HeLa cell plasma membranes

HeLa cell plasma membranes

Experimental Cell Reserrrch 109 (1977) 53-61 HeLa III. CELL PLASMA MEMBRANES Incorporation of [3H]F~~~~e into Plasma Membrane Glycoproteins durin...

778KB Sizes 3 Downloads 48 Views

Experimental Cell Reserrrch 109 (1977) 53-61

HeLa III.

CELL

PLASMA

MEMBRANES

Incorporation of [3H]F~~~~e into Plasma Membrane Glycoproteins during the Cell Cycle S. JOHNSEN Institute

of Medical

Biology,

and T. STOKKE

University

of Tromse,

Tromse,

Norway

SUMMARY HeLa cells grown in suspension culture were synchronized by amethopterin block and thymidine reversal. The cells were labelled with [3H]fucose at various phases in the cell cycle. The incorporation of [3H]fucose was determined in various subcellular fractions including isolated plasma membranes. The plasma membrane proteins were separated by SDS-gel electrophoresis and the distribution of radioactivity in the gels was determined and compared with the protein staining. Radioactivity was registered in about 12 bands, most of which correlated with stained bands. The incorporation of [3H]fucose took place in all examined phases of the cell cycle, and only small changes in the rate of incorporation in individual glycoproteins were observed, when different phases were compared.

Studies on synchronized cells in culture have revealed that synthesis of certain macromolecules such as DNA, histones [I] and certain “peak’ enzymes [2] takes place at discrete periods in the cell cycle. It is likely that the observed cyclic variations in morphology of the cell surface [3], in the expression of lectin receptor sites and antigens [4, 51, and in electrophoretic mobility of cells [6] reflect corresponding changes in synthesis of plasma membrane components or their incorporation into the plasma membrane. Few studies have been published on the biosynthesis of the components of the plasma membrane during the cell cycle. Gerner et al. [7] measured the incorporation of leucine, glucosamine and choline into the plasma membrane of synchronized KB cells and found maximal incorporation of all

three precursors just after mitosis. Nowakowski et al. [8] followed the incorporation of fucose into plasma membranes of synchronized HeLa cells and found maximal incorporation in late S phase of the cycle. Protein, glycoprotein and lipid syntheses have been studied in total cellular membranes from synchronized L5178Y (mouse lymphoma) and KB cells [9, IO], and glycoprotein synthesis in cells arrested in metaphase has also been compared with that of cells not in metaphase [I 11. As part of a study of variations in the HeLa cell plasma membrane during the cell cycle [12, 131, we investigated the synthesis during cell cycle of individual glycoproteins present in such membranes. The incorporation of [3H]fucose was used to measure the rate of glycoprotein synthesis. Fucose has been shown to be neither metabolized nor con-

54

Johnsen and Stokke

verted to other sugars, but is incorporated, unchanged, as fucosyl residues, into glycoproteins [14, 151. We observed [3H]fucose incorporation into several plasma membrane glycoproteins occurring in all examined phases of the cell cycle, with only minor deviations in rate and specificity. MATERIALS

