Effect of Magnesium on Replication of Rhinovirus HGP'

JOURNAL OF VIROLOGY, June 1967, p. 489-493 Copyright 1967 American Society for Microbiology Vol. 1, No. 3 Printed in U.S.A. Effect of Magnesium on Replication of Rhinovirus HGP' MILAN FIALA' AND GEORGE E. KENNY Departmen1t of Preventive Medicinie, Un1iversity of Washin/gtonI School of Medicine, Seattle, Washi,igton 98105 Received for publication 16 February 1967 It is known that plaque formation by some rhinoviruses is greatly enhanced by increasing the concentration of MgCl2. The mechanism of this action was studied by investigating the effects of MgCI2 on rhinovirus HGP adsorption, growth, clumping, thermal stability, and cell susceptibility to viral cytopathic effect. The latent period was at most 7 hr, whether virus was propagated in cells maintained in 0.8 or 30 mm MgCl2, but virus release was 8- to 310-fold greater in the presence of 30 mm MgCl2, depending on initial multiplicity. Intracellular virus content appeared unaffected by Mg++ and reached maximal yield (plateau phase) at about 10 hr. Viral adsorption was increased when cells were maintained in 30 mm Mg++. It is likely that the two effects of magnesium, enhanced adsorption and increased virus release, both contribute to enhancement of plaque formation. We have previously reported the enhancing effect of magnesium and calcium on rhinovirus plaque formation in human heteroploid cell monolayers (2), an effect similar to that observed on plaque formation by certain enteroviruses in primary monkey kidney cells (7, 8). Plaque formation was also shown to be enhanced by diethylaminoethyl (DEAE)-dextran. The effect of magnesium on phases of the growth cycle of rhinovirus HGP in human heteroploid cells was studied to determine how plaque enhancement occurred. This study demonstrates two effects of 30 mm magnesium chloride on the growth of rhinovirus HGP in HeLa cells: enhancement of virus adsorption and release. MATERIALS AND METHODS HGP virus (rhinovirus type 2) was kindly made available by D. Hamre. This rhinovirus was propagated in M HeLa cells in medium supplemented with 30 mm Mg++. Cell cultures. M HeLa cells were grown on glass in Eagle's minimal essential medium (MEM) with 10% fetal bovine serum (MEMFBS1O). Two million cells were seeded in 5 ml of growth medium in a 60-mm plastic petri dish (Falcon Plastics, Los Angeles, Calif.) and were used for plaque assay the following day Ṗlaque assay. The overlay for plaque assay was I Some of the data in this paper were presented in a preliminary communication: Federation Proc. 25: 492, 1966. 2 Present address: Department of Research Medicine, Hospital of the University of Pennsylvania, Philadelphia 19104. modified since the last report (2) by the incorporation of 30 jug of DEAE-dextran per ml in addition to the following previously described components (in final concentrations): 30 mm MgCl2, penicillin (100 units/ ml), streptomycin (100 Ag/ml), and nystatin (10 units/ml), 2% fetal bovine serum, 5.6 mm bicarbonate and 0.4% agarose (SeaKem; Bausch & Lomb, Inc., Rochester, N. Y.) in MEM with Hank's balanced salt solution (BSS) as saline solution. Petri dishes were incubated at 34 C in 2.5% CO2 in air and were stained at about 2.5 days with 1% gentian violet in 20% alcohol (3). Virus diluent (V-BSS) consisted of 0.2%; fetal bovine serum in BSS. The ph was adjusted to 7.1 to 7.2 with 0.83 M tris(hydroxymethyl)aminomethane (Tris) solution (ph 10) and 0.8 N HCI. For some experiments, V-BSS was modified by incorporation of 30 mm MgCl2 (V-BSS Mg 30). The divalent cationfree diluent was GKNP (glucose-potassium-sodiumphosphate solution), essentially BSS without magnesium and calcium salts. Preparationi of high-titered virus. One-day-old monolayers of M HeLa cells containing 1.5 X 107 cells in 150 by 25 mm plastic petri dishes (Falcon Plastics) were infected with HGP virus at a multiplicity of 0.05 to 0.1 plaque-forming unit (PFU) of virus per cell. After adsorption of virus for I hr, overlay was added to cells. Virus was harvested 2 days after infection of cells by sliding overlay off the cells, adding 1 ml of V-BSS to the cell layer, and, after 10 min, aspirating this fluid. Media for maintenanice. The maintenance medium was either MEM with 2%7o fetal bovine serum and 5.6 mm bicarbonate (ph after equilibration with 2.5%o CO2 in air was 7.3; MEMFBS2) or the same medium containing additional 30 mm MgCI2 (MEMFBS2 Mg 30). 489

