TRE JOU~NAI. OF RI~LOGICAL CEIEMISTRY Vol. 246, No. 8, Issue of April 25, pp. 2574-2583, 1971

Printed in U.S.A.

Size Distribution of Membrane Proteins of Rat Liver and

Their Relative Rates of Degradation*

(Received for publication, October 26, 1970)

PETER J. DEHLINGER~ AND ROBERT T. SCHIMKE

From the Department of Pharmacology, Stanford University School of Medicine, Stanford, California 94305

SUMMARY

The proteins associated with various membrane fractions, including smooth and rough endoplasmic reticulum, plasma membrane, and mitochondria of rat liver, were solubilized with sodium lauryl sulfate and separated by electrophoresis on sodium lauryl sulfate acrylamide gels. Smooth and rough endoplasmic reticulum have similar electrophoretic patterns. All other fractions have characteristic patterns in which no single or predominant protein occurs. The smooth endoplasmic reticulum fraction was fractionated by use of Triton X-100, which solubilized approximately 50% of the membrane proteins. The patterns of Triton X-100~soluble and -insoluble proteins are distinctive.

The relative rates of degradation of the proteins were studied by use of a double isotope technique in which 14C- leucine was administered to intact rats 4 days prior to ad- ministration of aH-leucine. As indicated by the differences in 3H:14C ratios, there is a marked heterogeneity of turnover of proteins of both the endoplasmic reticulum and plasma membrane.

There is a general correlation between the relative degrada- tion rates of membrane proteins and the molecular size of the protein (subunit) as separated by electrophoresis in sodium lauryl sulfate acrylamide gels. This general rela- tionship was found also for the soluble protein fraction of rat liver, whether fractionated by sodium lauryl sulfate acryl- amide gels, or by Sephadex G-ZOO columns in the presence of sodium lauryl sulfate.

A model in which membrane proteins associate and dis- sociate from the protein-lipid complex is favored to explain the heterogeneity of degradation, and the general correla- tion of size with degradation rate.

Several studies have shown that the protein components of liver microsomes and plasma membrane are turning over rela- tively rapidly (l-3). This finding has raised the question of whether the membrane is degraded and synthesized as a unit, or whether individual protein components can be introduced

* This work was supported by Grant GM 14931 from the Na- tional Institutes of Health.

$ Recipient of National Institutes of Health Training Grant GM 0712.

and removed from the membrane at different rates (asyn- chronously). Studies on the degradation rate constants of individual membrane proteins, cytochrome bS (1, 2), cytochrome, c reductase (1, 2), and NAD-glycohydrolase (4) indicate a heterogeneity of half-lives of these purified proteins ranging from 60 to 80 hours (cytochrome c reductase (2)) to 17 days (NAD- glycohydrolase (4)). The problems of isolation and identifica- tion of only a few protein components of membranes led Arias, Doyle, and Schimke (1) to determine relative rate constants of degradation of proteins solubilized from endoplasmic reticulum and separated by chromatography on DEAE-cellulose in the presence of Triton X-100. These studies likewise indicated marked heterogeneity of turnover of such solubilized proteins.

In the present study we have examined the relative degradation rates of protein constituents of rat liver microsomal and plasma membrane fractions as they are separated by electrophoresis on acrylamide gels in the presence of sodium lauryl sulfate. This method offers several advantages in the study of membrane turnover: (a) essentially all of the membrane proteins are solu- bilized by the detergent and can be resolved well by gels; (b) major protein bands can be identified and studied for unusual turnover characteristics; (c) since migration in SLS’ acrylamide gels is dependent on protein size (5), correlations between molec- ular size and degradation rate constants can be made; (d) dif- ferences in the protein components of various membrane types can be readily assessed.

METHODS

Cell Fractionation--All fractionation procedures were per- formed on livers from male, white Sprague-Dawley rats weighing 120 to 150 g each. The animals were fasted 18 hours before killing. Livers were perfused through the spleen with 10 ml of cold 0.25 M sucrose before excision, and homogenized with a glass homogenizer and Teflon pestle, three strokes at 700 rpm. For the isolation of plasma membrane, a Dounce homogenizer was used. The method of Leighton et al. (6) was used for isolating mitochondria and peroxisomes, except that the iso- pycnic centrifugation was done in 13-ml swinging bucket tubes (SB-283 rotor) in an International centrifuge, model B-60, rather than a zonal rotor. The method of Neville (7) as modified by Emmelot et al. (8) was used for preparing the plasma mem- brane fraction. Three microsomal fractions were prepared as

1 The abbreviations used are: SLS, sodillm laurpl sulfate; RER, rough endoplasmic reticulum, SER, smooth eadoplasmic reticulum; LKR, light endoplasmic reticulum; TX, Triton X-100; ER, endoplasmic reticulum.

2574

by guest on April 27, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Issue of April 25, 1971 P. J. Dehlinger and R. T. Xchimke 2575

described by Dallner (9): the rough endoplasmic reticulum fraction (density in 20 mM C&l greater than 1.174), designated RER; the predominantly smooth endoplasmic reticulum fraction (density in 10 mM MgC12 between 1.154 and 1.174), designated SER; and a so-called light endoplasmic reticulum fraction (density in 10 mM MgC12 less than 1.154), designated LER.

For preparing the soluble liver fraction, a perfused liver was homogenized with 1 volume (v/w) of 0.05 M Tris-glycine buffer, pH 8.9. The homogenate was centrifuged at 10,000 x g for 20 min and the resulting supernatant spun at 105,000 x g for 1 hour. The 105,000 x g supernatant was drawn off and passed through a Sephadex G-25 column, 1.5 x 20 cm, which had been equilibrated with the homogenization buffer. This procedure effectively removed free amino acid from the supernatant frac- tion. The resulting material was frozen until used.

Enzyme Assays-The following enzyme markers were used to monitor purity of membrane fractions (6). Catalase, a peroxi- somal enzyme, was assayed by following the disappearance of HzOz at 240 rnp (10) ; cytochrome oxidase was assayed by the oxidation of reduced cytochrome c (11). Both assays were recorded with a Gilford model 2000 spectrophotometer. Alka- line nitrophenylphosphatase activity, characteristic of plasma membrane (8), was measured by appearance of p-nitrophenol at pH 8.9. In this assay, the sample was previously incubated at 37” in 0.05 M Tris-HCl, pH 8.9, for 30 min to inactivate glucose 6-phosphatase (12) prior to the addition of the substrate, p- nitrophenylphosphate. Glucose 6-phosphatase was assayed by the appearance of inorganic phosphate from glucose 6-phos- phate at pH 6.5 (13).

