key: cord-0707440-cmhkjosa
authors: Kraeft, Stine-Kathrein; Chen, Daniel S.; Li, Hsin-Pai; Chen, Lan Bo; Lai, Michael M.C.
title: Mouse Hepatitis Virus Infection Induces an Early, Transient Calcium Influx in Mouse Astrocytoma Cells
date: 1997-11-25
journal: Experimental Cell Research
DOI: 10.1006/excr.1997.3768
sha: 856f85d3ac19c7b58c6a8bccc21fbbb80b325c34
doc_id: 707440
cord_uid: cmhkjosa

Abstract Mouse hepatitis virus (MHV), a murine coronavirus, utilizes murine carcinoembryonic antigens as receptors. The events that follow virus–receptor binding and eventually lead to virus entry are poorly understood. We studied the possible effects of MHV infection on intracellular calcium in a mouse astrocytoma cell line. Using the calcium-sensitive dye fluo-3 and confocal laser scanning microscopy, we found that MHV strain JHM induced an immediate (within 20 s) and transient (lasting no longer than 2 min) calcium increase in about 5% of the infected cells. The calcium increase was blocked by antibodies against the viral spike protein, suggesting that it was specifically triggered by the interaction of the viral spikes with cells. It was also inhibited by L-type calcium channel blockers and was not detected in calcium-free medium, suggesting that the calcium increase was caused by calcium influx from the extracellular medium. Studies of the kinetics of viral replication by immunofluorescence staining of the viral nucleocapsid protein revealed that at 3 h postinfection there was roughly the same percentage of cells (5%) that produced the viral protein as the percentage of cells that had responded with a calcium signal. This finding and the virus dilution studies together suggest that calcium responders may represent cells that had been infected with multiple viruses and undergone rapid viral replication. Furthermore, calcium channel blockers, including verapamil and cadmium chloride, and the calcium chelator EGTA inhibited virus infection. Therefore, the transient intracellular calcium increase reported here may be an early signaling event associated with virus infection.

Murine coronavirus mouse hepatitis virus (MHV) is tors. The events that follow virus-receptor binding an enveloped virus with a large, 31-kb, positiveand eventually lead to virus entry are poorly understranded RNA genome that is known to cause hepatitis, stood. We studied the possible effects of MHV infection vasculitis, encephalitis, and demyelination in mice [1] . on intracellular calcium in a mouse astrocytoma cell MHV utilizes members (principally biliary glycoproline. Using the calcium-sensitive dye fluo-3 and confoteins) of the murine carcinoembryonic antigen (CEA) cal laser scanning microscopy, we found that MHV family as receptors, with the prototypic receptor being strain JHM induced an immediate (within 20 s) and MHVR 1 [2] . Other identified murine receptors include transient (lasting no longer than 2 min) calcium in-MHVR 2 , BGP2, and bCEA [3] [4] [5] . The virus binding to crease in about 5% of the infected cells. The calcium the receptors is mediated by the viral envelope spike increase was blocked by antibodies against the viral spike protein, suggesting that it was specifically trig-(S) protein. It is not known how S protein-receptor gered by the interaction of the viral spikes with cells. interaction leads to virus entry by either endocytosis It was also inhibited by L-type calcium channel blockor membrane fusion [6] . Conceivably, virus binding ers and was not detected in calcium-free medium, sugtriggers a series of signal transduction events that culgesting that the calcium increase was caused by calminate in virus entry; however, the precise molecular cium influx from the extracellular medium. Studies of mechanism involved has not yet been characterized.

