Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE
Environmental Processes
Inactivation mechanisms of human and animal rotaviruses by solar UVA and visible light Elbashir Araud, Joanna L. Shisler, and Thanh H. Nguyen Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.7b06562 • Publication Date (Web): 19 Apr 2018 Downloaded from http://pubs.acs.org on April 19, 2018
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 32
Environmental Science & Technology
1
Inactivation mechanisms of human and animal rotaviruses
2
by solar UVA and visible light
3 4
Elbashir Araud1, Joanna L. Shisler2, and Thanh H. Nguyen1
5 6
Department of Civil and Environmental Engineering, University of Illinois at Urbana-
7
Champaign, Urbana, Illinois, USA1; Department of Microbiology and Department of
8
Pathobiology, University of Illinois at Urbana-Champaign, Urbana, Illinois, USA2
9 10
1 ACS Paragon Plus Environment
Environmental Science & Technology
11
ABSTRACT
12
Two rotavirus (RV) strains (sialidase-resistant Wa and sialidase-sensitive OSU) were irradiated
13
with simulated solar UVA and visible light in sensitizer-free phosphate buffered solution
14
(lacking exogenous reactive oxygen species (ROS)) or secondary effluent wastewater (producing
15
ROS). Although light attenuated for up to 15% through the secondary effluent wastewater
16
(SEW), the inactivation efficacies increased by 0.7 log10 for Wa and 2 log10 for OSU compared
17
to those in sensitizer-free phosphate buffered solution (PBS) after 4 hrs of irradiation. A binding
18
assay using magnetic beads coated with porcine gastric mucin containing receptors for
19
rotaviruses (PGM-MB) was developed to determine if inactivation influenced RV binding to its
20
receptors. The linear correlation between the reduction in infectivity and the reduction in binding
21
after irradiation in sensitizer-free solution suggests that the main mechanism of RV inactivation
22
in the absence of exogenous ROS was due to damage to VP8*, the RV protein that binds to host
23
cell receptors. For a given reduction in infectivity, greater damage in VP8* was observed with
24
sialidase-resistant Wa compared to sialidase-sensitive OSU. The lack of correlation between the
25
reduction in infectivity and the reduction in binding, in SEW, led us to include RNase treatment
26
before the binding step to quantify virions with intact protein capsids and exclude virions that
27
can bind to the receptors but have their capsid permeable after irradiation. This assay showed a
28
linear correlation between the reduction in RV infectivity and RV-receptor interactions,
29
suggesting that RV inactivation in SEW was due to compromised capsid proteins other than the
30
VP8* protein. Thus, rotavirus inactivation by UVA and visible light irradiation depends on both
31
the formation of ROS and the stability of viral proteins.
32 33 2 ACS Paragon Plus Environment
Page 2 of 32
Page 3 of 32
Environmental Science & Technology
34
INTRODUCTION
35
Group A rotaviruses (RVs) are the major etiological agent of acute gastroenteritis in infants
36
worldwide and accounted for 215,000 deaths in children in 2013.1 It has as low an infectious
37
dose of fewer than 10 particles and is shed at high titers (1010 to 1012 particles per gram of stool)
38
from infected persons.2-4 Despite the success of vaccines against RVs in the reduction of severe
39
gastroenteritis among infants, RVs are still of major public health importance and the leading
40
cause of diarrhea in children younger than five years old, globally.2, 5 RV outbreaks still occur
41
worldwide due to the high genetic diversity of RVs and the lack of cross-protection.2, 5 There are
42
eight groups of rotaviruses, referred to as A, B, C, D, E, F, G, and H. Humans are primarily
43
infected by group A rotavirus, which can be further divided into different serotypes.6 The outer
44
capsid glycoprotein (VP7) and the spike protein (VP4) differentiate RVs into 14 G
45
(Glycoprotein) serotypes and 27 different P (Protease sensitive) genotypes.6 Currently, five
46
serotypes (G1–G4, and G9) are the predominant circulating viruses, accounting for almost 95%
47
of strains worldwide.7 In addition, reassortment of RV strains between animal and human has
48
been reported.8-13 Thus, RVs are highly diverse, both antigenically and genetically.
