Can. J. Biochem. 1977.55:1233-1236. Downloaded from www.nrcresearchpress.com by UNIVERSITY OF MICHIGAN on 11/12/14. For personal use only.
NOTES
An improved procedure for the purification of plasma membranes from Dictyostelium discoideum Department of Microbiology, University of British Columbia, Vancouver, B.C., Canada V6T I W5 Received May 12, 1977 Accepted October 4, 1977 GILKES,N. R. & WEEKS,G. (1977) An improved procedure for the purification of plasma membranes from Dictyostelium discoideum. Can. J . Biochem. 55, 1233-1236 A novel procedure was recently described for the purification of plasma membranes of Dictyostelium discoideum (Gilkes, N. R. & Weeks, G. (1977) Biochim. Biophys. Acta 464, 142-156). Considerable enrichment of plasma membrane marker enzymes was achieved, but since purified mitochondrial and endoplasmic reticulum fractions were unavailable, it was not possible to accurately assess the contamination level of these organelles. We have therefore slightly modified the plasma membrane preparation procedure, improving purification, and have prepared partially purified mitochondrial and endoplasmic reticulum fractions. The data suggest that the contamination of the plasma membranes by endoplasmic reticulum membranes is no greater than lo%, and probably considerably less. No mitochondrial contamination is detectable.
Introduction
Highly purified Dictyostelium discoideum plasma membranes are desirable for the study of the molecular changes that occur at the cell surface during the differentiation of this organism. A novel method was recently described for the purification of plasma membranes from D. discoideum (1). Analysis of the plasma membrane marker enzymes, alkaline phosphatase and 5'-nucleotidase, indicated that considerable purification had been attained. Furthermore, there was only slight mitochondrial contamination as assessed by measuring succinate dehydrogenase activity. However, evaluation of possible contamination of our plasma membrane preparation by endoplasmic reticulum membranes proved difficult because pure preparations of endoplasmic reticulum were unavailable for comparison. Furthermore, the NADPH-cytochrome c reductase, an endoplasmic reticulum enzyme in rat liver (2), appeared to be partially located in the mitochondria of D. discoideum (I), complicating the estimation of contamination. This study was undertaken firstly as an attempt to improve the original plasma membrane purification procedure by inclusion of a differential centrifugation step, and secondly to prepare fractions enriched for endoplasmic reticulum and mitochondrial memABBREVIATIONS: PMSF, phenylmethylsulphonyl fluoride; sucrose-Tris-PMSF, 8.6% (wlv) sucrose, 5 mM Tris-C1 (pH 7.4) PMSF-saturated buffer. 'Author to whom reprint requests should be addressed.
branes in order to provide a more reliable estimate of the level of internal membrane contamination. Materials and Methods Cells of Dictyostelium discoideum, Ax-2, were grown in HL5 medium, and harvested and washed as described previously (1). The method of preparing the plasma membranes was similar to that described previously (I), except that a differential centrifugation step was included in the procedure. Washed cells (= 2 x lot0)were resuspended in sucroseTris-PMSF at a cell density of lo8 cells/ml. Cells were broken by vigorous stimng in the presence of glass beads (I), and the homogenate centrifuged at 700 g for 10 min to remove unbroken cells (Fig. 1). The 700 g supernatant was centrifuged at 8200 rpm for 10 min in a Sorval SS34 rotor and the resulting pellet (8K pellet) resuspended in sucrose-Tris-PMSF. The supernatant was recentrifuged at 30000 rpm for 30 min in a Spinco type 30 rotor and the resulting pellet (30K pellet) resuspended in sucroseTris-PMSF (Fig. 1). The resuspended 8K and 30K pellets were layered onto separate discontinuous sucrose gradients as described previously (1) except that no EDTA was added. Three gradients were used for each pellet. The gradients were centrifuged at 24 000 rpm for 16 h in a Spinco SW27 rotor. Two distinct bands of membranous material (PMl and PM2) and a bottom fraction (BF) were obtained for both samples (Fig. 1). Gradient fractions were collected and processed as described previously (1). Alkaline phosphatase, 5'-nucleotidase, succinate dehydrogenase, and NADPH-cytochrome c reductase were assayed by previously published procedures (3, 4). Glucose-6-phosphatase was assayed as described by Hiibscher and West (5) with inhibition of alkaline and acid
C A N . J . BIOCHEM. VOL. 5 5 . 1977
HOMOGENATE OF
Can. J. Biochem. 1977.55:1233-1236. Downloaded from www.nrcresearchpress.com by UNIVERSITY OF MICHIGAN on 11/12/14. For personal use only.