Isotope Administration-Uniformly labeled i*C-leucine (300 mCi per mM) and 4,5-3H-leucine (2000 mCi per mM) were ob- tained from Schwarz BioResesrch. The isotope was diluted to 2 ml in 0.15 M NaCl and injected intraperitoneally in two l-ml doses. In the double label experiments animals were fasted 18 hours, then given 100 PCi of 14C-leucine. Food was restored 4 hours after isotope administration. Four days later this schedule was repeated; this time the animal was given 250 PCi of %leucine and killed 4 hours later. The rationale for killing the animals 4 hours after isotope administration is based on the finding that the maximum specific radioactivity of liver microsomes occurs within 60 min of injection. From this time until about 4 hours after injection the specific radioactivity decreases rapidly to about 50% of the maximum radioactivity, and subsequently achieves a more gradual rate of decrease (1). Ray, Lieberman, and Lansing have also made a similar observa- tion for SER (14). This pattern of isotope decay is observed with both RER and SER fractions, and also with four sub- fractions of SER which we can obtain by isopycnic centrifugation of SER on a 20 to 45y0 (w/v) sucrose gradient.2 Presumably the rapid decrease of labeling during the 4 hours after isotope administration is caused by the passage of serum proteins through the endoplasmic reticulum (15), and is not the result of actual loss of membrane proteins.

With the isotope doses indicated above, about 5 mg of protein were required for gel electrophoresis experiments in order to obtain levels of 3H and i*C radioactivity that were three to four times background. A single 120-g rat was adequate for isolating 5 mg of material for the studies involving the proteins of SER and the supernatant fraction. It was necessary to pool the livers of four rats to obtain sufficient plasma membrane.

z P. J. Dehlinger and It. T. Schimke, unpublished observations.

Gel Electrophoresis-The membrane fractions were prepared for electrophoresis by sonic treatment in 0.1% SLS for 20 set with a Branson sonifier at a power setting of 1. Solubilization procedures were performed at room temperature. Labeled SER was prepared for large scale electrophoresis as follows: 0.15 M KC1 was added to the pelleted SER fraction (9) and sonically disrupted as described above until completely resuspended (about 15 set). This material was centrifuged at 105,000 x g for 1 hour and the supernatant discarded. This wash procedure was repeated with distilled water. The washed SER fraction was then resuspended in 1% Triton X-100 (v/v) to a final protein concentration of about 5 mg per ml, and sonically dis- rupted as described above for 20 sec. This material was cen- trifuged at 105,000 x g for 1 hour. The supernatant, designated TX-soluble SER, contained approximately 55% of the washed microsomal protein. This fraction was made 0.1% in SLS (w/v), sonically disrupted for 20 set, and applied directly to the gels. The membrane fraction insoluble in 1% Triton X-109 required 2 to 3 min of sonic treatment in 0.5% SLS for solubiliza- tion. It was not necessary to reduce the SLS concentration of this material before electrophoresis.

Protein separation by acrylamide gel electrophoresis was carried out on discontinuous gels according to the procedure described by Grossfeld (16). Discontinuous gradient gels consist of regions of varying concentration of acrylamide, thus allowing for greater resolution of proteins of heterogeneous size. The gel st’ock solutions were similar to those used by Davis (17) for electrophoresis in Tris-glycine buffer, except that stocks for 12, 10, 7.5, and 4% acrylamide were prepared. The upper reservoir buffer (pH 8.9) consisted of 6 g of Tris and 4 g of glycine per liter; the lower reservoir buffer (pH 8.1) consisted of 12 g of Tris and 50 ml of 1 N HCl per liter. The gels consisted of one part of acrylamide stock to three parts of 0.5 M Tris-HC1 buffer, pH 8.9, containing 0.1% ammonium persulfate (w/v) and 0.08% N , N , N’ , N’-tetramethylethylenediamine (v/v). Detergent was not incorporated into the gel itself, since SLS in the sample and upper buffer migrated into the gel ahead of the sample proteins. The gel solutions were carefully layered in the glass tubes, one over the other, before polymerization occurred. This resulted in a more gradual transition at the gel interfaces. Polymeriza- tion was retarded by addition of 0.154” (w/v) potassium fer- ricyanide to the acrylamide stocks.

We employed gel gradients of three types, the compositions of which are given in Table I. The polymerized resolving gel was overlayered with stacking gel (16) which was polymerized by exposure to light. Gels of two sizes were employed. Gels, 5 x 75 mm, were used for display of protein patterns (analytical gels). Large (preparative) gels, 19 x 75 mm, were employed for radioactive samples, since such gels were capable of resolving an amount of protein (about 5 mg) required to obtain sufficient radioactivity for accurate counting. After washing the stacking gel surface with the upper reservoir buffer, samples of 200 to 300 pg of protein in 50 to 100 ~1 were applied to the small gels, and 5 mg in 1 to 2 ml, to the large gels.

The samples were mixed with a small amount of sucrose and tracking dye (bromphenol blue) and overlayered with upper buffer. Electrophoresis was performed in a Hoefer electropho- resis chamber. A current of 0.1 ma per small gel (1.5 ma per large gel) was applied until the tracking dye entered the stacking gel, at which time the current was raised to 0.2 ma per small gel (3 ma per large gel). When the dye had entered the re-

by guest on April 27, 2020

http://ww

w.jbc.org/

Dow

nloaded from

2576 Relative Turnover of Membrane Proteins Vol. 246, No. 8

TABLE I

Volume of acrylamide solutions used in preparing gradient gels

Lower case letters indicate 5 X 75 mm gels (analytical gels); upper case letters indicate the corresponding gradients of 19 X 75 mm gels (preparative gels). Gel a was used in comparing the electrophoretic patterns of six subcellular fractions. Gels b and B were employed in the double label studies on protein constituents of TX-insoluble SER, plasma membrane, and the supernatant fraction. Electrophoresis of the double labeled proteins of the TX-soluble SER were performed on Gels c and C.

5X75mm a.......... b c

19 X 75 mm A B c

Volume of acrylamide solutions

4% / 7.5% / 10% / 12%

0.1 0.1 0.1

1.4 1.4 1.4

ml

0.2 0.G 0.1 1.3 1.4

2.6 8.0 1.4 17.0

18.5

0.G

8.0

solving gel the current was raised 0.1 ma per small gel per hour (1.5 ma per large gel per hour) until a final current of 0.5 ma per small gel (7.5 ma per large gel) was reached. The samples were run at low current to avoid bowing of the protein bands. The small gels were stained with 1% (w/v) of Amido Schwartz in 10% acetic acid for 2 hours and destained in 10% acetic acid. The large gels were frozen on a slab of Dry Ice and stored at 18” until sliced.