49
RV-contaminated water and food are the main source of RV outbreaks.14 In both developed and
50
developing countries, infectious particles or genomes of human or animal RV are found in
51
drinking water sources.15-19 The contamination of the drinking water source is attributed to
52
inadequately treated wastewater.20 Also, it is troublesome that RV is detected on lettuce irrigated
53
with surface water in the US.21 Thus, the risk of RV outbreaks may increase when contaminated
54
surface water comes into contact with vegetables and seafood. To complicate matters, newly
55
emerging RV strains due to intra-genogroup reassortment are reported constantly and may also
56
add to current or future RV outbreaks.8, 22 Therefore, prevention of RV infection and outbreaks 3 ACS Paragon Plus Environment
Environmental Science & Technology
Page 4 of 32
57
can be achieved by improving wastewater disinfection and by knowing the stability of RV in
58
wastewater receiving surface water so that minimal contamination of food or drinking water
59
sources can be planned.20, 23
60
Solar inactivation of viruses is an important process that controls the water quality in a number
61
of cases including point of use water treatment (solar disinfection; SODIS); waste stabilization
62
ponds to treat municipal wastewater; reuse of treated wastewater for irrigation; discharge of
63
treated wastewater to surface water bodies, which are also used for recreation.24-26 In solar
64
disinfection, pathogen inactivation relies on solar irradiation, which directly or indirectly targets
65
the microbes (endogenous) or causes reactions between the microbes and the reactive oxygen
66
species (ROS) formed from natural sensitizers in water (exogenous).27, 28 Most solar disinfection
67
studies use indicator bacteria and bacteriophages as surrogates for human enteric viruses.27,
68
However, this is not a practical surrogate for RVs.29 Because there is strain-dependent
69
susceptibility of RVs to solar disinfection (e.g., the porcine RV strain OSU is more resistant
70
than the human RV strain Wa)30 and to disinfection by high pressure.31 These findings suggest
71
that it is important to understand disinfection mechanisms of real pathogens instead of
72
surrogates.
73
Binding to host cell is the first essential step in virus infectivity. The RV VP4 spike protein
74
attaches to host cell glycans, including gangliosides (N-acetylneuraminic acid (Neu5Ac) and N-
75
glycolylneuraminic acid (Neu5Gc)) and histo-blood group antigens (HBGA). VP4 consists of
76
two components VP8* and VP5. Located on the top of VP4, VP8* is essential for the early stage
77
of the virus cell attachment, while VP5 is involved in the later stages of the cell membrane
78
penetration and virus entry. Some RV strains (referred to as sialidase-sensitive) attach to sialic
79
acid moieties on the surface of the cell. This attachment for sialidase-sensitive rotaviruses 4 ACS Paragon Plus Environment
28
Page 5 of 32
Environmental Science & Technology
80
requires VP8* to mediate this interaction. Other RV strains, referred to as sialidase-resistant, use
81
Neu5Ac and histo-blood antigen groups instead of sialic acid to bind to the host cells. Because of
82
these two different binding mechanisms, these VP8* amino acid sequence is slightly different.
83
Specifically, for VP8* binding to sialic acid domain, a common grove on the VP8* domain with
84
two key amino acids at 157 and 187 position is important.32 In addition, amino acids at positions
85
R101, Y189, and S190 are strictly conserved for rotavirus strains that are sialidase-sensitive. For
86
some sialidase-resistant strains with VP4 genotype P4, P6, P8 and P11 the amino acid at R101,
87
the key binding amino acid to sialic acid, is substituted with amino acids with hydrophobic side
88
chains. This change disrupts the formation of the hydrogen bond and the interaction with the
89
sialic acid. Replacing of amino acids residues Y155, Y188, and Y189 with other amino acid with
90
hydrophobic side chains further precludes the binding to the sialic acid.33 However, knowledge
91
on how RV capsid proteins may dictate resistance or susceptibility to inactivation is limited.
92
To understand the stability of the RV capsid under solar irradiation, this study aimed to
93
determine the mechanisms controlling inactivation of two RV strains representing the most
94
predominant human and animal group A rotavirus genotypes. In addition to measuring the
95
inactivation efficacy of these RV strains in solutions without exogenous ROS or in secondary
96
effluent of wastewater (SEW) treatment, we also developed and calibrated binding assays
97
combined with RNase treatment to determine the extent of damage to the RV capsid caused by
98
solar UVA and visible light irradiation. Our findings demonstrated that design of UVA and
99
visible light disinfection or prediction of rotavirus survival in solar UVA and visible light must
100
consider both the formation of ROS and the stability of viral proteins.
101
MATERIALS AND METHODS
5 ACS Paragon Plus Environment
Environmental Science & Technology
102
Rotavirus stock preparation. RV strains from two different serotypes were used in this study.
103
The Wa strain (serotype G1P[8]) and OSU strain (serotype G5P[7]) were obtained from ATCC
104
catalogue numbers VR2018TM and VR892TM, respectively. All strains were propagated in MA-
105
104 cells cultured in Eagle's minimum essential medium (MEM) supplemented with 0.5 µg/ml
106
trypsin (Sigma, St. Louis, MO), as previously described.34 RVs were purified and concentrated
107
by sucrose cushion and centrifugation. Purified RVs were aliquoted and stored at -80 °C until
108
used. RV plaque assays were conducted as described previously31, in which MA-104 cellular
109
monolayers were used and plaques were visualized by using crystal violet staining.