AND METHODS

Culture methods and synchronization HeLa S, cell suspensions were grown in modified Eagle’s minimum essential spinner-medium containing 10% calf serum, and synchronized by blocking DNA synthesis with amethopterin and releasing with thymidine as previously described [ll, 121. The synchrony was controlled by measuring the incorporation of [3H]thymidine into acid-insoluble material and counting the cell number at regular intervals after reversa‘i with thymidine. The [3H]thymidine incorporation was measured each hour during S phase by incubating IO5 cells in culture medium supplemented with 0.1 &i/ml [3H]thymidine (spec. act. 21 Cilmmol, labelled at the methyl group) for 20 min. After incubation the cells were cooled on ice, sedimented at 1500 g min and washed twice in cold culture medium. Acid-insoluble material was orecinirated with ice-cold 5% trichloroacetic acid (TCA): The precipitate was collected on glass fibre filter (Whatman GF/C) and washed with 3x5 ml 5% TCA ((PC) and 1X5 ml ethanol (20°C). The filter was dried and incubated with 200 ~1 Soluene (Packard) at 50°C for 1 h. Ten ml scintillation fluid, containing 5 g PPO and 0.05 g dimethyl-POPOP per litre of toluene, was added and the radioactivity was counted. The cell number was determined every second hour by electronic cell counting. The viability of the cells was checked under the microscope after staining with trypan blue. In some cases the mitotic index was determined as previously described [ 121. As shown in fig. 1 maximal r3H]thymidine incorporation occurred 3-4 h after thymidine reversal. Cell division took place from 7-14 h after reversal, when 70-80% of the cells divided. About 10 h after reversal the mitotic index reached the maximal value of 20-25 %. The cells were harvested 1 h (early S phase), 4 h (middle S phase), 7 h (late S phase), 10 h (early mitosis) and 13 h (late mitosis) after thymidine reversal. In one experiment cells from two independently synchronized populations were harvested at 4 and 10 h after reversal and subsequently mixed. In each experiment 500 ml cultures of 2~ lo” cells/ml were used. Mycoplasma contamination was absent, judging from microbioiogicat and biochemical criteria. In the microbiological test, cells and medium were transferred to PPLO agar (Difco) and incubated aerobically and anaerobically for 2 weeks at 37°C. The biochemical test was carried out as described by Schneider & Stanbridge [16], using the incorporation ratio of [3H]uridine to [3H]uracil into DNA.

Radioactive

labelling

The cells were grown in the presence of L-[3H]fucose (spec. act. 1.0 Cilmmol, labelled at C,) for 1 or 24 h before harvesting, as specified under “Results”. In experiments where plasma membrane proteins were separated by gel electrophoresis the final concentration was 4 &i/ml; otherwise 1 pCi/ml. When mitotic cells and cells in S phase were mixed the mitotic cells were labelled beforehand with [“Hlfucose for 1 h, whereas cells in S phase were grown for 1 h in the presence of L-[‘Clfucose (spec. act. 60 mCi/mmol, labelled at C,) at a final concentration of 0.8 &i/ml. In preliminary experiments, the cells were labelled with [3H]fucose for 3 h before harvesting. Simultaneously the cells were labelled with [‘“Clleucine (spec. act. 2UOO mCi/mmol, universally labelled) at a final concentration of 1 &i/ml. The labelled compounds were obtained from The Radiochemical Centre Ltd, Amersham.

Isolation

of plasma membranes

HeLa cell plasma membranes were prepared as described earlier [12]. Briefly, the cells were broken in a Dounce homogenizer to produce large plasma membrane ghosts. The nuclei were sedimented at 680 gmin, and the ulasma membranes left in the supernatant then sed;mented at 8000 gmin. They were purified further by isopycnic centrifugation in a linear 30-50% w/w sucrose gradient. The washed membranes were finally suspended in 100 ~1 distilled water. The yield and purity of the plasma membranes were estimated by determining the recovery and relative specific activity of ouabain-sensitive ATPase [ 121. The yield was usually 15-20%, but slightly lower when plasma membranes were prepared f;om-mitotic cells rather than from S phase cells. This effect probably arises from the greater fragility of mitotic ceil plasma membranes. The specific activity of ouabain-sensitive ATPase in the plasma membrane fraction was about 15-fold higher than that in the whole cell homogenate. From each fraction obtained during membrane isolation, aliquots were withdrawn for determination of protein, according to Lowry et al. [17], and total and acid-precipitable radioactivity. The total radioactivity was -determined after addition of 10 ml Diluene (Packard) scintillation fluid. Radioactivity in acidinsoluble material was determined after precipitation with ice-cold 5 % TCA. The precipitate was collected on glass tibre filters and the radioactivity determined as described for [3H]thymidine.

