Association of Viral Particles and Viral Proteins ... - Journal of Virology

5 downloads 154 Views 2MB Size Report
Apr 7, 1982 - Altenburg, B. C., D. Y. Graham, and M. K. Estes. 1980. Ultrastructural study of rotavirus replication in cultured cells. J. Gen. Virol. 46:75-85. 4.
Vol. 44, No. 3

JOURNAL OF VIROLOGY. Dec. 1982, p. 983-992 0022-538X/82/120983-10$02.00/0 Copyright ©D 1982, American Society for Microbiology

Association of Viral Particles and Viral Proteins with Membranes in SAl1-Infected Cells CARMEN SOLER, CLAUDETTE MUSALEM, MARIA LORONO, AND ROMILIO T. ESPEJO* Instituto de Investigaciones Biomedicas, Universidad Nacional Aut6noma de Mexico, Mexico 20, D.F., Mexico Received 7 April 1982/Accepted 9 August 1982

Electron microscopy after negative staining of SAl1-infected cell homogenates revealed that most of the viral particles are associated with membrane-like material. Many of the particles seemed to be fully enveloped in a membrane. This association could also be detected by the observed cosedimentation of viral proteins and cell membranes. Pulse-chase experiments showed that viral glycoproteins rapidly associate with membranes, whereas most of the structural proteins appearing in the soluble fraction immediately after the pulse were slowly chased into the membrane fraction. The membranes could be further fractionated into at least four fractions differing in density and containing a different distribution of viral proteins. Also, the distribution of label into each of these membrane fractions changed after long chase periods. The inhibition of glycosylation with tunicamycin yielded viral particles without an outer layer, but did not affect the described association with membranes. The possible relationship of this finding to the maturation of the virion is discussed. cells (9). These enveloped particles have also been observed in the preparations of orbiviruses (10, 21), which are closely related to rotaviruses in composition and morphology. In this paper, we describe the association of immature particles and SAl1 glycoproteins with membranes and discuss the possibility that the observed association is a step in the virus acquisition of the thin outer layer of protein, contained in the virion.

Tissue culture-adapted simian rotavirus SAl 1 (19) has become a model system for rotaviruses. These members of the Reoviridae family cause enteritis in the young of many mammalian species (14, 15). SAl virions are composed of an RNA genome and a double-layered capsid. The genome is made up of 11 segments of doublestranded (ds) RNA (23). The capsid consists of at least five protein classes; three of them (VP1, VP2 and VP6) make the inner capsid, and two others (VP3 and VP7) form the outer capsid (12). At least 12 viral polypeptides can be distinguished in infected cells. Two of them, VP7 and NCVP5, seem to be glycosylated with N-glycosidic residues of the high-mannose type. This is suggested by their sensitivity to tunicamycin and endo-N-acetylglucosaminidase H (12). Of the 12 viral polypeptides observed in infected cells, one (pNCVP5) is the apoprotein of the noncapsid viral glycoprotein NCVP5 (4). These findings make SAll a unique model for studying the role of N-glycoproteins in the replication of nonenveloped viruses. It is well known that enveloped viruses acquire their membrane by budding through one of the cell membranes and that in this process the glycoproteins play a major role. By analogy, SAll glycoproteins may have a role in virus morphogenesis because immature particles seem to bud into the lumen of the endoplasmic reticulum (3, 22). In fact, SAll virions have been observed enveloped by membranes in negatively stained lysates of infected

MATERIALS AND METHODS Cells, virus, and media. MA104 cells were grown as monolayers in 125-cm2 disposable flasks in Eagle minimal essential medium (MEM), supplemented with 10% fetal calf serum, penicillin (100 ,ug/ml), streptomycin (100 ,ug/ml), and gentamicin (160 ,g/ml). Simian rotavirus (SAl1) was obtained as previously described (12). Infection of cells, radioactive labeling, and processing of infected cells. MA104 confluent monolayers were washed twice with phosphate-buffered saline and infected with SAl1 virus in the presence of trypsin (10 ,ug/ml). In experiments involving tunicamycin (a gift of P. L. Hamill of Lilly Research Laboratories), the drug was added to the cells during the adsorption period at a concentration of 5 ,ug/ml and was kept the same during all subsequent steps. After 1 h of adsorption at 37°C the inoculum was removed, the monolayer was washed with phosphate-buffered saline, and the cells were incubated at 37°C for 4 h in MEM without serum and containing 1/20 the normal amount of methionine. At that time, 100 iCi of L-[35S]methionine (500 Ci/ mmol; New England Nuclear Corp.; Boston, Mass.) 983