110
Sialic acid sensitivity assays. Rotavirus strains with different compositions of the binding
111
proteins VP8* recognized different host cell receptors.35 To determine whether two strains of RV
112
that differed in their mechanism of binding to the host cell receptors have different inactivation
113
mechanisms, we selected two RV strains (OSU and Wa) that were likely to recognize different
114
host cell receptors. To determine the sensitivity of the selected strains toward sialidase, we
115
compared the titer of each strain after incubation with either untreated or NA-treated cellular
116
monolayers as follows: MA104 cells in 6-well plates were washed twice with 2 mL of serum-
117
free Minimum Essential Medium (MEM). Cellular monolayers were incubated in 1 mL MEM
118
containing or lacking 40 mU/ml neuraminidase (Sigma, St. Louis, MO ) for 1 hr at 37 °C and 5%
119
CO2. Next, monolayers were washed twice with 2 mL of cold MEM, and then ten-fold dilutions
120
of viral solutions with initial titers of 2.5×106 and 1.0×107 PFU/mL for Wa and OSU strains,
121
respectively, were inoculated onto different monolayers. Infected monolayers were incubated on
122
ice for 60 min, with agitation every 10 to 15 min. Monolayers were washed twice with pre-
123
warmed MEM medium to remove unattached viruses. Next, monolayers were overlaid with 2.5
124
ml of 2X Eagle minimum essential medium (MEM) containing 2.5% agarose, Anti-Anti (10,000
6 ACS Paragon Plus Environment
Page 6 of 32
Page 7 of 32
Environmental Science & Technology
125
units/mL of penicillin, 10,000 µg/mL of streptomycin, and 25 µg/mL of Gibco Amphotericin B)
126
(Thermo Scientific, Rockford, IL), and 1µg/ml trypsin (Sigma, St. Louis, MO). These plates
127
were incubated at 37 ˚C and 5% CO2 and examined for plaques at 3 days post-infection.
128
Comparison of VP4 gene sequence of OSU and Wa strains. To confirm the differences on the
129
sialic acid binding sites on the VP8* protein, the VP4 gene of OSU and Wa strains were
130
amplified by RT-PCR and sequenced. Genomic RNA of OSU or Wa strain was extracted from
131
140 µl of the virus stock using the QIAamp Viral RNA Mini Kit (Qiagen, Valencia, CA),
132
according to the manufacturer’s manual. Based on GenBank published nucleotide sequences of
133
OSU and Wa, two different pairs of primers for VP4 of OSU and Wa were designed and used in
134
One-Step Reverse Transcription-PCR (RT-PCR) (Table 1S). The amplified VP4 band was
135
visualized by 1% agarose gel electrophoresis and purified using QIAquick PCR Purification Kit,
136
according to the manufacturer’s manual. To confirm identity of the studied strains, the purified
137
VP4 was sequenced, using as set of primer walking (Table 1S), at the High-Throughput
138
Sequencing and Genotyping Unit at the University of Illinois at Urbana-Champaign. Sequence
139
conversion and alignment were conducted using BLAST (NCBI).
140
Sampling and characterization of wastewater. Wastewater from secondary clarified effluent
141
of the Northeast Wastewater Treatment Plant (NEP) in Urbana, Illinois was collected as
142
previously described by Dong et al.36 This secondary effluent wastewater (SEW) was used to
143
study exogenous inactivation of RVs triggered by UVA and visible light irradiation. The total
144
organic carbon (TOC) of 16.5 mg/L, ammonia of 7.3 mg/L, and nitrate of 0.2 mg/L were
145
detected.36 All experiments with SEW were conducted at its pH of 8.1. Absorbance of the SEW
146
used for irradiation experiments is shown in Figure 4S.
7 ACS Paragon Plus Environment
Environmental Science & Technology
147
Solar UVA and visible light disinfection experiments. A sunlight simulator (Atlas Suntest(r)
148
XLS+, photosimulator, Chicago, IL) with a xenon arc lamp was used for virus inactivation
149
experiments. UVA/visible light with wavelength of 320 nm and above and intensity setting of
150
400 W m−2 was used in all irradiation experiments. The light density and the wavelength cut of
151
the sunlight simulator was measured using a StellarNet Inc. spectrometer. Because UVB was
152
largely absent in the solar spectra collected in IL (Figure 1S and Table 2S) and the attenuation
153
of UVB was at least twice more than UVA in water with comparable dissolved organic matter to
154
wastewater using the correlations developed by Roberts et al.,37 solar UVA and visible light
155
irradiation is more relevant. For this reason, we conducted the study with solar spectra without
156
UVB. To confirm the removal of light under the UVA/visible (320 nm), Newport 320 nm cutoff
157
filters (Atlas MTS, Cat.56052372) wer used. These filters were put on top of all reactors for solar
158
disinfection and radical formation measurement. These reactors were made of 100-ml glass
159
beakers, with edges covered with a dark tape to prevent the light penetration and reflection. The
160
solution depth in the reactor was 1 cm. The irradiation experiments were conducted at 21 °C
161
maintained by a circulated water bath. A magnetic stirrer was used for continuous stirring the
162
solution in the reactors. The RV titer in each reactor was 105 PFU/ml. Titers were determined
163
from aliquots withdrawn from the reactor at different time points. Dark controls were used in
164
parallel with samples under the same conditions as a negative control, to confirm that the
165
reduction of the virus titer was only due to the irradiation. At least three replicates were
166
conducted with the same wastewater that was collected in one day. This wastewater was kept at
167
4 ˚C in the dark and used for disinfection experiments within 2 weeks from the collection date.
