Chem.-Biol. Interactions, 11 (1975) 207-224
0 Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands
THE GOLGI COMPLEX III. THE EFFECTS OF PUROMYCIN GLYCOPROTEIN SYNTHESIS
ON ULTRASTRUCTURE
JENNIFER M. STURGESS, MARDIE M. MITRANIC
AND
207
AND
MARIO A. MOSCARELLO
The Departments of Pathology and Biochemistry, Research Institute, Hospital for Sick Children, 5.55 University Avenue, Toronto, Ontario (Canada M5G 1X8)
(Received February 4th, 1974) (Revision received January lOth, 1975) (Accepted January 23rd, 1975)
SUMMARY
The effect of puromycin has been investigated on protein and glycoprotein synthesis and on ultrastructure of the Golgi complex from rat liver. Incorporation of [14C]leucine into protein in Golgi fractions and into serum proteins was depressed rapidly after puromycin treatment. In the serum proteins, incorporation returned to normal levels at 2 h whereas in Golgi fractions it continued to rise to 200% of the control levels at 3 h and was still elevated at 24 h after puromycin treatment. Incorporation of [W]glucosamine into glycoprotein was depressed in Golgi and sekum fractions in a similar manner but slightly later than that of leucine. Leucine labelled material found at 3 h was a poor acceptor for carbohydrate, since [Wlglucosamine incorporation was not elevated above control values. Galactosyl transferase activity was not depressed in the Golgi membranes and, at 3 h, was elevated implying that an adequate supply of enzyme was available at all times. The activity of the galactosyl transferase in serum appeared to bc depressed suggesting that transport of enzyme from Golgi complex to serum was defective. Ultrastructural changes in the Golgi complex were observed to occur rapidly after puromycin treatment. The cistecnae became irregular, compressed, and degenerated progressively from central region towards the periphery. Irregular tubular structures formed at the expense of cisternal membrane and showed accumulation of low density lipoprotein. Vesiculation and degenerative changes of the GO& membranes continued from 2-l 2 h while more typical arrangements of the Golgi complex were observed between 24-48 h. The morphological changes correlated with changes in glycoprotein synthesis.
Abbreviation: TCA, trichloroacetic acid.
208 INTRODUCTION
The action of puromycin on the synthesis of secretory protein has been well documented and it has been shown that puromycin acts at the ribosomal level to release nascent peptide chains by replacing aminoacyl transfer RNAl. This exchange inhibits completion of protein molecules. Ultrastructural and biochemical studies2 have shown that injection of puromycin causes disaggregation of polyribosomes both in vitro and in vivo but this effect is reversed rapidly. In other intracellular membrane systems, morphological changes based on electron microscopic examination of ultrathin sections, have been described in the Golgi complex in various cell types including the livers, the pancreas and thyroids, and intestinal eplthelial cells 4. The reported changes, including regression and increased vacuolation of the Golgi apparatus have been attributed to the inhibition of protein synthesis. However, there is some evidence to suggest a separate mechanism of puromycin action responsible for changes in intracellular membranes, which is distinct from that controlling synthesis and transport of secretory proteinss. Puromycin appears also to have an effect on the assembly of sugars into glycoproteins. In the thyroid gland, puromycin completely inhibits the synthesis of the protein portion of the thyroglobulin molecule whereas it has no effect or causes only partial inhibition of carbohydrate assembly in soluble thyroglobulin and particle-bound thyroglobulin respectively 6. The degree of inhibition of assembly of different carbohydrates varies and appears to be related to their arrangement in the oligosaccharide chain, so that those adjacent to the peptide core are affected to a greater extent than more terminal sugars 6~~.Since carbohydrates are assembled into the oligosaccharide chain mainly in the Golgi complex by a mechanism fundamentally different from that responsible for the synthesis of the peptide chains, it is possible that puromycin has an additional effect on the Golgi membranes which influences carbohydrate incorporation. The purpose of this study has been to investigate the effects of puromycin on the structural arrangement of the Golgi complex and on glycoprotein biosynthesis. Morphological changes after puromycin treatment have been evaluated by electron microscopy in isolated Golgi-rich fractions and in liver. The findings have been correlated with incorporation studies in the same membrane preparations using specific radioactive precursors for protein and oligosaccharide biosynthesis and assays of galactosyl transferase, an enzyme required for carbohydrate assembly which is specific to the Golgi complex. MATERIALS
AND METHODS
Experimental procedure Male Wistar rats, 200-220 g fasted for 15 h were used in these studies. A single injection of 19 mg puromycin hydrochloride pH 7.0 was given to each rat via the jugular vein: controls received a single injection of saline. After completion of injections, each rat was anesthetised with ether, exsanguin-
209 ated, a sample of blood collected, and the liver removed. A slice of liver was resected for electron microscopy and the remainder used to isolate the Golgi membranes. At each time interval, Golgi membranes were isolated from at least four rats: two were treated with puromycin and two were controls. The control and treated rats were paired, being maintained under identical conditions and receiving equivalent amounts of food and water. One rat of each pair was injected intravenously with puromycin; the other received an equal volume of saline. The Golgi fractions were prepared from each pair of rats simultaneously and the time at which injections were given was adjusted so that all rats were sacrificed at the same time (11:OO a.m.). 3so~af~u~of GoZgifructio~s
The Golgi fractions were isolated separately from each rat liver using techniques described [email protected] of the isolated Golgi membranes were used for electron microscopy, for assay of galactosyl transferase and for studies on incorporation of radioactive precursors. Ellectron microscopy
From each rat used in this study, samples of liver tissue and of the isolated Golgi fractions were examined by electron microscopy. At least 2 rats were investigated at each time interval: 0, 1,2,3, 12 and 24 h after puromycin. Tissue and pelleted membrane samples were fixed in 3 % glutaraldehyde in 0. I M sodium phosphate buffer, pH 7.4, and post-fixed in 1% osmium tetroxide in veronal-acetate buffer. After dehydration in graded ethanol solutions, liver samples were polymerized in Epon epoxy resin, and membrane samples in Spurr low viscosity epoxy resin. The methods for preparation of samples for electron microscopy have been described in detail previously9, Galactosyl transferuse assay
A portion of each Golgi fractioa prepared from rat liver at 0, 1,2,3,12 and 24 h after injection of puromycin was assayed for galactosyi transferase using jV-acetyl glucosamine as acceptor 10. The enzyme activity was calculated as dpm~mg protein per 2 h incubation. Galactosyl transferase was used to monitor purification and recovery of Golgi membranes from the homogenates. The purification factor (PF = specific activity of isolated Golgi membranes/specific activity of the original homogenate) was obtained for each Golgi preparation. These factors varied from 60-120 in different preparations and corresponded to 40-83 % recovery of galactosyi transferase activity from the total homogenate, These values were in the same range as those obtained with normal rat liver so that puromycin did not alter the separation of Golgi membranes in this procedure, Incorporation of radioactiveprecursors
The incorporation of L-[U-WZ]leucine (specific activity 278 mCi/mmole; New England Nuclear) and ~-[l-l~C]~lucosamine (specific activity 7.5 mCi/mmole;
210
