Membranes of M a m m a r y Gland. X I . Marker E n z y m e Distribution Profiles for Membranous Components f r o m Bovine M a m m a r y Gland C. R. BAUMRUCKER l and T. W. KEENAN Department of Animal Sciences Purdue University West Lafayette, IN 47907
ABSTRACT
cellular organization. Recently methods have been developed to fractionate mammary tissue homogenates. Methods for preparation of fractions enriched in nuclear membranes (1), endoplasmic reticulum (13), Golgi apparatus (13 to 15), plasma membranes (17), and mitochondria (10) have been developed. Assessment of purity of preparations is on morphological criteria and on the concentration of putative marker enzymes in the fractions. To be reliable moniters of tissue fractionation, such marker enzymes must be localized specifically in or at least be highly concentrated in one particular membrane type. Such localization has been shown with a limited number of enzymes in mammary gland (16). Many investigations have been predicated on the assumption that marker enzymes for the extensively characterized rat liver system would be valid for mammary gland. Literature of cell fractionation is replete with examples of the invalidity of many such assumptions. We present results of an analytical study of the distribution of known or assumed membrane marker enzymes in mammary homogehates fractionated by equilibrium density gradient centrifugation. Such distribution patterns have facilitated greatly isolation of membranous components from rat liver (4, 5), and we hope that our study will facilitate future development of mammary cellular fractionation methods.
Enzyme distribution profiles of clarified bovine mammary homogenates separated by equilibrium centrifugation on linear sucrose gradients suggested that several of the commonly utilized marker enzymes for rat liver are also valid markers for mammary cellular components. These marker enzymes include: Succinate dehydrogenase (mitochondria), nicotinamide adenine dinucleotide phosphate cytochrome c reductase and, to a lesser extent, retenone insensitive nicotinamide adenine dinucleotide cytochrome c reductase (endoplasmic reticulum), galactosyl transferase (Golgi apparatus), 5~-nucleotidase (plasma membranes), uric acid oxidase (microbodies), and acid phosphatase (lysosomes). Rotenone sensitive nicotinamide adenine dinucleotide cytochrome c reductase and sodium, potassium, magnesium-stimulated adenosine triphosphatase were widely distributed among subcellular fractions and are not valid marker enzymes. The boyant densities determined for the above fractions should aid in design of methods to obtain enriched sources of these components for analysis. INTRODUCTION To gain a detailed understanding of the origin of milk, it is necessary to determine where within the cell the various constituents of this fluid are synthesized or assembled and packaged for discharge. Collection of such information would be facilitated by studying fractions enriched in the various units of
Received December 12, 1974. ~Present address: Dairy Science Department, University of Illinois, Urbana 61801.
EXPERIMENTAL PROCEDURES Mammary tissue from lactating Holstein cows at slaughter was transported to the laboratory on ice. Unless specified otherwise, all subsequent operations were at 0 to 4 C. Tissue was minced with scissors and homogenized in a buffer of 50 mM Tris-HC1, pH 7.5, 25 mM KC1, and 5 mM MgC12 (designated TKM buffer hereafter) at a ratio of 1 g to 4 ml. Homogenization was with a Polytron 20 ST (Brinkman
1282
MEMBRANE MARKER ENZYMES Instruments) for 20 s at medium speed. The homogenate was filtered sequentially through 2, 4, 6, and 8 double layers of cheesecloth and then centrifuged at 800 × g for 10 min in a swinging bucket rotor to remove whole cells, nuclei, and debris. The supernatant was recovered, filtered through 8 double layers of cheesecloth to remove floating lipids and centrifuged at 650 ×g for 10 min. After filtration, the protein content of this clarified homogenate was adjusted to 5 to 20 mg/ml (18) with TKM buffer. Linear gradients of 20 to 55% (1.084 to 1.262 g/ml) sucrose were generated in 40 ml cellulose nitrate tubes with an ISCO Dialgrad Gradient Pump or a Model 570 gradient former (Instrument Specialties Co.). Gradients were formed at a rate of 2 ml/min to a final volume of 38 to 39 ml. One milliliter of the clarified homogenate was layered over the gradients which were then centrifuged at 113,000 × g in a Beckman L3-50 centrifuge with a SW-27 rotor. After centrifugation, gradients were fractionated into a total of 32 fractions of 1.2 ml each with an ISCO Model 640 fractionator. Absorbance of the fractions at 280 nm was monitored and recorded with a flow-through photometer. Fractions were diluted with TKM buffer, and particulate material was collected by centrifugation at 176,000 × g for 30 min. Pellets were suspended in the appropriate enzyme assay medium, and protein concentration was determined (18). All enzyme assays were performed under conditions of linear kinetics with respect to enzyme concentration and reaction time. Cytochrome c reductase (NADH, EC 1.6.99.3; NADPH, EC 1.6.2.3) was determined by monitoring the reduction of Wtochrome c at 550 nm in a recording spectrophotometer at 30 C without or with (4 x 10 -7 M) rotenone (19). A molar extinction coefficient of 21.1 mM -1 cm -I for cytochrome c was used. Adenosine triphosphatase (ATPase, EC 3.6.1.4) and 5'-nucleotidase (AMP substrate, EC 3.1.3.5) were assayed according to Emmelot et al. (8). Succinate dehydrogenase (EC 1.3.99.1) was assayed as s u c cinate-2-(p-indophenyl)-3-(p-nitrophenyl)-5phenyhetrazolium reductase (25). Galactosyl transferase (EC 2.4.1.22) was assayed with glucose as acceptor in the presence of ~-lactalbumin as described (15). Uric acid oxidase (EC 1.7.3.3) (9) and acid phosphatase (EC 3.1.3.2)
