SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References

This study determined the sensitivity of vaccinia virus, an orthopox virus commonly used as a surrogate for variola virus (etiological agent of smallpox), exposed to UVB radiation emitted by a solar simulator, or to direct natural sunlight. The data obtained indicate that: (1) the virucidal effect of natural sunlight can be mimicked adequately by an artificial light source with similar spectral characteristics in the UVB, (2) viral sensitivity to UVB or to solar radiation can be correlated with experimental data previously obtained with UVC, (3) the correlation factor between virus inactivation by solar radiation (measured at 300 ± 5 nm) and by UVC (254 nm) is between 33 and 60, and (4) the sensitivity of viruses either dry on glass surfaces or in liquid suspension is similar when in the presence of similar amounts of cellular debris and growth media. The findings reported in this study should assist in estimating the threat posed by the persistence of virus during epidemics or after an accidental or intentional release.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References

The persistence in the environment is of key importance in assessing the risk of infectious viruses broadcast by contagious patients during epidemics or released from biological weapons during a biological attack. Virus inactivation occurs in darkness by the effects of physical factors in the environment, mainly drying, where the physical integrity of the viral particle is apparently the target [1, 2]. However, inactivation in darkness is slow [1] and exposure to solar ultraviolet (UV) radiation is a primary means of virus inactivation in the environment. The radiation with wavelengths between 290 and 320 nm (designated UVB) accounts for most microbicidal effects of natural sunlight [3], but the virucidal effect of daily exposure to sunlight has not been fully characterized. Safety and security constraints have hindered exposing highly virulent viruses to natural sunlight and gathering the data needed to assess the risk of environments contaminated with high-consequence (also known as biothreat) viruses [4]. The relative difficulty in exposing infectious agents to direct sunlight has resulted in few studies on virus persistence outdoors [5-7]. In contrast to the limited information available on the viral sensitivity to natural sunlight, substantial data exist on the sensitivity of viruses to UVC (radiation of 254 nm, also known as germicidal UV), which although absent in solar radiation reaching the earth surface, is frequently used to inactivate viruses in laboratory settings and hospitals and other critical public environments ([3, 8, 9] reviewed in [10]). The relatively abundant data on viral sensitivity to UVC radiation has been previously extrapolated to several biothreat viruses by a model based on the type, size and strandedness of the nucleic acid genomes of the different virus families [10]. This model also involved calculation of the virus-inactivating effective spectrum (normalized to 254 nm as a function of wavelength) for radiation in the UVB region [10]. This action spectrum correlates fluences at different UVB wavelengths with that at 254 nm resulting in similar virucidal inactivation. UVC sensitivities predicted by this model were later confirmed experimentally after irradiating Ebola and Lassa viruses with UVC inside high-containment laboratories [11], but correlation of sunlight to UVC sensitivities has not yet been established.

We selected for study vaccinia virus, classified within the Poxviridae family some of whose members are associated with high-fatality rates and person-to-person transmission [12, 13]. In particular, the etiological agent of smallpox, variola virus that is an orthopoxvirus, causes 20–30% mortality [14] and persists infectious for many days in darkness in dried crusts from skin lesions as well as in fluid from vesicles [15]. Vaccinia virus is also an orthopoxvirus that has been used to vaccinate against smallpox [16], and is often used as a less-virulent human experimental surrogate of variola virus, due to the limited availability and associated risk of the latter.

In this study with the virus in conditions simulating environmental contamination, i.e., with growth medium and cellular debris, we wanted to establish whether viral inactivation produced by UVB from a solar simulator under controlled laboratory conditions could mimic the effect of natural sunlight in the environment. We also wanted to measure viral inactivation sensitivity to UVB and compare that to UVC inactivation as previously reported by us [11], either validating or correcting the previous viral inactivation model based on relatively well-known viral UVC sensitivities [10]. The findings from the studies reported here should allow a more accurate correlation between the UVC and UVB sensitivities measured experimentally with predictions of viral inactivation by natural sunlight of viruses relevant to public health and biodefense.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
UVB Radiation