984

SOLER ET AL.

per ml was added. Time zero of infection was considered the time after adsorption, when medium was added. After a 15-min pulse period, the cells were washed three times with warmed MEM and were further incubated for the desired chase period. Cells which were only pulse-labeled were placed on ice, washed, and fractionated immediately as described below. At the end of the chase period, the cells were washed four times with cold calcium and magnesiumfree phosphate-buffered saline and once with ET buffer (0.02 M Tris, 0.003 M EDTA [pH 8.2]). Labeling with [32P]phosphate was performed as described for L[35S]methionine, except that MEM containing 1/20 the normal amount of phosphate and 20 ,uCi of [32P]phosphate ([32P]orthophosphoric acid, carrier free; New England Nuclear Corp.) per ml was used. Homogenization and fractionation of infected cells. The method of homogenization and fractionation was essentially that described by Erwin and Brown (11). The washed cells (8 x 106) were suspended in 2 ml of ET buffer containing 190 ,ug of phenylmethylsulfonyl fluoride (PMSF; Sigma Chemical Co., St. Louis, Mo.) per ml. The cells were then allowed to detach, and the cell suspension was homogenized by 25 strokes of a Dounce homogenizer. Cellular breakage was monitored with a light microscope. For fractionation, 1.2 ml of the cell homogenate was layered on a discontinuous gradient made of 1.0 ml of CsCl (1.5 g/cm3), 6.0 ml of 60% sucrose (wt/vol), and 8 ml of 10% sucrose (wt/vol) (all solutions were made in ET buffer containing PMSF at 190 ,ug/ml). Centrifugation was carried out in a Sorvall HB-4 swinging rotor at 12,000 rpm for 60 min at 5°C. Fractions (0.2 ml) were subsequently collected from the bottom of the centrifuge tube, and 10-,ul samples of each fraction were trichloroacetic acid precipitated onto Whatman filters and counted. For further analysis, the fractions containing the membranous material were pooled, diluted with 1 volume of ET buffer, and overlaid on a discontinuous gradient containing 0.5 ml of CsCl (1.5 g/cm3), 1.5 ml of 50% sucrose, 1.5 ml of 44% sucrose, 3.0 ml of 39% sucrose, 3.0 ml of 34% sucrose, and 2.5 ml of 29% sucrose (all sucrose solutions are given as weight to weight in ET buffer containing PMSF at 190 ,g/ml). Centrifugation was performed in a Beckman SW40 rotor at 36,000 rpm for 90 min at 5°C. At the end of the run, 0.2-ml fractions were collected from the bottom of the tubes, and the radioactivity of 20 ,ul was determined as described above. To check that the association observed was not produced during homogenization, mock-infected cells were mixed with [35S]methionine-labeled partially purified virus just before homogenization, and the homogenate was subsequently analyzed. Partially purified virus was obtained by Freon treatment of cells suspended in homogenization buffer 24 h after SAl1 infection. Extraction of viral particles. Samples of the pooled fractions containing the membranous material, obtained after low-speed centrifugation, were extracted with an equal volume of trifluorotrichloroethane (Freon). The interface obtained after the first extraction was subsequently extracted twice with ET buffer. The material present in the interface and aqueous phase was analyzed by gel electrophoresis and velocity sedimentation. For velocity sedimentation, the material present in the aqueous phase was layered over a linear