1283
(24) activities also were determined. In this paper relative activities were compared by assigning a value of 100% to that fraction with highest specific activity. RESULTS
Sucrose gradient density equilibrium fractionation methods were adapted from the pioneering work of DeDuve (4, 5). Theoretically there is little limit to the number of components or enzymes that can be measured in experiments of this type, making possible comparisons of numerous frequency distribution curves. Utilizing the concept of biochemical homogeneity advanced by Claude (3), density equilibrium data may be interpreted in much the same manner as are chromatograms. If frequency distribution curves of two or more enzymes or constituents differ significantly, it may be concluded tentatively that these enzymes or constituents belong to different cellular particles (6). However, when such enzymes or constituents coincide in gradients, they may be localized in the same cellular fraction. Such coincidence does not prove coexistence within the same fraction, however. With respect to supposed marker enzymes, such coincidence must be shown before the eligibility can be determined. Fig. 1 gives the protein distribution profile for a typical density equilibrium fractionation of clarified bovine mammary homogenate. The initial absorption peak (Peak 1, fractions 1 to 6) probably coincides with released soluble proteins and membranous fragments that either
I0 .8
Z
II
< 0
6
11I IV 12
~
gr 18
24
30
36
FRACTION NUMBER
FIG. 1. Distribution of protein ( .) in clarified bovine mammary homogenates separated by equilibrium centrifugation in linear sucrose density gradients. Journal of Dairy Science Vol. 58, No. 9
1284
BAUMRUCKER ET AL.
did not penetrate or that barely penetrated into a 20% sucrose solution. Within the sucrose gradient five major peaks were discernible; at fractions 6 to 8 (Peak II), at fractions 9 and 10 (Peak 11I), at 11 and 12 (Peak IV), at 14 and 15 (Peak V), and the extended absorption peak in fractions 19 to 27 (Peak VI). In addition, shoulders were at fractions 10 and 11 (Shoulder A), 14 and 15 (Shoulder B), and 16 and 17 (Shoulder C). The final absorption peak at fractions 32 and 33 is not a true protein absorption but rather represents the interface between the gradient b o t t o m (55% sucrose) and the lifting solution of 60% sucrose. Succinic dehydrogenase and rotenone sensitive NADH cytochrome c reductase are widely used markers for mitochondria from various tissues, and the utility- of these activities as bovine mammary mitochondrial markers has been discussed (10, 16). Both succinate dehydrogenase and rotenone sensitive NADH cytochrome c reductase activities were maximal in fraction 21 of the linear sucrose gradient (Fig. 2). With both, amounts of activity were significant in fractions 3 and 4; this is attributable to fragments produced during homogenization or subsequent manipulation. Small peaks of rotenone sensitive NADH cytochrome c reductase activity at fractions 8, 17, and 18 and the amount of this activity in fractions 10 to 15 suggest that this enzyme is not localized specifically in mitochondria, although it appears to be concentrated therein. Nonspecific localization of this NADH dehydrogenase has been reported for rat liver membrane fractions (21, 23). Nevertheless, the coincidence of succinate de-
IO0
hydrogenase and the majority of the NADHcytochrome c reductase with Peak VI identifies this as the mitochondrial containing region of the gradient. Both NADPH cytochrome c reductase and rotenone-insensitive NADH cytochrome c reductase exhibited maximal activities in fractions 11 to 13 (Fig. 3). These enzymes are used widely as markers for liver endoplasmic reticulum. There was also NADPH cytochrome c reductase activity in fractions 8 and 14, coinciding with Peak III or Shoulder A and Peak V. Rotenone insensitive NADH, but not NADPH, cytochrome c reductase activity was in fractions 2 to 6. This may be explained by the discovery of rotenone insensitive NADH cytochrome c reductase in Golgi apparatus and outer mitochondrial membranes (21, 23) whereas the NADPH dehydrogenase appears to be absent from these latter membrane fractions. These observations, coupled with previous demonstrations that NADPH cytochrome c reductase is enriched in rough endoplasmic reticulum fractions from bovine mammary gland, identify Peak IV as a region of the gradient enriched in endoplasmic reticulum. The enzyme 5'-nucleotidase, a widely used plasma membrane marker, was concentrated in gradient fractions 6 to 8 (Fig. 4). This region, corresponding to Peak II, was the only region of the gradient where significant 5'-nucleotidase activity was detected. This, and the known concentration of 5'-nucleotidase activity on plasma membrane-enriched fractions (11) and in the apical plasma membrane derived milk fat globule membrane (12) suggest that Peak II is
n
~
l!
I>
>-
>U
i
0
"
6
12
,oo
18
24
BO
56
FRACTION N U M B E R
FIG. 2. Distribution of succinic dehydrogenase (o), rotenone sensitive NADH cytochrome c reductase (e), and protein ( ) in linear sucrose density gradients. Journal of Dairy Science Vol. 58, No. 9
bJ
50
o
0
6
12
FRACTION
18
24
50
56
NUMBER
FIG. 3. Distribution of rotenone insensitive NADH (e) and NADPH (o) cytochrorne c reductases and protein ( ~ ) in linear sucrose density gradients.