490 FIALA AND KENNY J. VIROL. Preparation of overlay extract. Plaque assay overlay medium was poured into centrifuge tubes, allowed to solidify at 4 C overnight, broken with a spatula, and then centrifuged at 34,000 X g for 60 min. The supernatant fluid represented the overlay extract. Intracellular virus. Intracellular virus was harvested as follows. After being washed four times, infected cell monolayers were removed from the plastic petri dish surface with 0.17% trypsin and a rubber scraper. The resulting cell suspensions were kept at 3 C for 8 to 24 hr and then disrupted with an ultrasonic oscillator (Sonifier, Branson Instruments Inc., Stamford, Conn.) for 45 sec. Cellular debris was immediately removed by centrifugation at 1,020 X g for 10 min. The supernatant fluid represented the intracellular virus. RESULTS Effect of Mg++ on adsorption. Because of the well-known effect of Mg++ and Ca++ on adsorption of bacteriophage (5) and poliovirus (1), adsorption of HGP rhinovirus was studied with respect to magnesium concentration. In five experiments, an enhanced plating efficiency was observed when virus was diluted in V-BSS Mg 30 rather than in V-BSS. The numbers of PFU adsorbed from V-BSS Mg 30 in 1 hr, expressed as percentage of PFU adsorbed from V-BSS, were 167, 186, 124, 153, and 224%. All of these results except 124% were significantly different at the 0.01 % level. The results of one such an experiment are shown on Fig. 1. The quantitative dependence of adsorption at room temperature (23 C) on the concentration of magnesium was investigated. Virus was diluted in GKNP to which were added increasing amounts of MgCl2 to establish concentrations t 701 z 60-3 50- _40 20- : 10- Time in minutes FIG. 1. Effect of MgC2 on adsorption of rhinovirus HGP. Estimated 50 PFU of virus diluted in V-BSS (squares) or V-BSS Mg 30 (circles) were added to monolayers containiiig 2 millionz M HeLa cells. Infected monolayers were incubated at room temperature (about 23 C), and at 5, 15, 30, and 60 mimi were washed with 5 ml of V-BSS. The cells were then overlaid aiid iiicubated as in a plaque assay. from 0 to 100 mm Mg+, and adsorption was measured at 20 and 60 min. When a parabola was fitted to observed plaque numbers by the regression method, it was found that both linear and quadratic regressions on MgC12 concentration were significant (P < 0.01). The Mg++ concentration for optimal adsorption lies between 10 and 50 mm, with a maximum at about 35 mm (Fig. 2). When the effect of Mg++ concentration on adsorption was studied with overlay extract as viral diluent, the results were closely similar to the results with V-BSS described above. Effect of Mg++ on virus yield. A possible explanation for plaque enhancement of Mg++ might be that virus yield was increased in cells exposed to 30 mm Mg++. A thorough investigation of this possibility was feasible when hightitered virus pools containing up to 3 x 108 PFU per ml were obtained by the method of harvesting virus under agar overlay. When extracellular yields of virus were determined in HeLa cells infected with initial multiplicities of 26, 8, and 0.7 PFU per cell; the differences between 30 and 0.8 mm MgC12 in extracellular virus yields were 310-, 65-, and 8-fold, respectively (Table 1). For comparison with conditions existing during a virus plaque assay, virus yields were also determined under the plaque assay overlay during a multicyclic growth experiment. One hundred PFU of virus were adsorbed for 2.5 hr, and cells were then overlaid with agarose overlay contain- V)60~~~~~ 0 02030 50 75 100 Mg Cl2 concn. (mm) in virus diluent FIG. 2. Relation1ship of viral adsorption to Mg++ concentrationt. Cell monolayers of 2 million M HeLa cells in 60-mm plastic petri dishes were washed once with the same diluent to be used for adsorption; 0.1 ml of HGP virus containing about SO PFU was then added to monolayers in appropriate diluent and incubated at room temperature (approximately 23 C) for 20 or 60 mmn Monlolayers were then washed on7ce with 4 ml of 0.2%o fetal boviine serum in GKNP (ph 7.3), overlaid, and incubated. Open circles =numbers of plaques adsorbed in 20 miin; dark circles = iiumbers of plaques adsorbed in7 60 min; broken curve = curve fitted by regressionl method to points at 20 miin; solid curve = curve fitted to poinlts at 60 miii.

