031-4038. Cancer Res Aileen F. Knowles, Jose F. Leis and Nathan O. Kaplan MelanomasTransplantable Human Astrocytoma, Oat Cell Carcinoma, and Isolation and Characterization of Plasma Membranes from

Updated version

http://cancerres.aacrjournals.org/content/41/10/4031

Access the most recent version of this article at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/41/10/4031To request permission to re-use all or part of this article, use this link


http://cancerres.aacrjournals.org/content/41/10/4031http://cancerres.aacrjournals.org/cgi/alertsmailto:[email protected]://cancerres.aacrjournals.org/content/41/10/4031http://cancerres.aacrjournals.org/

Documents

Isolation and Characterization of Plasma Membranes from ......T355, and M21) contained the purified endoplasmic reticulum. The collected membrane fractions were diluted 2- to 3-fold