The source for UVB radiation was a solar simulator (Model 16S-300-UV Simulator; Solar Light Company, Glenside, PA), which incorporated a 300 W xenon lamp as the UV source. The radiation was filtered by the manufacturer to simulate the summer noon-time solar spectrum at intermediate latitude (40o N), in the UVA and UVB part of the spectrum (290–380 nm, with relatively little visible or infrared radiation at the output port) and with an intensity (at convenient distances for experimental exposures) claimed by the manufacturer to be two to three times the solar intensity. On the lamp housing was a right-angle mirror that deflected the horizontally oriented beam downwards. The circular radiation beam was well defined, hence there was little chance of unintentional exposure from this source. The flux produced by the solar simulator was measured with a PMA2100 radiometer (Solar Light Company) provided by the manufacturer with a band-pass filter that transmitted radiation centered at 300 nm with a full-width at half maximum of 10 nm, thus measuring mainly UVB radiation in the 295–305 nm range. In addition, an RPS900 Spectroradiometer (International Light Technologies, Peabody, MA) was used to obtain the spectrum from the solar simulator. The spectroradiometer was calibrated for wavelengths between 250–400 nm and a UG11 280–380 nm bandpass filter was added to reduce measurements resulting from scattered visible and infrared radiation. The virus samples were irradiated at a distance from the solar simulator that provided a relatively uniform beam that encompassed the entire UV-transparent Petri dish (nominal 5 cm diameter, Lumox UV-permaeable 50TC, sterile, cat. no. 96077440; Greiner Bio-One, Gottingen, Germany) containing the virus samples. The UV-transparent half of the Lumox dish had a 100% transmittance between 290 and 310 nm as measured with a spectrophotometer. The UVB flux was ca 2 W m−2 as measured by the PMA2100 radiometer at the place where the samples were located. Any radiation loss due to reflectance from the covered samples was discounted by similarly covering the radiometer sensor with the Lumox UV-transparent cover. The uniformity of the radiation beam was determined by placing a mask with a square 2 × 2 mm2 hole over the center of the detector head. Within 5 mm of the center of the beam, there was ca 3.5% variation in beam intensity; within 10 mm, 12% variation. Therefore, the virus samples were placed within 5 mm of the center of the special UV-transparent Petri dishes and the Petri dish placed in the center of the radiation beam. To provide more complete uniformity the Petri dishes containing the virus samples were rotated 90° every 2 min during the exposure. Since, the exposure times were measured in 10s of minutes, this provided reasonably uniform exposure for all samples.

Solar exposure

The set-up used to expose vaccinia virus to direct sunlight was similar to that previously described for exposing virulent bacteria [17]. Instant solar flux was monitored throughout the exposure period with the sensor of a PMA2100 radiometer placed onto crushed ice in the vicinity of the samples and measuring at 300 ± 5 nm as it was previously demonstrated that the virucidal effect of solar radiation should peak at or slightly above 300 nm [10].

One additional dish placed on the same ice container with the samples had distilled water covering the tip of a flexible temperature digital sensor that monitored temperature inside the dish and outside in the ambient. Exposure to direct sunlight was carried out between June 28 and September 22, 2010 in northern Germany at 53o 0′ N latitude and 10o 8′ E longitude. Experiments were carried out on days that were clear in the morning and noon, with some clouds beginning to appear, sometimes before the longest exposure times were concluded. Fluence was integrated by the dosimeter even on those occasions when a cloud passed between the sun and the samples in the dishes. Experiments were terminated in two occasions when heavy clouds began approaching. Data from cloudy skies were not utilized. The solar noon included in the graphs was obtained from raising and setting times for the sun in Hamburg (nearest city with tabulated values).

Viruses

Two vaccinia strains were studied that showed similar sensitivities to irradiation. The WR strain of vaccinia virus (a strain adapted to tissue culture by NIH scientists) was propagated on Vero cell monolayers as previously described [11], and a viral stock with a concentration of 1 × 109 pfu mL−1 was used in UVC and UVB radiation experiments. Vaccinia virus strain Elstree B5/Lister (WHO reference strain; [18]) was grown in Baby Hamster Kidney (BHK) cells at 37°C in 4% CO2 atmosphere resulting in viral stocks with concentrations ranging from 2 × 106 to 4 × 106 TCID50 as titrated in BHK cell monolayers as previously described [19-21]. The media where virus was grown and exposed was: Eagle's minimum essential medium (EMEM) containing 10% fetal calf serum (FCS) and 0.1% each of penicillin and streptomycin.

Irradiation of samples to UVC and UVB

The light source and dosimetry of UVC radiation as well as the exposure procedure were all as detailed previously [11]. An aliquot (3 μl) of stock virus was deposited onto each of 24 sterile glass cover slips (5 × 5 mm2, prepared as previously described [22], allowing the sample to air dry for ca 5 min at 25°C and 30% relative humidity. Samples were considered dry when liquid was no longer observed on the slide. Sets of three cover slips were randomized and placed into the middle of inverted UV-transparent 50 mm Petri dishes that allowed UV transmission through the gas-permeable bottom. Special care was taken in arranging the non overlapping specimens to be near the center of the Petri dish. After all UV exposures to UVB were completed, the cover slips containing the dried, UV-irradiated virus samples were individually placed in sterile micro centrifuge tubes containing 200 μL of phosphate buffered saline (pH 7.2) and the viruses eluted for 5 min by agitation by moderate vortex for 10 s as previously described [2, 11]. The amount of vaccinia viruses eluted and recovered from each carrier after various UVB irradiation doses was determined by serial dilution and subsequent infection of Vero cells in 25 mm well plates. The exposures with the sun simulator were in the order of minutes, with the longest time that the viruses might be on the cover slips in the experiments being less than 1 h after drying. The inactivation of unirradiated virus on cover slips kept in the dark during 1 h after drying was less than 5% of vaccinia virus in the original sample. Thus, the recovery after drying and eluting the viruses from glass slides was higher than 95%.