VOL. 1, 1967 EFFECT OF MAGNESIUM ON RHINOVIRUS GROWTH 491 TABLE 1. Effect of3o m.m. Mg++ anzd virus multiplicity on extracellular virus yielda Virus yield (PFU/cell) at indicated time Fold increase inoculum ~~~~~~~hr in yield in 30 at 13 multiplicity 30 mm Mg'+ 0.8 mm Mg++ over 0.8 mm (PFU/cell) - Mg ' 13 hr 23 hr 13 hr 23 hr 26 24.8 42 0.08 13.5 310 8 2.6 37 0.04 3.2 65 0.7 0.16 2 0.02 0.8 8 a Cell monolayers containing 800,000 M HeLa cells were washed once with MEMFBS2 and infected with HGP virus contained in 0.1 ml at initial multiplicities of 0.7, 8, and 26 PFU per cell. After 2 hr of adsorption at 34 C, the cells were washed three times with V-BSS. Warm MEMFBS2 medium containing MgCl2 as indicated in the table was then added, and cells were incubated at 34 C. The supernatant fluid sampled at various intervals represented the extracellular virus and was stored at -20 C for 24 to 28 hr until assayed at 3.16- and 10-fold dilution steps. ing either 0.8 mm MgCI2 or 30 mm MgCl2. The infected monolayers were harvested after 24, 45, and 74 hr by removing overlay and washing cells with 1 ml of V-BSS. Virus yields under overlay with 30 mm MgCL2 were consistently higher than those obtained under unsupplemented overlay. The maximal difference existed at 74 hr; at that time, 26 times as much virus was present under overlay with 30 mm MgCl2 as under overlay with 0.8 mm MgC12. Effect of Mg++ on the intracellular and extracellular virus yield in a single cycle virus growth experiment. A virus growth experiment was performed with the following conditions. Virus input was 1.6 x 106 PFU, number of cells was 8 X 105 (virus-cell multiplicity was 2:1), and medium was either MEMFBS2 or MEMFBS2 Mg 30. Cells and supematant fluid were assayed at indicated time periods to determine virus yields (Fig. 3). Because two cycles of freezing and thawing released only about one-tenth the amount of virus released by sonic oscillation, the latter method was adopted for measuring intracellular virus. The intracellular growth of virus in either MEMFBS2 or MEMFBS 2 Mg 30 was similar. The eclipse of intracellular virus (defined by the intercept of the logarithmic increase of virus growth with time axis) lasted 4 to 4.5 hr. The latent period (defined as the distance on the time axis from infection to the point of interception of a line drawn through the logarithmic portion of virus growth) was about 7 hr. However, owing to the presence of about 0.1 % of noneclipsed virus, increase of extracellular virus was noted only at 9 hr. At 14 hr, extracellular virus released from cells exposed to 30 mm MgC12 was 24-fold higher than in cells in 0.8 mm MgC12. Possible effect of Mg++ on dispersion of virus. It was considered possible that the increase in plating efficiency shown by rhinovirus HGP in the presence of Mg++ was due to dispersion of clumped virus by Mg++. This hypothesis was tested by a dilution series experiment. The following serial dilutions were prepared: (i) stock virus was diluted 1:100 in V-BSS, and a portion of this dilution was again diluted in V-BSS (labeled V-BSS/V-BSS dilution series); (ii) stock virus was diluted 1:100 in V-BSS Mg 30, and then a portion was diluted 1 :100 in V-BSS; (iii) stock virus was diluted 1:100 in V-BSS Mg 30, and then a portion 414 07 106 105-10 4 2 4 6 8 10 I. 12 14 _ I Hours after infection FIG. 3. Effect of MgCl2 on extra- and intracellular yields of rhinovirus HGP in M HeLa cells. Cell monolayers containting 8 X 105 cells were washed once with MEMFBS2 anid infected wit/i HGP virus contained in 0.1 ml at ani initial multiplicity of 2 PFU of virus per cell. The virus was adsorbed for 45 miii at 34 C, and the cells were then washed three times with V-BSS. Either MEMFBS2 or MEMFBS2 Mg 30 (3 ml) was added to cells; cells were incubated at 34 C and sampled at various intervals for intra- and extracellular virus. Solid line = intracellular virus yields; broken line = extracellular virus yields; open? circles = MEMFBS2 Mg 30 medium; dark circles = MEMFBS2 medium.