168
All the treated viruses were kept at -80 ˚C till all the plaque assays where conducted. The
169
irradiation was monitored throughout the study. Experiments with two viruses were conducted
8 ACS Paragon Plus Environment
Page 8 of 32
Page 9 of 32
Environmental Science & Technology
170
either at the same time or one right after the other to reduce variability among experimental
171
conditions. The initial titer in the inactivation solutions (Co) was determined using samples
172
immediately after the viral stock was added and mixed with the inactivation solutions and right
173
before irradiation. As shown in Figure 1S showing the spectra of the simulator and natural
174
sunlight, 4 hrs irradiation by the simulator is equivalent to 4 hrs of sunlight at 1 PM in IL. A
175
single peak at around 100 nm was observed using dynamic light scattering to indicate
176
monodispersitiy of the viral suspension in SEW.
177
Detection of oxygen reactive species (ROS) produced in wastewater. Phenol and Furfural
178
alcohol (FFA) were used as probes to detect hydroxyl (˙OH) and singlet oxygen (1O2) radicals,
179
respectively.31 Twenty milliliter of SEW samples were spiked with 100 µM of corresponding
180
probes and irradiated under the same conditions as for UVA and visible light disinfection
181
experiments (above) for 4 hrs. Samples of 400 µL were withdrawn every hour and the decay of
182
different probes was detected by HPLC Reverse-phase Agilent series 1200 (Agilent
183
Technologies) with Eclipse Plus C18 (3.5 µm) column was used for probes detection. The
184
mobile phase was water and acetonitrile at a ratio of 50:50 for phenol and 40:60 for FFA, with
185
flow rate of 0.3 ml/min and injection volume of 20 µl. The detection wavelength was 268 and
186
216 nm for phenol and FFA, respectively. ROS measurement was conducted with SEW without
187
spiked RV.
188
PGM-MB binding assay. To determine the effect of UVA and visible light irradiation on RV
189
binding to its receptor, assays were developed that discriminated between virions that retained or
190
lost their ability to bind to their receptors. Porcine gastric mucin (PGM) was bound to magnetic
191
beads (MB), as described previously for norovirus binding.38-40
192
MagnaBind carboxyl-derivatized beads (Thermo Scientific, Rockford, IL) was washed three time 9 ACS Paragon Plus Environment
Briefly, one milliliter of
Environmental Science & Technology
193
with PBS buffer. A magnetic separation rack (New England Biolabs, Ipswich, MA) was used to
194
collect beads after each wash. A ten milligram of type III mucin from porcine stomach supported
195
by 0.5 to 1.5% sialic acid (Sigma, St. Louis, MO) was dissolved in one milliliter of conjugation
196
buffer (0.1 M MES (2-(N-morpholino) ethanesulfonic acid), 0.9% NaCl, pH 4.7) and added to
197
the washed MagnaBind carboxyl-derivatized beads. Then 100 µL of 10 mg/mL 1-ethyl-3-[3-
198
dimethylaminopropyl] carbodimide hydrochloride (EDC) was immediately added as conjugation
199
buffer to the beads and the mucin. The mixture was rotated for 30 min at room temperature at 8
200
rpm. To remove excess mucin, beads were washed three times with PBS. A magnetic separation
201
rack (New England Biolabs, Ipswich, MA) was used to collect beads after each wash. Finally,
202
beads were re-suspended in 1 mL phosphate buffer solution (PBS) containing 0.005% sodium
203
azide, and beads were stored at 4 °C till used.
204
For each binding assay, 50 µL of each sample (with or without RNase treatment) and 70 µL of
205
Porcine Gastric Mucin beads (PGM-MB) were added to a 1.5 mL low adhesion centrifuge tube.