I
D. discoideum
CELLS
C e n t r i f u g e d , 700 g , 1 0 min
PELLET (unbroken c e l l s ) discarded
LOW SPEED SUPERNATANT
I C e n t r i f u g e d , 8200 rpm, 1 0 min
I
I
8K PELLET
SUPERNATANT
I Resuspended . Applied
I
t o sucrose gradient, c e n t r i f u g e d , 24 000 rprn,l6 h
C e n t r i f u g e d , 30 000 rpm, 30 min
M
I 4
J
30K PELLET
\\\\\\8K
I
39.3% 8K PM2
y \\\\\ b
41.0Z
44.0%
Resuspended. Applied t o sucrose gradient (composition a s f o r 8K p e l l e t ) . Centrifugeu, 24 000 rpm, 16 h
r
I
.\\\\\\
8K BF (bottom f r a c t i o n )
SUPERNATANT discarded
x \\,
I
30K pH1 30K PH2
L
30K BF FIG. 1 . Schematic representation of the preparation of various membrane fractions from Dicfyostelium discoideum homogenate.
1235
NOTES
Can. J. Biochem. 1977.55:1233-1236. Downloaded from www.nrcresearchpress.com by UNIVERSITY OF MICHIGAN on 11/12/14. For personal use only.
TABLE1. Distribution of enzyme activities during the purification of plasma membranes of Dictyostelium discoideum
Fractiona
Protein recovered, mg
Alkaline phosphatase Specificb Yield,' % activity
5'-Nucleotidase
Succinate dehydrogenase
NADPHcytochrome c reductase
Specific Yield, activity %
Specific Yield, activity %
Specific Yield, activity %
Glucose-6phosphatase Specific, activity
-
Yield,
%
Low-speed supernatant 1006.3 8K pellet 284.4 30K pellet 175.0 3.5 8K PM1 8K PM2 7.1 8K BF 38.5 30K PMl 7.2 30K PM2 4.5 30K BF 19.0 NOTE:ND, not detectable. °For fraction designation, see Fig. 1 and Materials and Methods. bspecific activities are given as nanomoles per minute per milligram protein. cYields are the enzyme activity of each fraction as a percent of the original total cell-free extract (low-speed supernatant) activity.
phosphatase activities by 5 mM EDTA and 2 mM KF respectively. Approximately 1 mg of rmmbrane protein was required for each assay and incubations were conducted for 20 min at 37°C. Liberated phosphate was determined by the method of Ames (6). Since both substrate and enzyme contained appreciable free phosphate, appropriate controls were run with each assay.