New England Nuclear) were studied separately as markers for protein and glycoprotein biosynthesis respectively. For each radioactive precursor, two rats were studied at each time interval. Incorporation of L-leucine was followed at 0, 1,2,3, 12 and 24 h after puromycin injection. A single dose of 5 ,&i L-leucine was given intravenously to each rat 60 min before sacrifice. D-Glucosamine incorporation was investigated at 0, 40 and 100 min and at 2, 12 and 24 h after puromycin. A single injection of 5 yCi D-glucosamine was given intravenously to each rat 40 min before sacrifice. The Golgi and serum fractions were isolated and extracted as follows: Low molecular weight material was removed from the Golgi fraction and from serum after precipitation of the high molecular weight material with an equal volume of 10% TCA. The IC_4 precipitate was extracted with 0.6 N perchloric acid and the high molecular weight (glycoprotein) fraction was precipitated with phosphotungstic acid 11. The glycoprotein fraction isolated in this manner was referred to as seromucoid. For each fraction, protein was determined on one aliquotls and another aliquot was used for radioactive counting in a Nuclear Chicago Mark I Scintillation Counter. For counting, 100 ~1 was dissolved in 1.3 ml of hyamine hydroxide and 10 ml of toluene scintillation fluid (0.4 % PPO and 0.03 % POPOP) was added. The radioactivity was expressed as specific activity dpm/mg protein. RESULTS
The effect of puromycin on the incorporation of L-[W4C]leucine proteins
into Golgi and serum
The incorporation of L-[r4C]leucine was investigated in fractions extracted from isolated Golgi membranes and from serum at intervals 0, 1, 2, 3, 12 and 24 h after treatment with puromycin. The variation in specific activity (dpmlmg protein) of Golgi and serum fractions after puromycin treatment is summarized in Table I. Each number represents the specific activity of a treated rat expressed as a percentage of the specific activity of the paired control. At each time interval, the percent specific activities are given from two pairs of rats. Serum. At 1 h after puromycin treatment, the specific activity of the serum seromucoid fraction was depressed to roughly 28% of the control, representing a change in specific activity from 450 dpm/mg protein in the control to 148 dpm/mg protein in the treated animal. At 2 h it rose to about 60 % of the control value but remained lower at about 75% of the control value at 3 to 12 h. At 24 h the specific activity of one of the test animals (450 dpm/mg protein) was higher than that of one of the paired control (337 dpmlmg protein). The specific activity of the serum residue, i.e. the fraction insoluble in 0.6 iV perchloric acid, was depressed markedly at 1 h after puromycin to less than 15% of the control value (specific activity of puromycin treated was 73 dpm/mg protein and control was 480 dpm/mg protein), rose to 50% of the control value at 2 h, and approached the control levels at about 3 h. Golgi fractions. At 1 h after puromycin treatment, the specific activity of the seromucoid fraction of Golgi membranes, i.e. the high molecular weight fraction
211 TABLE I INCORPORATION
OF [l”C]LEUCINE
INTO GOLGI FRACTIONS AND
SERUM FOLLOWING TREATMENT
WITH
PUROMYCIN
Time altep ~~~ro~~~~~i~ (h)
Specific activity ( $, obtained ~o~~froI~ Goigi seromucoid
G&i residm
Serum sermmoid
Serum residue
1
64 50
28 -
22 33
11 15
2
89 109
89 101
61 56
56 49
3
194 252
90 120
74 54
75 98
1L
187 145
79 97
75 94
92 90
24
129 125
76 96
78 133
113 93 ~-
soluble in 0.6 N perchloric acid, was depressed 50-64 y0 compared with the controls. At 2 h the specific activity increased to almost normal levels of 3933 dpm/mg protein compared with the control values of 4431 dpmlmg protein. However, the activity continued to rise to about 200 % of the control values at 3 h after puromyein. At 12 h the specific activity fell slightly but remained at 130% of the control at 24 h after puromycin. These data implied that there was an accumulation of [W]leucinelabelled material in the Golgi seromucoid fraction at 3 to 24 h after puromycin treatment. The Golgi residue, i.e., the fraction insoluble in 0.6 N perchloric acid showed a decreased specific activity at 1 h (28% of the control). The incorporation of [W)leucine returned to normal levels at 2 h and remained in the normal range after this time. Eflect o~puro~ly~i~on the i~co$por~fion of o-f i- W]gtucosanline infoGolgiand serum pro feins
The incorporation of D-[l-‘4C]ghKOSRIIline into perchloric acid-soluble and insoluble fractions of Golgi membranes isolated from rat liver and serum, was investigated at 40 and 100 min and at 3, 12, and 24 h after injr-:tion of ~~ornycin. The variation in specific activity in Golgi and serum fractions after puromycin is summarized in Table 11.Each number represents the specific activity of the puromycintreated rats expressed as a percentage of the paired controls; at each time the percent specific activities are given for at least 3 pairs of rats. Serum. At 40 min after injection of puromycin, the specific activity of the serum seromucoid fraction was 1774 dpm/mg protein compared with 4246 dpm/mg protein in the control, so that the specific activity of the puromycin-injected animal was
212 TMLE
11
INFRACTION
OF [‘~GLU~
E IKlo
GOLGI FRACXION!J AND SERUM FGLLOWING
TRRATMEkrT
WiTH PUROMYCIN
Timesfter
Specificactivity ( $6of paired control)
40 min
49 -
38 37 39
52 42 62
SO 43 57
lOOmin
53 71 -
86 78 72
35 23 47
61 -
3h
88 90 88
ma 88 98
85 70 fo0
79 95 -
12h
94 99 90
IO8 107 107
90 9s 85
143 95 -
24h
112 78 98
103 105 -
80 89 115
140 133 -
only 42% of the control level. In repeated experiments, the specific activity of the puromycin-injected animals was 52 and 62% of the paired control respectively. The lowest specific activities were observed 100 min after puromycin. After this time, the specific activity of the serum seromucoid of the treated animals returned to near control 1eveI and remained at this Ievel up to 24 h. The specific activity of the serum residue fraction was depressed at 40 and 100 min but was near control values at 3 h. At 12 and 24 h the specific activity of the control was 730 dpm~mg protein while that of the treated animal was 973 dpmfmg protein representing 133% of the control.
At 40 min after puromy~in, the specific activity of the Golgi serom~~id was depressed from 28 750 dpmfmg protein in the controt to I4 010 in the treated animal, representing 49 % of the specific activity of the paired control. At 100 min after puromycin, the specific activity of the treated animals was 53 and 71% of their respective controls. At 3 h, the specific activity returned to controt values and remained at this tevel. At 40 min after puromycin treatment, the specific activity of the Golgi residue fraction, i.e. the fraction insoluble in 0.6 N perchloric acid, was reduced to 38% of the control values. This represented a fall in the specific activity from the control level of I I 800 dpmlmg protein to 44W dpm/mg protein in the puromycin treated rat. At 100 min, the specific activity of the three pairs of puromycin treated rats was 86,
213 TABLE III GAtACT(MIYLTRANsFERAEACTIVIrY OF~~OLATED~~OLGIFRACTIONS AND SERUM AT VARIOUS TIMES TREATMENTOFTHEANlMALSWlTHPUROMYCtN
AFTER
-. I
140 154 97
63 95 115
2
111 &77 114
KM 87 117
3
170 170 161
58 68 67
12
151 100 -
102 152 123
24
98 88 -
79 81 100
78 and 72% of the controls. At 3, 12 and 24 h after puromycin, the specific activities were similar to the control levels. The effect of puromycir on galactosyltranSferasein Golgifractionsand serum
The activities of galactosyl transferase in Golgi fractions and in serum, at various times after injection of puromycin are shown in Table III. There was a trend towards elevated galactosyl transferase activity in the Golgi fractions at 1 h, but this increase was particularly marked at 3 h after puromycin when the specific activity was 7.5 10s dpmlmgprotein per 2 h incubationcompared with 50dpm/mgprotein/2 h in the controls. This represented an increase of 170%. Corresponding to these increases there was a decreased galactosyl transferase activity in serum particularly at 3 h where only 58-68 ‘A of the control activity was measured. l
The eflect of puromycinon the morphologyof the Golgiapparatus The changes in structural arrangement of the Golgi complex were investigated in sections of rat liver and in isolated Golgi fractions at intervals from 1 to 24 h after a single injection of puromycin. The isolated Golgi fractions were examined both by negative staining and in ultrathin sections. Golgi complex in hver sections The earliest changes observed in the Golgi complex appeared within 30 to 60
Fig. 1. Sections through rat hepatocytes 30 min after injection of puromycin. (a) To show narrowing of cisternae (0 of the Golgi complex. The tubular network (T) is more prominent. Mag. 37 000. (b) To show degeneration and vesiculation of normal cisternal structure (arrows) of the Golgi complex. Portions of the tubular network (r) appear in cross section. Mag. 37000.