For sectioning, a large gel was placed in the shallow trough of a Plexiglas mount and a slab of Dry Ice was placed over the mount to keep the gel frozen during sectioning. The gel was attached to the end of a l-mm thread screw which was used to advance the gel along its length. Gel slices 1 mm thick were cut as the gel advanced through the cutting plane of a razor blade held perpendicularly to the long axis of the gel. Each slice was placed in a scintillation vial containing 0.5 ml of 30% Hz02, capped tightly, and incubated at 37” for 48 hours. Following this means of dissolving the gel, 0.5 ml of “Hyamine” (New England Nuclear) and 10 ml of modified Bray’s solution (18) were added. The scintillation mixture cleared after being shaken at 37” overnight. The samples were counted in a Pack- ard Tri-Carb model 332C. A counting efficiency of approximately 15’% for tritium and 55% for 14C was obtained for all concen- trations of acrylamide present in the sample slices (i.e. from 4 to 12% acrylamide). Quenching was monitored routinely by a channels ratio method. Differential quenching from vial to vial was not encountered. Spillover of i4C counts into the tritium channel was subtracted from total counts in the tritium channel to yield net tritium counts. This spillover was about 35oj, of the total counts present in the i4C channel. There was no spillover of tritium counts into the 14C channel.

The small stained gels were scanned in a Gilford model 2000 spectrophotometer equipped with a gel-scanning attachment at a wave length of 600 rnl.c. The isotope-labeling pattern was superimposed on the optical density tracing by matching main features of the short term isotope incorporation pattern with corresponding features of the optical tracing. Once the corre-

spondence between the two patterns was established, the 3H : 14C ratios were plotted on the background of optical density profiles. Typical data are those of Figs. 4 to 9. Proteins used as molecular weight markers (see Figs. 1 and 2) were obtained from commer- cial sources. About 20 pg of marker protein were applied to analytical gels and electrophoresis was carried out as described for membrane proteins.

Column Chromatography-Proteins of liver supernatant were fractionated on Sephadex G-200 in the presence of SLS. Se- phadex G-200 was equilibrated with 0.05 M Tris-glycine buffer, pH 8.9, containing 0.1% SLS, identical with the upper buffer used in gel electrophoresis.

A sample of the supernatant fraction containing 30 mg of protein was heated for 5 min at 100” in the presence of 0.1% SLS before application to the column. Dextran blue and bromphenol blue were added to the sample to serve as end point markers. The flow rate was adjusted to about 12 ml per hour. Fractions of 5 ml were collected, and 1 ml of 50% (w/v) of trichloracetic acid was added to each fraction. The resulting precipitates were collected on glass filters, and incubated in 0.5 ml of “NCS” (Amersham-Searle) solubilizer for 3 hours at 37”. Ten milliliters of scintillation fluid (5 g of 2,5-diphenyloxazole and 0.3 g of 1,4-bis[2-(4-methyl-5-phenyloxazolyl)]benzene per liter of toluene) were added to the vials and counted as described above for gel slices. The counting efficiency in this method was approximately 357, for tritium and 63% for I%. Spillover of 1% into the tritium channel, above 13y0 of total W counts, was subtracted from total tritium counts to yield net tritium counts.

The column was calibrated with dextran blue (void volume), conalbumin (78,000 daltons), ovalbumin (45,000 daltons), and cytochrome c (13,000 daltons). The marker proteins were obtained from commercial sources.

Immunologic Techniques-Rabbit antiserum against whole rat serum and control rabbit serum were obtained from GIBCO, Grand Island, New York. The lyophilized antiserum was recon- stituted with 1 o/0 Triton X-100 in water. The antiserum was as effective in precipitating rat serum in the presence of 1% Triton X-100 as in its absence. In an experiment designed to determine the equivalence point of precipitation of microsomal proteins solu- bilized by Triton X-100, 50+1 aliquots of TX-soluble SER, con- taining 0.26 mg of protein per aliquot, were added to increasing volumes of antiserum in 0.4 ml (total volume) plastic “microfuge” tubes (Beckman Instruments). The volume of antiserum ranged from 0 to 250 ~1. In each tube the total serum volume was ad- justed to 250 ~1 by the addition of an appropriate volume of con- trol rabbit serum in 1% Triton X-100. The mixture was incu- bated for 2 hours at 37” and the resulting precipitate pelleted by centrifugation in a Beckman model 152 microfuge. Incubation longer than 2 hours was avoided, since upon longer incubation nonspecific precipitation of microsomal protein occurred. The precipitate was washed two times with 300 ~1 of 1% Triton X-100. The tips of the tubes (containing the precipitates) were cut off, placed in glass scintillation vials, and incubated with 0.5 ml of NC8 solubilizer overnight. Scintillation counting was per- formed as for the Sephadex G-200 column fractions. The microsomal material used in this experiment was obtained from a rat which had received 250 PCi of 3H-leucine intraperitoneally 4 hours before killing.

The TX-soluble SER proteins and the immunoprecipitate from this fraction were prepared for electrophoresis by heating the samples at 100” for 5 min in the presence of O.lyc SLS and 50

by guest on April 27, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Issue of April 25, 1971 P. J. Dehlinger and R. T. Xchimke 2577

TABLE II MEMBRANE FRACTIONS

Enzyme distributions in six subcellular fractions of rat liver Preparation of fractions and assays is described under “Meth-

ods.” That fraction containing the highest specific activity (units per mg of protein) was set at 100% and specific activity in other fractions was expressed as a percentage of that value. Un- detectable activity is denoted by dashes. These values are typ- ical of four-cell fractionation experiments.

Percentage of maximal specific activity of enzymes

Fraction

Peroxisomes . . . . . . Mitochondria. Plasma membrane. Rough ER. Smooth ER. Light ER. . . * -

% . 100

11.6 1.9 1.9 1.0 9.8

% 14.5

100.0 9.7 6.4 1.6 4.5

- - 10.4 5.2

100.0 5.4 12.6 100.0 30.3 80.0 40.5 19.0

%

mM dithiothreitol. We followed the method of Kabat and Mayer (19) for Ouchterlony double diffusion techniques, except that 1 s Triton X-100 was incorporated into the agar medium.

RESULTS

Electrophcretic Patterns of Proteins from Subcellular Fractions of Rat Liver-The degree of cross-contamination among the six subcellular fractions was determined from the distribution of the enzyme markers. Typical results are shown in Table II. The peroxisomal fraction contained about 15% of the specific activity of cytochrome oxidase that was measured for the mito- chondrial fraction. Plasma membrane was slightly contami- nated with mitochondria, but contained no peroxisomes (catalase) and little endoplasmic reticulum (glucose 6-phosphatase). The degree of plasma membrane contamination in the microsomal fractions is more difficult to evaluate since alkaline nitrophenyl- phosphatase may be an authentic component of endoplasmic reticulum, but present at a lower specific activity than in plasma membrane. More likely the ER fractions include cell mem- brane fragments as well as vesicles of ruptured endoplasmic reticulum. The progression of increasing specific activities of alkaline nitrophenylphosphatase, and decreasing specific ac- tivity of glucose 6-phosphatase among the plasma membrane and endoplasmic reticulum fractions suggests that the less dense microsomal fractions are increasingly rich in plasma membrane fragments.