Results
Our original preparative procedure for plasma membranes ( I ) has been modified firstly by inclusion of a differential centrifugation step, and secondly by omission of EDTA from the sucrose gradients. The differential centrifugation step involved the centrifugation of the cell-free extract (low-speed supernatant) at 8200 rpm to produce a 'mitochondrial' pellet (8K pellet) followed by the centrifugation of the 8200-rpm supernatant at 30 000 rpm to produce a 'microsomal' pellet (30K pellet). The plasma membrane marker enzymes, alkaline phosphatase and 5'-nucleotidase, were more enriched in the 30K pellet than in the 8K pellet, and the levels of the contaminating membrane enzymes were lower (Table 1). The 8K-pellet and 30K-pellet fractions were then applied to separate sucrose density gradients without EDTA. The plasma membrane enrichment of the 30K PMl fraction on the basis of alkaline phosphatase activity was approximately, 58-fold, while as assessed by 5'-nucleotidase activity, it was 36-fold, both considerably greater than that observed in the original study (1). The presence of EDTA in the sucrose gradients inhibits the activities of both alkaline phosphatase and 5'-nucleotidase (unpublished observations), which in part explains the lower enrichment values reported in the earlier study. The improved enrichment is not due to differences between the specific activities of the marker enzymes in the cell homogenate of the original study and those activities of the cell-free extract (Table 1).
We have routinely found that the specific activities of alkaline ~hosvhataseand 5'-nucleotidase are similar in cell homogenates and cell-free extracts (11, suggesting that the low-speed centrifugation does non-plasma membrane 'Onstituents. In the experiment shown in Table 1, all the material recovered from the 30K gradients was assayed. The recovery of 5'-nucleoiidase was loo%, while the total alkaline phosphatase activity recovered was found to be approximately 1.6 times greater than that applied (data not shown), suggesting an activation of this enzyme as a result of the gradient centrifugation. Thus, the very high enrichment of the PMl fraction as assessed by alkaline phosphatase activity is almost certainly overestimated. The 30K PMl fraction has slightly higher enrichments of the plasma membrane marker enzymes and also lower levels of succinate dehydrogenase, NADPH-cytochrome c reductase, and glucose-6phosphatase than the 8K PMl fraction, suggesting that the removal of the 8200-rpm pelletable material prior to the layering on the sucrose gradient produces a purer plasma membrane preparation. The attempt to produce enriched mitochondrial and endoplasmic reticulum fractions by a differential centrifugation step prior to the layering on sucrose gradient was partially successful. The presumptive mitochondrial fraction (8K pellet) and the presumptive endoplastic reticulum fraction (30K pellet) were further purified by the sucrose density centrifugation procedure (Fig. 1). The mitochondrial fraction (8K BF) was considerably enriched for succinate dehydrogenase and the endoplasmic reticulum fraction (30K BF) was enriched for NADPH-cytochrome c reductase. There was however considerable cross contamination. The 8K BF was almost eight-fold enriched for succinate dehydrogenase, but was also five-fold enriched for NADPH-cytochrome c reductase; the 30K BF was
1236
CAN. J. BIOCHEM. VOL. 55, 1977
Can. J. Biochem. 1977.55:1233-1236. Downloaded from www.nrcresearchpress.com by UNIVERSITY OF MICHIGAN on 11/12/14. For personal use only.
enriched five-fold for NADPH-cytochrome c reductase, but also three-fold for the succinate dehydrogenase. These relatively low enrichment values for the succinate dehydrogenase and NADPHcytochrome c reductase can be explained in part by the inactivation of these enzymes during centrifugation. All the minor fractions were collected and assayed, and it was found that only approximately 40% of the original activity was recoverable (data not shown). Since there were indications that the NADPHcytochrome c reductase was not exclusively located on the endoplasmic reticulum of D. discoideum (I), glucose-6-phosphatase, an endoplasmic reticulum enzyme of at least some mammalian cells (7), was also assayed. However, since the 8K pellet and 8K BF material had higher specific activities and recoveries than the 30K pellet and the 30K BF material, it appears that this enzyme is also a poor marker for the endoplasmic reticulum of D. discoideum, a result that is consistent with other recent findings (8, 9). The recovery of membrane-bound glucose-6phosphatase activity from the sucrose density gradients was also low. Discussion