FIN. 2. Sectlon through rat hepatocyte to show the Golg~ complex at 60 mm after Injection of puromycin. The ctsternae (arrows) are compressed and n-regular m outhne and the mtracisterna! spaces reduced. An increase is Seen in small vesictes(T). Mag 33 000
mm after injection of puromycin (Fags. I and 7). The cisternae were flattened, less well-defined and contained less osmlophihc material. The intercisternal space was reduced so that the stack of cisternae appeared compressed sE:d th+zindlvildual cisternae difficult to distinguish. The number of small vesicles, including both smoothsurfaced and coated vesicles, associated wrth the crsternae was significantly increased in the Golgi zone. No significant changes were observed in the larger vacuoles at this time. Cytoplasm surrounding the Golgi zone was more densely granular than that in the normal cell. Accompanying the changes in the Golgi complex were structural changes in the rough endoplasmlc reticulum, including loss of organized arrtlys of cisternae, dilatation and vesiculation of cisternae and a reduction in number of bound ribosomes. Between 2 and 3 h after puromycin, rough endoplasmic reticulum returned to its normal arrangement in most cells, while the Golgi complex consisted of irregular collections of degenerate cisternae, large vacuoles and small vesicles. In most cells, the cisternae were curved and distorted with myelin-type figures frequently associated with the central portion (Fig. 3). Small vesicles were numerous and were found throughout the entire Golgi zone. Large vacuoles increased both in number apd size between 2 and 12 h due to fragmentation and vesiculation of dilated cisternae.
F-g. 3. Sections through rat hepatocytes to show the Go&i complex at 3 h after poromycin. show irregular {arrow) and dilated tisternae. ~yelin-f~ figures (M) are seen freq~ntly central region of the cisternae of the Gofgi complex. Mag. 33 000. (6) To show detail of the type figures in the cisternae of the Goigi complex, SV, secretory vesicle. T, small vesicles. Mag
(al To in the myelin 65 000.
217
Fig. 4. Section through rat hepatocyte 48 h after puromqcm. The Golgi complex shows more typtcal arrangement of 3 to 4 parallel clsternae (ATOWS) ulth regular small vesicles (7) and some larger vesicles (SV). Mag 30 000
At 24 h, the Golg~ complex consisted of 1-2 short flattened clsternae together with small vesicles and large vacuoles. Between 24-48 h, the Golg~ membr;anes were similar in appearance to the normal cell with 3-4 parallel cisternae associated with regular arrays of small vesicles (Fig. 4).
In negatively stained preparations, the most typIcal arrangement of the Golgi complex isolated from normal rat h\er was the series of symmetrical plate-hke structures each with a central saccule or cisterna surrounded by a regular network ol fine tubules (Fig. 5). Wlthin 30 min after purom}cm treatment, most of the Golg~ membranes appeared very irregular in form, folded Into a cup-shape and distorted (Fig. 6). The size of the central sac or cisterna of each plate wa5 reduced in size. had a poorly defined outline, and membranous blebs 200-300 nm III diameter were seen projecting from the surface. The electron density of the cisternae was less than that in the normal Golgi complex and the most central area was frequently vacuolated. which suggested that the normal content of the cisternae had been markedly reduced. The fenestrated tubules surrounding each cisterna increased, presumably at the
Fig. 5. Negatively stained preparation of Golgi membranes isolated from normal rat liver. The Golgi a series of 3 superimposed plate-like structures, each with a central cisterna surrounded by a network of fine tubules. l.Sy< sodium phosphotungstate, pH 7.0. Mag. 40000.
complexconsists of
expense of cisternal membranes since the increased number of tubules was accompanied by a significant reduction in cisternal size. The tubular fenestration lacked the regular arrangement typical of the normal Golgi plate due to irregular size of both the tubules and the intertubular spaces. Proliferation of the tubular structures progressed further at the expense of cisternal membranes so that the overall diameter of most Golgi plates remained in the normal range. At I h after puromycin, the plates were folded and showed large projections of membrane devoid of electron dense material. The number ofelectron-lucent granules, presumably low density lipoprotein in nature, increased both in the peripheral tubules and in membrane-bound vesicles of the Golgi complex (Fig. 7). At 2,3 and 12 h, the isolated Golgi fractions contained predominantly fragmented tubules and vesicles. The plates of the Golgi complex were reduced in size with littit cisternal structure and irregular fenestrated tubules with electron-lucent material. At 2-3 h, abnormally large vesicles were numerous and contained some spherical electron-lucent granules. In ultrathin sections of the isolated Golgi fractions, the membranes of the Golgi complex were arranged as a stack of 3-4 parallel cisternae (Fig. 8). At 0.5 to 1 h after
Frg. 6. Negatively stained preparatton of Golg~ membranes isolated from rat liver 30 mm after mjection of buromycin. The crsternae (C) are less electron-dense, showmg vacuolatron and formatton of large membranous projecttons from their surface (arrows) The arrangement of tubular networks (T) IS less regular than m the normal Golgi complex. I .5 7; sodmm phosphotungstate, pH 7.0. Mag.
40 000. puromycin treatment, the cisternae became progressively less osmiophilic, more distorted and vacuolated, particularly in the central region (Fig. 9). Degeneration of cisternae occurred more rapidly at the maturing face than at the forming face. The number of small vesicles increased and there was an increase particularly in number of coated vesicles associated with the Golgi complex. At 2 to 3 h after puromycin, the isolated Golgi membranes appeared mainly as short cisternal fragments, numerous tubulus containing osmiophilic granules, and large vacuolar structures containing a few granules. Associated with the Golgi membranes, there was an increased number of coated vesicles, autophagic vacuoles and lysosomal structures. At 24 and 48 h, the Golgi consisted mainly of a series of 3 or 4 short osmiophilic cisternae, associated wrth vesicles and vacuoles as seen in the preparations from normal liver. Fragmented groups of Golgi membranes were observed at 24 h, but these represented less than 30% of the total membranes in the fraction. tWXXJSSION
A single injection of puromycin inhibited protein synthesis in the rat liver.