Sodium dodecylsulphate electrophoresis

gel

The membrane suspension (100 ~1) was adjusted to pH 9.0 by adding 10 ~1 0.4 M borate buffer, pH 9.0, and the proteins were solubilized and reduced by addine 20 ~LLI10% sodium dodecvlsulphate (SDS) and 5 /*I 2lmercaptoethanol followed by incubation in boiling water for 2 min. The sample was dialysed at room temperature for 16 h against 300 ml 0.0625 M TrisHCl, pH 6.8, 0.1% SDS, 1% 2-mercaptoethanol and submitted to disc electrophoresis according to

HeLa

Fir. 1. Abscissa: time after reversal (hours); ordinate: (/&) SH radioactivity (cpmx lOma); (right) cell no. (X 10-5/ml). X-X, “H radioactivity; O-O, cell count. Variations in cell number and incorporation of [3H]thvmidine into acid-insoluble material in a cell culture synchronized by amethopterin-thymidine. Reversal with thvmidine at time 0. Samples of IO5 cells were labelled with [3H]thymidine for 20 min and incorporation was determined as described in the text.

Laemmli [ 181 on 160x5 mm separating gels of 8 % acrylamide and 2.7% cross-linkage. A 100 ~1 sample containing about 50-100 pg protein was layered on each gel. The electrophoresis was run at room temperature with 0.5 mA/gel and stopped when the marker, bromophenol blue, was 5 mm from the bottom of the gel (about 20 h). The gels were stained with Coomassie Brilliant Blue according to Bjerklid et al. [ 191. After staining, the gels were scanned at 560 nm in a Gilford spectrophotometer with gel scanning attachment. Molecular weights of the membrane proteins were estimated from a standard curve (fig. 2) obtained by submitting the following marker proteins to electrophoresis under the above conditions: Human haemoglobin (Sigma); ovalbumin (Sigma); reduced and unreduced human gammaglobulin (Nutritional Biochemicals); lactoperoxidase (Sigma); glucose oxidase (Worthington); human fibrinogen (Kabi); bovine serum albumin (Sigma); and oligomers of bovine serum albumin cross-linked by glutaraldehyde [20].

Counting ofradioactivity

cell plasma

membranes.

III

55

poration into acid-insoluble material -of the labelled fucose. The cells were cooled on ice and washed three times with cold, serum-free culture medium. The cell pellet was suspended in I ml ice-cold 0.4 M perchloric acid (PCA) and incubated for 20 min on ice. The precipitate was sedimented at 64000 gmin, the supernatant decanted and neutralized with 0.5 ml of a solution of 0.72 M KOH containing 0.6 M KHCO,. Potassium perchlorate was sedimented and the supematant decanted. The radioactivity in PCA-precioitated material was determined after solubilizat&n in 0.5 ml Soluene at 50°C for I h and addition of 10 ml toluene-based scintillation fluid. The radioactivity in soluble material was counted in IO ml Diluene. The PCA-soluble material was chromatographed on a Dowex-I anion exchanger, equilibrated with IO mM phosphate buffer pH 7.0. The material was eluted by a continuous gradient of KC1 from 0 to 2 M, in the same phosphate buffer. Fractions of I.5 ml were collected and radioactivity in each fraction was counted after addition of IO ml Diluene. L-fl-3Hlfucose and GDP-L[U-“‘Clfucose were chromaiographed under identical conditions to identify their elution positions in the chromatogram.

Extraction of lipids Synchronized cell populations were labelled with I &i/ml of f3Hlfucose for 60 min. About 10’ cells were harvested, washed 3 times and suspended in 1 ml water. Two hundred ~1 cell suspension was mixed with 3.8 ml chloroform-methanol 2: 1 v/v, stoppered, and incubated at 37°C for 30 min. The oreciuitate was separated by centrifugation and washed &with I ml 0.88 % KCl/chloroform/methanol 3 : 38 : 19 v/v/v. The combined supematant was extracted with 0.2 vol

in gels

The gels were frozen and each gel cut with razor blades into about 80 slices 2 mm in length. The slices were dried, and the radioactive material eluted with 200 ~1 Soluene (Packard) at 50°C for 3 h. Six ml scintillation fluid (Dimilume, Packard) was added to each vial. In double-label experiments, 3H and r4C were counted in separate channels and the values in the 3H channel were corrected for overflow of ‘“C disintegrations (about IO %).