Free calcium functions as an intracellular second cence staining of the viral nucleocapsid protein remessenger in various pathways. As a prerequisite for vealed that at 3 h postinfection there was roughly the this function, the cell must precisely regulate its calsame percentage of cells (5%) that produced the viral cium level. A sustained increase in intracellular free protein as the percentage of cells that had responded calcium is often a sign of increased membrane permewith a calcium signal. This finding and the virus diluability, loss of energy, and finally cell death. Viruses tion studies together suggest that calcium responders have been shown to influence intracellular free calcium may represent cells that had been infected with multiin various ways. An immediate calcium response after ple viruses and undergone rapid viral replication. Furvirus or envelope protein binding has been reported thermore, calcium channel blockers, including verafor human immunodeficiency virus (HIV) and influenza pamil and cadmium chloride, and the calcium chelator virus [7] [8] [9] [10] [11] viral proteins with the cell membrane, leading to calcium influx from the extracellular medium [12, 13] .

To understand the mechanism of MHV entry, we studied the possible effects of MHV on intracellular 1 On leave of absence from Institute of Clinical Chemistry and Pathobiochemistry, University of Rostock, Ernst-Heydemannfree calcium in a mouse astrocytoma cell line using the Strasse 6, D-18055 Rostock, Germany. calcium-sensitive dye fluo-3 and confocal laser scan-2 To whom correspondence and reprint requests should be adning microscopy. We found that MHV induced an im- the ''responding area'' was done based on the sum of all images rather the cells. This increase was caused by calcium influx than using a single time point because the cells responded to the through the cell membrane and was sensitive to the Laddition of the virus at different time points. Finally, the ''responding type calcium channel blockers nifedipine and verapcellular area'' was determined by subtraction of area (II) from the amil. Furthermore, calcium channel blockers inhibited responding area (III), and expressed as percentage of the total cellular area (I).

virus infection. Therefore, calcium influx may repre-Immunofluorescence. DBT cells in eight-well coverglass tissue sent the earliest detectable molecular event following culture chambers (10 4 cells/well) were infected with MHV (at a m.o.i. virus infection. of 5) in serum-free MEM at 37ЊC for 1 h. After washing, the cells were incubated at 37ЊC for an additional 3 h in MEM supplemented as described above. Immunofluorescence staining of the cells was

performed after fixation in 2% paraformaldehyde in PBS, followed by permeabilization with 0.5% Triton X-100 in PBS, using a monoclonal Cell culture, virus, and antibodies. DBT cells, a murine astrocyantibody against the N protein [19] and a rhodamine-conjugated toma cell line [14] , were grown in Eagle's minimal essential medium anti-mouse secondary antibody (Jackson Immuno Research, West (MEM) supplemented with heat-inactivated 7.5% newborn calf se-Grove, PA). Images of 512 1 512 pixels were recorded with a confocal rum, 1% glutamine, 1% penicillin/streptomycin, and tryptose phoslaser scanning microscope and printed as described above. The perphate broth. In this study, MHV strain JHM-DL, propagated in DBT centage of stained cells was calculated essentially as described under cells under serum-free conditions, was used [15] . MEM and Joklikstatistical analysis. modified MEM were obtained from GIBCO Life Technologies (Gaith-Virus inhibition studies. DBT cells were grown in 60-mm plates ersburg, MD). Monoclonal antibodies specific for the S and nucleocapto 80% confluence, washed with MEM, and incubated with various sid (N) protein of JHM have been described earlier [16] .

inhibitors at the desired concentration for 2 h at 37ЊC. Subsequently, Calcium measurements and imaging. To determine intracellular cells were washed with MEM, and virus mixed with inhibitors (in calcium, 10 4 cells/well were plated in eight-well coverglass tissue 500 ml total volume) was added to the cells at a m.o.i. of 0.05 to 1 culture chambers (Nunc Inc., Naperville, IL) 24 h before measureand incubated at 37ЊC for 2 h. The cells were washed again with ments. Cells were loaded with 2 mM fluo-3/AM (Molecular Probes, MEM and incubated with MEM containing 1% newborn calf serum Eugene, OR) in serum-free MEM at room temperature for 20 min, and the inhibitors for an additional 7 h at 37ЊC. The supernatant and fluorescence measurements of intracellular calcium ([Ca 2/ ] i ) was harvested at 9 h postinfection and used for plaque assay on DBT were performed within 2 h after dye loading. Chambers were cells to determine virus titers. The concentrations of the inhibitors mounted on the microscope stage of an inverted confocal laser scanused were predetermined to be free of cytotoxicity. ning microscope (LSM410; Zeiss, Germany) equipped with an external argon-krypton laser. Time series of images of whole microscopic fields were taken and stored for further analysis. In each experiment,