Fig. 7. Negatively stained preparation of Golgi membranes isolated from rat her 60 mm after puromycm. The plates of Golg~ complex are folded into cup-shaped structures and show proliferatlon of Irregular tubule networks(T) at the expense of cisternal membrane (0. Irregular membranous sacs project from cIsternal membranes (arrows) and there IS an increase in low density lipoprotein granules (L). 1.5% sodium phoFphotungstate, pH 7.0. Mag. 40 000.
cursor, the maximum inhibition of serum When leucine was used as radioactive protein synthesis was detected at I h, but was not completely recovered until 12 h. The maximum decrease of incorpor: tton of glucosamine into serum glycoproteins was observed at 2 h after puromycin and appeared to be affected for a longer period of time than leucine incorporation. This sequence was in accordance with what is known about the assembly of glucosamine into glycoproteins. Most of the glucosamine was added to the completed polypeptide after its release from the ribosome. Therefore, the maximum inhibition observed with glucosamine would have been expected at a later time than that with leucine. A reduction of leucine-labelled protein was detected early in the Golgi apparatus, which presumably resulted from a reduction of newly synthesized protein. This rapid reduction of leucine incorporation into proteins was accompaniecl by a reduction in glucosamine incorporation, most marked in the Golgi fractions within the first hour after injection of puromycin. Followiog the initial inhibition of protein and glycoprotein assembly, the incorporation of precursors approached normal levels. However, there was a dramatic
Fig. 8. Ultrathin section through pelleted Golgl fraction isolated from normal liver. The membranes of the Golgi complex show the typical arrangement III series of 3-4 parallel or concentric clsternae (C) associated with small vesicles (T) and larger secretory vesicles (SV) Mag 27 000.
increase in leucine-labelled protein in the Golgi complex at 3 h. The appearance of this material in the seromucoid fraction extracted from the Golgi membranes implied that it was glycoprotein in nature or that it consisted of small peptide molecules. Its accumulation might have reflected a compensation mechanism resulting in overproduction of protein after the initial inhibition by puromycin. However, the incorporation of glucosamine did not change, suggesting that the leucine-labelled protein was either not being glycosylated or was only partially glycosylated. It seemed likely therefore that the accumulating protein was not a good acceptor for carbohydrate. Studies of galactosyl transferase showed that early alterations in glucosamine incorporation were not the result of loss of enzyme activity. The only consistent change in the levels of this enzyme in the Golgi fractions appeared at 3 h, when enzyme activity was significantly higher than control values. This indicated that the enzyme level was not a limiting factor in the synthesis of glycoprotein so that the newly synthesized leucine-labelled material did not appear to be acting as an acceptor for carbohydrate. Alternatively, the action of puromycin might have resulted from a specific inhibition of one of the glycosyl transferase enzymes other than galactosyl
Fig, 9. Section through pelleted Colgi membranes ifdated from rat liver I h after puromycin. The Golgi complex shows increasing curvature of the cisternae (0) with progressive vacuoIation and degeneration of central portion of the cisterna (arrows) and proliferation of fenestrated tubules (T). Mag. 27ooO.
transferase. However, the appearance of normal levels of glucosamine-labelled fractions in serum at 3 h after puromycin indicated that glycoprotein was being assembled in the normal manner and that the activities of glycosyl transferases were normal. It rvas noteworthy that an efevated galactosyl t~nsferase activity was found in the Golgi fractions wh:le the activity of serum enzyme was significantly lower than normal. This suggested that, while synthesisof the enzyme was continuing, there was some alteration in its transport to serum. Although little has been reported regarding the relationship between the liver and serum enzymes, these findings suggested that the levels of galactosyl transferase in serum probably depended on continuing synthesis and secretion of engme from the liver cell, and that a steady release of enzyme is necessary to maintain the activity in serum, In comparison with the changes of glycoprotein synthesis and secretion, the most dramatic structural changes appeared in the Golgi complex during the first three hours after puromy~in injection. These changes, followed in isolated Golgi fractious and in ultrathin sections from the same liver, were compared with the struct-
223 ure of the Golgi apparatus from the normal rat hepatocyte which has been described in detail previouslys. The effects of puromycin treatment were observed very rapidly in the Golgi complex. Within half an hour after injection of puromycin, changes appeared in the zone surrounding and including the Golgi complex at the same time that changes appeared in the rough endoptasmic reticulum. The sequence of changes in the Golgi complex affected firstly the central region of the GoIgi plate, the cisternae and then the fenestrated tubules. There was a progressive vesiculation and disorganization of the cisternae and tubular structures during the first three hours after puromycin. This was accompanied by accumulation of low-density lipoprotein material and later by increasing numbers of coated vesicles, autophagic vesicles, and Iysosomal structures in association with the Golgi complex. At this stage, 3 to 12 h after injection of puromycin, there was vesiculation and degradation of many of the Golgi membranes with frequent appearance of myelin-type figures in Golgi complex of the liver. Reorganization or regeneration of membranes occurred so that at 24 h and 48 h, more typical arrangement was seen in isolated Golgi fractions and in liver sections. This recovery of the normal structural arrangement of the Gofgi system in most cells appeared later than the time at which glucosamine incorporation returned to normal levels. However, the morphological features might have reflected the altered enzyme and protein transport phenomena seen between 3-12 h. In summary, puromycin caused a rapid inhibition of protein synthesis, followed by inhibition of oligosaccharide assembly into glycoprotein molecules and a secondary phenomena of accumulation and apparently defective transport of materials from the Golgi complex. More recent studies being carried out on isolated membrane fractions in vitro have shown that puromycin does bind to Golgi membranes. This binding resulted in inhibition of galactosyl transferase activity, indicating that puromycin did have a direct effect or the Golgi complex in addition to Its known effects at the ribosomal levells. The implications of these findings are important to an understanding of the mechanism of glycoprotein synthesis and on the regulation of both protein and glycoprotein secretion.
REFERENCES 1 M. A. Darken, Puromycrnmhtbitton of protein synthesrs, Phutwmcol. Rev., 16 (1964) 223-243. M, Reid, l-f.Shinozukaand H. Sidransky,Polyr&osomaldlsaggregatroninducedby puromycm and Its reversalwith time, L&. fnv&., 23 (1970)119427, 3 A. Weinstock,Cytotoxiceffectsof puromycinon the Golgt apparatus of p~~c~tIc acinarcells, hepatocytesand amclobfasts, J. ~i~?oc/i~i?f.Cyroehm., 18 (1970) 875-886. 4 H. I. Friedman and R. R. Cardell, Effects of puromycrn on the structure of rat intestinal epithel2 I.
ial cells during fat absorption, J. CeN Bloi., 52 (1972) 15-4& 5 J.
D. Jamieson and G. E. Palade, Intracellulartransport of secretory proteins in pancreatrc
exocrine cells, 111.Dissociation of intracellular transport from protein synthesis, J. C’t>llBiol., 39 (1968) 580-588. 6 R. G, Spiro and M, J. Spiro, Giycoprotein biosynthesis: Studies on thyroglobulin, J. Biol. Chetjt., 241 (1966) 1271-1282.
7 A. Herscovics, Biosynthesis of thyrogIobulin. Incorporation of [I-WJmannose and [4, S-aH%Jleucine into soluble proteins by rat thyroids in vitro, Biuchem. J.., 112 (1969) 709-719. 8 R. G. Spiro, Glycoproteins, Ann. Rev. Biuchem, 39 (1970) 599638. 9 1. M. Sturgzss, E. Katona and M. A. Moscarello, The Golgi complex, 1. Isolation and ultrastructure in normal rat liver, J. Membrane Bid., 12 (1972) 367-384. 10 H. Schachter, I. Jabbal, R. L. Hudgin, L. Pinteric, E. J. McGuire and S. Roseman, Intracellular localization of liver sugar nucleotide glycoprotein glycosyltransferases in a Golgi-rich fraction, J. Biol. Chem., 245 (1970) 10!&1100. II J. M. Sturgess, M. Mitranic and M. A. Moscarello, The incorporation of D-glucosamine-sH into the Golgi complex from rat liver and into serum glycoproteins, B&hem. Biophys. Res. Commun., 46 (1972) 1270-1277. 12 0. H. Lowry, N. J. Rosebrough, A. L. Farr and R. J. Randall, Protein measurement with the folin phenol reagent, J. Biof. Chem.. 193 (1951) 265-275. 13 M. Treloar, J. M. Sturgess and M. A. Moscarello, An efkct of puromycin on galactosyltransferase of Golgi-rich fractions from rat liver, J. Biol. Gem., 249 (1974) 6628-6632.