J. VIROL.

sucrose gradient (15 to 45% [wt/vol]) made in TSM buffer (0.01 M Tris-hydrochloride, 0.15 M NaCl, 0.001 M MgCl2; pH 8.2) and centrifuged in a Beckman SW40 rotor at 36,000 rpm for 90 min at 5°C. Fractions (0.2 ml) were collected from the bottom of the tubes and counted as described above. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Analysis of the polypeptide composition of the different samples was carried out by the method of Laemmli (17) in 12% polyacrylamide gels. Samples were boiled for 2 min with Laemmli sample buffer and stacked at 10 mA/gel, and electrophoresis was continued at 20 mA/gel. Gels were first fixed and stained with Coomassie brilliant blue and after destaining were processed for fluorography by the method of Bonner and Laskey (5). Negative-contrast electron microscopy. Cell homogenates and subcellular membrane fractions obtained as described above were used for electron microscopic observations. A drop of the sample was placed for 1 min on a 400-mesh Formvar-coated copper grid covered with a carbon film, the excess was drawn off with a piece of filter paper, a drop of the contrasting agent (2% phosphotungstate, pH 7.0) was added to the sample, and the excess liquid was quickly removed with a piece of filter paper. After being dried, the grid was examined in a JEOL 1000B electron microscope at a magnification of 30,000, using an operating voltage of 80 kv.

RESULTS Electron microscopy. An examination of negatively stained homogenates of infected cells showed that most of the viral particles were associated with membrane-like material. In the cell homogenates obtained 9 h postinfection (Fig. 1), particles were usually in large aggregates which consisted of both cores and viruses without an outer layer. Both types of particles were observed either full or empty. The viruses without an outer layer were observed either partially associated with (Fig. 1B and C), or completely surrounded by (Fig. 1A and D), membrane-like material. This association had been reported previously in infected cells ruptured by osmotic shock (9). On the other hand, cores were not found directly associated with the putative membranes (arrow in Fig. 1C). Some of these aggregates were found inside membrane vesicles (Fig. 1D), reminiscent of those observed in stained thin sections of SAilinfected cells (3). An examination of infected cell homogenates obtained 22 h postinfection revealed essentially the same general features, except for the presence of free and associated viruses which had acquired the thin outer layer. The observation of complete virus late after infection indicates that the failure to observe particles with an outer layer at earlier times is not due to the presence of EDTA in the homogenization buffer, which can cause the loss of the outer layer of virions in other conditions.

MEMBRANE ASSOCIATION IN SAl1-INFECTED CELLS

VOL. 44, 1982

AL

*

*.l

B

I..

--

t

X.

is

,Q.

t

985

,D

.: t t-

D

C

If V

Ci

4wi

I

y ...

-A.

p.,i'v -

FIG. 1. Electron micrograph of negatively stained homogenates of MA104 cells obtained 9 h after infection with SAil. Infection, cell homogenization, and negative staining were performed as described in the text. Symbols: *, viruses without outer layer; t cores. ,

The examination of negatively stained homogenates of mock-infected cells mixed with partially purified viruses did not show viral particles

associated with membrane-like material. Distribution of SAil viral proteins. The subcellular fractionation of [35S]methionine-labeled infected cells was performed to determine the possible association between cellular membranes and SAil viral proteins. SAil-infected cells were pulse-labeled at 4 h postinfection for

15 min. At this time, the synthesis of host protein was inhibited, and most of the label was incorporated into viral proteins. The cells were homogenized either immediately or after a 5-h or 18-h chase period. The initial fractionation of the subcellular components was accomplished by low-speed centrifugation in a discontinuous density gradient. Under these conditions, the recovery of TCA-insoluble [35S]methionine was between 90 and 100%, and most of the label was

986

SOLER ET AL.