The plasma membrane purification procedure described in this report is only a slight modification of that already published (1). However, the inclusion of the differential centrifugation step prior to the layering on the sucrose gradient improves the purity of the resulting plasma membrane preparation, although the membrane yield is almost halved. Therefore this additional step is only recommended if the purity of the preparation is crucial. In order to accurately measure the levels of internal membrane contamination in purified plasma membrane preparations, pure preparations of the internal membrane are needed. Thus, the ratio of the specific activity of the marker enzyme in the pure plasma membrane to the specific activity of the same enzyme in a pure preparation of the internal membrane accurately represents the proportion of contaminating internal membrane in the plasma membrane preparation. However, partially pure preparations of internal membranes can be used for the estimation of the maximum levels of contamination. Thus, for NADPH-cytochrome c reductase, the ratio of the specific activity of the purest plasma membrane fraction (1 .O nmol/(min/mg protein)) to that of the purest endoplasmic reticulum fraction (33.0 nmol/(min/mg protein)) indicates a maximum contamination of only 3%. Similarly for glucose-6phosphatase, the ratio of the specific activity of the plasma membrane fraction (2.6 nmol/(min/mg protein)) to that of the fraction most enriched for this enzyme (26.6 nmol/(min/mg protein)) indicates a maximum contamination of approximately 10%. However, since the 8K BF material is also highly enriched for succinate dehydrogenase, the mem-
brane containing glucose-6-phosphatase is far from pure and the 10% contamination value is an overestimation. Since the plasma membranes contain no detectable succinate dehydrogenase (Table 1) or cytochrome oxidase (unpublished observations), the mitochondria1contamination is negligible. Other data are consistent with the idea that the contamination of the plasma membranes by internal membranes containing either glucose-6-phosphatase or NADPH-cytochrome c reductase is at most 10% and probably considerably less. Thus, the specific activities of NADPH-cytochrome c reductase and glucose-6-phosphatase in the 30K PM2 fraction are twice as high as those in the 30K PM 1 fraction. If these components represented large levels of contamination, then the specific activities of alkaline phosphatase and 5'-nucleotidase would be considerably lower in the PM2fraction than in the PM1 fraction. For example, if the contamination due to these components represented 20% of the 30K PM1 plasma membrane preparation, then the contamination of the 30K PM2 fraction should be 40%. Thus, the proportion of genuine plasma membrane in the 30K PMl and 30K PM2 fractions would be 80% and 60% respectively, and the ratios of the specific activities of alkaline phosphatase and 5'nkleotidase of the 30K PM1 fraction to that of the PM2 would be 1.33. In fact, the ratio is actually 1.03 for alkaline phosphatase and 1.12for 5'-nucleotidase (Table l), suggesting a maximum contamination in the PM1 fraction of between 3% and 10%. These data are therefore also consistent with the idea that the endoplasmic reticulum contamination is at most 10% and probably considerably less. A similar argument can also be made with regard to our previously published data (1). Acknowledgements
We wish to thank Kathy LaRoy for skilled technical assistance and Dr. Claire Weeks for critically reading the manuscript. The work was supported by a grant from the National Cancer Institute of Canada. 1. Gilkes, N. R. & Weeks, G. (1977) Biochim. Biophys. Acta 46#, 142-156 2. Phillips, A. H. & Langdon, R. G. (1962) J. Biol. Chem. 237,2652-2660 3. Lee, A., Chance, K., Weeks, C. & Weeks, G. (1975) Arch. Biochem. Biophys. 171,407-417 4. Scottocasa, G. L., Kuylenstiena, B., Ernster, L. & Bergstrand, A. (1967) J. Cell. Biol. 32,415-438 5. Hiibscher, G. & West, G. R. (1965) Nature (London) 205,799-800 6. Ames, B. N. (1966) Methods in Enzymology (Neufeld, E. F . & Ginsburg, V., eds.), vol. 8 , pp. 115-118, Academic Press, New York 7. Goldfischer, S., Essner, E. & Novikoff, A. B. (1964) J. Histochem. Cytochem. 12,72-95 8. Green, A. A. & Newell, P. C. (1974) Biochem. J. 140, 3 13-322 9. McMahon, D., Miller, M. & Long S. (1977) Biochim. Biophys. Acta 465,224-24 1
NOTES