0 Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands
THE GOLGI COMPLEX III. THE EFFECTS OF PUROMYCIN GLYCOPROTEIN SYNTHESIS
ON ULTRASTRUCTURE
JENNIFER M. STURGESS, MARDIE M. MITRANIC
AND
207
AND
MARIO A. MOSCARELLO
The Departments of Pathology and Biochemistry, Research Institute, Hospital for Sick Children, 5.55 University Avenue, Toronto, Ontario (Canada M5G 1X8)
(Received February 4th, 1974) (Revision received January lOth, 1975) (Accepted January 23rd, 1975)
SUMMARY
The effect of puromycin has been investigated on protein and glycoprotein synthesis and on ultrastructure of the Golgi complex from rat liver. Incorporation of [14C]leucine into protein in Golgi fractions and into serum proteins was depressed rapidly after puromycin treatment. In the serum proteins, incorporation returned to normal levels at 2 h whereas in Golgi fractions it continued to rise to 200% of the control levels at 3 h and was still elevated at 24 h after puromycin treatment. Incorporation of [W]glucosamine into glycoprotein was depressed in Golgi and sekum fractions in a similar manner but slightly later than that of leucine. Leucine labelled material found at 3 h was a poor acceptor for carbohydrate, since [Wlglucosamine incorporation was not elevated above control values. Galactosyl transferase activity was not depressed in the Golgi membranes and, at 3 h, was elevated implying that an adequate supply of enzyme was available at all times. The activity of the galactosyl transferase in serum appeared to bc depressed suggesting that transport of enzyme from Golgi complex to serum was defective. Ultrastructural changes in the Golgi complex were observed to occur rapidly after puromycin treatment. The cistecnae became irregular, compressed, and degenerated progressively from central region towards the periphery. Irregular tubular structures formed at the expense of cisternal membrane and showed accumulation of low density lipoprotein. Vesiculation and degenerative changes of the GO& membranes continued from 2-l 2 h while more typical arrangements of the Golgi complex were observed between 24-48 h. The morphological changes correlated with changes in glycoprotein synthesis.
Abbreviation: TCA, trichloroacetic acid.
208 INTRODUCTION
The action of puromycin on the synthesis of secretory protein has been well documented and it has been shown that puromycin acts at the ribosomal level to release nascent peptide chains by replacing aminoacyl transfer RNAl. This exchange inhibits completion of protein molecules. Ultrastructural and biochemical studies2 have shown that injection of puromycin causes disaggregation of polyribosomes both in vitro and in vivo but this effect is reversed rapidly. In other intracellular membrane systems, morphological changes based on electron microscopic examination of ultrathin sections, have been described in the Golgi complex in various cell types including the livers, the pancreas and thyroids, and intestinal eplthelial cells 4. The reported changes, including regression and increased vacuolation of the Golgi apparatus have been attributed to the inhibition of protein synthesis. However, there is some evidence to suggest a separate mechanism of puromycin action responsible for changes in intracellular membranes, which is distinct from that controlling synthesis and transport of secretory proteinss. Puromycin appears also to have an effect on the assembly of sugars into glycoproteins. In the thyroid gland, puromycin completely inhibits the synthesis of the protein portion of the thyroglobulin molecule whereas it has no effect or causes only partial inhibition of carbohydrate assembly in soluble thyroglobulin and particle-bound thyroglobulin respectively 6. The degree of inhibition of assembly of different carbohydrates varies and appears to be related to their arrangement in the oligosaccharide chain, so that those adjacent to the peptide core are affected to a greater extent than more terminal sugars 6~~.Since carbohydrates are assembled into the oligosaccharide chain mainly in the Golgi complex by a mechanism fundamentally different from that responsible for the synthesis of the peptide chains, it is possible that puromycin has an additional effect on the Golgi membranes which influences carbohydrate incorporation. The purpose of this study has been to investigate the effects of puromycin on the structural arrangement of the Golgi complex and on glycoprotein biosynthesis. Morphological changes after puromycin treatment have been evaluated by electron microscopy in isolated Golgi-rich fractions and in liver. The findings have been correlated with incorporation studies in the same membrane preparations using specific radioactive precursors for protein and oligosaccharide biosynthesis and assays of galactosyl transferase, an enzyme required for carbohydrate assembly which is specific to the Golgi complex. MATERIALS
AND METHODS
Experimental procedure Male Wistar rats, 200-220 g fasted for 15 h were used in these studies. A single injection of 19 mg puromycin hydrochloride pH 7.0 was given to each rat via the jugular vein: controls received a single injection of saline. After completion of injections, each rat was anesthetised with ether, exsanguin-
209 ated, a sample of blood collected, and the liver removed. A slice of liver was resected for electron microscopy and the remainder used to isolate the Golgi membranes. At each time interval, Golgi membranes were isolated from at least four rats: two were treated with puromycin and two were controls. The control and treated rats were paired, being maintained under identical conditions and receiving equivalent amounts of food and water. One rat of each pair was injected intravenously with puromycin; the other received an equal volume of saline. The Golgi fractions were prepared from each pair of rats simultaneously and the time at which injections were given was adjusted so that all rats were sacrificed at the same time (11:OO a.m.). 3so~af~u~of GoZgifructio~s
The Golgi fractions were isolated separately from each rat liver using techniques described [email protected] of the isolated Golgi membranes were used for electron microscopy, for assay of galactosyl transferase and for studies on incorporation of radioactive precursors. Ellectron microscopy
From each rat used in this study, samples of liver tissue and of the isolated Golgi fractions were examined by electron microscopy. At least 2 rats were investigated at each time interval: 0, 1,2,3, 12 and 24 h after puromycin. Tissue and pelleted membrane samples were fixed in 3 % glutaraldehyde in 0. I M sodium phosphate buffer, pH 7.4, and post-fixed in 1% osmium tetroxide in veronal-acetate buffer. After dehydration in graded ethanol solutions, liver samples were polymerized in Epon epoxy resin, and membrane samples in Spurr low viscosity epoxy resin. The methods for preparation of samples for electron microscopy have been described in detail previously9, Galactosyl transferuse assay
A portion of each Golgi fractioa prepared from rat liver at 0, 1,2,3,12 and 24 h after injection of puromycin was assayed for galactosyi transferase using jV-acetyl glucosamine as acceptor 10. The enzyme activity was calculated as dpm~mg protein per 2 h incubation. Galactosyl transferase was used to monitor purification and recovery of Golgi membranes from the homogenates. The purification factor (PF = specific activity of isolated Golgi membranes/specific activity of the original homogenate) was obtained for each Golgi preparation. These factors varied from 60-120 in different preparations and corresponded to 40-83 % recovery of galactosyi transferase activity from the total homogenate, These values were in the same range as those obtained with normal rat liver so that puromycin did not alter the separation of Golgi membranes in this procedure, Incorporation of radioactiveprecursors
The incorporation of L-[U-WZ]leucine (specific activity 278 mCi/mmole; New England Nuclear) and ~-[l-l~C]~lucosamine (specific activity 7.5 mCi/mmole;
210