MEMBRANE MARKER ENZYMES
1285
tO0 >-
> ~'-
75
w ig
~ so Z w >..
25
%
Z w w
i
0
so
12
kL
18
24
30
O0
36
6
12
LL
FRACTION NUMBER
18
24
30
56
FRACTION NUMBER
FIG. 4. Distribution of 5'-nucleotidase activity (o) and protein ( ) in linear sucrose density gradients.
FIG. 6. Distribution of galactosyl transferase (lactose synthesis) (o) and protein ( ) in linear sucrose density gradients.
the primary plasma membrane containing region of the gradient. In contrast to 5Lnucleotid ase, Na +, K +, Mg+2 activated ATPase was distributed widely throughout the gradient (Fig. 5). While monovalent ion activated ATPases are suitable plasma membrane markers with other systems, results in Fig. 5 rule out their localization exclusively in plasma membrane from bovine mammary gland. Ion activated ATPases have been found in intracellular endomembrane fractions from bovine mammary gland (2). Lactose synthetase, or more properly the galactosyl transferase involved in lactose synthesis, was concentrated in gradient Peaks I and II (Fig. 6). This enzyme is a valid marker for Golgi apparatus from mammary gland (13, 15) and rat liver (22). Golgi apparatus appears to concentrate in these regions of the gradient. It is probable that intact dictyosomes occur
with Peak II and dictyosomal fragments occur with Peak I, particularly since homogenization conditions were not optimal for stabilization of Golgi apparatus (20). The distribution of acid phosphatase in gradients is in Fig. 7. While lysosomes had not yet been isolated from mammary gland, acid phosphatase is a widely used rat liver lysosomal marker enzyme (7). Major activity in acid phosphatase was in gradient fractions 13 to 15 and 19 to 21. The latter coincides with the region occupied by mitochondria (Fig. 2). This is not surprising since lysosomes and mitochondria normally parallel each other during density gradient isolations (6). The region of acid phosphatase aligning with Peak V and Shoulder B seems to be a more promising density area (1.165 g/ml) for attempts to isolate mitochondria-free lysosomes for characterization. The two areas of acid phosphatase activity may be
75
75
~ "' 8
as
w rr
0
LL
~W 2o
so
\
,
b
b2
0
6
12
18
24
3(D
36
FRACTION NUMBER
FIG. 5. Distribution of Na+, K+, Mg+2-activated adenosine triphosphatase activities (o) and protein ( • ) in linear sucrose density gradients,
w ¢r
u_
0
6
12
18
24
30
36
FRACTION NUMBER
FIG. 7. Distribution of acid phosphatase activi~ (o) and protein ( ~ ) in linear sucrose density gradieuts. Journal of Dairy Science Vol. 5 8, No. 9
1286
BAUMRUCKER ET AL.
due to a separation of primary and secondary lysosomes. However, the distribution pattern for acid phosphatase appears to justify its exploitation as a marker for monitering the purification of lysosomes. Uric acid oxidase, a common liver microbody (peroxisome) marker (23), was concentrated in the more dense region of Peak VI (Fig. 8). Other areas of high activity in this enzyme were the top of the gradient and that area corresponding to Shoulder C. If uric acid oxidase proves to be a valid mammary microbody marker, and these results suggest that it may be, there are two regions of the gradient in which microbodies are enriched. DISCUSSION
Results support succinic dehydrogenase and 5'-nucleotidase as marker enzymes for bovine mammary mitochondria and plasma membranes, respectively, and of the gatactosyltransferase of lactose synthetase as a Golgi apparatus marker. In addition, results suggested the potential validity of the following enzymes as markers for respective cellular fractions: NADPH cytochrome c reductase and, to a lesser extent, rotenone insensitive NADH cytochrome c reductase for endoplasmic reticulum; uric acid oxidase for microbodies; and acid phosphatase for lysosomes. Rotenone sensitive NADH cytochrome c reductase and especially Na +, K +, Mg+2-activated ATPase activities were distributed widely in density gradients and, thus, associated with more than one cellular component. These activities are not valid markers for
IOC >1--
75 b.I 50 Z t,i
~_
25
z ~A 0 w i,
O~ "^
6
12
18
24
30
36
FRACTION NUMBER
FIG. 8. Distribution of uric acid oxidase activity (o) and protein ( ) in linear sucrose density gradients. Journal of Dairy Science Vol. 58, No. 9
mitochondria and plasma membrane, respectively, as they are in the more widely studied liver system. If presumed marker enzymes assayed prove to be valid markers for the various fractions, the following ranges represent the boyant densities of the units of subcellular organization discussed herein: Endoplasmic reticulum, 1.152 to 1.169 g/ml; plasma membranes and Golgi apparatus, 1.120 to 1.146 g/ml; mitochondria, 1.183 to 1.215 g/ml;lysosomes, 1.157 to 1.173 g/ml and 1.183 to 1.207 g/ml; and microbodies, 1.161 to 1.187 g/ml and 1.210 to 1.243 g/ml. These data should aid investigators desiring only an enriched preparation of cellular fractions. In addition, this information should help in design of methods of isolation of previously uncharacterized mammary subcellular fractions and may serve as a basis for improvement of present techniques used in fractionation of mammary homogenates. ACKNOWLEDGMENT
This research was supported by Grant 25110 from the National Science Foundation. Keenan is a Research Career Development Awardee (GM 70596) of the National Institute of General Medical Science. Purdue University AES Journal Paper No. 5742. REFERENCES
1. Baumrucker, C. R., and T. W. Keenan. 1974. Membranes of mammary gland. VIII. Isolation and composition of nuclei and nuclear membrane from bovine mammary gland. J. Dairy Sci. 57:24. 2. Baumrucker, C. R., and T. W. Keenan. 1975. Membranes of mammary gland. X. Adenosine triphosphate-dependent calcium accumulation by Golgi apparatus from bovine mammary gland. Exp. Cell Res. 90:253. 3. Claude, A. 1946. Fractionation of mammalian liver cells by differential centrifugation. J. Exp. Med. 84:51. 4. DeDuve, C. 1964. Principles of tissue fractionation. J. Theoret. Biol. 6:33. 5. DeDuve, C. 1971. Tissue fractionation. Past and present. J. Cell Biol. 50:20D. 6. DeDuve, C. 1965. The separation and characterization of subcellular particles. Harvey Lectures 59:49. 7. DeDuve, C. 1959. Lysosomes, a new group of cytoplasmic particles. Page 128 in T. Hayashi, ed. Subcellular particles. Ronald Press, New York. 8. Emmelot, P., C. J. Bos, E. L. Benedetti, and P. Rumke. 1964. Studies on plasma membranes. I.
