of poliovirus particles observed in spray preparations (11) ,, the particles were seen to be highly uniform and spherical, 27 my. in diameter. The infrequency of the linear organization, or their purification methods, may explain these results. It is proposed that the inlaid arrangement of poliovirus particles in filamentous protein material can occur in the cytoplasm of infected cells, either as a brief passing stage in the usual particle development or as an form of the fully developed “aberrant” virus particle. It seems of importance to determine whether this type of particle is infectious or whether it represents an independent, noninfectious form, as has been claimed for the filamentous form of influenza virus by Morgan et al. (12).
Holland et al. (1) showed that hypertonic salt washings of HeLa cells did not inactivate poliovirus ribonucleic acid (RNA j . They suggested, therefore, that the rapid loss of infectivity from hypertonic solutions in contact with HeLa cells was due to adsorption of the infectious RNA by the cells. It seemed to us, however, that the presence of ribonucleases at the HeLa cell surface which were inhibited but not completely blocked by high saIt concentrations, accounted more plausibly for their results. This opinion was strengthened when experiments very similar to those recently reported by Sprunt et al. (2) revealed the presence of an inactivator of infectious RNA in isotonic washings of HeLa cell REFERENCES monolayers. In this note we present evidence that RNA is hydrolyzed at the HeLa 1. STUART, D. C., JR., and FOGH, J., Ezptl. Cell Research 18,37&381 (1959). cell surface, that the ribonuclease activity 2. FOOH, J., and STUART, D. C., JR., Virology 11, comes from the cells and not from the me308-311 (1960). dium in which they were grown, and that 3. FOGH, and STUART, D. C., JR., Virology 9, 705 the rate of inactivation of infectious RNA 708 (1959). is equivalent to that obtained with about /t. FOGH, J., and STUART, D. C., JR., Federation lo-l6 g of ribonuclease per HeLa cell. Proc. 19,404 (1960). HeLa cells were grown in a medium con5. STUART, D. C., JR., FOGH, J., and PLACER, H., sisting of 10% horse serum, 10% calf serum, Virology l&321-324 (1960). 0.5% lactalbumin hydrolyzate, and 0.1% 6. STUART, D. C., JR. and FOGH, J., Virology 13, yeast extract in Earle’s saline unt,il they 177-190 (1961). 7’. RIFKIND, R. A., GODMAN, G. C., HOWE, C., formed confluent monolayers in 60-mm petri MORGAN, C., and ROSE, H. M., Virology 12, dishes. Alternatively, monolayers were 331-334 (1960). formed by seeding petri dishes with washed 8. TAYLOR, A. R., SHARP, D. G., BEARD, J. W., HeLa cells and incubating for 24 hours in DINGLE, J. H., and FELLER, A. E., J. Immunol. NCTC-109, a serum-free medium. The 47,261-282 ( 1943). monolayers were washed three times and in9. MOSLEY, V. M., and WYCKOFF, R. W. G., cubated for 30 minutes with about 10 pg of Nature 157,263 (1946). P52-labeled HeLa RNA in 1.0 ml per plate 10. WILLIaMS, R. C., KASS, S. J., and KNIGHT, C. A., of an isotonic or hypertonic salt solution. Virology 12,48-58 (1960). Exposure to hypertonic saline for 30 min11. SCHWERDT, C. E., WILLIAMS, R. C., STANLEY, W. M., SCHAFFER, F. L., and MCCLAIN, M. E., utes did not result either in loss of capacity Proc. Sot. Exptl. Biol. Med. 86, 310-312 t.o support viral ‘growth or detachment of (1954). cells from the monolayers. Following incu22. MORGAN, C., ROSE, H. M., and MOORE, D. H,, bation the supernatant liquid was removed J. Exptl. Med. 104, 171-182, (1956). and dialyzed for 2 hours against saturated J#RGEN FOGH ammonium sulfate solution. The HeLa Division of Experimental Pathology Sloan-Kettering Walker Laboratory Xye, New York Received
‘This work was done in part under contract AT( 11-11-34 between the Atomic Energy Commission and the University of California. Dr. Irving Rappaport supplied the TMV RNA and a stimulating discussion.
DISCUSSION TABLE HYDROLYSIS BY HELA
Per cent of RNA-P= dialyzable following incubation at 37°C for 30 minutes under the indicated conditions SerumNo cells, low salt
0.21, 0.55, 0.50 Per cent
HeLa cells, high salt
4.9, 5.9,/ 8.5 j
cells, low salt I4.7,
of RNA-P” adsorbed to cells following incubation at 37°C for 30 minutes
Serum-free HeLa cells, low salt
HeLa cells, low salt 9.1,
The data at the bottom of Table 1 show that, the percentage of RNA-p2 bound by cells is too small to require an upward rcvision in the figures for the percentage dialyzable shown at the top of Table 1. The low uptake and slow hydrolysis observed in hypertonic solutions constitute a major bar to RNA infect,ivity. The rapid hydrolysis in isotonic solutions makes it unlikely that any intact RNA is bound by the cells; the RNAPSZ bound is probably largely inorganic phosphate released in the hydrolysis of RNA. Ribonuclease activity was also found in cell-free, isotonic washings of HeLa cells, a result showing that the hydrolysis is clue in part to enzymes in solution released by the HeLa cells. The amount. of ribonuclease present at the HeLa cell surface was estimated by comparing the rate of loss of infectivity of to-
a The individual numbers within each block come from different experiments. All experiments were done with radioactive HeLa RNA at a concentration of about 10 pg/ml at pH 7.2. “High salt” is 1.2 M NaCl in 0.04 M phosphate. “Low salt” is Hanks’ balanced salt solution minus Ca and Mg. “Serum-free HeLa cells” refers to cells incubated for 24 hours in serum-free, chemtally defined medium. “Growth medium” is described in text.