Exposure to direct sunlight

Unfortunately, safety regulations related to aerosolization prevented us from using virus dried onto a surface for exposure to sunlight outdoors. Thus, we could expose outdoors only virus suspended in liquid. A 100 μL aliquot of the viral stock (in undiluted growth media or diluted with phosphate buffer saline, PBS [pH 7.2]) was dispensed in the middle of a Petri dish, covered with the UV-transparent lid and the edges very safely sealed with parafilm. All samples (each in a separate Petri dish) were placed over crushed ice inside a foam ice box and covered with an opaque lid. At predetermined intervals relative to noon sunlight, samples were uncovered, exposed to direct unfiltered sunlight before covering them with aluminum foil and transporting them for analysis. A set of samples was covered during the whole experiment, serving as unexposed controls. The amount of virus surviving direct sunlight exposure was determined by the TCID50 method on BHK cell monolayers [19-21]. Other aliquots of the same exposed sample were tested by either (1) Real-Time PCR where viral DNA extracted and purified from irradiated or from non irradiatred controls was amplified as previously described [23, 24], or (2) by enzyme-linked immunoassays (ELISA) carried out on irradiated and non irradiated control virions incubated with the same monoclonal antibodies and conditions detailed in previous publications [23, 24].

All virus specimens were protected from ambient radiation before exposure to UVB or sunlight. In addition, the Petri dishes with the irradiated samples were kept in the dark until all exposures were performed. A flexible mini thermometer immersed in water within an additional UV-transparent dish placed among the samples monitored the ambient temperatures during the duration of each experiment. In any case, the solar virucidal effect was always determined in comparison to random samples kept in the same exposure container, but in the dark as null irradiation controls.

Data analysis

Survival curves were constructed by plotting the surviving percentage (ratio of the titer of an irradiated sample to the titer of the unirradiated control × 100) versus UV fluence on a semi logarithmic graph. The slopes of the resulting survival curves obtained from the linear portions of the graphs were used to calculate D37 values. The D37 value equals the reciprocal of the slope and corresponds to the UV fluence that produced on average one lethal hit to the virus, resulting in a survival of 37%. D37 can be calculated by dividing the fluence that inactivates 1 log10 virus load (as obtained from the linear portion of the graphs) by 2.3 (the natural logarithmic base). A lower value for the D37 indicates a higher sensitivity to inactivation by UV radiation.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References

Characterization of solar simulator output

The spectrum of the solar simulator used in this study was measured in the UVB region with a RPS900 spectroradiometer and the recorded fluxes are shown in Fig. 1. The UVB region of the solar spectrum measured in the central United States in early June (June 5, 2007) is shown for comparison also in Fig. 1. Multiplying the solar irradiances by 10 allowed nearly perfect overlap of the solar and simulator spectra. Thus, the graphs of the intensities produced by the solar simulator and recorded from the sun at various wavelengths were similar in shape, but of different intensity, with the solar irradiance measured in early June being ca 10-fold lower than the output from the solar simulator at the distance at which samples were placed in this study. It should be noted that essentially only direct solar radiation was measured because of the baffled construction of the spectroradiometer detector, i.e., little indirect, scattered radiation from the sky reached that detector. As indirect radiation accounts for nearly half the total solar radiation reaching the earth [25-27], the actual UVB output of the solar simulator is only 4–5 times higher than the total sun irradiance, rather than the 10 times factor applied to overlap both spectra as shown in Fig. 1. A four- to five-fold difference observed with the spectroradiometer was supported by additional data gathered with another UVB radiometer, which measures the direct solar radiation and a large portion of the indirect solar radiation. This second UVB radiometer (PMA2100, filtered to measure 300 ± 5 nm Full-width, half maximum [FWHM]) recorded 0.37 W m−2 UVB from sunlight, which is 4.5-fold lower than the 1.65 W m−2 measured from the solar simulator with the spectroradiometer.

image

Figure 1. Combination spectra. Spectra (in the left y-axis) measured by the RPS900 spectroradiometer at the wavelengths (in the x-axis) of radiation emitted by a solar simulator (triangles) and received by direct sunlight in Hays, Kansas, latitude 38o 54′ N, longitude 99o 19′ W (circles, after multiplying each data point by 10) are presented. The 254 nm equivalent spectrum for solar simulator radiation (diamonds; right y-axis), obtained by multiplying the above solar simulator spectrum by the 254 nm normalized action spectrum for virus inactivation, as described by Lytle and Sagripanti [10].

Download figure to PowerPoint

The overlapping curves in Fig. 1 thus indicate that similar viral inactivation should be attained with comparable irradiation levels from either the sun simulator or natural sunlight, if UVB radiation is the only, or at least major, solar component contributing to virus inactivation. When the solar simulator spectrum was multiplied by the 254-equivalent action spectrum for virus inactivation [10], the resultant graph showed the expected effectiveness spectrum (normalized to 254 nm) for virus inactivation for the solar simulator peaking at 302–303 nm (Fig. 1).