MEMBRANE MARKER ENZYMES
9.
10.
11.
12.
13.
14.
15.
16.
17.
Chemical composition and enzyme content of plasma membranes isolated from rat liver. Biochim. Biophys. Acta 90:126. Henry, R. J., C. Sorbel, and J. Kim. 1957. Modified carbonate-phosphotungstate methods for the determination of uric acid and comparison with spectrophotometric uricase. Amer. J. Clin. Pathol. 28:152. Huang, C. M., and T. W. Keenan. 1971. Membranes of mammary gland. I. Bovine mammary mitochondria. J. Dairy Sci. 54:1395. Huang, C. M., and T. W. Keenan. 1972. Membranes of mammary gland. II. 5'-Nucleotidase activity of bovine mammary plasma membranes. J. Dairy Sci. 55:862. Huang, C. M., and T. W. Keenan. 1972. Preparation and properties of 5'-nucleotidases from bovine milk fat globule membranes. Biochim. Biophys. Acta 274:246. Keenan, T. W., C. M. Huang, and D. J. Morre. 1972. Membranes of mammary gland. V. Isolation of Golgi apparatus and rough endoplasmic reticulum from bovine mammary gland. J. Dairy Sci. 55:1577. Keenan, T. W., and C. M. Huang. 1972. Membranes of mammary gland. VI. Lipid and protein composition of Golgi apparatus and rough endoplasmic reticulum from bovine mammary gland. J. Dairy Sci. 55:1586. Keenan, T. W., D. J. Morre, and R. D. Cheetham. 1970. Lactose synthesis by a Golgi apparatus fraction from rat mammary gland. Nature 228:1105. Keenan, T. W., D. J. Morre, and C. M. Huang. 1974. Membranes of mammary gland. Page 191 in B. L. Larson and V. R. Smith, eds. Lactation: A comprehensive treatise, Vol. II. Academic Press, New York. Keenan, T. W., D. J. Morre, D. E. Olson, W. N.
18.
19. 20.
21.
22.
23.
24.
25.
1287
Yunghans, and S. Patton. 1970. Biochemical and morphological comparison of plasma membrane and milk fat globule membrane from bovine mammary gland. J. Cell Biol. 44:80. Lowry, O. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall. 1951. Protein determination with the Folin phenol reagent. J. Biol. Chem. 193:265. Mahler, H. R. 1965. DPNH cytochrome c reductase (animal). Methods Enzymol. 2:688. Morre, D. J., R. L. Hamilton, H. H. Mollenbauer, R. W. Mahley, W. P. Cunningham, R. D. Cheetham, and V. S. LeQuire. 1970. Isolation of a Golgi apparatus-rich fraction from rat liver. I. Method and morphology. J. Cell Biol. 44:484. Morre, D. J., T. W. Keenan, and C. M. Huang. 1974. Membrane flow and differentiation: Origin of Golgi apparatus membranes from endoplasmic reticulum. Page 107 in F. Clementi, B. Ceccarelli, and J. Meldolesi, eds. Advances in cytopharmacology, Vol. 2. Raven Press, New York. Morre, D. J., L. M. Merlin, and T. W. Keenan. 1969. Localization of glycosyl transferase activities in a Golgi apparatus-rich fraction isolated from rat liver. Biochem. Biophys. Res. Comm. 37:813. Morre, D. J., W. N. Yunghans, E. L. Vigil, and T. W. Keenans. 1975. Isolation of organelles and endomembrane components from rat liver: Biochemical markers and quantitative morphometry. In E. Reid, ed. Methodological developments in biochemistry, Vol. 4. Longrnan, North-Holland. (In press.) Nyquist, S. E., and D. J. Morre. 1971. Distribution of UDP-glucuronyl transferase among cell fractions from rat liver. J. Cell Physiol. 78:9. Pennington, R. J. 1961. Biochemistry of drystrophic muscle. Mitochondrial succinate-tetrazolium reductase and adenosine triphosphatase. Biochem. J. 80:649.