Cellular uptake and incorporation of [3Hlfucose Synchronized cell populations were labelled with I pCi/ml of f3Hlfucose for 60 min. At short intervals ahquots of about 3~ IO6 cells were harvested in order to determine the uptake, metabolization and incor-

I 0.z

0.4

cl..4

0.8

1.0

2. Abscissa: mobility relative to bromophenol blue; ordinate: log mol. wt. Correlation between Ion mol. wt and migration relative to bromophenol blue in SDS-gel ele&ophoresis. I, Fibrinogen; 2, bovine serum albumin, trimer; 3, gamma globulin; 4, bovine serum albumin, dimer; 5, glucose oxidase; 6, lactoperoxidase; 7, bovine serum albumin, monomer; 8, gamma globulin heavy chain; 9, ovalbumin; 10, gamma globulin light chain; II, haemoglobin monomer. Fig.

Exp

Cdl

Ru

109 (1977)

56

Johnsen

and Stokke

3. Abscissa: time of incubation (min); ordinate: cpm (X 10m3). (a) Total activity; (b) activity in acidprecipitable material. Uptake and incorporation of [3H]fucose into HeLa cells. 2~ lo7 cells were incubated at 3PC with [3H]fucose (1 &i/ml) for 60 min. Aliquots of 2x lo6 cells were withdrawn at indicated time points and the radioactivity in total and acid-insoluble material was determined as described in the text.

Fig.

0.88 % KCl. The phases were separated and the lower phase was washed with the new upper phase, according to Folch et al. [21]. The upper phase was fractionated by chromatography on Sephadex LH-20 equilibrated with 0.88% KCI-methanol 1 : 1 v/v. The column was calibrated with cytochrome c , [IIC]GPD fucose and [3H]fucose in 0.88 % KCl-methanol 1 : 1 v/v. The radioactivity in the upper and lower phase and in fractions collected from the Sephadex LH-20 column was determined in Diluene. The quenching in chloroform-containing samples was measured and the results were corrected accordingly.

RESULTS [3H]Fucose uptake and incorporation in synchronized HeLa cells

Exogenous [3H]fucose was taken up and incorporated into acid-insoluble material by HeLa cells in suspension culture. The rate of uptake was constant for at least 60 min (fig. 3a), and intracellular [3H]fucose was rapidly converted to [3H]GDP-fucose, which constituted 70-80% of the acid-soluble labelled pool (fig. 4). With the exception of an initial delay of 10-15 min, the labelled precursor was incorporated into acid-insoluble material at a constant rate over the 60 min observation period (fig. 3 b). The uptake rates were nearly identical for E.r/> Cell Rcs 109 (1977)

Fig. 4. Abscissa: fraction no.; ordinate: (leff) cpmi 0.5 ml fraction; (right) KC1 cont. (moles/l). Distribution of radioactivity in acid-soluble pool after labelling of HeLa cells with [3H]fucose. 2x 10” cells were labelled for 60 min with [3H]fucose (1 pCi/ ml), harvested and extracted with 0.4 M PCA. Acidsoluble material was applied to a Dowex-1 column and eluted with increasing concentrations of KC1 in 10 mM phosphate buffer, pH 7.0. Fractions of 1.5 ml were collected and their radioactivity determined. About 80% of applied radioactivity was recovered in the eluate. The elutions of fucose and GDP-fucose are indicated.

cells labelled and harvested in different phases in the cell cycle and so were the incorporation rates (table 1). 8.0-8.4 cpm/pg cellular protein was incorporated in cells pulse-labelled during the first 10 h after thymidine reversal. This constituted 3841 % of the radioactivity taken up by the cells. Cells labelled 12-13 h after reversal incorporated 9.1 cpm/pg protein, which was 46 % of the total cellular uptake. The pulse-labelled cells were extracted with chloroform-methanol and the phases separated according to Folch et al. [21]. Table 1. Uptake and incorporation of [“HIfucose in synchronized HeLa cells Pulse (hours)

Uptake”

Incorp.”

O-l 3-4 6-7 9-10 12-13

22.2 21.4 20.6 19.7 19.8

8.4 8.4 8.4 8.0 9.1

a Cpm/pg protein (mean of three experiments).