Solar simulator inactivation of vaccinia virus

Vaccinia virus dried onto glass surfaces was inactivated by UVB radiation from a solar simulator according to the kinetics shown in Fig. 2. The inactivation kinetics data fit a biphasic graph with a relatively small inflection appearing near 1000 J m−2. The biphasic survival curves of vaccinia virus after UVB irradiation were similar in shape to the biphasic survival of vaccinia virus in response to 254 nm UVC [11], shown also in Fig. 2. The initial slope of the biphasic survival curve indicated a D37 of 360 J m−2; the final slope had a D37 of 690 J m−2. Extrapolation of the final slope indicated that 80% of the virus population in the inoculum was inactivated with the initial kinetics and the 20% remainder was less sensitive (larger D37). The initial slope of the biphasic survival curve of vaccinia virus dried onto surfaces after UVC exposure indicated a D37 of 6.0 J m−2 and the final slope a D37 of 21 J m−2. Extrapolation of the final slope to zero fluence occurred at 7% surviving virus, indicating that this percentage of the virus particles in the innoculum contributed the final segment of the UVC survival curve. When the UVC and the solar simulator inactivation data were plotted in the same figure with the different fluence ranges adjusted to show the final slopes of the survival curves as parallel, the two graphs were non coincident only because the extrapolation values of the two survival curves were different.

image

Figure 2. Sensitivity of vaccinia virus to UVB and UVC. The relative survival of vaccinia virus dried onto glass carriers (in the y-axis) is shown after exposure to the sun simulator for various periods of time providing the fluences indicated in the upper x-axis (dark squares) or to UVC at fluences indicated in the lower x-axis (empty diamonds). The symbols correspond to the averages of 2–5 (simulator, UVB) or of 6–12 data points (UVC) with standard error of the mean (SEM) shown as brackets when extending beyond the symbols. Fine dotted and thick lines represent the initial and final slopes of the curve, respectively. The protected virus population responsible for the final inactivation slope was obtained by extrapolation (fine trace line) to Fluence = 0.

Download figure to PowerPoint

The D37 for the final slope of the UVB inactivation was ca 33 times larger on a J m−2 basis than the D37 obtained for the UVC inactivation. Thus, the experimentally determined correlation factor of UVB (measured at 300 ± 5 nm) to UVC (254 nm) was 33. A similar value was obtained by comparing the fluence ranges in Fig. 2.

Solar irradiation of vaccinia virus

There are previous reports suggesting that the effect of natural sunlight cannot be fully mimicked by artificial sources of UVB [28-33]. Therefore, in spite of the spectral similarities in the UVB region shown in Fig. 1, the actual virucidal potency of natural sunlight in the environment was compared to that observed after exposure to the sun simulator in the laboratory. All experiments where viruses were exposed to direct sunlight were conducted between July 19 and September 22, 2010 in northern Germany. The solar flux was measured at various times before and after solar noon with representative results shown in Fig. 3A. To account for any short-term variation in flux, the corresponding fluences were continuously integrated automatically by the PMA2100 radiometer reading at 300 ± 5 nm with its sensor placed among the dishes with the viral samples. The data in Fig. 3A indicate that the relatively constant flux received within 2 h before and 2 h after solar noon resulted in fluences that varied nearly linearly with time within this period. During this time period, the temperature rose with insolation in a pattern that lagged solar noon, whereas the samples kept over crushed ice remained at temperatures between 8 and 20°C all this time as seen in Fig. 3B.

image

Figure 3. Sunlight dosimetry: panel (A) depicts the solar flux measured at 300 ± 5 nm (as circles in the left y-axis) measured at the time of the day indicated in the x-axis. The fluences integrated at the time of the day (as 24 h) indicated in the x-axis by dark squares are represented in the right y-axis. Panel (B) presents the temperature (in the left y-axis) recorded at the times of the day (as 24 h) indicated in the x-axis by a flexible micro probe immersed in water inside an exposure dish placed on top of ice inside an insulated cooler used in the irradiation set-up (dark circles). Ambient temperatures (empty squares) recorded by another probe connected to the same thermometer are plotted in the right y-axis.

Download figure to PowerPoint

The inactivation by direct sunlight of vaccinia virus suspended in growth media with an optical density of 0.607 appeared biphasic with a measurable final slope presented in Fig. 4. The final slope of the kinetics corresponded to a D37 value of 410 J m−2, intersecting the y-axis between 40% and 50%.

image

Figure 4. Comparative virucidal inactivation efficiency of sunlight and solar simulated radiation. The survival of vaccinia virus suspended in liquid growth media either undiluted (dark circles) or diluted 1:20 in PBS (empty circles) and exposed to sunlight during noon time for periods resulting in the fluencies measured at 300 ± 5 nm indicated in the x-axis is compared to the survival of vaccinia virus dried in growth media after irradiation with UVB from a solar simulator (empty squares) at similar fluences. The data points correspond to the average of two independent experiments with triplicate samples with SEM represented by brackets when extending beyond the averages.