An improved procedure for the purification of plasma membranes from Dictyostelium discoideum Department of Microbiology, University of British Columbia, Vancouver, B.C., Canada V6T I W5 Received May 12, 1977 Accepted October 4, 1977 GILKES,N. R. & WEEKS,G. (1977) An improved procedure for the purification of plasma membranes from Dictyostelium discoideum. Can. J . Biochem. 55, 1233-1236 A novel procedure was recently described for the purification of plasma membranes of Dictyostelium discoideum (Gilkes, N. R. & Weeks, G. (1977) Biochim. Biophys. Acta 464, 142-156). Considerable enrichment of plasma membrane marker enzymes was achieved, but since purified mitochondrial and endoplasmic reticulum fractions were unavailable, it was not possible to accurately assess the contamination level of these organelles. We have therefore slightly modified the plasma membrane preparation procedure, improving purification, and have prepared partially purified mitochondrial and endoplasmic reticulum fractions. The data suggest that the contamination of the plasma membranes by endoplasmic reticulum membranes is no greater than lo%, and probably considerably less. No mitochondrial contamination is detectable.
Introduction
Highly purified Dictyostelium discoideum plasma membranes are desirable for the study of the molecular changes that occur at the cell surface during the differentiation of this organism. A novel method was recently described for the purification of plasma membranes from D. discoideum (1). Analysis of the plasma membrane marker enzymes, alkaline phosphatase and 5'-nucleotidase, indicated that considerable purification had been attained. Furthermore, there was only slight mitochondrial contamination as assessed by measuring succinate dehydrogenase activity. However, evaluation of possible contamination of our plasma membrane preparation by endoplasmic reticulum membranes proved difficult because pure preparations of endoplasmic reticulum were unavailable for comparison. Furthermore, the NADPH-cytochrome c reductase, an endoplasmic reticulum enzyme in rat liver (2), appeared to be partially located in the mitochondria of D. discoideum (I), complicating the estimation of contamination. This study was undertaken firstly as an attempt to improve the original plasma membrane purification procedure by inclusion of a differential centrifugation step, and secondly to prepare fractions enriched for endoplasmic reticulum and mitochondrial memABBREVIATIONS: PMSF, phenylmethylsulphonyl fluoride; sucrose-Tris-PMSF, 8.6% (wlv) sucrose, 5 mM Tris-C1 (pH 7.4) PMSF-saturated buffer. 'Author to whom reprint requests should be addressed.
branes in order to provide a more reliable estimate of the level of internal membrane contamination. Materials and Methods Cells of Dictyostelium discoideum, Ax-2, were grown in HL5 medium, and harvested and washed as described previously (1). The method of preparing the plasma membranes was similar to that described previously (I), except that a differential centrifugation step was included in the procedure. Washed cells (= 2 x lot0)were resuspended in sucroseTris-PMSF at a cell density of lo8 cells/ml. Cells were broken by vigorous stimng in the presence of glass beads (I), and the homogenate centrifuged at 700 g for 10 min to remove unbroken cells (Fig. 1). The 700 g supernatant was centrifuged at 8200 rpm for 10 min in a Sorval SS34 rotor and the resulting pellet (8K pellet) resuspended in sucrose-Tris-PMSF. The supernatant was recentrifuged at 30000 rpm for 30 min in a Spinco type 30 rotor and the resulting pellet (30K pellet) resuspended in sucroseTris-PMSF (Fig. 1). The resuspended 8K and 30K pellets were layered onto separate discontinuous sucrose gradients as described previously (1) except that no EDTA was added. Three gradients were used for each pellet. The gradients were centrifuged at 24 000 rpm for 16 h in a Spinco SW27 rotor. Two distinct bands of membranous material (PMl and PM2) and a bottom fraction (BF) were obtained for both samples (Fig. 1). Gradient fractions were collected and processed as described previously (1). Alkaline phosphatase, 5'-nucleotidase, succinate dehydrogenase, and NADPH-cytochrome c reductase were assayed by previously published procedures (3, 4). Glucose-6-phosphatase was assayed as described by Hiibscher and West (5) with inhibition of alkaline and acid
C A N . J . BIOCHEM. VOL. 5 5 . 1977
HOMOGENATE OF
Can. J. Biochem. 1977.55:1233-1236. Downloaded from www.nrcresearchpress.com by UNIVERSITY OF MICHIGAN on 11/12/14. For personal use only.