two images of nonstimulated cells (baseline calcium) were recorded before addition of virus, antibodies, or control liquids. MHV prepared and thereby obtain the maximal fluorescence (F max ). Subsequently, the minimal fluorescence (F min ) was measured after addition of an cells an immediate increase in fluo-3 fluorescence in excess of EGTA (5 mM). Fluorescent intensities (F) were then transthe cytoplasm as well as in the nucleus that ranged lated into [Ca 2/ ] using the equation [Ca 2/ ] Å K d (F-F min /F max -F), where between 1.5-and 4.7-fold (mean, 2.9; SD, 0.8; n Å 26 K d is the dissociation constant [18] . sity values for the nuclear compartment ( Fig. 2A) . How-with approximately 15% responding cells. A higher virus titer (m.o.i. of 10) only marginally increased the ever, transformation of the fluorescence intensities into [Ca 2/ ] i after calibration of the respective areas (see Ma-percentage of responding cells (data not shown), but decreasing virus titers resulted in an exponential re-terials and Methods) demonstrated that baseline [Ca 2/ ] i is slightly lower in the nucleus than that in the duction of the number of responding cells. For example, at a m.o.i. of 1, the percentage of responding cells was cytoplasm (Fig. 2B) . After the addition of MHV, the [Ca 2/ ] i is 10-fold higher than that in control cells. A reduced to 0.29% (SD, 0.63; n Å 11), and a further dilution to 0.5 pfu/cell practically failed to evoke any similar difference was observed in both the nucleus and the cytoplasm (Fig. 2B) .

calcium response in DBT cells. This result suggests that multiple virus particles per cell may be necessary There were some variations in the pattern of [Ca 2/ ] i increases in individual cells in response to the addition for inducing [Ca 2/ ] i changes. In all subsequent experiments, MHV infection was performed at a m.o.i. of 5, of MHV. In most of the cells, [Ca 2/ ] i rose sharply and transiently as a single peak, which subsided to the rest-and every experiment was repeated at least five times. ing level after approximately 20 s and remained at that Specificity of the MHV-induced calcium increase. level thereafter (Fig. 3A) . In some cells, multiple oscil-To assess the specificity of the [Ca 2/ ] i increases oblations of [Ca 2/ ] i were observed (Fig. 3B) . In either case, served following the addition of MHV to DBT cells, we [Ca 2/ ] i spikes usually occurred within 20 s after the first attempted to rule out the possibility that calcium addition of the virus. Occasionally, however, late reincreases were caused by components of media. Cells sponders were observed, in which the sharp increase were incubated with the conditioned medium from of [Ca 2/ ] i did not occur until 1 or 2 min after the addi-DBT cell culture; no change in intracellular calcium tion of virus. Nevertheless, in all cases, cells returned was observed (Table 1) . Furthermore, a related coroto the resting calcium level within 4 min; no longer navirus, bovine coronavirus, which belongs to the same lasting changes in [Ca 2/ ] i could be observed (data not serogroup as MHV but cannot infect DBT cells [6] , did shown).

not induce any changes in the intracellular calcium The percentage of DBT cells that responded, exlevel, suggesting that this calcium change was specifipressed as ''responding cellular area'' (see Materials cally associated with MHV. To further demonstrate and Methods), ranged from 3 to 15%, depending on the this specificity, MHV was preincubated with a monoexperiment, with a mean of 5.04% (SD, 5.7; n Å 14). clonal antibody against the JHM spike protein (a-S protein mab) that is known to neutralize MHV infec- Figure 1A shows a time series of a selected field of view Note. Fluo-3-loaded cells were incubated with conditioned medium or infected with MHV or BCV at a m.o.i. of 5. Blocking experiments were performed by preincubation of MHV with the indicated antibodies at 37ЊC for 1 h before addition to fluo-3-loaded cells. From a series of 16 images recorded over 160 s, the percentage of responding cells (expressed as % cellular area responding) was calculated.