New England Nuclear) were studied separately as markers for protein and glycoprotein biosynthesis respectively. For each radioactive precursor, two rats were studied at each time interval. Incorporation of L-leucine was followed at 0, 1,2,3, 12 and 24 h after puromycin injection. A single dose of 5 ,&i L-leucine was given intravenously to each rat 60 min before sacrifice. D-Glucosamine incorporation was investigated at 0, 40 and 100 min and at 2, 12 and 24 h after puromycin. A single injection of 5 yCi D-glucosamine was given intravenously to each rat 40 min before sacrifice. The Golgi and serum fractions were isolated and extracted as follows: Low molecular weight material was removed from the Golgi fraction and from serum after precipitation of the high molecular weight material with an equal volume of 10% TCA. The IC_4 precipitate was extracted with 0.6 N perchloric acid and the high molecular weight (glycoprotein) fraction was precipitated with phosphotungstic acid 11. The glycoprotein fraction isolated in this manner was referred to as seromucoid. For each fraction, protein was determined on one aliquotls and another aliquot was used for radioactive counting in a Nuclear Chicago Mark I Scintillation Counter. For counting, 100 ~1 was dissolved in 1.3 ml of hyamine hydroxide and 10 ml of toluene scintillation fluid (0.4 % PPO and 0.03 % POPOP) was added. The radioactivity was expressed as specific activity dpm/mg protein. RESULTS
The effect of puromycin on the incorporation of L-[W4C]leucine proteins
into Golgi and serum
The incorporation of L-[r4C]leucine was investigated in fractions extracted from isolated Golgi membranes and from serum at intervals 0, 1, 2, 3, 12 and 24 h after treatment with puromycin. The variation in specific activity (dpmlmg protein) of Golgi and serum fractions after puromycin treatment is summarized in Table I. Each number represents the specific activity of a treated rat expressed as a percentage of the specific activity of the paired control. At each time interval, the percent specific activities are given from two pairs of rats. Serum. At 1 h after puromycin treatment, the specific activity of the serum seromucoid fraction was depressed to roughly 28% of the control, representing a change in specific activity from 450 dpm/mg protein in the control to 148 dpm/mg protein in the treated animal. At 2 h it rose to about 60 % of the control value but remained lower at about 75% of the control value at 3 to 12 h. At 24 h the specific activity of one of the test animals (450 dpm/mg protein) was higher than that of one of the paired control (337 dpmlmg protein). The specific activity of the serum residue, i.e. the fraction insoluble in 0.6 iV perchloric acid, was depressed markedly at 1 h after puromycin to less than 15% of the control value (specific activity of puromycin treated was 73 dpm/mg protein and control was 480 dpm/mg protein), rose to 50% of the control value at 2 h, and approached the control levels at about 3 h. Golgi fractions. At 1 h after puromycin treatment, the specific activity of the seromucoid fraction of Golgi membranes, i.e. the high molecular weight fraction
211 TABLE I INCORPORATION
OF [l”C]LEUCINE
INTO GOLGI FRACTIONS AND
SERUM FOLLOWING TREATMENT
WITH
PUROMYCIN
Time altep ~~~ro~~~~~i~ (h)
Specific activity ( $, obtained ~o~~froI~ Goigi seromucoid
G&i residm
Serum sermmoid
Serum residue
1
64 50
28 -
22 33
11 15
2
89 109
89 101
61 56
56 49
3
194 252
90 120
74 54
75 98
1L
187 145
79 97
75 94
92 90
24
129 125
76 96
78 133
113 93 ~-
soluble in 0.6 N perchloric acid, was depressed 50-64 y0 compared with the controls. At 2 h the specific activity increased to almost normal levels of 3933 dpm/mg protein compared with the control values of 4431 dpmlmg protein. However, the activity continued to rise to about 200 % of the control values at 3 h after puromyein. At 12 h the specific activity fell slightly but remained at 130% of the control at 24 h after puromycin. These data implied that there was an accumulation of [W]leucinelabelled material in the Golgi seromucoid fraction at 3 to 24 h after puromycin treatment. The Golgi residue, i.e., the fraction insoluble in 0.6 N perchloric acid showed a decreased specific activity at 1 h (28% of the control). The incorporation of [W)leucine returned to normal levels at 2 h and remained in the normal range after this time. Eflect o~puro~ly~i~on the i~co$por~fion of o-f i- W]gtucosanline infoGolgiand serum pro feins
The incorporation of D-[l-‘4C]ghKOSRIIline into perchloric acid-soluble and insoluble fractions of Golgi membranes isolated from rat liver and serum, was investigated at 40 and 100 min and at 3, 12, and 24 h after injr-:tion of ~~ornycin. The variation in specific activity in Golgi and serum fractions after puromycin is summarized in Table 11.Each number represents the specific activity of the puromycintreated rats expressed as a percentage of the paired controls; at each time the percent specific activities are given for at least 3 pairs of rats. Serum. At 40 min after injection of puromycin, the specific activity of the serum seromucoid fraction was 1774 dpm/mg protein compared with 4246 dpm/mg protein in the control, so that the specific activity of the puromycin-injected animal was
212 TMLE
11
INFRACTION
OF [‘~GLU~
E IKlo
GOLGI FRACXION!J AND SERUM FGLLOWING
TRRATMEkrT
WiTH PUROMYCIN
Timesfter
Specificactivity ( $6of paired control)
40 min
49 -
38 37 39
52 42 62
SO 43 57
lOOmin
53 71 -
86 78 72
35 23 47
61 -
3h
88 90 88
ma 88 98
85 70 fo0
79 95 -
12h
94 99 90
IO8 107 107
90 9s 85
143 95 -
24h
112 78 98
103 105 -
80 89 115
140 133 -
only 42% of the control level. In repeated experiments, the specific activity of the puromycin-injected animals was 52 and 62% of the paired control respectively. The lowest specific activities were observed 100 min after puromycin. After this time, the specific activity of the serum seromucoid of the treated animals returned to near control 1eveI and remained at this Ievel up to 24 h. The specific activity of the serum residue fraction was depressed at 40 and 100 min but was near control values at 3 h. At 12 and 24 h the specific activity of the control was 730 dpm~mg protein while that of the treated animal was 973 dpmfmg protein representing 133% of the control.
At 40 min after puromy~in, the specific activity of the Golgi serom~~id was depressed from 28 750 dpmfmg protein in the controt to I4 010 in the treated animal, representing 49 % of the specific activity of the paired control. At 100 min after puromycin, the specific activity of the treated animals was 53 and 71% of their respective controls. At 3 h, the specific activity returned to controt values and remained at this tevel. At 40 min after puromycin treatment, the specific activity of the Golgi residue fraction, i.e. the fraction insoluble in 0.6 N perchloric acid, was reduced to 38% of the control values. This represented a fall in the specific activity from the control level of I I 800 dpmlmg protein to 44W dpm/mg protein in the puromycin treated rat. At 100 min, the specific activity of the three pairs of puromycin treated rats was 86,
213 TABLE III GAtACT(MIYLTRANsFERAEACTIVIrY OF~~OLATED~~OLGIFRACTIONS AND SERUM AT VARIOUS TIMES TREATMENTOFTHEANlMALSWlTHPUROMYCtN
AFTER
-. I
140 154 97
63 95 115
2
111 &77 114
KM 87 117
3
170 170 161
58 68 67
12
151 100 -
102 152 123
24
98 88 -
79 81 100
78 and 72% of the controls. At 3, 12 and 24 h after puromycin, the specific activities were similar to the control levels. The effect of puromycir on galactosyltranSferasein Golgifractionsand serum
The activities of galactosyl transferase in Golgi fractions and in serum, at various times after injection of puromycin are shown in Table III. There was a trend towards elevated galactosyl transferase activity in the Golgi fractions at 1 h, but this increase was particularly marked at 3 h after puromycin when the specific activity was 7.5 10s dpmlmgprotein per 2 h incubationcompared with 50dpm/mgprotein/2 h in the controls. This represented an increase of 170%. Corresponding to these increases there was a decreased galactosyl transferase activity in serum particularly at 3 h where only 58-68 ‘A of the control activity was measured. l
The eflect of puromycinon the morphologyof the Golgiapparatus The changes in structural arrangement of the Golgi complex were investigated in sections of rat liver and in isolated Golgi fractions at intervals from 1 to 24 h after a single injection of puromycin. The isolated Golgi fractions were examined both by negative staining and in ultrathin sections. Golgi complex in hver sections The earliest changes observed in the Golgi complex appeared within 30 to 60
Fig. 1. Sections through rat hepatocytes 30 min after injection of puromycin. (a) To show narrowing of cisternae (0 of the Golgi complex. The tubular network (T) is more prominent. Mag. 37 000. (b) To show degeneration and vesiculation of normal cisternal structure (arrows) of the Golgi complex. Portions of the tubular network (r) appear in cross section. Mag. 37000.