J. V IROL.

A

X

sucrose 10 %

sucrose 60 %

60-

E C, U)

50-

4030-

20-

5

I0

15

20

25

30

35

40

45

FRACTION NUMBER

FIG. 2. Distribution of SA1l viral proteins. (A) Centrifugation at 12,000 rpm on a discontinuous gradient of [35S]methionine-pulse- and pulse-chase-labeled SAl-infected cells, homogenated as described in the text. Incorporation after pulse and chases did not vary by more than 10%. A total of 10 and 34% of the incorporated label was found in the medium after the 5- and 18-h chases, respectively. Each gradient was loaded with the same amount of radioactive material. The original gradient composition is shown at the top of the figure. When the sucrose concentration in the collected samples was determined, the middle band was found in 50% sucrose. The recovery of radioactivity after centrifugation was between 90 and 100%. Symbols: 0, 15-min pulse cell homogenate; 0, 5-h pulse-chase cell homogenate; A, 18-h pulse-chase cell homogenate. (B) Sodium dodecyl sulfate-I 2%Y polyacrylamide gel electrophoresis of polypeptides obtained upon the fractionation of SAl 1-infected MA104 cells. Abbreviations: H, cell homogenate; S, soluble fractions; M. membrane fractions of pulse, 5-h. and 18-h chase. For electrophoresis, the three or four fractions containing the higher amount of label of each peak were pooled, and the gels were subsequently loaded with the appropriate amount of radioactivity to be able to observe the different viral proteins. One thousandth of the total amount of homogenate loaded in the gradient was employed for electrophoresis in the pulse and chases. A total of 8/1,000, 11/1,000, and 15/1,000 of the material recovered in the soluble fractions was employed in the pulse, 5-h chase, and 18-h chase, respectively. A total of 20/1,000, 20/1,000 and 15/1,000 of the material recovered in the membrane fractions was employed in the pulse, 5h chase, and 18-h chase, respectively.

distributed in two main components: a slightly turbid band close to the 60 to 10% sucrose interface and a soluble fraction at the top of the gradient (Fig. 2A). Most of the label appearing at the bottom of the gradient did not correspond to viral proteins. The distribution of the label in the infected cell homogenates obtained after the pulse or after the subsequent chase suggested that part of the viral proteins initially observed

in a soluble form are slowly incorporated into the membranous material found close to the top of the 60% sucrose layer. More than 90% of the label was released from this material when the homogenate was treated with Triton X-100 (data not shown). The treatment of [35S]methioninelabeled infected cells with Freon at 24 h postinfection left most of the viral proteins and viral particles in soluble form in the aqueous phase.

MEMBRANE ASSOCIATION IN SAil-INFECTED CELLS

VOL. 44, 1982

18h Chose

9u' hose SJ

S

v W

987

VMH

S

M

IV

:Vpi

--

-

VP2-3' NCVP1 NCVP2 VP6p36

NCVP3-

pV P7o:-

NCVPe t6s b ,. o

FIG. 2B

When this material was mixed with mock-infected cells and subsequently homogenized and centrifuged as above, 10 to 12% of the label sedimented together with the membrane fraction. This amount of trapping or absorption could be significant when interpreting the results obtained after the pulse, but it probably does not change significantly the results obtained after the chase, when the amount of label found in the membrane fraction was greater than 75%. Gel electrophoresis analysis of the radioactively labeled proteins of the homogenates of pulse- and pulse-chase-labeled cells, showed the same composition reported by Dyall-Smith and Holmes (8) and by Arias et al. (4) for SA1linfected cells, except for minor differences which will be discussed below. Significant differences were observed in the distribution and composition of viral proteins in the membrane and soluble fractions obtained after centrifugation (Fig. 2B). Nonstructural proteins, previously identified as NCVP1 and NCVP5 (4), were found in the membrane fraction of both pulseand pulse-chase-labeled cells, whereas most structural viral proteins, notably VP2, VP3, and VP6, appeared in large amounts in the soluble fraction immediately after the pulse and were chased to the membrane fraction after further incubation of the infected cells. Nonstructural proteins NCVP3 and NCVP4 behaved like structural proteins VP2, VP3, and VP6, although to a lesser degree. p36 and p26 (8), possible precursors of VP7 and NCVP5 (4), and a protein migrating between NCVP3 and NCVP4 were observed only in pulse-labeled cells, sedimenting together with the membrane fraction. The differences in the amount of label and polypeptide composition in the membrane and