HeLa cell plasma membranes. III

Table 2. Incorporation lipid-soluble cells (I

material

of [3Hlfucose into of synchronized HeLa

Pulse (hours)

Chloroformmethanol extract (%*o)

phase (%)

Lower phase m’c)

O-1 3-l 67 9-10 12-13

83.5 80.3 76.3 75.8 72.8

53.5 53.9 52.0 48.2 47.6

34.0 30.6 29.6 27.8 26.0

Upper

fI The values arc given as percentages of total radioactivity recovered in the cells (means of duplicate samples).

The radioactivity recovered in the total extracted material and its distribution into the upper and lower phase is recorded in table 2. Most of the radioactivity was recovered in the upper water-methanol phase and further fractionation by chromatography on Sephadex LH-20 showed that this radioactive material eluted exclusively as fucase and GDP-fucose (fig. 5). The radioactive material recovered in the lower phase was not further characterized.

Fig. 5. Abscissa: fraction no.; ordinate: cpm/fraction. Chromatography on Sephadex LH-20 of radioactive material present in the upper water-methanol phase after Folch separation of chloroform-methanol extract from HeLa cells incubated for 1 h with [3H]fucose (1 @/ml). Upper phase extract from lo6 cells contained 3 800 cpm, 92 % of which was recovered in the eluate. The elutions of cytochrome c; GDP-fucose and fucose are indicated.

57

Table 3. Distribution of incorporated [3H]fucose over subcellular fractions a Cells labelled in S phase*

Cells labelled in mitosis’

Fraction

Total act.”

Spec. act.e

Total act.”

Spec. act.?

Cell homogenate 680 g supernatant 680 g pellet 8 000 g supernatant 8 000 g pellet Plasma membranes

325 201 55 126 39 6.5

8.4 9.2 7.1 7.6 16.9 26.0

403 241 88 173 40 6.0

8.0 8.2 6.2 7.2 13.5 24.1

n Mean of three experiments. b One hour pulse (4 &i/ml) given 3 h after thymidine reversal. c One hour pulse (4 &i/ml) given 9 h after thymidine reversal. d Acid-insoluble radioactivity, cpm. p Cpm/pg protein.

Incorporation of [3Hlfucose into subcellular fractions

Labelling the cells for 1 h with [3H]fucose resulted in a distribution of TCA-precipitable radioactivity over the subcellular fractions obtained from the plasma membrane isolation as shown in table 3. 15 2.0 % of the total incorporated radioactivity was recovered in the plasma membrane fraction, and the specific activity obtained was about 25 cpm/pg protein, a 2-3-fold increase above the whole cell homogenate. Radioactivity recovered in the 680 gmin pellet and the 8000 gmin supernatant was partly due to plasma membrane material, as shown by the presence of ouabain-sensitive ATPase in these fractions [12]. However, some of the radioactive material in these fractions was apparently unrelated to, but could still originate from, the plasma membranes. Thus, separation by gel electrophoresis of the proteins in the 8000 gmin supernatant revealed incorporation of [3H]fucose into several glycoproteins in addition to those glycoproteins present in the plasma membrane fractions. /+/I Cd/ Rr

109 (/977)

58

Johnsen

and Stokke

Fig. 6. Abscissa: mobility relative to bromophenol blue; ordinate: AsGo nm. SDS-polyacrylamide gel electrophoresis of plasma membrane proteins (50 pg) prepared from HeLa cells exposed to [3H]fucose (4 @I/ml) for 1 h before harvesting in middle S phase (4 h after reversal with thymidine). Optical density tracing after staining with Coomassie Brilliant Blue. The positions of radioactive bands shown in fig. 7 are indicated by arrows.