Fig. 1 shows the electrophoretic patterns obtained with the six subcellular fractions. The peroxisomal, mitochondrial, plasma membrane, and RER and SER fractions have rather distinctive and characteristic patterns that are highly reproduci- ble. The SER and RER patterns are indistinguishable. The LER fraction appears to be a composite of the pattern of plasma membrane and endoplasmic reticulum. The distinctive patterns of plasma membrane and the endoplasmic reticulum, as well as peroxisomal and mitochondrial fractions, in general corroborate the results with enzyme markers indicating little contamination, except in the so-called LER fraction.

Although each fraction resulted in a number of protein bands when displayed on SLS acrylamide gels, the question is raised as to what extent the bands represent dissociated protein sub-

- _ Cytochrome C (13.000)

L PW Mito PM RER SER LER

FIG. 1. Acrylamide gel electrophoresis patterns of subcellular fractions of rat liver. Peroxisomes (Per), mitochondria (Mito), plasma membrane (PM), rough endoplasmic reticulum (RER), smooth endoplasmic reticulum (SER) , and light endoplasmic reticulum (LER) were isolated and solubilized in 0.1% SLS as outlined under “Methods.” Approximately 225 rg of protein of each fraction were applied to analytical gels of Composition a (Table I).

units, and to what extent they are undissociated protein com- plexes which exist in the native membrane, or protein aggregates which form when membranes are treated with SLS. There is ample documentation that SLS, in the presence of a reducing agent such as mercaptoethanol, is effective in dissociating soluble oligomeric proteins, and that the electrophoretic mobility of the dissociated subunits in the presence of SLS is dependent on subunit molecular weight (5, 20). With the gel electrophoresis system described under “Methods,” we have verified these findings on several known soluble oligomeric proteins, including hemoglobin, catalase, and lactate dehydrogenase. However, it cannot be assumed from this that the relatively insoluble mem- brane fractions will be completely dissociated under these con- ditions. Several factors which might prevent complete disso- ciation of the membrane proteins are considered below.

We have ruled out extensive protein aggregation on the basis of the observation that essentially all of the applied membrane sample enters the resolving gel, as judged visually on the stained analytical gels and by the distribution of radioactive counts on the large gels (see Fig. 5). Incomplete dissociation caused by disulfide linkages between protein subunits was examined by two methods: the SER was prepared for electrophoresis by heating the membrane sample at 100” for 5 min in the presence of 0.1% SLS and 50 mM dithiothreitol. In a second experiment, the SER fraction was reduced with mercaptoethanol in the presence of 8 M urea, pH 8.5, and carboxymethylated with excess iodo- acetic acid (21), prior to solubilization in 0.1% SLS. In both cases, the electrophoretic patterns were similar to that of the SER fraction seen in Fig. 1. In addition, the same electro- phoretic patterns were obtained when SER had been first washed with chloroform-methanol (2: 1) or ethanol-ether (3 : 1) to remove lipid. Nor was the electrophoretic pattern affected when the SER fraction was precipitated with trichloracetic acid prior to solubilization in SLS. Thus it appears that factors which might be expected to prevent complete dissociation of the membrane proteins, such as disulfide linkages or protein-lipid interactions, had no effect on the electrophoretic distribution of SER pro- teins. These considerations suggest that membrane proteins

by guest on April 27, 2020

http://ww

w.jbc.org/

Dow

nloaded from

2578 Relative Turnover of Membrane Proteins Vol. 246, No. 8

are indeed resolved on the basis of monomeric units by the electrophoresis procedure employed in these studies. The possibility that some of the protein bands result from stacking at the gradient discontinuities in the gel is unlikely, since only at the 4 to 7% acrylamide interface does a protein band exist (Fig. 1).

The general features seen in Fig. 1, including the distinctive patterns and the lack of common, major protein species, agree with previous electrophoretic studies comparing mitochondria and microsomes (22, 23) and various membranes (plasma mem- brane, smooth and rough ER, mitochondria, and Golgi mem- brane) of bovine liver (24). Another consistent finding has been the presence of major protein components in the molecular weight range of 45,000 to 70,000 in all subcellular membrane fractions (22, 23). In agreement with several studies which indicate similarities between rough and smooth microsomal

SUBFRACTIONATION OF SMOOTH MICROSOMES

-- - -I -L.J TOIOI TX-Sduble TX-Insoluble TX-Soluble TX-Insoluble

FIG. 2. Acrylamide gel electrophoretic patterns of total SER and TX-soluble and TX-insoluble subfractions. Fractionation of SER on the basis of solubility in Triton X-100 is described under “Methods.” The three gels at the left were obtained with gels of Composition a (Table I). The two gels at the right show the same material on gels of Composition c (TX-soluble SER) and b (TX- insoluble SER). Approximately 225 pg of protein were applied to each gel.

500 0

0 ul ANTISERUM

FIG. 3. Precipitation of rat serum proteins by rabbit antirat serum. TX-soluble SER was prepared as described under “Meth- ods.” To 50 ~1 of TX-soluble SER (265 pg of protein) increasing volumes of rabbit antirat serum in 1% Triton X-100 were added. The total number of counts contained in the fractions (3250 cpm) was determined by precipitation of 50 ~1 of TX-soluble SER with 10% trichloracetic acid. Washing and counting procedures are described under “Methods.”

fractions as based on enzyme activities (9, 25, 26) we find no difference between the electrophoretic patterns of these two fractions. Also noteworthy is the correspondence between the marker protein catalase and a major band of the peroxisomal fraction, of which catalase may represent up to 15% of the total protein (27).

The smooth endoplasmic reticulum fract.ion was further fractionated on the basis of solubility of proteins in 1% Triton X-100 as outlined under “Methods.” This fractionation was undertaken to increase the resolution of proteins and to determine the extent to which the membrane componenbs as displayed on acrylamide gels represent plasma proteins in transit through endoplasmic reticulum. This latter point becomes of particular importance in the studies on turnover of the membrane com- ponents. The electrophoretic patterns of TX-soluble and -insoluble fractions, each comprising approximately 50% of the total microsomal protein, are shown in Fig. 2. The two fractions have distinctive patterns; most of the prot,eins in the TX-soluble fraction have molecular weights greater than 5O,OOO,’

6

I

2 ”

,

FRACTION NUMBER

i 100

FIG. 4. Electrophoretic patterns of immunoprecipitate from SER. Serum proteins present in the TX-soluble SER subfraction were precipitated by a rabbit anti-rat serum antibody preparation. The immunoprecipitate obtained from 5 mg of TX-soluble SER treated with an excess of antibody (see Fig. 3) was washed, solu- bilked, and electrophoresed on a large gel, Composition C (Table I), and subsequently fractionated for radioactive counting as de- scribed under “Methods.” Identical electrophoresis was per- formed with 5 mg of TX-soluble SER protein. The animal from which the SER was prepared had been injected with 250 pCi of 3H-leucine 4 hours prior to death. The upper box shows the optical density scan of an analytical gel to which 225 rg of TX-soluble SFR were added. The lower box shows the radioactivity patterns of the TX-soluble SER (----) and the immunoprecipitate from this fraction (- - -).

by guest on April 27, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Issw of April 25, 1971 P. J. Dehlinyer and R. T. Xchimlce 2X9

whereas the TX-insoluble fraction contains proteins of a more heterogeneous molecular weight distribution. The presence of 1 y;, Triton X-100 in the SER fraction did not alter the elec- trophoretic pattern of the sample on SLS gels. A gel of pre- dominantly 7%) acrylamide was effective in fractionating the TX-soluble proteins. A gel containing predominantly 10% acrylamide was used to display more effectively the proteins of the TX-insoluble fraction.