the soluble fraction, observed between pulse and chases, are not due to a difference in the integrity of the infected cell at different times postinfection; the same distribution of label obtained for the pulse performed 4 h postinfection was observed when the pulse was done 9 h postinfection. Nine hours postinfection corresponds to the time of cell processing of the previously described 5-h chase (data not shown). The effectiveness of the subcellular fractionation procedure employed was substantiated by the different distribution of the cellular proteins between membrane and soluble fraction observed after Coomassie blue staining of the gel (data not shown). The analysis of SAl1-infected cells by the electrophoresis of double-stranded RNA labeled with 32p, in a pulse-chase experiment performed as described above, indicated that viral RNA sediments together with the membrane fraction. Proteins found in the soluble fraction were not assembled in viral particles. When this fraction, from the gradient shown in Fig. 2A, was analyzed by velocity sedimentation, no detectable viral particles were found. However, viral particles were observed in the membrane fractions of the pulse-chase-labeled cells, after extraction with trichlorotrifluoroethane and subsequent sedimentation in a 15 to 45% sucrose gradient. This is a common procedure employed to release SAl virus from infected cells for further purification. Of the total radioactivity present in the membrane fractions, 25% remained in the trichlorotrifluoroethane interface, and only 5% was recovered as viral particles. Viral particles were released into the medium of the chase. Their presence was determined by

988

SOLER ET AL.

J. VIROL.

CsCl sucrose sucrose

sucrose 39%

sucrose 34%

1

sucrose 29%

7.

6-

5. 4-

1.5

go x

E

a.

U

c,, U'

0.5-

20 30 40 FRACTION NUMBER FIG. 3. Different density components of the membrane fraction of SAl1-infected cells. (A) Centrifugation at 36,000 rpm on a discontinuous gradient of [35S]methionine-labeled membrane fraction of SAl1-infected cells obtained as described in the legend to Fig. 2A. For centrifugation, the three or four fractions containing the highest amount of label were pooled; 200 ,ul of these was then diluted to 1/2 with ET buffer to lower the sucrose concentration, and they were subsequently centrifuged. The recovery of radioactivity in the gradient was greater than 95% except for the 18-h chase, which was 76%. For electrophoresis, the two or three fractions of each peak containing the highest amount of radioactivity were pooled. Between 10 and 100 RI of each pool, containing between 3,500 and 5,500 cpm, was loaded in the gel. The original gradient composition is shown at the top of the figure. Symbols: 0, 15-min-pulse-labeled membranes; *, 5-h pulse-chase membranes; A, 18-h pulse-chase membranes. (B) Sodium dodecyl sulfate-1 2% polyacrylamide gel electrophoresis of polypeptides of the different density components shown in (A). Ml through M5 are given in order of decreasing density of pulse, 5-h. and 18-h 0

chase.

velocity sedimentation analysis of the material which precipitated with polyethylene glycol. After the 5-h and 18-h chase period, 19 and 50%, respectively, of the incorporated radioactivity present in the medium was found as polyethylene glycol-precipitable material. However, only 15% of the label in this material sedimented as viral particles after the 18-h chase. Density heterogeneity of the membrane fraction. By centrifugation at high speed in a severalstep discontinuous density gradient, the membrane fraction was resolved into at least four different density components (Fig. 3A). The

distribution of incorporated radioactivity in the cells after the pulse and the 5-h chase was similar; however, after the 18-h chase, most of the label sedimented with the heavier components. The analysis of the labeled proteins present in the different components, called Ml through M5 in order of decreasing density, showed the presence of most viral proteins in all of them, although the relative amount of each protein class varied greatly (Fig. 3B). MS, for example, was markedly enriched for VP6. The viral protein composition of the different fractions changed with time during the chase.

fi

MEMBRANE ASSOCIATION IN SA1l-INFECTED CELLS

VOL. 44, 1982

B VPI VP2

MI -_

Pulse M2

VP3: o

_

NCVP2- 4 VP6 -4-

4

NCVPI

v36 NCVP3-

NCVP4 26 NCVP5 -

It

M3

M4

5 h Chose M3. M4 M5

M5

MI M2

.,

989

18 h Chase MI

M2

M5

:1, .100

--O -A

..