Journal of Dairy Science Vol. 58, No. 9
ABSTRACT
cellular organization. Recently methods have been developed to fractionate mammary tissue homogenates. Methods for preparation of fractions enriched in nuclear membranes (1), endoplasmic reticulum (13), Golgi apparatus (13 to 15), plasma membranes (17), and mitochondria (10) have been developed. Assessment of purity of preparations is on morphological criteria and on the concentration of putative marker enzymes in the fractions. To be reliable moniters of tissue fractionation, such marker enzymes must be localized specifically in or at least be highly concentrated in one particular membrane type. Such localization has been shown with a limited number of enzymes in mammary gland (16). Many investigations have been predicated on the assumption that marker enzymes for the extensively characterized rat liver system would be valid for mammary gland. Literature of cell fractionation is replete with examples of the invalidity of many such assumptions. We present results of an analytical study of the distribution of known or assumed membrane marker enzymes in mammary homogehates fractionated by equilibrium density gradient centrifugation. Such distribution patterns have facilitated greatly isolation of membranous components from rat liver (4, 5), and we hope that our study will facilitate future development of mammary cellular fractionation methods.
Enzyme distribution profiles of clarified bovine mammary homogenates separated by equilibrium centrifugation on linear sucrose gradients suggested that several of the commonly utilized marker enzymes for rat liver are also valid markers for mammary cellular components. These marker enzymes include: Succinate dehydrogenase (mitochondria), nicotinamide adenine dinucleotide phosphate cytochrome c reductase and, to a lesser extent, retenone insensitive nicotinamide adenine dinucleotide cytochrome c reductase (endoplasmic reticulum), galactosyl transferase (Golgi apparatus), 5~-nucleotidase (plasma membranes), uric acid oxidase (microbodies), and acid phosphatase (lysosomes). Rotenone sensitive nicotinamide adenine dinucleotide cytochrome c reductase and sodium, potassium, magnesium-stimulated adenosine triphosphatase were widely distributed among subcellular fractions and are not valid marker enzymes. The boyant densities determined for the above fractions should aid in design of methods to obtain enriched sources of these components for analysis. INTRODUCTION To gain a detailed understanding of the origin of milk, it is necessary to determine where within the cell the various constituents of this fluid are synthesized or assembled and packaged for discharge. Collection of such information would be facilitated by studying fractions enriched in the various units of
Received December 12, 1974. ~Present address: Dairy Science Department, University of Illinois, Urbana 61801.
EXPERIMENTAL PROCEDURES Mammary tissue from lactating Holstein cows at slaughter was transported to the laboratory on ice. Unless specified otherwise, all subsequent operations were at 0 to 4 C. Tissue was minced with scissors and homogenized in a buffer of 50 mM Tris-HC1, pH 7.5, 25 mM KC1, and 5 mM MgC12 (designated TKM buffer hereafter) at a ratio of 1 g to 4 ml. Homogenization was with a Polytron 20 ST (Brinkman
1282
MEMBRANE MARKER ENZYMES Instruments) for 20 s at medium speed. The homogenate was filtered sequentially through 2, 4, 6, and 8 double layers of cheesecloth and then centrifuged at 800 × g for 10 min in a swinging bucket rotor to remove whole cells, nuclei, and debris. The supernatant was recovered, filtered through 8 double layers of cheesecloth to remove floating lipids and centrifuged at 650 ×g for 10 min. After filtration, the protein content of this clarified homogenate was adjusted to 5 to 20 mg/ml (18) with TKM buffer. Linear gradients of 20 to 55% (1.084 to 1.262 g/ml) sucrose were generated in 40 ml cellulose nitrate tubes with an ISCO Dialgrad Gradient Pump or a Model 570 gradient former (Instrument Specialties Co.). Gradients were formed at a rate of 2 ml/min to a final volume of 38 to 39 ml. One milliliter of the clarified homogenate was layered over the gradients which were then centrifuged at 113,000 × g in a Beckman L3-50 centrifuge with a SW-27 rotor. After centrifugation, gradients were fractionated into a total of 32 fractions of 1.2 ml each with an ISCO Model 640 fractionator. Absorbance of the fractions at 280 nm was monitored and recorded with a flow-through photometer. Fractions were diluted with TKM buffer, and particulate material was collected by centrifugation at 176,000 × g for 30 min. Pellets were suspended in the appropriate enzyme assay medium, and protein concentration was determined (18). All enzyme assays were performed under conditions of linear kinetics with respect to enzyme concentration and reaction time. Cytochrome c reductase (NADH, EC 1.6.99.3; NADPH, EC 1.6.2.3) was determined by monitoring the reduction of Wtochrome c at 550 nm in a recording spectrophotometer at 30 C without or with (4 x 10 -7 M) rotenone (19). A molar extinction coefficient of 21.1 mM -1 cm -I for cytochrome c was used. Adenosine triphosphatase (ATPase, EC 3.6.1.4) and 5'-nucleotidase (AMP substrate, EC 3.1.3.5) were assayed according to Emmelot et al. (8). Succinate dehydrogenase (EC 1.3.99.1) was assayed as s u c cinate-2-(p-indophenyl)-3-(p-nitrophenyl)-5phenyhetrazolium reductase (25). Galactosyl transferase (EC 2.4.1.22) was assayed with glucose as acceptor in the presence of ~-lactalbumin as described (15). Uric acid oxidase (EC 1.7.3.3) (9) and acid phosphatase (EC 3.1.3.2)