I
D. discoideum
CELLS
C e n t r i f u g e d , 700 g , 1 0 min
PELLET (unbroken c e l l s ) discarded
LOW SPEED SUPERNATANT
I C e n t r i f u g e d , 8200 rpm, 1 0 min
I
I
8K PELLET
SUPERNATANT
I Resuspended . Applied
I
t o sucrose gradient, c e n t r i f u g e d , 24 000 rprn,l6 h
C e n t r i f u g e d , 30 000 rpm, 30 min
M
I 4
J
30K PELLET
\\\\\\8K
I
39.3% 8K PM2
y \\\\\ b
41.0Z
44.0%
Resuspended. Applied t o sucrose gradient (composition a s f o r 8K p e l l e t ) . Centrifugeu, 24 000 rpm, 16 h
r
I
.\\\\\\
8K BF (bottom f r a c t i o n )
SUPERNATANT discarded
x \\,
I
30K pH1 30K PH2
L
30K BF FIG. 1 . Schematic representation of the preparation of various membrane fractions from Dicfyostelium discoideum homogenate.
1235
NOTES
Can. J. Biochem. 1977.55:1233-1236. Downloaded from www.nrcresearchpress.com by UNIVERSITY OF MICHIGAN on 11/12/14. For personal use only.
TABLE1. Distribution of enzyme activities during the purification of plasma membranes of Dictyostelium discoideum
Fractiona
Protein recovered, mg
Alkaline phosphatase Specificb Yield,' % activity
5'-Nucleotidase
Succinate dehydrogenase
NADPHcytochrome c reductase
Specific Yield, activity %
Specific Yield, activity %
Specific Yield, activity %
Glucose-6phosphatase Specific, activity
-
Yield,
%
Low-speed supernatant 1006.3 8K pellet 284.4 30K pellet 175.0 3.5 8K PM1 8K PM2 7.1 8K BF 38.5 30K PMl 7.2 30K PM2 4.5 30K BF 19.0 NOTE:ND, not detectable. °For fraction designation, see Fig. 1 and Materials and Methods. bspecific activities are given as nanomoles per minute per milligram protein. cYields are the enzyme activity of each fraction as a percent of the original total cell-free extract (low-speed supernatant) activity.
phosphatase activities by 5 mM EDTA and 2 mM KF respectively. Approximately 1 mg of rmmbrane protein was required for each assay and incubations were conducted for 20 min at 37°C. Liberated phosphate was determined by the method of Ames (6). Since both substrate and enzyme contained appreciable free phosphate, appropriate controls were run with each assay.
Results
Our original preparative procedure for plasma membranes ( I ) has been modified firstly by inclusion of a differential centrifugation step, and secondly by omission of EDTA from the sucrose gradients. The differential centrifugation step involved the centrifugation of the cell-free extract (low-speed supernatant) at 8200 rpm to produce a 'mitochondrial' pellet (8K pellet) followed by the centrifugation of the 8200-rpm supernatant at 30 000 rpm to produce a 'microsomal' pellet (30K pellet). The plasma membrane marker enzymes, alkaline phosphatase and 5'-nucleotidase, were more enriched in the 30K pellet than in the 8K pellet, and the levels of the contaminating membrane enzymes were lower (Table 1). The 8K-pellet and 30K-pellet fractions were then applied to separate sucrose density gradients without EDTA. The plasma membrane enrichment of the 30K PMl fraction on the basis of alkaline phosphatase activity was approximately, 58-fold, while as assessed by 5'-nucleotidase activity, it was 36-fold, both considerably greater than that observed in the original study (1). The presence of EDTA in the sucrose gradients inhibits the activities of both alkaline phosphatase and 5'-nucleotidase (unpublished observations), which in part explains the lower enrichment values reported in the earlier study. The improved enrichment is not due to differences between the specific activities of the marker enzymes in the cell homogenate of the original study and those activities of the cell-free extract (Table 1).