Download figure to PowerPoint

To determine any effect of the media in solar inactivation of vaccinia virus in suspension, viruses suspended in growth medium at various dilutions in sterile distilled phosphate buffered saline were investigated. Viruses diluted 1:20 in PBS had an optical density of 0.004 and exposure to sunlight resulted in the inactivation kinetics shown in Fig. 4. The D37 of the final inactivation slope corresponded to 280 J m−2. Extrapolation of the final slope intersected the y-axis at 30–40%, indicating that 30–40% of the population appeared relatively more resistant to sunlight even when the colored media was considerably diluted in transparent PBS. Virus exposed to direct sunlight at dilutions between 1:20 and undiluted decayed with intermediate D37 values (data not shown).

As, it was not practical to irradiate the virus with solar radiation at the higher fluence levels possible with the very intense solar simulator radiation, only data obtained by irradiation with the solar simulator at lower fluences could be compared to the survival curves of inactivation by sun-irradiation of virus in liquid cell culture medium undiluted or 20-fold diluted (Fig. 4). The survival levels after solar irradiation fell close to the initial slope of the survival curve produced by solar simulator radiation. Although safety constrains precluded exposing virus to sunlight and to simulated sunlight under identical conditions, the survival curves corresponding to sun-irradiated virus in undiluted liquid culture medium and solar simulator-irradiated virus in dried culture medium are approximately parallel in Fig. 4. The slope of inactivation in media diluted 1:20 was somewhat steeper than the other two kinetics.

Considering the solar flux received in June near 39o latitude (where a substantial portion of the world population and most of the U.S. population lives), as well as the solar simulator data and the UVC equivalent solar flux (0.00444 w m−2), we estimated the exposure times (at solar noon for example in Kansas at 39o latitude) necessary for reducing the virus load to various levels. Inactivation of vaccinia virus to survival levels of 10%, 1% and 0.1% (surviving/initial = n/no = 0.10, 0.01 and 0.001) should require exposure times of 1.25, 3.75 and 6.56 h, respectively. Additional fluxes and exposure times for other latitudes and times of the year could be calculated as described in [10].

We investigated whether the substantial viral inactivation produced by direct solar radiation could be readily correlated with nucleic acid and/or antigenic damage as reflected by impairment of analytical tests targeting those molecules. Three identical aliquots from each of the same replicated samples exposed to direct sunlight were titrated, amplified by vaccinia genome specific PCR techniques, or subjected to vaccinia specific ELISA. The signal produced by PCR amplification of virus irradiated through the range studied up to a solar fluence of 1400 J m−2 remained constant and indistinguishable from those obtained testing unexposed (fluence 0) controls (data not shown). At the same solar fluence of 1400 J m−2, vaccinia virus in undiluted media drop to the minimal level of survival detection (a 2 log10 reduction in infectious virions as measured by reduction in TCID50). Lacking detectable levels of DNA damage in our test, we investigated whether perhaps a more sensitive molecular component could be associated to viral inactivation. Relatively, high levels of solar radiation (up to 1400 J m−2) did not alter the antigenic determinants targeted by the monoclonal antivaccine antibodies used in our ELISA system (data not shown).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References

Solar inactivation of vaccinia virus

The main purpose of this study was to determine the efficacy with which solar radiation can inactivate vaccinia virus, which is a surrogate for variola major virus (a highly virulent virus and a possible bioterrorism agent). The survival curves presented in Fig. 4 demonstrate that such inactivation occurred rather rapidly for the vaccinia virus, in liquid culture medium that included 10% serum and cellular debris, when exposed to sunlight even at a relatively northern latitude, which limited total exposure to 1500 J m−2 UVB as measured at 300 ± 5 nm. Control experiments further demonstrated that the experimental conditions (samples in sealed Petri dishes on ice) prevented a rise in temperature to levels that could inactivate virus, indicating that the virucidal effect of sunlight did not result from thermal effects.

Virus suspended in liquid growth medium (with 10% serum plus cellular debris) and exposed to sunlight decayed with biphasic inactivation kinetics. Biphasic inactivation kinetics, as we observed for virus in liquid suspension by sunlight, were also found for viruses dried on surfaces after irradiation with UVB of spectral properties similar to sunlight radiation (Fig. 2) and after exposure of dried virus to monochromatic radiation of 254 nm wavelength [11]. Our findings are similar to the biphasic inactivation kinetics in dried medium previously reported for UVC inactivation of Ebola and Lassa viruses [11]. Biphasic inactivation by sunlight was previously reported also for a bacterium, Burkholderia pseudomallei, in liquid suspension [17].

The biphasic survival curves obtained here after exposure to sunlight, UVB or UVC light differ from previously published linear curves describing the effect of 254 nm UV radiation on vaccinia virus in liquid transparent buffer suspensions [33-36]. We attribute the biphasic nature of the present survival curves to varying degrees of protection that the proteins and other components of the growth medium present in our study provide against the radiation [10, 11].