1283
(24) activities also were determined. In this paper relative activities were compared by assigning a value of 100% to that fraction with highest specific activity. RESULTS
Sucrose gradient density equilibrium fractionation methods were adapted from the pioneering work of DeDuve (4, 5). Theoretically there is little limit to the number of components or enzymes that can be measured in experiments of this type, making possible comparisons of numerous frequency distribution curves. Utilizing the concept of biochemical homogeneity advanced by Claude (3), density equilibrium data may be interpreted in much the same manner as are chromatograms. If frequency distribution curves of two or more enzymes or constituents differ significantly, it may be concluded tentatively that these enzymes or constituents belong to different cellular particles (6). However, when such enzymes or constituents coincide in gradients, they may be localized in the same cellular fraction. Such coincidence does not prove coexistence within the same fraction, however. With respect to supposed marker enzymes, such coincidence must be shown before the eligibility can be determined. Fig. 1 gives the protein distribution profile for a typical density equilibrium fractionation of clarified bovine mammary homogenate. The initial absorption peak (Peak 1, fractions 1 to 6) probably coincides with released soluble proteins and membranous fragments that either
I0 .8
Z
II
< 0
6
11I IV 12
~
gr 18
24
30
36
FRACTION NUMBER
FIG. 1. Distribution of protein ( .) in clarified bovine mammary homogenates separated by equilibrium centrifugation in linear sucrose density gradients. Journal of Dairy Science Vol. 58, No. 9
1284
BAUMRUCKER ET AL.
did not penetrate or that barely penetrated into a 20% sucrose solution. Within the sucrose gradient five major peaks were discernible; at fractions 6 to 8 (Peak II), at fractions 9 and 10 (Peak 11I), at 11 and 12 (Peak IV), at 14 and 15 (Peak V), and the extended absorption peak in fractions 19 to 27 (Peak VI). In addition, shoulders were at fractions 10 and 11 (Shoulder A), 14 and 15 (Shoulder B), and 16 and 17 (Shoulder C). The final absorption peak at fractions 32 and 33 is not a true protein absorption but rather represents the interface between the gradient b o t t o m (55% sucrose) and the lifting solution of 60% sucrose. Succinic dehydrogenase and rotenone sensitive NADH cytochrome c reductase are widely used markers for mitochondria from various tissues, and the utility- of these activities as bovine mammary mitochondrial markers has been discussed (10, 16). Both succinate dehydrogenase and rotenone sensitive NADH cytochrome c reductase activities were maximal in fraction 21 of the linear sucrose gradient (Fig. 2). With both, amounts of activity were significant in fractions 3 and 4; this is attributable to fragments produced during homogenization or subsequent manipulation. Small peaks of rotenone sensitive NADH cytochrome c reductase activity at fractions 8, 17, and 18 and the amount of this activity in fractions 10 to 15 suggest that this enzyme is not localized specifically in mitochondria, although it appears to be concentrated therein. Nonspecific localization of this NADH dehydrogenase has been reported for rat liver membrane fractions (21, 23). Nevertheless, the coincidence of succinate de-
IO0
hydrogenase and the majority of the NADHcytochrome c reductase with Peak VI identifies this as the mitochondrial containing region of the gradient. Both NADPH cytochrome c reductase and rotenone-insensitive NADH cytochrome c reductase exhibited maximal activities in fractions 11 to 13 (Fig. 3). These enzymes are used widely as markers for liver endoplasmic reticulum. There was also NADPH cytochrome c reductase activity in fractions 8 and 14, coinciding with Peak III or Shoulder A and Peak V. Rotenone insensitive NADH, but not NADPH, cytochrome c reductase activity was in fractions 2 to 6. This may be explained by the discovery of rotenone insensitive NADH cytochrome c reductase in Golgi apparatus and outer mitochondrial membranes (21, 23) whereas the NADPH dehydrogenase appears to be absent from these latter membrane fractions. These observations, coupled with previous demonstrations that NADPH cytochrome c reductase is enriched in rough endoplasmic reticulum fractions from bovine mammary gland, identify Peak IV as a region of the gradient enriched in endoplasmic reticulum. The enzyme 5'-nucleotidase, a widely used plasma membrane marker, was concentrated in gradient fractions 6 to 8 (Fig. 4). This region, corresponding to Peak II, was the only region of the gradient where significant 5'-nucleotidase activity was detected. This, and the known concentration of 5'-nucleotidase activity on plasma membrane-enriched fractions (11) and in the apical plasma membrane derived milk fat globule membrane (12) suggest that Peak II is
n
~
l!
I>
>-
>U
i
0
"
6
12
,oo
18
24
BO
56
FRACTION N U M B E R
FIG. 2. Distribution of succinic dehydrogenase (o), rotenone sensitive NADH cytochrome c reductase (e), and protein ( ) in linear sucrose density gradients. Journal of Dairy Science Vol. 58, No. 9
bJ
50
o
0
6
12
FRACTION
18
24
50
56
NUMBER
FIG. 3. Distribution of rotenone insensitive NADH (e) and NADPH (o) cytochrorne c reductases and protein ( ~ ) in linear sucrose density gradients.