As can be seen in Fig. 2, a major protein band of the TX- soluble fraction has migratory propert.ies similar to serum al- bumin. This raises the question of the extent to which measure- ments of protein turnover in endoplasmic reticulum reflect the passage of serum proteins through the ER. Several techniques were used to identify secretory protein released from SER by treatment with Triton X-100. The studies were performed with microsomes from animals that had been injected with Wleucine 4 hours before killing. The results indicate that at most 12%

of the radioactivity solubilized by 1% TX-100 represents serum proteins. One such technique was the use of an antiserum against rat serum. Preliminary results with Ouchterlony double diffusion and quantitative precipitation techniques indicated that 1 y0 Triton X-100 did not alter the ability of this antiserum to precipitate serum components. As shown in Fig. 3, this antiserum was capable of precipitating labeled SER proteins to a plateau of 12% of the radioactivity solubiliaed by 1% Triton

t A.A. INCORPORATION

3H - 1000 ‘4c --

FIG. 5. Electrophoretic pattern of radioactivity and optical density of TX-insoluble HER proteins. An animal received a simultaneous inject,ion of 50 pCi of 14C-leucine and 250 &i of 3H- leucine, and was killed 4 hours later. The TX-insoluble SER was prepared for electrophoresis and counting as described under “Methods,” by using Gel compositions b and B (Table I). This experiment is a control to indicate the number of counts obtained, and the range of 3H:14C ratios that result from experimental error. The brackets indicate 1 S.D. The upper box shows the optical scan of the stained analytical gels; the two lower boxes show actual ?H and 1% levels and the 3H:l4C ratios obtained.

X-100. Addition of antiserum beyond 0.4 ml antiserum per mg of SER TX-soluble protein did not further increase the amount of labeled protein precipitated. That such precipitated protein is compa,rable to serum proteins was indicated by the esistence of common precipitation bands between serum and solubilized membrane proteins as evidenced by Ouchterlony double diffusion (not shown). Fig. 4 shows the SLS acrylamide gel-labeling pattern of the immunoprecipitates compared wit,h that of the total TX-soluble fraction from which the immunoprecipitate was derived. Four radioactive protein peaks are observed in the solubilized precipitate, one of which is a protein (or group of proteins) of molecular weight 60,000 to 70,000, and presumed to represent serum albumin. The identity of the other proteins is not known. We have used the method of Campbell, Green- gard, and Kernot (28) for isolating serum albumin from liver microsomes and have found that the serum albumin isolated constitutes at most 2.5% of the TX-soluble microsomal radio- activity and 0.40/, of the TX-insoluble radioactivity. These experiments, then, indicate that secretory proteins, which are presumed to be trapped within vesicles during homogenization, and not released by the washing procedures employed, con- stitute only a small portion of the proteins being studied by isotope labeling.

Turnover of Membrane Subfractions-We have employed a double isotope technique first described by Arias et al. (1) for measuring the relative turnover rates of the protein constituents of the Triton X-100 subfractions of SER and plasma mem- brane. In this method two isotopic forms of the same amino acid, 3H- and 14C-leucine, are used to establish two time points on the decay curve resulting from degradation of the protein. In the experiments report,ed here, an animal was injected initially

- 1.8

- I.4 f I I

4.

- 1.0

L I 1 I I I I

0 20 40 60 80 100 FRACTION NUMBER

FIG. 6. Relative degradation rates of TX-soluble SER proteins. 3H-Leucine (250 pCi) was administered 4 days after administration of 100 PC1 of 14C-leucine. Protein (225 rg) was applied to an analytical gel, Composition c; 5 mg to a preparative gel, Composi- tion C. The 3H:14C ratios were matched to the optical scan as described in the text. The control 3H-14C values are those from the animal of Fig. 5, where the brackets indicate 16.D.

by guest on April 27, 2020

http://ww

w.jbc.org/

Dow

nloaded from

2580 Relative Turnover of Membrane Proteins Vol. 246, No. 8

1.6 .6 n I I OD -

I I I t 1 I 0 20 40 60 00 100

FRACTION NUMBER

FIG. 7. Relative degradation of TX-insoluble SER proteins. The proteins used from this experiment are labeled as described in Fig. 6. The control *H:W ratios are those from the animal of Fig. 5. The protein peak is in the molecular weight range of about 50,000.

with W-leucine, followed 4 days later by administration of aH-leucine, with killing after 4 hours. For any protein species the level of 3H radioactivity indicates short term incorporation, whereas the 14C radioactivity indicates the amount of decay after 4 days. Proteins that are turning over rapidly, i.e. syn- thesized rapidly (high 3H counts) and degraded rapidly (low 1°C counts) will have high aH:14C ratios. If all proteins are turning over at the same rate the ratios will be constant. The assumptions and discussion of the validity of the method are discussed by Arias et al. (1).3

To determine the limit of statistical error inherent in the double label method as applied to gel electrophoresis, a standard control experiment was performed in which an animal was given both W- and aH-leucine simultaneously. The results are seen in Fig. 5, which shows the optical density tracing, the pattern of radioactivity, the typical counts obtained, and the standard error of ratios obtained in such a control experiment. The results of the experiments to determine the relative turnover rates among the proteins of the two Triton X-166 subfractions of SER are given in Figs. 6 and 7. The upper frames show the optical gel scan of the stained analytical gels and the aH:14C ratios determined from the preparative gels. The control 3H : W ratios are those from the control experiment just described. Several features of the aH :I% ratio patterns are common to both of the SER subfractions.

1. There is a marked heterogeneity in the ratios among the

3 A comparison of rate constants of degradation, Kd, among pro- teins, is given by In 3H:l”C ratios. Although it is theoretically possible to obtain absolute rate constants of degradation by this method, knowing the total initial incorporation of radioactivity into the protein (as well as in the specific protein) (4)) we have not chosen to do so, since such estimates are grossly inaccurate be- cause of the fact that we are using the reutilizable amino acid leucine (see Reference 29 for discussion of this problem).

01 v ’ ’ ’ ’ ’ 0 ’ ’ J

400

g200

0

, I

0 20 40 60 80 FRACTION NUMBER

FIG. 8. Relative degradation of proteins of the plasma mem- brane fraction. Animals were given 250 pCi of aH-leucine 4 days after receiving 100 pCi of W-leucine. The plasma membrane fraction was prepared from the livers of four double labeled rats. Conditions of electrophoresis are similar to those in Figs. 4 to 7. Gels of Compositions b and B were used.

protein constituents, again supporting the concept of hetero- geneity of degradation rate constants. There is a general correlation between the 3H:14C ratios and the migration distance in the gel. Because migration in SLS acrylamide gel reflects subunit molecular weight (5, u)), we infer that turnover rates, in general, are related to subunit molecular size.