Our results showing a biphasic decay after UVB irradiation of vaccinia virus in either dry medium or liquid medium suspensions correlates to the environmental behavior of non irradiated variola major virus whose infectivity in dried scabs or from vesicles' fluid diminished also in a biphasic fashion, first fairly rapidly and then remaining detectable at low titers for many months [15]. These previous and our findings together suggest that vaccinia and variola major viruses in the dark or exposed to sunlight should be expected to decay by similar biphasic kinetics whether in liquid suspensions or within vesicles, or dried on surfaces or in scabs.

The conditions used in this study suggests that at least a two-fold protection from sunlight can be expected for viruses mixed with proteins and cellular debris as expected in human secretions from infected patients or in certain batches produced in the laboratory. Accordingly, the time estimated for daylight to inactivate viruses in the environment should be increased (at least two-fold) to compensate for the apparently more resistant population of virions.

The lower fluence required to inactivate virus diluted in PBS than in undiluted medium is consistent with the respective absorbance of solar radiation by the suspensions, i.e., the growth medium absorbed some of the radiation, as expected, protecting some of the virus particles from the radiation. Thus, growth medium supplemented with 10% FCS and containing cellular debris under our experimental conditions reduced the effective UVB fluence by about 32% of that producing a similar effect in virus exposed in (transparent) PBS. These results on a viral population appearing more resistant to UVB likely due to shielding by proteins and cellular debris is also supported by our data obtained after exposure to UVC which, being more penetrating than UVB, resulted in fewer virions being protected (7% after UVC exposure) than those shielded from sunlight (40–50%) or artificial UVB (10–20%).

Initially, we thought the rather extensive viral inactivation that we observed after solar radiation could be sufficient to produce damage to the viral genome that would be detectable by PCR. No such inactivation of PCR activity was found at UVB fluencies up to 1400 J m−2. This was likely due to a relative target size for virus infectivity (yet to be determined) being much larger than the target size for PCR activity (145 base pairs).

Comparing solar radiation with radiation from a solar simulator

It is unlikely that highly infectious organisms will be ever allowed outside high-containment laboratories to measure their sensitivity to sunlight. Therefore, we studied the effect of radiation from a solar simulator as a surrogate for natural sunlight. The solar simulator used in these experiments produced a UVB spectrum that was quite similar to the solar UVB spectrum, although with a greater intensity under the conditions employed in the laboratory (Fig. 1). Thus, the spectral similarity in Fig. 1 indicates that radiation from the sun simulator should be an experimentally adequate surrogate for UVB in natural sunlight, but that needed to be validated in our biological system. When virus survival curves were compared up to the maximum fluence available from the sun during our experiments, the graphs (Fig. 4) indicated similar inactivation potency, with a D37 of 410 J m−2 obtained for solar irradiation of virus in liquid suspension and a D37 of 360 J m−2 obtained for virus dried on glass plates after irradiation with the solar simulator. These remarkably similar D37 values indicate that (1) comparable doses of radiation from sunlight and a solar simulator having a UVB spectrum similar to sunlight produce essentially the same virus inactivation, (2) the liquid or dried state of the virus yielded only minor difference, if any, in UV sensitivity, and (3) the main virucidal activity of sunlight is within the UVB region and any additional virucidal effect produced by sunlight outside this region must be relatively minor. These findings demonstrate that future experiments to determine the environmental survival of highly pathogenic viruses (either in liquid suspensions or dried in surfaces) could be more safely and conveniently carried out in the laboratory with artificial light from a source spectrally similar to sunlight rather than exposing to natural sunlight outdoors, where safety concerns are high and many factors can affect the reliability of consistent solar UVB, e.g., clouds, pollution, time of year, etc.

Test of the model that predicts solar inactivation of viruses

Previous modeling of the virus persistence in the environment included calculation for the UVB and UVC wavelength regions of an action spectrum that was normalized to the virucidal effect of 254 nm UVC radiation [10]. This normalized model together with the solar simulator spectrum reported here predicted that solar UVB radiation centered around 300 nm was 1/41 as effective as 254 nm UVC radiation, on a fluence (J m−2) basis, i.e., 41 times more UVB than UVC radiation would be required to inactivate viruses to the same level. Lacking alternative approaches, this factor UVB/UVC was used in making biodefense calculations and in risk assessment of influenza epidemics [37], but accuracy of the model had not been previously determined experimentally. As the solar simulator produced much more intense UVB than the sun, radiation from the simulator was used to yield a virus survival curve over an extended range of survival similar to that produced by 254 nm UVC radiation (Fig. 2). Comparison of the initial slopes indicated that UVC was 60 times more effective, whereas comparison of the final slopes indicated UVC was 33 times more effective. Both values are well within the accuracy range predicted for the modeled value of 41, given the variations inherent in experimental data. We consider the level of agreement between the UVB/UVC transformation factors previously predicted from radiometric measurements and obtained here experimentally to be a reasonable validation of the solar virucidal model previously described [10]. The transformation (between 33 and 60) correlating quantitatively the virucidal efficacy of UVB to that of UVC allows predicting environmental inactivation of viruses not yet exposed to sunlight from the relatively abundant previously available data obtained after UVC irradiation in the laboratory [10, 38].