MEMBRANE MARKER ENZYMES
1285
tO0 >-
> ~'-
75
w ig
~ so Z w >..
25
%
Z w w
i
0
so
12
kL
18
24
30
O0
36
6
12
LL
FRACTION NUMBER
18
24
30
56
FRACTION NUMBER
FIG. 4. Distribution of 5'-nucleotidase activity (o) and protein ( ) in linear sucrose density gradients.
FIG. 6. Distribution of galactosyl transferase (lactose synthesis) (o) and protein ( ) in linear sucrose density gradients.
the primary plasma membrane containing region of the gradient. In contrast to 5Lnucleotid ase, Na +, K +, Mg+2 activated ATPase was distributed widely throughout the gradient (Fig. 5). While monovalent ion activated ATPases are suitable plasma membrane markers with other systems, results in Fig. 5 rule out their localization exclusively in plasma membrane from bovine mammary gland. Ion activated ATPases have been found in intracellular endomembrane fractions from bovine mammary gland (2). Lactose synthetase, or more properly the galactosyl transferase involved in lactose synthesis, was concentrated in gradient Peaks I and II (Fig. 6). This enzyme is a valid marker for Golgi apparatus from mammary gland (13, 15) and rat liver (22). Golgi apparatus appears to concentrate in these regions of the gradient. It is probable that intact dictyosomes occur
with Peak II and dictyosomal fragments occur with Peak I, particularly since homogenization conditions were not optimal for stabilization of Golgi apparatus (20). The distribution of acid phosphatase in gradients is in Fig. 7. While lysosomes had not yet been isolated from mammary gland, acid phosphatase is a widely used rat liver lysosomal marker enzyme (7). Major activity in acid phosphatase was in gradient fractions 13 to 15 and 19 to 21. The latter coincides with the region occupied by mitochondria (Fig. 2). This is not surprising since lysosomes and mitochondria normally parallel each other during density gradient isolations (6). The region of acid phosphatase aligning with Peak V and Shoulder B seems to be a more promising density area (1.165 g/ml) for attempts to isolate mitochondria-free lysosomes for characterization. The two areas of acid phosphatase activity may be
75
75
~ "' 8
as
w rr
0
LL
~W 2o
so
\
,
b
b2
0
6
12
18
24
3(D
36
FRACTION NUMBER
FIG. 5. Distribution of Na+, K+, Mg+2-activated adenosine triphosphatase activities (o) and protein ( • ) in linear sucrose density gradients,
w ¢r
u_
0
6
12
18
24
30
36
FRACTION NUMBER
FIG. 7. Distribution of acid phosphatase activi~ (o) and protein ( ~ ) in linear sucrose density gradieuts. Journal of Dairy Science Vol. 5 8, No. 9
1286
BAUMRUCKER ET AL.
due to a separation of primary and secondary lysosomes. However, the distribution pattern for acid phosphatase appears to justify its exploitation as a marker for monitering the purification of lysosomes. Uric acid oxidase, a common liver microbody (peroxisome) marker (23), was concentrated in the more dense region of Peak VI (Fig. 8). Other areas of high activity in this enzyme were the top of the gradient and that area corresponding to Shoulder C. If uric acid oxidase proves to be a valid mammary microbody marker, and these results suggest that it may be, there are two regions of the gradient in which microbodies are enriched. DISCUSSION
Results support succinic dehydrogenase and 5'-nucleotidase as marker enzymes for bovine mammary mitochondria and plasma membranes, respectively, and of the gatactosyltransferase of lactose synthetase as a Golgi apparatus marker. In addition, results suggested the potential validity of the following enzymes as markers for respective cellular fractions: NADPH cytochrome c reductase and, to a lesser extent, rotenone insensitive NADH cytochrome c reductase for endoplasmic reticulum; uric acid oxidase for microbodies; and acid phosphatase for lysosomes. Rotenone sensitive NADH cytochrome c reductase and especially Na +, K +, Mg+2-activated ATPase activities were distributed widely in density gradients and, thus, associated with more than one cellular component. These activities are not valid markers for
IOC >1--
75 b.I 50 Z t,i
~_
25
z ~A 0 w i,
O~ "^
6
12
18
24
30
36
FRACTION NUMBER
FIG. 8. Distribution of uric acid oxidase activity (o) and protein ( ) in linear sucrose density gradients. Journal of Dairy Science Vol. 58, No. 9
mitochondria and plasma membrane, respectively, as they are in the more widely studied liver system. If presumed marker enzymes assayed prove to be valid markers for the various fractions, the following ranges represent the boyant densities of the units of subcellular organization discussed herein: Endoplasmic reticulum, 1.152 to 1.169 g/ml; plasma membranes and Golgi apparatus, 1.120 to 1.146 g/ml; mitochondria, 1.183 to 1.215 g/ml;lysosomes, 1.157 to 1.173 g/ml and 1.183 to 1.207 g/ml; and microbodies, 1.161 to 1.187 g/ml and 1.210 to 1.243 g/ml. These data should aid investigators desiring only an enriched preparation of cellular fractions. In addition, this information should help in design of methods of isolation of previously uncharacterized mammary subcellular fractions and may serve as a basis for improvement of present techniques used in fractionation of mammary homogenates. ACKNOWLEDGMENT
This research was supported by Grant 25110 from the National Science Foundation. Keenan is a Research Career Development Awardee (GM 70596) of the National Institute of General Medical Science. Purdue University AES Journal Paper No. 5742. REFERENCES
1. Baumrucker, C. R., and T. W. Keenan. 1974. Membranes of mammary gland. VIII. Isolation and composition of nuclei and nuclear membrane from bovine mammary gland. J. Dairy Sci. 57:24. 2. Baumrucker, C. R., and T. W. Keenan. 1975. Membranes of mammary gland. X. Adenosine triphosphate-dependent calcium accumulation by Golgi apparatus from bovine mammary gland. Exp. Cell Res. 90:253. 3. Claude, A. 1946. Fractionation of mammalian liver cells by differential centrifugation. J. Exp. Med. 84:51. 4. DeDuve, C. 1964. Principles of tissue fractionation. J. Theoret. Biol. 6:33. 5. DeDuve, C. 1971. Tissue fractionation. Past and present. J. Cell Biol. 50:20D. 6. DeDuve, C. 1965. The separation and characterization of subcellular particles. Harvey Lectures 59:49. 7. DeDuve, C. 1959. Lysosomes, a new group of cytoplasmic particles. Page 128 in T. Hayashi, ed. Subcellular particles. Ronald Press, New York. 8. Emmelot, P., C. J. Bos, E. L. Benedetti, and P. Rumke. 1964. Studies on plasma membranes. I.