2. Not all proteins conform to this size degradation rate cor- relation. In Fig. 6 two definite 3H:14C peaks are seen. The nature of these proteins is not known, but on the basis of the exclusion of serum proteins as significant labels in these molecular weight ranges, it appears unlikely that they represent serum protein in transit. The major protein band in the TX-insoluble fraction is particularly interesting, since it is a protein (or group of proteins) with an unusually high turnover rate. Because of the studies of Omura and Sato (30) indicating that, the carbon monoxide-binding protein, P-456, is a major protein constituent of the endoplasmic reticulum, amounting to approximately 16% of total microsomal protein of rabbit liver, it seems possible that this major protein species (or a component of it) represents the hemoprotein P-450. In support of this, we have purified par- tially the CO-binding pigment (30) from rat liver microsomes and found that the fraction contains a protein whose migration rate on SLS gels is identical with the major protein band of the TX-insoluble fraction.4

The relative turnover rates of protein constituents of the plasma membrane are shown in Fig. 8. The results show several similarities with the SER subfractions. First, heterogeneity

4 P. J. Dehlinger and R. T. Schimke, manuscript in preparation.

by guest on April 27, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Issue of April 25, 1971 P. J. Dehlinger and R. T. Schimke 2581

A.A. INCORPORATION

FIG. 9. Relative degradation of proteins of the supernatant fraction of rat liver as separated by SLS-acrylamide gel electro- phoresis. The isotope administration schedule was similar to that described in Figs. 6 to 8. Supernatant protein (5 mg) prepared as outlined under “Methods” was electrophoresed. Methods of preparation, electrophoresis, and counting are outlined under “Methods.” Gels of Compositions c and C were used.

is indicated by the variation in 3H:14C ratios. Second, the general correlation between 3H:14C ratios and protein size is evident.

Turnover of Supernatant Proteins--The finding of a general correlation between subunit size and rate of degradation in the membrane fractions prompted an investigation of the charac- teristics of turnover of proteins in the supernatant fraction from rat liver. Experiments similar to those employed with the mem- brane fractions, but using supernatant fractions as the source of labeled protein, are shown in Fig. 9. The same general cor- relation between size and degradation rates exists with these proteins. Since in control experiments this systematic variation does not occur, it cannot be ascribed to a systemic error in counting. Furthermore the same results are obtained when the isotopes are reversed (i.e. when the aH-leucine is administered before the 14C-leucine), and when the animals are fed throughout the experiment, as opposed to the B-hour fast prior to isotope injection used in the experiments reported. The same results were obtained when the supernatant fraction was treated with 0.1% sodium lauryl sulfate and 50 mM dithiothreitol, indi- cating that disruption of the disulfide bonds has no effect on the size distribution of subunits.

We have also subjected the supernatant proteins to chro- matography on Sephadex G-200 in the presence of 0.1% SLS, as seen in Fig. 10. The difference in actual aH:14C ratios com- pared with the corresponding gel electrophoresis experiment (Fig. 9) is caused by the difference in counting efficiencies between the two counting procedures employed (see “Methods”). The

2.0

0 / I I I I

Dentran blue Gawltumin Ovolbwrdn Cytochmms C NoId) V6.000) (45.OooO) (13.000)

I 1 I I

8000

A.A. INCORPORATION

3H- 6000 l4c ---

B 4000

_:I ;:----;.A 30 40 50 60 70 0

FRACTION NUMBER

FIG. 10. Relative degradation of proteins of the supernatant fraction of rat liver as separated by chromatography on Sephadex G-200 in the presence of 0.1% SLS. The source of the labeled protein is the same as that used in Fig. 9. The control animal received a simultaneous injection of 50 pCi of 14C-leucine and 250 pCi of aH-leucine and was killed 4 hours after injection. Approxi- mately 30 mg of protein were applied to the column. The migra- tion rates of marker proteins were determined in a separate run. Details are given under “Methods.” --, 3H:14C ratios in degra- dation experiment; - - -, 3H:14C ratios in control experiment.

correlation between turnover rates and molecular weight is again evident, with an independent method for separating pro- teins on the basis of size.

DISCUSSION

The experiments reported in this paper were addressed to the question of how biologic membranes are degraded, and indirectly, how they are assembled. Studies of Arias et al. (l), Omura, Siekevitz, and Palade (2), and Kuriyama et al. (31) have indi- cated that the endoplasmic reticulum fractions of rat liver are turning over rapidly, with a mean half-life of approximately 2 to 3 days. In addition Arias et al. (1) and Widnell and Siekevitz (3) have found that plasma membrane turns over at the same over-all rate as the endoplasmic reticulum. The lipid moieties of the endoplasmic reticulum likewise display rapid turnover (2, 32). The turnover characteristics of membrane components can suggest one of two general models for the assembly and breakdown of cellular membranes.

1. All of the major components of the membrane are syn- thesized and degraded at the same rate, indicating that the membrane is synthesized and degraded as a multiprotein unit; or

2. Turnover rates of membrane components are heterogeneous, suggesting that the membrane is in a continual state of flux, in which individual component proteins are assembled at different

by guest on April 27, 2020

http://ww

w.jbc.org/

Dow

nloaded from

2582 RelaG3 Turnover of Membrane Proteins Vol. 246, No. 8

rates, and subsequently dissociate, or are degraded while re- maining attached to the membrane, at rates characteristic for the individual proteins.

Fundamental to this problem is the question of the protein composition of cellular membranes. Are cellular membranes composed of a major structural protein to which lipid and en- zymatic components are attached (33, 34), or are they a com- posite of enzymatic proteins in association with lipid elements in the membrane? In the former case the structural protein by itself might control the turnover rate of the membrane. The question of membrane composition has been investigated in several recent studies utilizing SLS acrylamide gel electrophoretic techniques similar to those employed in this study (22, 23). The findings of Fig. 1 are in agreement with these other investigations in suggesting that there is no protein which comprises a majority of the membrane protein, and no single major protein that is common to all membranes.

Our results show that membrane proteins as separated by SLS acrylamide gel electrophoresis are heterogeneous with respect to degradation rate constants. This finding is in agreement with our own earlier studies involving the isolated membrane pro- teins, cytochrome bs and cytochrome c reductase, and non- identified proteins solubilized by Triton X-100 and separated by chromatography on DEAE-cellulose (l), as well as the findings of Omura et al. (2) and Kuriyama et al. (31) on cytochrome bg and cytochrome c reductase turnover. In these studies the half-lives of cytochrome bs and cytochrome c reductase were of the order of 120 and 75 hours, respectively. More recently Bock, Siekevitz, and Palade (4) have studied a NAD-glyco- hydrolase solubilized from both endoplasmic reticulum and plasma membrane of rat liver, and have found its half-life to be approximately 17 days as isolated from either membrane fraction. All these studies support the concept that membranes are in a state of flux, in which the membrane constituents are degraded asynchronously. Furthermore studies by Arias et al. (1) and Kuriyama et al. (31) have shown that the rates of synthesis of the two identifiable proteins of membranes, cytochrome c re- ducttLse and cytochrome bg, can be affected differentially by administration of phenobarbital.