We have routinely found that the specific activities of alkaline ~hosvhataseand 5'-nucleotidase are similar in cell homogenates and cell-free extracts (11, suggesting that the low-speed centrifugation does non-plasma membrane 'Onstituents. In the experiment shown in Table 1, all the material recovered from the 30K gradients was assayed. The recovery of 5'-nucleoiidase was loo%, while the total alkaline phosphatase activity recovered was found to be approximately 1.6 times greater than that applied (data not shown), suggesting an activation of this enzyme as a result of the gradient centrifugation. Thus, the very high enrichment of the PMl fraction as assessed by alkaline phosphatase activity is almost certainly overestimated. The 30K PMl fraction has slightly higher enrichments of the plasma membrane marker enzymes and also lower levels of succinate dehydrogenase, NADPH-cytochrome c reductase, and glucose-6phosphatase than the 8K PMl fraction, suggesting that the removal of the 8200-rpm pelletable material prior to the layering on the sucrose gradient produces a purer plasma membrane preparation. The attempt to produce enriched mitochondrial and endoplasmic reticulum fractions by a differential centrifugation step prior to the layering on sucrose gradient was partially successful. The presumptive mitochondrial fraction (8K pellet) and the presumptive endoplastic reticulum fraction (30K pellet) were further purified by the sucrose density centrifugation procedure (Fig. 1). The mitochondrial fraction (8K BF) was considerably enriched for succinate dehydrogenase and the endoplasmic reticulum fraction (30K BF) was enriched for NADPH-cytochrome c reductase. There was however considerable cross contamination. The 8K BF was almost eight-fold enriched for succinate dehydrogenase, but was also five-fold enriched for NADPH-cytochrome c reductase; the 30K BF was
1236
CAN. J. BIOCHEM. VOL. 55, 1977
Can. J. Biochem. 1977.55:1233-1236. Downloaded from www.nrcresearchpress.com by UNIVERSITY OF MICHIGAN on 11/12/14. For personal use only.
enriched five-fold for NADPH-cytochrome c reductase, but also three-fold for the succinate dehydrogenase. These relatively low enrichment values for the succinate dehydrogenase and NADPHcytochrome c reductase can be explained in part by the inactivation of these enzymes during centrifugation. All the minor fractions were collected and assayed, and it was found that only approximately 40% of the original activity was recoverable (data not shown). Since there were indications that the NADPHcytochrome c reductase was not exclusively located on the endoplasmic reticulum of D. discoideum (I), glucose-6-phosphatase, an endoplasmic reticulum enzyme of at least some mammalian cells (7), was also assayed. However, since the 8K pellet and 8K BF material had higher specific activities and recoveries than the 30K pellet and the 30K BF material, it appears that this enzyme is also a poor marker for the endoplasmic reticulum of D. discoideum, a result that is consistent with other recent findings (8, 9). The recovery of membrane-bound glucose-6phosphatase activity from the sucrose density gradients was also low. Discussion
The plasma membrane purification procedure described in this report is only a slight modification of that already published (1). However, the inclusion of the differential centrifugation step prior to the layering on the sucrose gradient improves the purity of the resulting plasma membrane preparation, although the membrane yield is almost halved. Therefore this additional step is only recommended if the purity of the preparation is crucial. In order to accurately measure the levels of internal membrane contamination in purified plasma membrane preparations, pure preparations of the internal membrane