FIG. 3B

Whereas M3 and M4 contained all of the structural viral proteins, plus the nonstructural proteins NCVP3, NCVP4, and NCVP5 after pulse labeling, the fractions enriched in these same protein classes were M2 after the 5-h chase and Ml and M2 after the 18-h chase. Effect of tunicamycin. Since it has been demonstrated that tunicamycin blocks the glycosylation of SAl glycoproteins in infected cells (4), the effect of this inhibitor on the membrane association described above was examined. Homogenates of cells infected in the presence of tunicamycin were negatively stained and observed in the electron microscope at different times after infection. An association of viral particles with membrane-like material indistinguishable from that observed in the absence of the inhibitor was found (data not shown). The distribution of viral protein, obtained after subcellular fractionation by low-speed centrifugation in a discontinuous density gradient, was the same as that observed for cells infected in the absence of tunicamycin (cf. Fig. 4A and Fig. 2A). Three significant differences were observed after the gel electrophoresis of the different components (Fig. 4B): pNCVP5 was present in the membrane fraction instead of NCVP5; the protein migrating between NCVP3 and NCVP4, observed only in the pulse-labeled cells infected in the absence of tunicamycin, remained stable during a chase in the presence of the drug; and finally, p36 was not present in cells infected with tunicamycin. Further fractionation of the membrane-rich material by high-speed centrifugation or by trichlorotrifluoroethane treatment showed no effect of tunicamycin. However, analysis by sedimentation and subsequent gel electrophoresis of the labeled material, released into the medium after the 18-h chase in the presence of tunicamycin, showed only viral particles devoid of the outer layer protein, as expected.

DISCUSSION

The possible implication of cell membranes in the maturation of SAl1 virions has been previously suggested from observations of infected cells by thin-section electron microscopy (3, 22). The budding of SAl1 virus particles through endoplasmic reticulum into vesicles is a feature that seems to be common to all rotaviruses (1, 6, 15, 16, 18, 20). It has also been reported (22) that the viral particles become enveloped in the process; however, the resistance to chloroform and ether (13) of the infectious virions suggests that they no longer have a membrane envelope when they are mature, but have a thin outer layer necessary for infectivity, probably made up of only two proteins (12). The enveloped SAl1 particles previously reported in negatively stained infected cell lysates (9) and the ones described in this manuscript probably correspond to the enveloped particles observed in thin sections (22). Since the membranes seem to contain viral antigens (2, 7), it is possible that the outer capsid layer of the virions is derived from the observed envelope. The glycoproteins induced by SAl1 in the infected cells (4) are likely candidates for the modification of the cellular membranes permitting the budding of the virion. The association with membranes of the viral proteins described in this paper supports the involvement of virus-modified cell membranes in the maturation of SAl1 virions. The early incorporation of glycoproteins or their precursors into these membranes suggests that the glycoproteins modify the membranes to permit the slow association of other viral proteins. The method of fractionation used was chosen because it separates different kinds of membranes in infected BHK cells (11). Although it is not proven that the same applies to MA104 cells, it seems that the fractionation obtained is a property of the cell membranes and not of the

990

J. VIROL.

SOLER ET AL. CsCI

sucrose 10 %

sucrose

60 %I

A

15C

x E C-) U) ur LO

FRACTION NUMBER

Pulse B VPI

w

5 h. Chase 18 h. Chose S M M H S M H .1 .a.

S

(N,

a

VP2:

VP3

-

_

sin~~qww

_

NCVP2?VP6NCVP3- _ pVP7a- a NC VP4

h

I,. s,

p r,.

aeeea~e a

-

w

ea

_r

-r-if?