206
Sterilized PBS was added to this mixture for a total volume of 1 mL. This tube was gently
207
shaken by a bench shaker at room temperature for 30 min at 8 rpm. Beads were washed three
208
times with 1 mL PBS to remove unbounded particles. Beads were re-suspended in 140 µL of
209
DNase/RNase-Free distilled water before being subjected to RNA extraction. To avoid detecting
210
the noninfectious virus particles that had partially damaged capsid but still retained the ability to
211
bind to the beads, RNase treatment was used for another set of samples. Specifically, 2 µL of
212
RNase A/T1 (40 µg/ml RNase A and 100 U/ml RNase T) (Thermo Scientific, Rockford, IL) was
213
added to 50 µL samples collected from the UVA and visible light disinfection experiments with
214
wastewater and incubated at 37 °C for 30 min. After incubation with RNase, 2 µL of
215
SuperRNase inhibitor (sigma-Aldrich) was added to the samples and incubated for 20 min at
10 ACS Paragon Plus Environment
Page 10 of 32
Page 11 of 32
Environmental Science & Technology
216
room temperature to inactivate the residual RNase. Then the RNase-treated samples were
217
subjected to PGM-MB binding, RNA extraction, and RT-qPCR detection, as mentioned above.
218
Note that RNase A degrades RNA at C, and U residues and RNase T1 degrades the RNA at G
219
residue. The concentrations of RNase and RNase inhibitor were found through preliminary
220
experiments. Specifically, different concentrations of the RNase mixtures (80, 60, and 40 µg/ml
221
RNase A, and 200, 150, and 100 U/ml RNase T1) were added to the virus solution before the
222
binding assay to determine the right concentration allowing degradation of naked RNA and the
223
RNA inside the damaged capsid. At concentration higher than 40 µg/ml RNase A, and 100 U/ml
224
RNase T1, no RNA was detected by RT-qPCR. From 10 PFU/mL to 107 PFU/mL of either OSU
225
or Wa was used to establish the calibrations. This wide range was used to cover the detected
226
infectivity needed to assess binding ability of rotavirus after irradiation.
227
Quantification of RVs RNA by quantitative reverse transcription PCR (RT-qPCR). RT-
228
qPCR amplification was used to detect and quantify RV genes after UVA/visible light irradiation
229
or to calibrate the MBG-MB binding assays described above. The total RNA from untreated or
230
UVA-visible light irradiated RVs was extracted using the RNeasy kit (Qiagen). This RNA was
231
then reverse transcribed using the following sets of primers targeting the NSP3 genes of RV. For
232
Wa and OSU, the forward primer JVKF (5'- CAG TGG TTG ATG CTG AAG AT -3') and
233
reverse primer JVKR (5'- TCATTGTAATCATATTGAATACCCA -3'),41 were used to target
234
amplicon of 120 bp. A PCR machine (Applied Biosystems, Foster City, CA) was used for RT-
235
qPCR amplification. The targeted region of NSP3 was amplified using iTaq-Universal SYBR
236
Green iScript reverse transcriptase (Invitrogen) according to the manufacturer’s protocol.
237
Cycling parameters were as follows: 48 °C for 10 min for cDNA production, then inactivation of
238
RT and activation of PCR enzymes at 95 °C for 1 min. For qPCR, there were 40 cycles of 94 °C
11 ACS Paragon Plus Environment
Environmental Science & Technology
239
for denaturation for 10 s, 30 s for annealing and extension at 50°C. Standard curves were
240
generated by performing RT-qPCR reactions in parallel with samples containing a known
241
concentration of the plasmid encoding the NSP3 gene. The PCR efficiency ranged from 90-99%
242
with R2 of 0.97 to 0.99. Samples had a high enough concentration of RV that dilution 100 times
243
was usually needed. This high dilution level avoided PCR inhibition.