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References

The virucidal effect of natural sunlight can be mimicked adequately by an artificial light source with proven similar spectral characteristics in the UVB. Viral sensitivity to UVB or solar radiation can be correlated from experimental data obtained with UVC. The correlation factor between radiation measured at 300 ± 5 and 254 nm is between 33 and 60. The sensitivity of viruses either dry on surfaces or in liquid suspension is similar when in the presence of similar amounts of protein debris or growth media. More transparent media result in increased photocidal activity. The results of this study should assist in predicting the time whereby viruses remain a viable threat after natural broadcast from infected patients during epidemics or after an accidental or intentional release.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References

We gratefully acknowledge Dr. Birgit Hülseweh (WIS, Germany) by developing and providing the PCR method used here to detect vaccinia virus as well as by her critical review of the manuscript. We are grateful for the kind and excellent technical assistance of Katrin Böhling (WIS, Germany). The portion of this study involving natural sunlight was done at the WIS while Jose-Luis Sagripanti was participant in the US-German Technical Officers Exchange program (ESEP).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  • 1
    Fitzgibbons, J. E. and J.-L. Sagripanti (2008) Analysis of the survival of Venezuelan equine encephalomyelitis virus and possible viral simulants in liquid suspensions. J. Appl. Microbiol. 105, 14771483.
  • 2
    Sagripanti, J.-L., A. M. Rom and L. E. Holland (2010) Persistence in darkness of virulent Alphaviruses, Ebola Virus, and Lassa Virus deposited on solid surfaces. Arch. Virol. 155, 20352039.
  • 3
    Giese, A. C. (1976) Living with the Sun's Ultraviolet Rays. Plenum Press, New York.
  • 4
    Borio Borio, L., T. Inglesby, C. J. Peters, A. L. Schmaljohn, J. M. Hughes, P. B. Jahrling, T. Ksiazek, K. M. Johnson, A. Meyerhoff, T. O'Toole, M. S. Ascher, J. Bartlett, J. G. Breman, E. M. Eitzen, M. Hamburg, J. Hauer, D. A. Henderson, R. T. Johnson, G. Kwik, M. Layton, S. Lillibridge, G. J. Nabel, M. T. Osterhom, T. M. Perl, P. Russell and K. Tonat (2002) Hemorrhagic fever viruses as biological weapons: Medical and public health management. JAMA 287, 23912405.
  • 5
    Furusawa, Y., K. Suzuki and M. Sasaki (1990) Biological and physical dosimeter for monitoring solar UV-B light. J. Radiat. Res. 31, 189206.
  • 6
    Murphy, T. M. (1973) Inactivation of TMV-RNA by ultra-violet radiation in sunlight. Int. J. Radiat. Biol. 23, 519526.
  • 7
    Rontó, G., S. Gáspár and A. Bérces (1992) Phages T7 in biological UV dose measurement. J. Photochem. Photobiol. B. 12, 285294.
  • 8
    Calkins, J. and T. Thordadottir (1980) The Ecological significance of solar UV radiation on aquatic organisms. Nature 283, 563566.
  • 9
    Nicholson, W. L., A. C. Shuerger and P. Setlow (2005) The solar UV environment and bacterial spore UV resistance: Considerations for Earth-to-Mars transport by natural processes and human spaceflight. Mutat. Res. 571, 249264.
  • 10
    Lytle, C. D. and J.-L. Sagripanti (2005) Predicted inactivation of viruses of relevance to biodefense by solar radiation. J. Virol. 79, 1424414252.
  • 11
    Sagripanti, J.-L. and C. D. Lytle (2011) Sensitivity to ultraviolet radiation of Lassa, Vaccinia, and Ebola Viruses dried on surfaces. Arch. Virol. 156, 489494.
  • 12
    Fields, B. and D. M. Knipe (Ed) (1990) Fields Virology, 2nd edn, pp. 12451267. Raven Press, New York.
  • 13
    Knipe, D. M. and Howley P. M. (Ed) (2001) Fields Virology, 4th edn. Lippincott Williams and Wilkins, Philadelphia, PA.
  • 14
    Fenner, F. (1990) Poxviruses. In Fields Virology (Edited by D. M. Knipe, R. M. Chanock, M. S. Hirsch, J. Melnick, T. P. Monath and B. Roizman), pp. 21132133. Raven Press, New York.
  • 15
    Downie, A. W. and K. R. Dumbell (1947) Survival of variola virus in dried exudates and crusts from smallpox patients. Lancet 1, 550553.
  • 16
    Fenner, F. (1977) The eradication of smallpox. Prog. Med. Virol. 23, 121.
  • 17