VOL. 44, 1982

MEMBRANE ASSOCIATION IN SAil-INFECTED CELLS

association of these membranes with viral proteins because the same membrane fractions are obtained in mock-infected cells (data not shown). On the other hand, the peculiar fractionation obtained late after infection (22 h) (Fig. 3A) could be due to the heavy load of viral proteins, particles, and RNA associated with these membranes. The different viral protein composition observed in the membrane fractions obtained upon density fractionation (Fig. 3B) has no simple interpretation and could be the effect of the interplay of several factors: different association of viral proteins to the different cell membranes, sequential association by stage of maturation, modification of cell membrane density, cytopathic effect, etc. At present, the described fractionation is only a potentially useful technique to study the role of the different cell membranes in virus maturation. The observations of cells infected in the presence of tunicamycin are in agreement with previous studies by thin-section electron microscopy which showed apparent normal budding and the production of enveloped particles in the absence of glycosylation (N. Ikegami, K. Akatani, and M. Kimura, Abstr. 5th Int. Cong. Virol., 1981, p. 190; J. S. Tam, P. J. Middleton, and M. Petric, Abstr. 5th Int. Cong. Virol., 1981, p. 428). The drug would then possibly have an effect on the hypothetical conversion of the membrane envelope into the outer capsid layer. The gel electrophoresis of radioactively labeled proteins of homogenates of pulse- and pulse-chase-labeled cells showed two differences from previous reports of Dyall-Smith and FIG. 4. Distribution of SAil viral proteins in the presence of tunicamycin. (A) Centrifugation at 12,000 rpm on a discontinuous gradient of [35S]methionine-

pulse- and pulse-chase-labeled SAil-infected cells in the presence of tunicamycin. The original gradient composition is shown at the top of the figure. Symbols: 0, 15-min-pulse-labeled cell homogenate; *, 5-hpulse-chase-labeled cell homogenate; A, 18-h-pulsechase-labeled cell homogenate. (B) Sodium dodecyl sulfate-12% polyacrylamide gel electrophoresis of polypeptides obtained upon the fractionation of SAilinfected MA104 cells in the presence of tunicamycin. Abbreviations: H, cell homogenate; S, soluble fractions; M, membrane fractions of pulse, 5-h, and 18-h chase. The incorporation of label after pulse and chases did not vary by more than 10%. A total of 5 and 16% of the trichloroacetic acid-insoluble label was found in the medium after the 5- and 18-h chase, respectively. For electrophoresis, the three or four fractions containing the highest amount of label were pooled. The gel was loaded with 1/1,000 of cell homogenates; 10/1,000, 20/1,000, and 20/1,000 of the total soluble fraction for pulse, 5-h chase, and 18-h chase, respectively; and 10/1,000, 10/1,000, and 5/1,000 of the total membrane fraction for pulse, 5-h chase, and 18-h chase, respectively.

991

Holmes (8) and Arias et al. (4). (i) The failure to observe VP7, very noticeable in our previous reports (4, 12), is probably due to the constant selection of SAl1, having a faster migrating glycoprotein-VP7a-instead of VP7. VP7a was unnoticeable in our earlier preparations of purified virus, but its presence increased continuously to such an extent that in our last preparations it was found in a larger proportion than VP7. We are presently studying different clones of our SAl1. (ii) The pulse-labeled infected cells (Fig. 2B) and the tunicamycin-treated labeled infected cells (Fig. 4B) show, compared with previous results (unpublished data), the presence of several additional bands between those corresponding to polypeptides VP7 and NCVP4. The polypeptides rendering these additional bands are possible VP7 and VP7a precursors. Further studies on the intracellular distribution of viral proteins and their precursors will be undertaken with different preparations of cloned SAl1 virus. ACKNOWLEDGMENTS It is a pleasure to acknowledge the excellent technical assistance of P. Romero with the electron microscopy and of E. Huerta in tissue culture. We also thank E. Morett for help and criticism in some experiments. Tunicamycin was a kind gift of P. L. Hamill of Lilly Research Laboratories. This work was partially supported by grant POSANAL 800462 from the Programa Nacional Indicativo de Salud of the Consejo Nacional de Ciencia y Tecnologia. LITERATURE CITED 1. Adams, W. R., and L. Kraft. 1967. Electron microscopic study of the intestinal epithelium of mice infected with an agent of epizootic diarrhea of infant mice. Am. J. Pathol. 51:39-60. 2. Altenburg, B. C., D. Y. Graham, and M. K. Estes. 1979. Ultrastructural immunocytochemistry of rotavirus infected cells, p. 41-42. In Proceedings of the Electron Microscopy Society of America 37th Annual Meeting. Electron Microscopy Society of America, San Antonio. Tex. 3. Altenburg, B. C., D. Y. Graham, and M. K. Estes. 1980. Ultrastructural study of rotavirus replication in cultured cells. J. Gen. Virol. 46:75-85. 4. Arias, C. F., S. L6pez, and R. T. Espejo. 1982. Gene protein products of SAl1 simian rotavirus genome. J. Virol. 41:42-50. 5. Bonner, W. N., and R. A. Laskey. 1974. A film detection method for tritium labeled proteins and nucleic acids in polyacrylamide gels. Eur. J. Biochem. 46:83-88. 6. Carpio, M. M., L. A. Babiuk, V. Misra, and R. M. Blumenthal. 1981. Bovine rotavirus cell interactions: effect of virus infection on cellular integrity and macromolecular synthesis. Virology 114:86-97. 7. Chasey, D. 1980. Investigation of immunoperoxidaselabeled rotavirus in tissue culture by light and electron microscopy. J. Gen. Virol. 50:195-200. 8. Dyall-Smith, M. L., and I. H. Holmes. 1981. Comparisons of rotavirus polypeptides by limited proteolysis: close similarity of certain polypeptides of different strains. J. Virol. 40:720-728. 9. Els, H. J., and G. Lecatsas. 1972. Morphological studies on simian virus SAIl and the 'related' 0 agent. J. Gen. Virol. 17:129-132. 10. Els, H. J., and D. W. Verwoerd. 1969. Morphology of bluetongue virus. Virology 38:213-219.