are needed. Thus, the ratio of the specific activity of the marker enzyme in the pure plasma membrane to the specific activity of the same enzyme in a pure preparation of the internal membrane accurately represents the proportion of contaminating internal membrane in the plasma membrane preparation. However, partially pure preparations of internal membranes can be used for the estimation of the maximum levels of contamination. Thus, for NADPH-cytochrome c reductase, the ratio of the specific activity of the purest plasma membrane fraction (1 .O nmol/(min/mg protein)) to that of the purest endoplasmic reticulum fraction (33.0 nmol/(min/mg protein)) indicates a maximum contamination of only 3%. Similarly for glucose-6phosphatase, the ratio of the specific activity of the plasma membrane fraction (2.6 nmol/(min/mg protein)) to that of the fraction most enriched for this enzyme (26.6 nmol/(min/mg protein)) indicates a maximum contamination of approximately 10%. However, since the 8K BF material is also highly enriched for succinate dehydrogenase, the mem-
brane containing glucose-6-phosphatase is far from pure and the 10% contamination value is an overestimation. Since the plasma membranes contain no detectable succinate dehydrogenase (Table 1) or cytochrome oxidase (unpublished observations), the mitochondria1contamination is negligible. Other data are consistent with the idea that the contamination of the plasma membranes by internal membranes containing either glucose-6-phosphatase or NADPH-cytochrome c reductase is at most 10% and probably considerably less. Thus, the specific activities of NADPH-cytochrome c reductase and glucose-6-phosphatase in the 30K PM2 fraction are twice as high as those in the 30K PM 1 fraction. If these components represented large levels of contamination, then the specific activities of alkaline phosphatase and 5'-nucleotidase would be considerably lower in the PM2fraction than in the PM1 fraction. For example, if the contamination due to these components represented 20% of the 30K PM1 plasma membrane preparation, then the contamination of the 30K PM2 fraction should be 40%. Thus, the proportion of genuine plasma membrane in the 30K PMl and 30K PM2 fractions would be 80% and 60% respectively, and the ratios of the specific activities of alkaline phosphatase and 5'nkleotidase of the 30K PM1 fraction to that of the PM2 would be 1.33. In fact, the ratio is actually 1.03 for alkaline phosphatase and 1.12for 5'-nucleotidase (Table l), suggesting a maximum contamination in the PM1 fraction of between 3% and 10%. These data are therefore also consistent with the idea that the endoplasmic reticulum contamination is at most 10% and probably considerably less. A similar argument can also be made with regard to our previously published data (1). Acknowledgements
We wish to thank Kathy LaRoy for skilled technical assistance and Dr. Claire Weeks for critically reading the manuscript. The work was supported by a grant from the National Cancer Institute of Canada. 1. Gilkes, N. R. & Weeks, G. (1977) Biochim. Biophys. Acta 46#, 142-156 2. Phillips, A. H. & Langdon, R. G. (1962) J. Biol. Chem. 237,2652-2660 3. Lee, A., Chance, K., Weeks, C. & Weeks, G. (1975) Arch. Biochem. Biophys. 171,407-417 4. Scottocasa, G. L., Kuylenstiena, B., Ernster, L. & Bergstrand, A. (1967) J. Cell. Biol. 32,415-438 5. Hiibscher, G. & West, G. R. (1965) Nature (London) 205,799-800 6. Ames, B. N. (1966) Methods in Enzymology (Neufeld, E. F . & Ginsburg, V., eds.), vol. 8 , pp. 115-118, Academic Press, New York 7. Goldfischer, S., Essner, E. & Novikoff, A. B. (1964) J. Histochem. Cytochem. 12,72-95 8. Green, A. A. & Newell, P. C. (1974) Biochem. J. 140, 3 13-322 9. McMahon, D., Miller, M. & Long S. (1977) Biochim. Biophys. Acta 465,224-24 1