    Sagripanti, J.-L., A. Levy, J. Robertson, A. Merritt and T. J. Inglis (2009) Inactivation of virulent Burkholderia pseudomallei by sunlight. Photochem. Photobiol. 85, 978986.
  • 18
    Czerny, C. P. and H. Mahnel (1990) Structural and functional analysis of orthopoxvirus epitopes with neutralizing monoclonal antibodies. J. Gen. Virol. 71, 23412352.
  • 19
    Greiser-Wilke, I. M., V. Moennig, O. R. Kaaden and R. E. Shope (1991) Detection of Alphaviruses in a genus-specific antigen capture enzyme immunoassay using monoclonal antibodies. J. Clin. Microbiol. 29, 131137.
  • 20
    Kaerber, G. (1931) Beitrag zur kollektiven Behandlung pharmakologischer Reihenversuche. Arch. Exp. Pathol. Pharmakol. 162, 480487.
  • 21
    Spearman, C. (1908) The method of “right and wrong” cases (constant stimuli) without Gauss formulae. Br. J. Psychol. 2, 227242.
  • 22
    Sagripanti, J.-L., M. Carrera, J. Insalaco, M. Ziemski, J. Rogers and R. Zandomeni (2007) Virulent spores of Bacillus anthracis and other Bacillus species deposited on solid surfaces have similar sensitivity to chemical decontaminants. J. Appl. Microbiol. 102, 1121.
  • 23
    Sagripanti, J.-L., B. Hülseweh, G. Grote, L. Voß, K. Böhling and H.-J. Marschall (2011) Microbial inactivation for safe and rapid diagnostics of biological threat agents. Appl. Environ. Microbiol. 77, 72897295.
  • 24
    Sagripanti, J.-L., H.-J. Marschall, L. Voss and B. Hülseweh (2011) Photochemical inactivation of alpha- and poxviruses. Photochem. Photobiol. 87, 13691378.
  • 25
    Ireland, W. and R. Sacher (1996) The angular distribution of solar ultraviolet, visible and near-infrared radiation from cloudless skies. Photochem. Photobiol. 63, 483486.
  • 26
    USDA UV-B Monitoring and research Program (2003) UV-B Radiation: Definition and Characteristics. Available at: http://uvb.nrel.colostate.edu/UVB/index.jsf. Accessed on 12 October 2011.
  • 27
    Ben-David, A. and J. L. Sagripanti (2010) A model for inactivation of microbes suspended in the atmosphere by solar ultraviolet radiation. Photochem. Photobiol. 86, 895908.
  • 28
    Rauth, A. M. (1965) The physical state of viral nucleic acid and the sensitivity of viruses to ultraviolet light. Biophys. J . 5, 257273.
  • 29
    Tyrrell, R. M. (1977) Solar dosimetry with repair deficient bacterial spores: Action spectra, photoproduct measurements and a comparison with other biological systems. Photochem. Photobiol. 27, 571579.
  • 30
    Turner, M. A. and R. B. Webb (1981) Comparative mutagenesis and interaction between near-ultraviolet and far ultraviolet radiation in Escherichia coli strains with different repair capabilities. J. Bacteriol. 147, 410417.
  • 31
    Munakata, N. (1993) Continual increase in biologically effective dose of solar UV radiation determined by spore dosimetry from 1980 to 1993 in Tokyo. J. Photochem. Photobiol. B. Biol. 31, 6368.
  • 32
    Riesenman, P. J. and W. L. Nicholson (2000) Role of the spore coat layers in Bacillus subtilis spore resistance to hydrogen peroxide, artificial UV-C, UV-B and solar UV radiation. Appl. Environ. Microbiol. 66, 620626.
  • 33
    Sutherland, J. C. (2002) Biological effects of polychromatic light. Photochem. Photobiol. 76, 164170.
  • 34
    Bossart, W., D. L. Nuss and E. Paoletti (1978) Effect of UV irradiation on the expression of vaccinia virus gene products synthesized in a cell-free system coupling transcription and translation. J. Virol. 26, 673680.
  • 35
    Lytle, C. D., S. A. Aaronson and E. Harvey (1972) Host-cell reactivation in mammalian cells. II. Survival of herpes simplex virus and vaccinia in normal human and xeroderma pigmentosum cells. Int. J. Radiat. Biol. 22, 159165.
  • 36
    Lytle, C. D., S. G. Benane and J. E. Stafford (1976) Host cell reactivation in mammalian cells. V. Photoreactivation studies with herpes virus in marsupial and human cells. Photochem. Photobiol. 23, 331336.
  • 37
    Sagripanti, J.-L. and C. D. Lytle (2007) Inactivation of influenza virus by solar radiation. Photochem. Photobiol. 83, 12781282.
  • 38
    Coohill, T. P. and J.-L. Sagripanti (2008) Overview of the inactivation by 254 nm ultraviolet radiation of bacteria with particular relevance to biodefense. Photochem. Photobiol. 84, 10841190.