992

SOLER ET AL.

11. Erwin, C., and D. T. Brown. 1980. Intracellular distribution of Sindbis virus membrane proteins in BHK-21 cells infected with wild-type virus and maturation defective mutants. J. Virol. 36:775-786. 12. Espejo, R. T., S. L6pez, and C. Arias. 1981. Structural polypeptides of simian rotavirus SAil and the effect of trypsin. J. Virol. 37:156-160. 13. Estes, M. K., D. Y. Graham, E. M. Smith, and C. P. Gerba. 1979. Rotavirus stability and inactivation. J. Gen. Virol. 43:403-409. 14. Flewett, T. H., and G. N. Woode. 1978. The rotaviruses. Arch. Virol. 57:1-23. 15. Holmes, I. H. 1979. Viral gastroenteritis. Prog. Med. Virol. 25:1-36. 16. Holmes, I. H., B. J. Ruck, R. F. Bishop, and G. P. Davidson. 1975. Infantile enteritis viruses: morphogenesis and morphology J. Virol. 16:937-943. 17. Laemmli, U. K. 1970. Cleavage of structural proteins during assembly of the head of bacteriophage T4. Nature

(London) 227:680-685. 18. Lecatsas, G. 1972. Electron microscopy and serological studies on simian viruses SAil and the "related" 0

J. VIROL. agent. Onderstepoort J. Vet. Res. 39:133-138. 19. Malherbe, H. H., and M. Strickland-Cholmley. 1967. Simian virus SAil and the related "O" agent. Arch. Gesamte Virusforsch. 22:235-245. 20. McNulty, M. S., W. L. Curran, and J. B. McFerran. 1976. The morphogenesis of a cytopathic bovine rotavirus in Madin-Darby bovine kidney cells. J. Gen. Virol. 33:5037508. 21. Palmer, E. L., M. L. Martin, and F. A. Murphy. 1977. Morphology and stability of infantile gastroenteritis virus: comparison with reovirus and bluetongue virus. J. Gen. Virol. 35:403-414. 22. Petrie, B. L., D. Y. Graham, and M. K. Estes. 1981. Identification of rotavirus particle types. Intervirology 16:20-28. 23. Rodger, S. M., and I. H. Holmes. 1979. Comparison of the genomes of simian, bovine, and human rotaviruses by gel electrophoresis and detection of genomic variation among bovine isolates. J. Virol. 30:839-846. 24. Saif, L. S., K. W. Theil, and E. H. Bohl. 1978. Morphogenesis of porcine rotavirus in porcine kidney cell cultures and intestinal epithelia cells. J. Gen. Virol. 39:205-217.