a Data are expressed as an arithmetic mean (and standard deviation) of 5 to 10 independent experiments. Calculation of significance compared to MHV-infected cells by t test. s, significant; ns, not significant; P, probability.

supplemented with 1.8 mM EGTA) was added to DBT cells preincubated in Joklik medium, no [Ca 2/ ] i response was observed ( Table 2 ), indicating that the MHV-mediated calcium rise in DBT cells was caused by Ca 2/ influx from the extracellular medium. To further assess the involvement of Ca 2/ channels in this process, cells were preincubated with the L-type channel block- ble 2). Similar results were obtained in the presence of cadmium chloride (20 mM), a divalent cation that blocks almost all types of calcium channels ( Table 2 ). In contrast, a sodium channel inhibitor, tetrodotoxin (1 mM), tion of DBT cells [16] ; the calcium increase was abolished ( Fig. 1B and Table 1 ). In contrast, preincubation of the virus with a monoclonal antibody directed against an internal viral protein (a-N protein mab) or When MHV (virus preparation in serum-free MEM Immunofluorescence of MHV infection in DBT cells. Since only an average of 5% of cells exhibited a measur- pressed the viral N protein (Fig. 4B) . Thus, the cells expressing viral N protein at 3 h postinfection may represent only a subset of virus-infected cells, probably did not inhibit the MHV-induced calcium increase (Taonly those in which virus replicated early. These cells ble 2). Taken together, our results suggest that the may have been infected with multiple virus particles. MHV-induced [Ca 2/ ] i increase in DBT cells is caused Alternatively, the N-protein-producing cells may repreby Ca 2/ influx through the cell membrane involving Lsent all of the virus-infected cells early in infection, type or related calcium channels. and the syncytia seen at 6 h p.i. may be the result of Effects of calcium channel blockers and EGTA on fusion between the virus-infected and the uninfected MHV infection. To investigate whether this transient cells. calcium increase is required for virus infection, we studied the effects of calcium channel blockers and the DISCUSSION calcium chelator EGTA on MHV infection in DBT cells. Cells were incubated in medium containing verapamil, Virus binding to host cell receptors is an essential cadmium chloride, or EGTA for 2 h before and throughfirst step of viral infection that triggers the internalizaout MHV infection. At 9 h postinfection, the media were tion of virus particles into cells. However, virus-recepharvested, and the titers of virus released from the tor binding is necessary but not sufficient for virus eninfected cells were determined by plaque assay. The try, as many cell types have receptors for MHV, and results showed that verapamil (at 10 mM) caused an yet, the virus cannot enter the cells [3, 21] . MHV recep-84.5% reduction of MHV titer produced by the virustors are expressed in many nontarget tissues, e.g., kidinfected cells; at 40 mM, the inhibition reached 97.7% ney, which are resistant to virus infection [22, 23] . (Table 3) . However, at a lower concentration (1 mM) of Therefore, additional molecular events must occur to verapamil, no significant inhibition of MHV titer was trigger virus internalization following virus binding. In noted (data not shown). Cadmium chloride, at the conthis paper, we have demonstrated that MHV infection centration (20 mM) that is required to block calcium induces an immediate and transient calcium increase channels [20] , also showed a significant inhibitory efin DBT cells. This response is specific for MHV and fect on virus infectivity. In contrast, the sodium chantriggered by an interaction between the MHV spike nel inhibitor tetrodotoxin (1 mM) did not have any effect protein and DBT cell surface molecules. Furthermore, on the MHV titer. Similar to the effect of calcium chanour results suggest that the calcium increase was the nel blockers, chelating of the extracellular calcium with result of the influx of calcium from the extracellular EGTA (1.8 mM) caused a 99% inhibition of the MHV medium through L-type or related calcium channels. A titer (Table 3 ). These results suggest that calcium insimilar calcium increase through influx from the extraflux may play an important role in virus infection. cellular medium has been reported to be induced by the HIV envelope protein gp120 in primary cultures of Calcium measurements at later stages of MHV infection. Virus-infected DBT cells were examined for neurons [7, 8] . Similarly for MHV, the transient calcium increase was also mediated by the viral spike changes of intracellular calcium at various time points for up to 15 h postinfection. No changes in fluo-3 fluo-protein and not by other components of cell products or media, since this response was blocked by the neu-rescence intensity could be found when compared to that in noninfected cells, suggesting that MHV does tralizing antibody and not induced by the conditioned media. not disturb calcium homeostasis over a longer period of time (data not shown).