244 245
Effect of visible light and UVA irradiation on RV genome. To determine the effect of visible
246
light and UVA irradiation on the RV genome, total RNA from 140 µL of purified Wa (initial
247
titer of 2.5 x 106 PFU/mL) was extracted using RNeasy kit (Qiagen) and mixed with 800 µL of
248
organic free phosphate buffered solution (PBS) containing 10 µl RNase-out (1:10
249
dilution) (Sigma). A portion of 110 µL of RNA, from the 940 µL mixture in PBS solution was
250
put in one of nine wells of a 96-well plate. The plate was covered by a Clear Adhesive Film (The
251
Applied Biosystems® MicroAmp®) to prevent evaporation (Figure. 2S). The sides of the plate
252
were also secured by regular tape to prevent evaporation. The control well was covered by dark
253
tape to prevent irradiation exposure. A Newport 320 nm cutoff filter (Atlas MTS, Cat.56052372)
254
was put on the top of the 96-well plate. The plate was irradiated in the sunlight simulator for 4
255
hrs. The temperature was controlled around 21 °C during the irradiation by a circulated water
256
bath. After 2 and 4 hrs of exposure, one well was covered by dark tape to stop the exposure to
257
the light at that time point. The change in the amount of irradiated or the control RNA was
258
quantified by two-step RT-qPCR (Applied Biosystems, Foster City, CA). VP4 segment of 2360
259
bp was selected for the reverse transcriptase step. Reverse primers (VP4OSU 2340-2360) and
260
(VP4Wa 2336-2358) (Table 3S) were used to produce cDNA in the size of 2,360 and 2,358 bp
261
from VP4 of OSU and Wa, respectively. Ten microliters of irradiated and unirradiated RNA 12 ACS Paragon Plus Environment
Page 12 of 32
Page 13 of 32
Environmental Science & Technology
262
were used for first strand amplification (cDNA) using reverse transcriptase enzyme (BioLabs,
263
New England). Ten microliters of the irradiated or unirradiated RNA samples were used for
264
synthesizing cDNA using reverse transcriptase enzyme (BioLabs, New England). The denaturing
265
step at 95 °C for 5 min was followed by two minute cooling on ice. Then 10 µL of reverse
266
master mix contained the follows: 2µL of reverse primer (10 µM), 2 µL of M-MuL buffer, 1 µL
267
of RNase-inhibitor, 1µL the RT enzyme, 1 µL of (10 µM) dNTP, and 3 µL of water were added
268
to each denatured RNA. The RT step cycling was 5 min at 25 °C, 45 °C for 60 min, and 62 °C
269
for 20 min, to inactivate the RT enzyme. For the qPCR step, 1.5 µL of the cDNA synthesized in
270
the reverse transcription step was used with either primer pair VP4OSU 89-109 and VP4OSU
271
173-195 or primer pair VP4Wa 202-225 and VP4Wa 119-140 to amplify a segment of 106 bp.
272
Primers sequences are listed in Table 3S. Cycling parameters were as follows: 95 °C for 2 min,
273
to activate the PCR enzyme. For qPCR, there were 40 cycles of 94°C for denaturation for 15 s,
274
and 50°C for 30 s for annealing and extension. General iTaq Universal SYBR Green and qPCR
275
kits from Applied Biosystems (Foster City, CA) were used for the qPCR step. The standard
276
curves for the qPCR step were generated by performing qPCR on six aliquots of the cDNA with
277
known concentration of the targeted fragment. Based on the standard curves, the qPCR
278
efficiency was 115% for OSU and 112% for Wa. Coefficients of determination R2 were 0.99 in
279
both cases.
280
Statistical analysis. For each experimental condition and for each type of measurement (either
281
inactivation or radical measurement), an experiment was performed at least three times
282
independently. Data showing inactivation kinetics were presented as log10 of the negative value
283
of the titer at a given irradiation time normalized by the titer at time zero. This parameter is
284
referred to as log10 reduction in plaque forming unite (PFU). The reduction in binding was
13 ACS Paragon Plus Environment
Environmental Science & Technology
285
presented in terms of log10 of the negative value of the RT-qPCR results after binding assay for a
286
given sample at a given irradiation time, normalized by the RT-qPCR results after binding assay
287
for the sample collected at time zero. For all possible linear correlations between a pair of two
288
variables, Pearson product-moment correlation analysis was conducted for all collected data. The
289
significance of the linear correlation was based on the value of the Pearson’s correlation, p-value,
290
and coefficient of determination R2.
291
RESULTS AND DISCUSSION
292
The effect of UVA and visible light irradiation on RV genome in sensitizer-free phosphate
293
buffered solution. One potential reason for RV inactivation by UVA-visible light irradiation
294
would be if visible light and UVA irradiation damaged the viral RNA, therefore preventing virus
295
replication. To explore this possibility, RNA from RV strain Wa and OSU was extracted from
296
untreated virions and then exposed to visible-UVA irradiation for 2 or 4 hrs. We exposed naked
297
viral RNA instead of whole virions because naked RNA should be more susceptible to
298
irradiation than RNA encapsidated inside the virions. Next, we conducted reverse transcription
299
of the VP4 gene, a medium-sized segment of RVs genome. We chose this segment as a
300
representative of genomic damage that might occur in other segments of the RV genome. We
301
used a modified protocol that would PCR amplify undamaged RV RNA to cDNA genome
302
fragments in the size of 2360 bp for OSU and 2358 bp for Wa. We then quantified the presence
303
of cDNA by using qPCR, targeting a 106-bp region of the 2,360 or 2,358-bp amplicons. We
304
tested 3 biological replication and 2 technical replicates for each irradiation condition. We found
305
no statistically significant difference between samples irradiated for 2 hr or 4 hr versus the dark
306
control (p=0.14 for OSU and p=0.29 for Wa). Thus, irradiation by solar UVA and visible light
307
did not degrade the RV genome. 14 ACS Paragon Plus Environment
Page 14 of 32
Page 15 of 32
Environmental Science & Technology
308 309
Differences in the VP8* residues important for sialic acid interactions. The above data
310
suggested that RV inactivation due to solar UVA and visible light irradiation most likely was due
311
to capsid protein transformation and/or degradation. However, it is not known what RV capsid
312
proteins are damaged due to endogenous or exogenous reactive oxygen species (ROS). To
313
answer this question, we examined the inactivation of two strains of RV that differ in their
314
mechanism of binding to the host cell receptors. We hypothesized that, if solar UVA and visible
315
light irradiation damaged these viral binding proteins, then RV would no longer be infectious.