MEMBRANE MARKER ENZYMES
9.
10.
11.
12.
13.
14.
15.
16.
17.
Chemical composition and enzyme content of plasma membranes isolated from rat liver. Biochim. Biophys. Acta 90:126. Henry, R. J., C. Sorbel, and J. Kim. 1957. Modified carbonate-phosphotungstate methods for the determination of uric acid and comparison with spectrophotometric uricase. Amer. J. Clin. Pathol. 28:152. Huang, C. M., and T. W. Keenan. 1971. Membranes of mammary gland. I. Bovine mammary mitochondria. J. Dairy Sci. 54:1395. Huang, C. M., and T. W. Keenan. 1972. Membranes of mammary gland. II. 5'-Nucleotidase activity of bovine mammary plasma membranes. J. Dairy Sci. 55:862. Huang, C. M., and T. W. Keenan. 1972. Preparation and properties of 5'-nucleotidases from bovine milk fat globule membranes. Biochim. Biophys. Acta 274:246. Keenan, T. W., C. M. Huang, and D. J. Morre. 1972. Membranes of mammary gland. V. Isolation of Golgi apparatus and rough endoplasmic reticulum from bovine mammary gland. J. Dairy Sci. 55:1577. Keenan, T. W., and C. M. Huang. 1972. Membranes of mammary gland. VI. Lipid and protein composition of Golgi apparatus and rough endoplasmic reticulum from bovine mammary gland. J. Dairy Sci. 55:1586. Keenan, T. W., D. J. Morre, and R. D. Cheetham. 1970. Lactose synthesis by a Golgi apparatus fraction from rat mammary gland. Nature 228:1105. Keenan, T. W., D. J. Morre, and C. M. Huang. 1974. Membranes of mammary gland. Page 191 in B. L. Larson and V. R. Smith, eds. Lactation: A comprehensive treatise, Vol. II. Academic Press, New York. Keenan, T. W., D. J. Morre, D. E. Olson, W. N.
18.
19. 20.
21.
22.
23.
24.
25.
1287
Yunghans, and S. Patton. 1970. Biochemical and morphological comparison of plasma membrane and milk fat globule membrane from bovine mammary gland. J. Cell Biol. 44:80. Lowry, O. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall. 1951. Protein determination with the Folin phenol reagent. J. Biol. Chem. 193:265. Mahler, H. R. 1965. DPNH cytochrome c reductase (animal). Methods Enzymol. 2:688. Morre, D. J., R. L. Hamilton, H. H. Mollenbauer, R. W. Mahley, W. P. Cunningham, R. D. Cheetham, and V. S. LeQuire. 1970. Isolation of a Golgi apparatus-rich fraction from rat liver. I. Method and morphology. J. Cell Biol. 44:484. Morre, D. J., T. W. Keenan, and C. M. Huang. 1974. Membrane flow and differentiation: Origin of Golgi apparatus membranes from endoplasmic reticulum. Page 107 in F. Clementi, B. Ceccarelli, and J. Meldolesi, eds. Advances in cytopharmacology, Vol. 2. Raven Press, New York. Morre, D. J., L. M. Merlin, and T. W. Keenan. 1969. Localization of glycosyl transferase activities in a Golgi apparatus-rich fraction isolated from rat liver. Biochem. Biophys. Res. Comm. 37:813. Morre, D. J., W. N. Yunghans, E. L. Vigil, and T. W. Keenans. 1975. Isolation of organelles and endomembrane components from rat liver: Biochemical markers and quantitative morphometry. In E. Reid, ed. Methodological developments in biochemistry, Vol. 4. Longrnan, North-Holland. (In press.) Nyquist, S. E., and D. J. Morre. 1971. Distribution of UDP-glucuronyl transferase among cell fractions from rat liver. J. Cell Physiol. 78:9. Pennington, R. J. 1961. Biochemistry of drystrophic muscle. Mitochondrial succinate-tetrazolium reductase and adenosine triphosphatase. Biochem. J. 80:649.
Journal of Dairy Science Vol. 58, No. 9