The immediate and transient nature of the MHV- induced calcium response in DBT cells suggests its role MHV N protein was similar to the percentage of cells that responded with an intracellular calcium increase. in signal transduction. The influx of calcium following virus binding may trigger subsequent signal transduc-These cells probably represent those in which the virus had undergone very rapid internalization and prolifer-tion events that activate the cellular machinery for endocytosis or cell fusion and eventually lead to viral en-ation, so that viral proteins could be detected at such an early time point. Further evidence for the involve-try. Earlier work has suggested that an increase in intracellular free calcium is necessary to produce Sen-ment of calcium in MHV infection came from the studies which showed that calcium channel blockers or dai-virus-induced cell fusion at the maximum rate [24] . However, this result was obtained by an indirect EGTA decreased virus yield (Table 3 ). Preliminary studies have further shown that these inhibitors did method using the Ca 2/ -activated photoprotein obelin, and calcium changes were monitored for the entire cell not affect the virus particle itself or the virus-cell binding, but acted on an early step of virus infection, proba-population and not for single cells as demonstrated in this study. Recently, the requirement of an intracellu-bly a virus entry step (D. S. Chen, H.-P. Li, and M. M. C. Lai, unpublished observation) . It should be lar calcium increase for the cellular entry of the bacterium Salmonella typhimurium was demonstrated [25] . noted that the cadmium concentration required for inhibition of virus production was the same as that for Alternatively, the MHV-induced calcium increase could be a contributing factor to MHV pathogenesis. inhibition of calcium influx; however, the concentration of verapamil required for inhibition of virus production For example, an HIV-induced calcium increase has been suggested to play a role in AIDS-related symp-(10 mM) was higher than that for inhibition of calcium influx (1 mM). Thus, other factors associated with cal-toms, such as lymphocyte activation and neuro-and enteropathy [7, 8, 10, 11] . In the case of influenza A cium influx may contribute to virus infection.

It is curious that only about 5% of cells responded virus, a virus-induced calcium increase appears to be important for the impairment of neutrophil functions with a detectable intracellular calcium increase. One would have expected a higher percentage of responders in influenza [9] .

We attempted to establish the fact that the MHV-at an m.o.i. of 5, if every infected cell had responded with a calcium increase. One likely possibility is that induced calcium response is required for virus infection. Such a connection was suggested from the result these responders represent cells which were infected with multiple virus particles. Indeed, virus dilution of the immunofluorescence analysis of the MHV-infected cells, which showed that, at 3 h postinfection, studies showed that there was a precipitous drop in the percentage of responders as the m.o.i. was decreased the percentage of individual cells that had expressed

Proc. Natl

This is consistent with the multiple-hit hypothesis

the calcium responders may represent virus-infected

Pavonly approximately 5% of the cells under the culture lotsky

be the result of fusion between the virus-infected and 10

In either scenario, there is an ap-11

and Thus far, MHV and HIV are the only two viruses that

The early and transient nature of the MHV-14

tween the percentage of calcium responders and early 15

infection by calcium channel blockers and EGTA sug-16

This work was supported by a fellowship of the DAAD (Germany)

Germany) for continuing support and stimu-22

Annual Meeting of the American Society for Cell Biology