The recovery of radioactivity in the plasma membrane fraction from cells in mitosis was slightly lower than that from cells in S phase, which merely reflects lower yield of plasma membranes from mitotic cells. Extension of the labelling time to 24 h caused a further displacement of radioactivity to the plasma membrane fraction and resulted in a specific activity which was 4-5-fold higher than that of the whole cell homogenate. [3H]Fucose incorporation into the plasma membrane glycoproteins

Gel electrophoresis of plasma proteins followed by staining

membrane with Coo-

massie Brilliant Blue revealed about 35-40 bands (fig. 6). The main pattern was unchanged in membrane preparations from various cell cycle phases. The proteins ranged, in apparent molecular weight, from 15 000 to more than 200 000, and the most prominent band had a relative mobility of about 0.6 corresponding to a molecular weight of 42 000. In a 3 h pulse [14C]leucine was apparently incorporated into all protein bands which were visible after staining and the radioactivity of the bands generally followed the staining intensities. The high number of bands made it difficult to resolve the fine pattern in the 14C radioactivity and minor protein bands lacking radioactivity or minor changes in specific radioactivity could not be excluded. [3H]Fucose was incorporated into about 12 bands, some of which corresponded well to major protein bands, while others could not with certainty be related to visible protein bands (fig. 7). Two radioactive bands (I, II) were regularly found at RM 0.1, probably corresponding to the two distinct protein bands with molecular weight around 250000 seen in the same area. About 3-5 poorly resolved bands appeared with RM from 0.2 to 0.4. The many protein bands in this part of the gel made it difficult to assign, with certainty, the radioactive bands to the protein bands. The main radioactive band (VIII) migrated with RM 0.4, suggesting a molecular weight of about 70000. It appeared in all

7. Abscissa: mobility relative to bromophenol blue; or&tare: cpm/gel slice. Incorporation of [3H]fucose into individual plasma membrane proteins. The polyacrylamide gel presented in fig. 6 was cut into slices of 2 mm and radioactivity was determined in each slice. A total of 8490 cpm was registered in the gel, which represented 91% of the applied radioac-

Fig.

tivity. Evp CdRes

109 (1977)

HeLa

plasma membrane preparations as a narrow band of high radioactivity and corresponded to a major protein band in the stained gels. Between RM 0.6 and 0.7 two closely positioned bands were recorded (X, XI) with apparent molecular weights around 40 000. The position of band X was not identical with the position of the major protein band at RM 0.6. Coelectrophoresis of [l”C]leutine- and [3H]fucose-labelled material in the same gel showed clearly that band X always migrated a little ahead of the [‘“Clleucine-labelled major protein band with R, 0.6.

When the cells were pulse-labelled and harvested in different phases of the cell cycle and the plasma membrane proteins separated by gel electrophoresis no major reproducible change in the pattern of fucose incorporation was observed. This conclusion was confirmed in an experiment where cells in S phase labelled with [14C]fucose and cells in mitosis labelled with [3H]fucose were mixed before plasma membrane isolation and gel electrophoresis. Measurements of 14C and 3H activities in the same gel revealed identical positions of the labelled bands and identical distribution of radioactivity over the bands. Minor deviations in the radioactive pattern occurred when cells from different phases were examined. Thus the ratio between incorporated fucose in band XI and band X was always higher in plasma membrane preparations from mitotic cells than from S phase cells. However, small variations in gel slicing might produce considerable changes in the assignment of counts between neighbouring bands. When cells were grown in the presence of [3H]fucose for 24 h before harvesting, the same pattern of radioactivity was obtained as after pulse labelling. This is consistent with a constant ratio through the cell cycle

cell plasma

membranes.

III

59

between the incorporation rates for all labelled plasma membrane glycoproteins. DISCUSSION HeLa cells grown in the presence of r3H]fucose incorporate the labelled compound into their glycoproteins. The label is enriched in the plasma membranes, and the extent of enrichment increases with the length of the labelling time. This can be explained by a movement of labelled glycoproteins from intracellular compartments to the plasma membrane. This view is consistent with previous studies on the biosynthesis of plasma membrane glycoproteins, in which it was shown by radioautography [22] and by subcellular fractionation [23] that exogenous fucose was initially incorporated in the Golgi apparatus and later appeared in the plasma membrane. The same conclusion was reached by Atkinson [24] using pulse-chase experiments with HeLa cells where it was found that incorporation of [3H]fucose in plasma membranes occurred later than incorporation in whole cells. Low enrichment in plasma membranes after 60 min pulse-labelling with [3H]fucose was found by Bosman [El, while Atkinson [24] observed a specific activity which was 7 times higher in the plasma membranes than in whole cells after being given a [3H]fucose pulse of 40 min. Separation of the plasma membrane proteins and glycoproteins on polyacrylamide gels revealed that there was [14C]leucine activity in all major protein bands and [3H]fucose activity in several bands. This indicates that when occurring in isolated plasma membranes, prepared from HeLa cells, none of the main proteins or glycoproteins are derived from serum in the culture medium, but does not exclude the possibility that they are so derived when present in