monolayers were then washed and the cells, the ammonium sulfate solution, and the original incubation medium were all assayed for P32 activity. Table 1 shows that S-876 of the RNA-P”” becomes dialyzable after 30 minutes’ incubation with either HeLa cells in isotonic saline or with the growth medium; but, as expected (2), hypertonic saline strongly inhibits the hydrolysis. The ribonuclease activity in the growth medium comes from the 20% serum fraction (3) ; the residual serum in the monolayers is estimated to be reduced to less than 0.02% by the washing procedure and, therefore, contributes very little to the observed ribonuclease activity. This is confirmed by the data in column 4, which show that cells maintained in serumfree medium are capable of hydrolyzing RNA.
COMPARISON OF RATE Loss OF TMV RNA FECTIVITY AT THE HEL.~ CELL SURFACE AND IN THE PRESENCE OF lo-i0 o/ML OF RIBONUCLEASE”
Per cent of initial infectivity after incubation 37°C under indicated conditions in the presence of 10-r” g/ml of ribonuclease
RNA concentration b-&ml)
lo-Minute incubation High salt
ZO-Minute incubation LOW
Per cent of initial infectivity after incubation 37°C under indicated conditions at HeLa cell surface
RNA concentration (w/ml)
0.38 -~____ 0.60
a See Table salt.
__-__ 56 1 for
bacco mosaic virus (TMV) RNA exposed to HeLa cells with the rate of loss caused by a crystalline ribonuclease preparation (Nutritional Biochemicals Corporation). Table 2 shows again that 1.2 M NaCl solutions strongly inhibit but do not block completely the ribonuclease activity. The data also show a roughly comparable rate of destruction of infectious RNA by the monolayers and by 1O-1o g per milliliter of ribonuclease. Since each monolayer contains about 2 x lo6 cells, the activity of the monolayers is equivalent to that exhibited by about lo-la g of ribonuclease per cell. In a single experiment, monolayers of Earle’s L cell, FL amnion cells, and a human lymph node line showed activities very similar to that of HeLa monolayers, which suggests that ribonuclease activity of this magnitude is widespread among mammalian cells in tissue culture. REFERENdS HOLLAND, J. J., HOYER, B. H., MCLAREN, and SYVERTON, J. T. J. Exptl. Med. 112,
L. C., 821-
839 (1960). SPRUNT, K., KOENIG, S., and ALEXANDER, H. E., Virology 13,135-138 (1961). HERRIOT, R. M., CONNOLLY, J. H., and GUPTA, S., Nature 189,817-820 (1961). AMOS NORMAN ROBERT Department University Los Angeles, Received
of California California April
N-Terminal and Influenza
The structure of influenza virus is of special interest since in these particles, together with a rather small amount of ribonucleic acid, there are at least two distinct antigens, hemagglutinating activity, and the enzyme neuraminidase. The soluble antigen associated with the RNA can be easily separated, but whether the other three functions are carried in one component or reside in separate structures is not
known. In order to carry out an investigation of structure and function in this group of viruses, one member was highly purified and a study of its protein moiety undertaken. The LEE strain of influenza B virus was grown in the allantoic sac, partially purified by adsorption-elution on erythrocytes, sedimented, and resuspended in 0.1 M sodium phosphate pH 7.1. Further purification was achieved by passage at 20” through a cellulose phosphate cation exchanger (Celex-P, Bio Rad Labs., California) in the same buffer. Impurities remained adsorbed to the column while the virus particles passed through, yields varying between 60 and 100%. Infectivity was not affected by this treatment. The purified virus preparations had hemagglutinin titers (1) of 105.0 to 105.2 HA units per milligram dry weight, and in close-packed array (Fig. 1) the particle diameter was about 83 m/l, corresponding closely to the value obtained by Williams (2) for freeze-dried influenza A virus. Attempts to purify four influenza A virus strains (SW, MEL, BEL, FMl) by this method failed, since none could be eluted from the column. Since a measure of N-terminal amino acid residues was expected to give an indication of both the number and minimal molecular weight of the viral proteins, a recently developed micro modification (3) of Edman’s method (4), capable of detecting as little as 10e4 pmole of N-terminal amino acid, was applied to the purified preparations of LEE. Two to three milligrams of virus was coupled with highly radioactive S36-phenyl isothiocyanate, and the N-terminal amino acids were split off as the S35-2-anilino-5thiaxolinone derivatives which, after separation, were converted to the corresponding S35-phenylthiohydantoins. Samples of unlabeled phenylthiohydantoins of appropriate amino acids were added as markers and the mixture chromatographed in solvents D, E, and F of Edman and Sjoquist (5). After drying, the positions of the carrier phenylthiohydantoins were marked out under ultraviolet light and the chromatograms scanned for radioact.ivity. Coincidence between a radioactive peak and a