316
Inactivation mechanisms of RV may be related to its ability to bind to the host cell surface. The
317
host cell receptors that are crucial for RV attachment and entry to the cell are well known 35. We
318
amplified and sequence the VP8* gene of two rotavirus strains Wa and OSU. When we aligned
319
the amino acid sequences of the VP8* gene of the rotavirus strains OSU and Wa, the amino acids
320
essential for sialic acid binding (101, 155, 188, and 190) were different, agreeing with reported
321
literature. In addition to comparing the amino acid sequence, we also experimentally determined
322
the sensitivity of these two strains toward sialidase by comparing the titer of each strain after
323
incubation with host cells with or without treatment with NA to remove sialic acid. As shown in
324
Table 1a, for the Wa strain, the titers of Wa RV propagated with cells not treated with NA were
325
from 2.9 to 3.4 fold higher than those propagated with cells treated with NA. This increase
326
suggested that the Wa strain is sialic acid-resistant because when sialic acid was removed, Wa
327
has better access to the host cell receptor (N-acetylgalactosamin). On the other hand, for the
328
OSU strain, the titers of RV infected cells treated with NA were from 0.2 to 0.3 fold smaller than
329
those of RV infected cells not treated with NA. This decrease suggests that when sialic acid was
330
removed from the host cells, OSU could not bind to the host cells and infect the cells, indicating
15 ACS Paragon Plus Environment
Environmental Science & Technology
Page 16 of 32
331
that OSU is sialic acid-sensitive. We further tested whether OSU and Wa, whose most outer
332
binding protein VP8* has different amino acid composition (Table 1b) and binding mechanisms
333
to host cells, have different sensitivities toward solar UVA and visible light irradiation in the
334
presence or absence of reactive oxygen species (ROS).
335 336 337
Table 1a. Results of sialidase sensitivity assay of Wa and OSU strains Strain
Ratio of RV infected cells not treated p-value (-NA compared with with NA over RV infected cells treated +NA) with NA
OSU
0.2; 0.2; 0.3
4×10-5
Wa
2.9; 3.4; 3.0
0.03
338 339
P-values are from t-test comparing the titers of rotavirus infected cells treated or not treated with NA.
340
Table 1b. Alignment of VP8* portion of the VP4, that attach to sialic acid, for OSU and Wa RV strain
OSU 61 Wa 61
Amino acid positions
VEPLLDGPYQPTTFNPPTSYWVLLAPTVEGVIVQGTNNTDRWLATILIEPNVQTTNRIYN 120 VEPILDGPYQPTTFTPPNDYWILINSNTNGVVYESTNNSDFWTAVVAIEPHVNPVDRQYT 120
OSU 121 LFGQQVTLSVENTSQTQWKFIDVSTTTPTGSYTQHGPLFSTPKLYAVMKFSGRIYTYNGT 180 Wa 121 IFGESKQFNVSNDSN-KWKFLEMFRSSSQNEFYNRRTLTSDTRFVGILKYGGRVWTFHGE 179 OSU 181 TPNATTGYYSTTNYDTVNMTSFCDFYIIPRNQEEKCTEYINHGLPPIQNTRNVVPVSLSA 240 Wa 180 TPRATTDSSSTANLNNISITIHSEFYIIPRSQESKCNEYINNGLPPIQNTRNVVPLPLSS 239 341 342 343
The differences in amino acid compositions are colored to show the differences in amino acids at positions R101, H155, Y188, Y189, and S190, between OSU (sialidase-sensitive) and Wa (sialidase-resistance).
344 16 ACS Paragon Plus Environment
Page 17 of 32
Environmental Science & Technology
345
PGM-MB binding assay. To determine if UVA-visible light irradiation affected VP4 protein
346
function for OSU or Wa, we developed a binding assay to quantify RV binding to cellular
347
receptors present in porcine gastric mucins coated on magnetic beads. The reason for developing
348
this method was that attempts at developing binding assays using host cells could not distinguish
349
between specific and non-specific binding of viruses because a high percentage of the RV
350
population binds non-specifically to host cells.42 In contrast, this PGM-MB binding assay uses
351
magnetic beads coated with porcine gastric mucins, which contains 0.5 to 1.5% sialic acid. These
352
mucins also contain integrins and histo-blood group antigens to bind to norovirus and Tulane
353
viruses.38 Thus, the presence of sialic acid, integrins, and histo-blood group antigens in the
354
porcine gastric mucins would allow for the quantification of binding properties of both RV
355
strains OSU and Wa. Moreover, the magnetic property of the beads facilitated the separation of
356
bound from unbound viruses as an indicator of transformation of VP8* residues important for
357
binding to host cell receptors. Virions (either with or without the irradiation treatment) were
358
incubated with mucin-coated beads, the beads were magnetically concentrated, and then
359
thoroughly rinsed before being subjected to RNA extraction. RT-qPCR was used to quantify the
360
virions with intact VP4, as an indicator of RVs that retained their ability to bind to host cell
361
receptors.