60

Johnsen and Stokke

minor quantities in or loosely adsorbed onto the membrane. It is possible, however, that [3H]fucose could be incorporated into glycoproteins not synthesized by the cells. Fucase occupies a terminal position in many glycoproteins. It could be attached by enzymes in the cell surface to incomplete or partially degraded carbohydrate chains after adsorption of glycoprotein to the plasma membrane from the environment. This mechanism requires fucosyl transferase to be located at the cell surface, a position where several other glycosyl transferase may well be located [25, 261. However, fucosyl transferases have only been detected in the smooth microsomal fraction [27, 281 and there is substantial evidence that fucosylation solely takes place in the Golgi apparatus [22]. The opposite situation, i.e. loss of plasma membrane components during the isolation procedure, is more difficult to exclude. However, to avoid disturbances in the membrane integrity mild conditions of homogenization and fractionation were used and these resulted in apparently intact plasma membrane ghosts. Moreover, glycoproteins seem generally to be more tightly integrated into membranes and require apolar solvents in order to be solubilized. Thus it is reasonable to assume that the labelled glycoproteins recovered in the plasma membrane fraction actually represent the fucose-containing glycoproteins present in the HeLa cell plasma membrane. Several glycoproteins, including the 70000 molecular weight one, could be labelled with lz51 using the lactoperoxidase method [29]. This is in accordance with the general assumption that proteins exposed on the external cell surface are always glycosylated. However, electrophoretic separation of the proteins in the cytoplasmic fraction (8 000 g supernatant) revealed [3H]El-i> Cell

Res 109 (l977)

fucose incorporation into other glycoproteins. These fucosylated glycoproteins might originate from the Golgi apparatus which is present in the 8000 g supernatant [12]. Thus, it is clear that after short labelling times, fucose incorporation is not confined to plasma membrane glycoproteins. We observed only minor changes in the incorporation rate of [3H]fucose during the cell cycle. More pronounced cell cycledependent changes have been reported for glycoprotein biosynthesis in different cell lines, both when examined in whole cells [9, lo] and in isolated plasma membranes [7]. Nowakowski et al. [S] observed maximal incorporation of [3H]fucose into TCAprecipitable material from HeLa cells in late S phase. However, these workers synchronized the cells by the double thymidine block method and to obtain increased incorporation of [3H]fucose into the plasma membranes the cells were concentrated before labelling. In our experience, exposure of synchronized cells to centrifugation and concentration disturbs the cell synchrony and for that reason we have avoided such operations in our experiments. Our tentative findings of an altered incorporation of [3H]fucose into glycoproteins (bands X and XI) in mitotic cell plasma membranes might reflect specific synthesis of fucosylation of one particular glycoprotein at mitosis. At least two fucosyl transferases with different glycoprotein acceptor specificities have been described for HeLa cells [27]. Alternatively it might reflect a modification (e.g. sialylation) of a fuco-glycoprotein already present. Chromatographic separation of fucoselabelled glycopeptides released from the surface of tibroblast cells has been shown to reveal different patterns when metaphase cells were compared with cells not in meta-