FIN. 2. Sectlon through rat hepatocyte to show the Golg~ complex at 60 mm after Injection of puromycin. The ctsternae (arrows) are compressed and n-regular m outhne and the mtracisterna! spaces reduced. An increase is Seen in small vesictes(T). Mag 33 000
mm after injection of puromycin (Fags. I and 7). The cisternae were flattened, less well-defined and contained less osmlophihc material. The intercisternal space was reduced so that the stack of cisternae appeared compressed sE:d th+zindlvildual cisternae difficult to distinguish. The number of small vesicles, including both smoothsurfaced and coated vesicles, associated wrth the crsternae was significantly increased in the Golgi zone. No significant changes were observed in the larger vacuoles at this time. Cytoplasm surrounding the Golgi zone was more densely granular than that in the normal cell. Accompanying the changes in the Golgi complex were structural changes in the rough endoplasmlc reticulum, including loss of organized arrtlys of cisternae, dilatation and vesiculation of cisternae and a reduction in number of bound ribosomes. Between 2 and 3 h after puromycin, rough endoplasmic reticulum returned to its normal arrangement in most cells, while the Golgi complex consisted of irregular collections of degenerate cisternae, large vacuoles and small vesicles. In most cells, the cisternae were curved and distorted with myelin-type figures frequently associated with the central portion (Fig. 3). Small vesicles were numerous and were found throughout the entire Golgi zone. Large vacuoles increased both in number apd size between 2 and 12 h due to fragmentation and vesiculation of dilated cisternae.
F-g. 3. Sections through rat hepatocytes to show the Go&i complex at 3 h after poromycin. show irregular {arrow) and dilated tisternae. ~yelin-f~ figures (M) are seen freq~ntly central region of the cisternae of the Gofgi complex. Mag. 33 000. (6) To show detail of the type figures in the cisternae of the Goigi complex, SV, secretory vesicle. T, small vesicles. Mag
(al To in the myelin 65 000.
217
Fig. 4. Section through rat hepatocyte 48 h after puromqcm. The Golgi complex shows more typtcal arrangement of 3 to 4 parallel clsternae (ATOWS) ulth regular small vesicles (7) and some larger vesicles (SV). Mag 30 000
At 24 h, the Golg~ complex consisted of 1-2 short flattened clsternae together with small vesicles and large vacuoles. Between 24-48 h, the Golg~ membr;anes were similar in appearance to the normal cell with 3-4 parallel cisternae associated with regular arrays of small vesicles (Fig. 4).
In negatively stained preparations, the most typIcal arrangement of the Golgi complex isolated from normal rat h\er was the series of symmetrical plate-hke structures each with a central saccule or cisterna surrounded by a regular network ol fine tubules (Fig. 5). Wlthin 30 min after purom}cm treatment, most of the Golg~ membranes appeared very irregular in form, folded Into a cup-shape and distorted (Fig. 6). The size of the central sac or cisterna of each plate wa5 reduced in size. had a poorly defined outline, and membranous blebs 200-300 nm III diameter were seen projecting from the surface. The electron density of the cisternae was less than that in the normal Golgi complex and the most central area was frequently vacuolated. which suggested that the normal content of the cisternae had been markedly reduced. The fenestrated tubules surrounding each cisterna increased, presumably at the
Fig. 5. Negatively stained preparation of Golgi membranes isolated from normal rat liver. The Golgi a series of 3 superimposed plate-like structures, each with a central cisterna surrounded by a network of fine tubules. l.Sy< sodium phosphotungstate, pH 7.0. Mag. 40000.
complexconsists of
expense of cisternal membranes since the increased number of tubules was accompanied by a significant reduction in cisternal size. The tubular fenestration lacked the regular arrangement typical of the normal Golgi plate due to irregular size of both the tubules and the intertubular spaces. Proliferation of the tubular structures progressed further at the expense of cisternal membranes so that the overall diameter of most Golgi plates remained in the normal range. At I h after puromycin, the plates were folded and showed large projections of membrane devoid of electron dense material. The number ofelectron-lucent granules, presumably low density lipoprotein in nature, increased both in the peripheral tubules and in membrane-bound vesicles of the Golgi complex (Fig. 7). At 2,3 and 12 h, the isolated Golgi fractions contained predominantly fragmented tubules and vesicles. The plates of the Golgi complex were reduced in size with littit cisternal structure and irregular fenestrated tubules with electron-lucent material. At 2-3 h, abnormally large vesicles were numerous and contained some spherical electron-lucent granules. In ultrathin sections of the isolated Golgi fractions, the membranes of the Golgi complex were arranged as a stack of 3-4 parallel cisternae (Fig. 8). At 0.5 to 1 h after
Frg. 6. Negatively stained preparatton of Golg~ membranes isolated from rat liver 30 mm after mjection of buromycin. The crsternae (C) are less electron-dense, showmg vacuolatron and formatton of large membranous projecttons from their surface (arrows) The arrangement of tubular networks (T) IS less regular than m the normal Golgi complex. I .5 7; sodmm phosphotungstate, pH 7.0. Mag.
40 000. puromycin treatment, the cisternae became progressively less osmiophilic, more distorted and vacuolated, particularly in the central region (Fig. 9). Degeneration of cisternae occurred more rapidly at the maturing face than at the forming face. The number of small vesicles increased and there was an increase particularly in number of coated vesicles associated with the Golgi complex. At 2 to 3 h after puromycin, the isolated Golgi membranes appeared mainly as short cisternal fragments, numerous tubulus containing osmiophilic granules, and large vacuolar structures containing a few granules. Associated with the Golgi membranes, there was an increased number of coated vesicles, autophagic vacuoles and lysosomal structures. At 24 and 48 h, the Golgi consisted mainly of a series of 3 or 4 short osmiophilic cisternae, associated wrth vesicles and vacuoles as seen in the preparations from normal liver. Fragmented groups of Golgi membranes were observed at 24 h, but these represented less than 30% of the total membranes in the fraction. tWXXJSSION
A single injection of puromycin inhibited protein synthesis in the rat liver.