362
We conducted two types of calibration using untreated virions to ensure that this assay was
363
sensitive enough to accurately quantify changes in the number of RV bound to the PGM-MB
364
beads. First, we conducted infectivity and RT-qPCR assays for Wa and OSU without the binding
365
step to determine the reliable range for both assays. For both Wa and OSU samples, we found a
366
linear correlation between the log10 of the infectivity and log10 of the targeted RNA, as shown in
367
Figure 1a. The Pearson’s coefficients, p-values, and coefficients of determination for both the
17 ACS Paragon Plus Environment
Environmental Science & Technology
Wa correlation and the OSU correlation were 0.98, 10-13, and 0.97, respectively. These two linear
369
correlations have similar slopes (0.75±0.03 for Wa and 0.74±0.03 for OSU), but different
370
intercepts (3.7±0.1 for Wa and 2.7±0.1 for OSU), probably due to the presence of a higher
371
number of non-matured Wa virions. The results of correlation analysis suggest that there was a
372
linear correlation between the genes detected by RT-qPCR and the PFUs present when 10 to 106
373
PFU/mL or 104 to 108 copies/mL for Wa and 100 to 107 PFU/mL or 103 to108 copies/mL for
374
OSU were present.
Log10 RNA in solutions
368
8
6
4
a 2
4
6
8
Log10 infectivity Human Wa (G1)
Porcine OSU (G5)
375
18 ACS Paragon Plus Environment
Page 18 of 32
376
Environmental Science & Technology
Log10 RNA detected after binding
Page 19 of 32
8
6
4
b 2
4
6
8
Log10 infectivity Human Wa (G1) Porcine OSU (G5)
377
Figure 1a. Correlations between log10 infectivity in PFU/mL and log10 of targeted RNA
378
determined by RT-qPCR in copies/mL for Wa and OSU without the binding assay. Figure 1b.
379
Binding assay calibration showing the correlations between log10 of the dilution factors and log10
380
of targeted RNA determined by RT-qPCR in copies/mL. All data above the detection limits for
381
infectivity and qPCR assays were used for the correlations.
382 383
Second, we conducted a binding assay for untreated Wa and OSU viruses to determine the range
384
of virions that can be used for the binding assay. For example, if too few virions were added to
385
the beads, then PCR assay results may not be consistently reliable to quantify the bound virions.
386
Oppositely, if too many virions were present, the number of available binding sites on the beads
387
would limit virion attachment, which could lead to faulty interpretation of results. As shown in
388
Figure 1b, a linear correlation was found between the infectivity of the untreated viral samples
389
and the PCR results obtained after the binding assays for the range of 10 to 106 PFU/mL and
390
104.3 to 107 RNA copies/mL for Wa, and 102 to 107.3 PFU/mL and 104.4 to 107.9 RNA copies/mL 19 ACS Paragon Plus Environment
Environmental Science & Technology
391
for OSU. The same correlation for both strains has the Pearson’s coefficient and the coefficient
392
of determination (R2) at 0.97 and 0.95, respectively. Based on the statistical analysis, we
393
conducted all binding assays for all virus strains within these ranges for infectivity and PCR
394
signals.
395 396
Degradation of chemical probes in sensitizer-free phosphate buffered solution (PBS) and
397
secondary effluent wastewater (SEW). To test the hypothesis that inactivation mechanisms of
398
RV due to UVA-Visible light irradiation depend on both RV genotype and the reactivity of
399
reactive oxygen species (ROS) with RV capsid, we conducted irradiation experiments using
400
either PBS or SEW. The lack of organic matter in PBS allowed the study of RV inactivation only
401
by direct effects of UVA-Visible light irradiation, while the formation of ROS from the organic
402
matter in SEW allowed the study of capsid damage due to the reaction between ROS with the
403
capsids. We first confirmed a lack of radicals present in PBS under irradiation conditions; there
404
was no significant change (p>0.05) in the concentrations of two probe compounds, phenol and
405
FFA, before and after the UVA-Visible light irradiation of PBS. In contrast, due to the presence
406
of organic matter, SEW had a quenching effect of 1.06 compared to PBS. There was a significant
407
change (P