HeLa cell plasma

phase [Ill. The metaphase cells released a higher proportion of glycopeptides of apparently higher molecular weight. A similar pattern was obtained after virus transformation of the fibroblasts. This suggests that the fucose-containing glycopeptides which are expressed on the cell surface of normal tibroblasts during mitosis are similar to those which are permanently expressed after virus transformation. Neuraminidase digestion of the glycopeptides resulted in identical chromatographic patterns for normal and transformed tibroblasts, thus indicating a higher content of sialic acid in glycopeptides from transformed or mitotic cells. A higher content of sialic acid in glycoproteins would presumably increase their rate of mobility in SDS-gel electrophoresis [30]. This could explain the observed increase in the ratio of radioactivity between band XI and band X found in mitotic cells. Cell cycle-dependent changes in the existence of one single glycoprotein species have been observed in a variety of Iibroblast strains. A high molecular weight (200 000-250 000) fucoglycoprotein exposed at the cell surface is lost when the cells enter mitosis [3 1, 321. No such relation was observed by us in the HeLa cells. This could be due to the epithelial or malignant origin of the cells, since transformed fibroblasts lack glycoprotein throughout the cell cycle. This work was supported in part by the Norwegian Research Council for Science and the Humanities.

REFERENCES 1. Robbins, E & Scharff, M, Cell synchrony (ed I L Cameron & G Padilla) p. 353. Academic Press, New York (1966). 2. Mitchison, J M. Science 165 (1969) 657.

membranes.

III

61

3. Porter, K R, Prescott, D & Frye, J, J cell biol 57 (1973) 815. 4. Noonan, K D, Levine, A J & Burger, M M, J cell biol58 (1973) 491. 5. Kuhns, W J &Bramson, S, Nature 219 (1968) 938. 6. Brent, T P & Forrester, J A, Nature 215 (1967) 92. 7. Gerner, E W, Glick, M C & Warren, L, J cell physiol75 (1970) 275. 8. Nowakowski, M, Atkinson, P H & Summers, D F, Biochim biophys acta 266 (1972) 154. 9. Bosmann, H B & Winston, R A, J cell biol 45 (1970) 23. 10. Chattejee, S, Sweeley, C C & Velicer, L F, Biothem biophys res commun 54 (1973) 585. 11. Glick, M C & Buck, C A, Biochemistry 12 (1973) 85. 12. Johnsen, S, Stokke, T & Prydz, H, J cell biol 63 (1974) 357. 13. - Exp cell res 93 (1975) 245. 14. Kaufman, R L & Ginsburg, V, Exp cell res 50 (1968) 127. 15. Bosmann, H B, Hagonian, A & Evlar. E H. Arch biochem biophys 130 i1969) 573. . 16. Schneider, E L & Stanbridge, E J, Methods in cell biology (ed D M Prescott) vol. 10, p. 277. Academic Press, New York (1975). 17. Lowry, 0 H, Rosebrough, N J, Farr, A L & Randall, R J, J biol them 193 (1951) 265. 18. Laemmli, U K, Nature 227 (1970) 680. 19. Bjerklid, E, Storm, E & Prydz, H, Biochem biophys res commun 55 (1973) 969. 20. Griffith, I P, Biochem j 126 (1972) 553. 21. Folch, J, Lees, M & Sloane Stanley, G H, J biol them 226 (1957) 497. 22. Bennett, G, Leblond, C P & Haddad, A, J cell biol 60 (1974) 258. 23. Riordan, J R, Mitranic, M, Slavik, M & Moscarello, M A, FEBS lett 47 (1974) 248. 24. Atkinson, P H, Methods in cell biology (ed D M Prescott) vol. 7, p. 157. Academic Press, New York (1973). 25. Roth,‘S, McGuire, E J & Roseman, S, J cell biol51 (1971) 536. 26. Bosmann, H B, Biochem biophys res commun 48 (1972) 523. 27. -Arch biochem biophys 145 (1971) 310. 28. Jabbal, I & Schachter, H, J biol them 246 (1971) 5154. 29. Gudjonsson, H & Johnsen, S. In preparation. 30. Segrest, J P, Jackson, R L, Andrews, E P & Marchesi, V T, Biochem biophys res commun 44 (1971) 390. 31. Hynes, R 0 &Bye, J M, Cell 3 (1974) 113. 32. Gahmberg, C G & Hakomori, S, Biochem biophys res commun 59 (1974) 283. Received January 10, 1977 Revised version received April 18, 1977 Accepted May 5, 1977

Eavp Cd/

Rr

IO9 (1977)