It should be noted that heterogeneity of turnover, by itself, does not preclude the possibility that the membrane fraction studied consists of several membrane types, each with a charac- teristic turnover rate. If this were the case an appropriate means of subfractionating the membranes into homogeneous types would eliminate the observed heterogeneity of turnover. We have attempted such a fractionation of the smooth endo- plasrnic reticulum by isopycnic centrifugation in 20 to 45% sucrose gradients. Four discernible fractions can be separated. These all had similar electrophoretic patterns on SLS acrylamide gels and similar mean rates of turnover as measured by the double isotope ratios of total membrane proteins2 By the criterion of density, then, we did not find heterogeneity of mem- brane types within the SER fraction.

The identity of the proteins displaying a heterogeneity of turnover in plasma membrane and endoplasmic reticulum as separated by sodium lauryl sulfate-acrylamide gels is, of course, essentially unknown. The experiments described in Figs. 3 and 4 suggest that these do not represent simply those proteins in transport to the serum. Since these results are based in large part on immunologic reactivity of the solubilized endoplasmic reticulum protein, we cannot discount at present the possibility

that protein destined for transport that is present in endoplasmic reticulum vesicles may not react with an antibody to the serum proteins. One major peak in the Triton X-insoluble fraction is of particular interest, since it constitutes such a large portion of the membrane protein, and has an “anomalously” high turnover rate (see Fig. 7). That this constitutes a major portion of the protein of this fraction, and has migratory properties similar to the solubilized form of the hemoprotein, cytochrome P-420, suggests that this band is, indeed, cytochrome P-450. This is supported further by our finding that isotopic amino acid incor- poration is preferentially stimulated in this protein band in response to phenobarbital.4 It is well established that the level of P-450 is increased in response to phenobarbit,al (35). The role which this protein may have in the regulation of turnover in ER is a question we are now investigating.

Heterogeneity of turnover and, in particular, the correlation between degradat.ion rate and molecular size of the protein com- ponents of membrane raise a number of questions concerning the mechanism of membrane degradation. Are the membrane proteins degraded in situ while associated with the membrane? Morrison and Neurath (36) have reported that erythrocyte membranes contain potent proteases, suggesting the presence of degradative enzymes within the membranes. Given the correlation between turnover rates and molecular size, this would require that the larger proteins (or subunits) are more accessible to the degradative enzymes present in the membrane. Such a membrane might be pictured as a core of smaller proteins enclustered by larger proteins. On the other hand, the degrada- tive enzymes may be in the so-called soluble fraction of the cell, such that those proteins more peripheral on the membrane might be degraded more readily. Trypsin treatment of the endo- plasmic reticulum preferentially releases the cytochrome bs,

presumably indicating that cytochrome b5 is more accessible. However cytochrome bs is actually turning over less rapidl!

(2). Perhaps a more attractive explanation of the heterogeneity of

membrane degradation involves the concept that the degradation proceeds through an init’ial dissociation of the protein from the membrane. Hence degradation of membrane proteins would proceed by the same protein degradative mechanism or mecha- nisms as the so-called soluble proteins. As a corollary of this proposal one might suggest that membrane proteins are actually in equilibrium between associated and dissociated forms. The heterogeneity of degradation, then, would be determined both by the equilibrium between associated and dissociated forms of the membrane protein, and by the normal controls of degrada- tion of soluble proteins. It is of note that for both soluble and membrane proteins, degradation rates are related to molecular size. Further, we have found the correlation between size and degradation rates to occur with ribosomal proteins.5 In this connection, it does appear that ribosomal proteins have a cyto- plasmic pool (37).

The notion that membrane proteins can have independent existence in the cytoplasm is not new. Studies on the genesis of chloroplasts (38) and mitochondria (39, 40) indicate that at least some membrane proteins are synthesized at sites in the cytoplasm and are subsequently incorporated into the developing membrane. Whether dissociated proteins are in true equilib- rium with those of the membrane is an intriguing problem, since

5 F. Dice and R. T. Schimke, manuscript in preparation.

by guest on April 27, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Issue of April 25, 1971 P. J. Dehlinger and R. T. Schimke 2583

if so, it would suggest that self-assembly of membranes could occur without specific enzymatic processes or the juxtaposition of specific ribosomes with a growing membrane.

The correlation between molecular weight and general tend- ency for more rapid degradation of proteins of membrane frac- tions and soluble proteins is a new finding concerning the prop- erties of protein degradation in rat liver. A number of studies have shown exponential decay or uptake of isotope into specific proteins (41-45), including the membrane-associated cytochrome b5 and cytochrome c reductase (2, 28), thereby implying that the degradation of any given molecule from among like molecules is a random event. This finding, plus the fact that specific pro- teins have characteristic half-lives, has led to the concept that it is the protein as a substrate which determines its rate of degradation (30). One simple interpretation of the correlation between size and degradation rate is that the larger molecule has a greater chance of a protease making an initial (and rate- limiting) “hit,” i.e. peptide bond cleavage.

An alternative possibility comes from the proposal of Segal et al. (46) that lysosomes act essentially as sieves, in which pro- teins may exit from lysosomes as well as enter and be degraded. If small molecules were to exit more readily from lysosomes, the general correlation between size and degradation rate could also be explained. It should be emphasized, however, that molecular size is by no means the sole determinant of degradation rate, since three soluble enzymes of rat liver, arginase (molecular weight 118,000, four subunits) (47), tyrosine aminotransferase (molecular weight 120,000, three subunits) (48), and lactate dehydrogenase-isozyme 5 (molecular weight 136,000, four sub- units) (49) have half-lives of 4 days (40), 13 hours (41), and 16 days (50), respectively. Perhaps these differences in rates of turnover relate to the rate of dissociation of these different proteins into subunits. Thus molecular size is only one of several parameters which determine the susceptibility of a protein to degradation, in addition to the nature of peptide bonds exposed, the conformation of the protein as affected by interactions with small molecules and macromolecular ligands, etc. Knowledge of the role each of these parameters plays, and of the actual mechanism of protein degradation, awaits localization and isolation of intracellular degradative enzymes.

1.

2.

3.

4.

5.

6.

7. 8.

9.

10.

REFERENCES

ARIAS, I. M., DOYLE, D., AND SCHIMKE, It. T., J. Biol. Chem., 244, 3303 (1969).

OMUR~, T., SIEKEVITZ, P., AND PALADE, G. E., J. Biol. Chem., 242, 2389 (1967).