Fig. 7. Negatively stained preparation of Golgi membranes isolated from rat her 60 mm after puromycm. The plates of Golg~ complex are folded into cup-shaped structures and show proliferatlon of Irregular tubule networks(T) at the expense of cisternal membrane (0. Irregular membranous sacs project from cIsternal membranes (arrows) and there IS an increase in low density lipoprotein granules (L). 1.5% sodium phoFphotungstate, pH 7.0. Mag. 40 000.
cursor, the maximum inhibition of serum When leucine was used as radioactive protein synthesis was detected at I h, but was not completely recovered until 12 h. The maximum decrease of incorpor: tton of glucosamine into serum glycoproteins was observed at 2 h after puromycin and appeared to be affected for a longer period of time than leucine incorporation. This sequence was in accordance with what is known about the assembly of glucosamine into glycoproteins. Most of the glucosamine was added to the completed polypeptide after its release from the ribosome. Therefore, the maximum inhibition observed with glucosamine would have been expected at a later time than that with leucine. A reduction of leucine-labelled protein was detected early in the Golgi apparatus, which presumably resulted from a reduction of newly synthesized protein. This rapid reduction of leucine incorporation into proteins was accompaniecl by a reduction in glucosamine incorporation, most marked in the Golgi fractions within the first hour after injection of puromycin. Followiog the initial inhibition of protein and glycoprotein assembly, the incorporation of precursors approached normal levels. However, there was a dramatic
Fig. 8. Ultrathin section through pelleted Golgl fraction isolated from normal liver. The membranes of the Golgi complex show the typical arrangement III series of 3-4 parallel or concentric clsternae (C) associated with small vesicles (T) and larger secretory vesicles (SV) Mag 27 000.
increase in leucine-labelled protein in the Golgi complex at 3 h. The appearance of this material in the seromucoid fraction extracted from the Golgi membranes implied that it was glycoprotein in nature or that it consisted of small peptide molecules. Its accumulation might have reflected a compensation mechanism resulting in overproduction of protein after the initial inhibition by puromycin. However, the incorporation of glucosamine did not change, suggesting that the leucine-labelled protein was either not being glycosylated or was only partially glycosylated. It seemed likely therefore that the accumulating protein was not a good acceptor for carbohydrate. Studies of galactosyl transferase showed that early alterations in glucosamine incorporation were not the result of loss of enzyme activity. The only consistent change in the levels of this enzyme in the Golgi fractions appeared at 3 h, when enzyme activity was significantly higher than control values. This indicated that the enzyme level was not a limiting factor in the synthesis of glycoprotein so that the newly synthesized leucine-labelled material did not appear to be acting as an acceptor for carbohydrate. Alternatively, the action of puromycin might have resulted from a specific inhibition of one of the glycosyl transferase enzymes other than galactosyl
Fig, 9. Section through pelleted Colgi membranes ifdated from rat liver I h after puromycin. The Golgi complex shows increasing curvature of the cisternae (0) with progressive vacuoIation and degeneration of central portion of the cisterna (arrows) and proliferation of fenestrated tubules (T). Mag. 27ooO.
transferase. However, the appearance of normal levels of glucosamine-labelled fractions in serum at 3 h after puromycin indicated that glycoprotein was being assembled in the normal manner and that the activities of glycosyl transferases were normal. It rvas noteworthy that an efevated galactosyl t~nsferase activity was found in the Golgi fractions wh:le the activity of serum enzyme was significantly lower than normal. This suggested that, while synthesisof the enzyme was continuing, there was some alteration in its transport to serum. Although little has been reported regarding the relationship between the liver and serum enzymes, these findings suggested that the levels of galactosyl transferase in serum probably depended on continuing synthesis and secretion of engme from the liver cell, and that a steady release of enzyme is necessary to maintain the activity in serum, In comparison with the changes of glycoprotein synthesis and secretion, the most dramatic structural changes appeared in the Golgi complex during the first three hours after puromy~in injection. These changes, followed in isolated Golgi fractious and in ultrathin sections from the same liver, were compared with the struct-
223 ure of the Golgi apparatus from the normal rat hepatocyte which has been described in detail previouslys. The effects of puromycin treatment were observed very rapidly in the Golgi complex. Within half an hour after injection of puromycin, changes appeared in the zone surrounding and including the Golgi complex at the same time that changes appeared in the rough endoptasmic reticulum. The sequence of changes in the Golgi complex affected firstly the central region of the GoIgi plate, the cisternae and then the fenestrated tubules. There was a progressive vesiculation and disorganization of the cisternae and tubular structures during the first three hours after puromycin. This was accompanied by accumulation of low-density lipoprotein material and later by increasing numbers of coated vesicles, autophagic vesicles, and Iysosomal structures in association with the Golgi complex. At this stage, 3 to 12 h after injection of puromycin, there was vesiculation and degradation of many of the Golgi membranes with frequent appearance of myelin-type figures in Golgi complex of the liver. Reorganization or regeneration of membranes occurred so that at 24 h and 48 h, more typical arrangement was seen in isolated Golgi fractions and in liver sections. This recovery of the normal structural arrangement of the Gofgi system in most cells appeared later than the time at which glucosamine incorporation returned to normal levels. However, the morphological features might have reflected the altered enzyme and protein transport phenomena seen between 3-12 h. In summary, puromycin caused a rapid inhibition of protein synthesis, followed by inhibition of oligosaccharide assembly into glycoprotein molecules and a secondary phenomena of accumulation and apparently defective transport of materials from the Golgi complex. More recent studies being carried out on isolated membrane fractions in vitro have shown that puromycin does bind to Golgi membranes. This binding resulted in inhibition of galactosyl transferase activity, indicating that puromycin did have a direct effect or the Golgi complex in addition to Its known effects at the ribosomal levells. The implications of these findings are important to an understanding of the mechanism of glycoprotein synthesis and on the regulation of both protein and glycoprotein secretion.
REFERENCES 1 M. A. Darken, Puromycrnmhtbitton of protein synthesrs, Phutwmcol. Rev., 16 (1964) 223-243. M, Reid, l-f.Shinozukaand H. Sidransky,Polyr&osomaldlsaggregatroninducedby puromycm and Its reversalwith time, L&. fnv&., 23 (1970)119427, 3 A. Weinstock,Cytotoxiceffectsof puromycinon the Golgt apparatus of p~~c~tIc acinarcells, hepatocytesand amclobfasts, J. ~i~?oc/i~i?f.Cyroehm., 18 (1970) 875-886. 4 H. I. Friedman and R. R. Cardell, Effects of puromycrn on the structure of rat intestinal epithel2 I.
ial cells during fat absorption, J. CeN Bloi., 52 (1972) 15-4& 5 J.
D. Jamieson and G. E. Palade, Intracellulartransport of secretory proteins in pancreatrc
exocrine cells, 111.Dissociation of intracellular transport from protein synthesis, J. C’t>llBiol., 39 (1968) 580-588. 6 R. G, Spiro and M, J. Spiro, Giycoprotein biosynthesis: Studies on thyroglobulin, J. Biol. Chetjt., 241 (1966) 1271-1282.
7 A. Herscovics, Biosynthesis of thyrogIobulin. Incorporation of [I-WJmannose and [4, S-aH%Jleucine into soluble proteins by rat thyroids in vitro, Biuchem. J.., 112 (1969) 709-719. 8 R. G. Spiro, Glycoproteins, Ann. Rev. Biuchem, 39 (1970) 599638. 9 1. M. Sturgzss, E. Katona and M. A. Moscarello, The Golgi complex, 1. Isolation and ultrastructure in normal rat liver, J. Membrane Bid., 12 (1972) 367-384. 10 H. Schachter, I. Jabbal, R. L. Hudgin, L. Pinteric, E. J. McGuire and S. Roseman, Intracellular localization of liver sugar nucleotide glycoprotein glycosyltransferases in a Golgi-rich fraction, J. Biol. Chem., 245 (1970) 10!&1100. II J. M. Sturgess, M. Mitranic and M. A. Moscarello, The incorporation of D-glucosamine-sH into the Golgi complex from rat liver and into serum glycoproteins, B&hem. Biophys. Res. Commun., 46 (1972) 1270-1277. 12 0. H. Lowry, N. J. Rosebrough, A. L. Farr and R. J. Randall, Protein measurement with the folin phenol reagent, J. Biof. Chem.. 193 (1951) 265-275. 13 M. Treloar, J. M. Sturgess and M. A. Moscarello, An efkct of puromycin on galactosyltransferase of Golgi-rich fractions from rat liver, J. Biol. Gem., 249 (1974) 6628-6632.