492 FIALA AND KENNY J. VIROL. was diluted 1:100 in V-BSS Mg 30. The V-BSS/ V-BSS and V-BSS Mg 30/V-BSS dilution series gave similar average PFU values, 37 and 36, respectively. The V-BSS Mg 30/V-BSS Mg 30 dilution series did, however, give a significantly (P < 0.05) higher mean value of 71 PFU, a result which was not unexpected because of the enhancing effect of Mg++ on plating efficiency. Effect of MgCl2 on thermal stability of virus. The slope of viral inactivation in V-BSS Mg 30 was significantly (P < 0.001) steeper than in V-BSS (Fig. 4). In conformity with the findings on viral adsorption, plaque counts were about 2.5 times higher when virus was diluted in V-BSS Mg 30 rather than in V-BSS. Effect of MgCI2 on stability of virus in presence of cell debris. It was possible that virus was protected by magnesium against a proteolytic attack by intracellular enzymes. Stein and Fischer (6) described stabilization by divalent metal ions of a-amylases toward proteolytic attack. Virus was diluted in V-BSS Mg 30 or V-BSS, and cell debris corresponding to 103 cells per ml was added to this viral suspension. The rates of inactivation were similar with or without cell debris. As described in the previous paragraph, virus assayed higher when diluted in V-BSS Mg 30. Time of addition of Mg++ overlay. It was also possible that MgCl2 enhanced the cytopathic effect of rhinovirus-infected cells, i.e., increased cell detachment and, hence, plaque formation. To test this hypothesis, monolayers infected with approximately 80 PFU of rhinovirus HGP were 1-1 1. 41 q) C4. Z) CL TABLE 2. Effect of delayed addition of MgCl2 Time after infection when the overlay MgCl2 concn Ag no. of Avg plaque was increased to 30 mm plaques diam Sto,ck virus was diluted 100-fold either in V-BSS Mg 30 creasing multiplicity. Since intracellular virus (open circles) or V-BSS (dark circles) and incubated at 37 C in 16 X 150 mm nionwettable polystyrene tubes yield did not appear to be increased by Mg++, it (Fa 'lcon Plastics). At hourly intervals, 0.1-ml samples was concluded that virus release was potentiated werpe plated. Stracight lines were fitted to observed points by 30 mm MgCI2. This is in agreement with the by i a The Mg++ concentration was 0.8 mm throughout. hr min 0 83 1.7 21 73 1.0 27 58 1.3 48 35 0.8 Controla 0 0 overlaid with 6 ml of overlay containing 0.8 mm MgCl2. At specific time intervals, a second 6-ml overlay containing 60 mm MgCl2 was added. Plaque numbers and mean plaque diameters were maximal when the Mg++ concentration was increased immediately, and fell off with delay in the addition of Mg++ (Table 2). Pretreatment of cells with 30 mmr Mg92. Cells were dispensed into petri dishes in either MEMFBS1O Mg 30 or MEMFBS1O and incubated for 17.5 hr. The medium was then removed, and the cells were infected with HGP virus and overlaid with overlay containing either 30 or 0.8 mm MgCl2. Although the plaques under overlay with 30 mm Mg++ appeared somewhat larger when the cells were pretreated with 30 mm Mg++, plaques were not observed with either treatment when the overlay contained only 0.8 mm MgCI9. DISCUSSION 60, ~~ ~~~ Possible mechanisms of magnesium effect were 0-_ investigated to explain the enhancement of plaque 50-8 0 ~~o o formation of some rhinoviruses with 30 mm 40- MgC2 contained in agarose-solidified overlays. 