WIDNELL, C. C., AND SIEKEVITZ, P., J. Cell Biol., 36, 142A (1967).

BOCK, K. W., SIEKEVITZ, P., AND PALADE, G. E., Fed. Proc., 29, (1970).

SHAPIRO, A. L., VINUELA, E., AND MAIZEL, J. V., JR., Biochim. Biophys. Res. Commun., 28, 815 (1967).

LEIGHTON, F., POOL, B., BEAUFAY, H., BAUDHUIN, P., COFFEY, J. W., FOWLER, S., AND DE DUVE, C., J. Cell Biol., 37, 483 (1968) .

NEVILLE, D. ill., JR., J. Biophys. Biochem. Cytol., 8,413 (1960). EMMELOT, P., Bos, C. J., BENEDETTI, E. L., AND RUMKE, P.,

Biochim. Biophys. Acta, 90, 126 (1964). DALLNER, G., Acta Pathot. Microbial. &and. Suppl., 166

(1963). BEERS, R. F., JR., .IND SIZER, I. W., 1. Biol. Chem., 196, 133

(1952).

11. COOPERSTEIN. S. J.. AND LAZAHOW. A.. J. Biol. Chem.. 189.

12.

13.

14.

15. 16. 17. 18.

19.

20. 21.

22.

23. 24.

25.

26.

27.

28.

29.

30. 31.

32.

33.

34.

35.

36.

37. 38.

39.

40. 41. 42. 43.

44.

45.

46.

47.

48.

49.

50.

665 (1951). ’ ’ I , , ,

NORDLIE, R. C., AND ARION, W. J., J. Biol. Chem., 239, 1680 (1964).

DE Duv~, C., PRESSMAN, G. C., GIBNETTO, R., W~TTIAUX, R., AND APPELMANS, F., Biochem. J., 60,604 (1955).

RAY, Y. K., LIEBERMAN, I., AND L.~NSING, A. I., &o&em. Biophys. Res. Commun., 31, 54 (1968).

PETERS, T., JR., J. Biol. Chem., 237, 1186 (1962). GROSSFELD, B., Ph.D. thesis, Stanford University, 1968. DAVIS, B. J., Ann. N. Y. Acad. Sci., 121,404 (1964). CAREY, N. H., AND GOLDSTEIN, A., Biochim. Biophys. Acta,

66, 346 (1962). KABAT, E. A., .4ND MAYER, M. M., Experimental immunochem-

istry, Charles C Thomas, Springfield, Illinois, 1961, p. 22. WEBER, K., AND OSBORN, M., J. Biol. Chem., 244, 4406 (1969). RENAUD, F. L., ROWE, A. J., AND GIBBONS, I. R., J. Cell Biol.,

36, 79 (1968). SCHNAITMAN, C. A., Proc. Nat. Acad. Sci. U. 8. A., 63, 412

(1969). KIEHN, E. D., .~ND HOLLAND, J. J., Biochemistry, 9,1729 (1970). ZAHLER, W. L., FLEISHER, B., AND FLEISH~R, S., Biochim.

Biophus. Acta. 203, 283 (1970). ORRE&&, S., ERICSSON, j. L.‘E., AND ERNSTER, L., J. Cell

Biol., 26, 627 (1965). GLAUMANN, H., VON DER DECKEN, A., AND DALLNEH, G., Life

Sci., 7, 905 (1968). POOLE, B., LEIGHTON, F., AND DE DUVE, C., J. Cell Biol., 41,

521 (1969). CAMPBELL, P. N., GREENGARD, O., AND KF:RNOT, B. A., Bio-

them. J., 74,107 (1960). SCHIMKE, R. T., in B. L. HORECKER AND E. R. STADTM~N

(Editors), Current topics in cellular regulation, Vol. 1, Aca- demic Press, New York, 1969, p. 77.

OMURA, T., AND SATO, R., J. Biol. Chem., 239,2370,2379 (1964). KURIYAMA, Y., OMURA, T., SUXEVITZ, P., AND PALADE, G. E.,

J. Biol. Chem., 244, 2017 (1969). HOLTZMAN, J. L., AND GILLETTE, J. R., J. Biol. Chem., 243,

3020 (1968). GREEN, D. E., TISDALE, H. D., CRIDDLX, R. S., CHEN, P. Y.,

AND BOCK, R. M., Biochem. Biophys. Res. Commun., 6, 102 (1961).

WOOD~ARD, D. O., AND MUNKRES, K. D., Proc. Xat. Acad. Sci. U. S. A., 66, 872 (1966).

GREIM, H., SCHENKMAN, J. B., KLOTZBUCHER, M., END RISM- MER, H., Biochim. Biophys. Acta, 201,20 (1970).

MORRISON, W. L., AND NEURATH, H., J. Biol. Chem., 203, 39 (1953).

WARNER, J. R., J. Mol. Biol., 19,383 (1966). HOOBER, J. K., SIEKEVITZ, P., *END PALADE, G. E., J. Biol.

Chem., 244, 2621 (1969). HENSON, C. P., PERLMAN, P., WEBEL, C. N., END MAHLER,

H. R., Biochemistry, 7, 1012 (1968). BEATTIE, D. S., J. Biol. Chem., 243, 4027 (1968). SCHIMKE, R. T., J. Biol. Chem., 239, 3808 (1964). KENNEY, F. T., Science, 166, 525 (1967). SCHIMKE, R. T., SWEENEY, E. W., AND BERLIN, C. M., J.

Biol. Chem., 240, 322 (1965). JOST, J.-P., KHAIRALL‘IH, E. A., AND PITOT, H. C., J. Biol.

Chem., 243, 3057 (1968). STICK, R. W., REXROTH, A. K., .IND STANGE, J. L., J. Biol.

Chem., 243, 3581 (1968). SEG.~L, H. L., MATSUZAWA, T., HAIDER, M., AND AURAI-IAM,

G. J., Biochem. Biophys. Res. Commun., 36, 764 (1969). HIRSCH-K• LB, H., AND GREENBERG, D. M., J. Biol. Chem.,

243, 6123 (1968). ~~~~~~~~~~~ F. A., AURICCHIO, F., TOMKINS, G. M., AND RIIXY,

L)., J. Biol. Chem., 244, 3618 (1969); Biochim. Biophys. Acta, 221, 307 (1970).

CASTELLINO, F. J., AND BARKER, R., Biochemistry, 7, 2207 (1968).

FRITZ, P. J., VESSELL, E. S., WHITE, E. L., AND PRUITT, K. M., Proc. Nat. Acad. Sci. U. 8. A., 62, 558 (1969).

by guest on April 27, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Peter J. Dehlinger and Robert T. SchimkeDegradation

Size Distribution of Membrane Proteins of Rat Liver and Their Relative Rates of

1971, 246:2574-2583.J. Biol. Chem. 

  http://www.jbc.org/content/246/8/2574Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/246/8/2574.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on April 27, 2020

http://ww

w.jbc.org/

Dow

nloaded from