8 The effect of magnesium was studied with respect to: 30,. * (i) phases of virus replication: adsorption, intracellular growth, release, and stability; and (ii) cell factors: growth and susceptibility to viral 20 cytopathic effect. However, owing to the lack of high-potency serum, studies of viral penetration were not possible. Magnesium may be important at this step, as shown for bacteriophage by Paranchych (4). 10 - During one-step growth-curve experiments, the 0 2 3 4 5 extracellular virus titers were from 8 to 310 times Time in hours higher in the presence of 30 mm MgCI2 than in 0.8 mm MgCl2, depending on the initial virus IG. 4. Thermal inactivation of HGP virus at 37 C. multiplicities, the difference increasing with in- the regressionz method. explanation proposed by Wallis and Melnick (7)

VOL. 1, 1967 EFFECT OF MAGNESIUM ON RHINOVIRUS GROWTH 493 for the enhancement of the susceptibility of monkey kidney cells to poliovirus by the addition of 25 mm MgC12. The observed effects of magnesium in potentiating plaque formation by rhinoviruses is most likely a result of the enhanced spread of virus owing to increased virus release. Our previous experience has indicated that our particular HeLa cells rapidly lose sensitivity for rhinovirus plaque assay with time after plating the cells; this effect appeared related to cell density. Accordingly, if a plaque is to occur, virus must spread rapidly and infect sufficient cells before insusceptibility arises due to "ageing." The finding that Mg++ enhances adsorption (plating efficiency) of rhinovirus HGP indicates not only that the virus is released more rapidly but also that adsorption is more nearly complete. Other effects of magnesium were studied in this system. Thermal inactivation at 37 C was not reduced by MgCI2. Magnesium was less effective when added 21 hr or later after infection. This would suggest that magnesium does not merely potentiate cytopathic effect without affecting the rate of viral spread. One dilution step in 30 mm MgCl2 was without effect on virus titer when followed by another dilution into 0.8 mm MgCl2. This finding does not support the hypothesis that increased plating efficiency is due to dispersion of clumped virus. ACKNOWLEDGMENTS This investigation was supported by Public Health Service research contract Ph-43-63-562, and by training grant TI-AI-206 from the National Institute of Allergy and Infectious Diseases. This work was stimulated by valuable discussions with N. Groman and E. Fischer. LITERATURE CITED 1. BACHTOLD, J. G., H. C. BUBEL, AND L. P. GEBHARDT. 1957. The primary interaction of poliomyelitis virus with host cells of tissue culture origin. Virology 4:582-589. 2. FIALA, M., AND G. E. KENNY. 1966. Enhancement of rhinovirus plaque formation in human heteroploid cell cultures by magnesium and calcium. J. Bacteriol. 92:1710-1715. 3. HOLLAND, J. J., AND L. C. McLAREN. 1959. Improved method for staining of monolayers for virus plaque counts. J. Bacteriol. 78:596. 4. PARANCHYCH, W. 1966. Stages in phage R17 infection: the role of divalent cations. Virology 28:90-99. 5. PUCK, T. T., A. GAREN, AND J. KLINE. 1951. The mechanism of virus attachment to host cells. I. The role of ions in primary reaction. J. Exptl. Med. 93:65-88. 6. STEIN, A. E., AND E. H. FISCHER. 1958. The resistance of a-amylases towards proteolytic attack. J. Biol. Chem. 232:867-879. 7. WALLIS, C., AND J. L. MELNICK. 1962. Magnesium chloride enhancement of cell susceptibility to poliovirus. Virology 16:122-132. 8. WALLIS, C., J. L. MELNICK, AND M.BIANCHI. 1962. Factors influencing enterovirus and reovirus growth and plaque formation. Texas Rept. Biol. Med. 20:693-702.