Solar disinfection (SODIS) and subsequent darkstorage of Salmonella typhimurium and Shigellaflexneri monitored by flow cytometry
Franziska Bosshard,1,2 Michael Berney,13 Michael Scheifele,1Hans-Ulrich Weilenmann1 and Thomas Egli1,2
1Eawag, Swiss Federal Institute of Aquatic Science and Technology, PO Box 611, CH-8600
2Institute of Biogeochemistry and Pollutant Dynamics, ETH Zu¨rich, 8092 Zu¨rich, Switzerland
Pathogenic enteric bacteria are a major cause of drinking water related morbidity and mortality indeveloping countries. Solar disinfection (SODIS) is an effective means to fight this problem. In thepresent study, SODIS of two important enteric pathogens, Shigella flexneri and Salmonellatyphimurium, was investigated with a variety of viability indicators including cellular ATP levels,efflux pump activity, glucose uptake ability, and polarization and integrity of the cytoplasmicmembrane. The respiratory chain of enteric bacteria was identified to be a likely target of sunlightand UVA irradiation. Furthermore, during dark storage after irradiation, the physiological state ofthe bacterial cells continued to deteriorate even in the absence of irradiation: apparently thecells were unable to repair damage. This strongly suggests that for S. typhimurium and Sh.
flexneri, a relatively small light dose is enough to irreversibly damage the cells and that storage ofbottles after irradiation does not allow regrowth of inactivated bacterial cells. In addition, we show
that light dose reciprocity is an important issue when using simulated sunlight. At high irradiation
intensities (.700 W m”2) light dose reciprocity failed and resulted in an overestimation of the
effect, whereas reciprocity applied well around natural sunlight intensity (,400 W m”2).
is crucial to understand the way in which SODIS damagesbacteria – and whether repair can occur.
The availability of safe drinking water is a key health issuein developing countries. The United Nations have declared
The effectiveness of SODIS has been proven by cultivation-
it a millennium development goal to reduce the number of
based techniques with Escherichia coli and some pathogenic
people without sustainable access to safe drinking water by
organisms (Acra et al., 1980; Berney et al., 2006b;
half by 2015. Solar disinfection (SODIS) is one of the
McGuigan et al., 1998; Wegelin et al., 1994), and recently
means to reach this goal. Its success is based on easily
we have applied cultivation-independent methods to
available and low-cost tools: one day of exposure to the sun
characterize the inactivation of E. coli by sunlight (Berney
of hygienically unsafe drinking water in PET bottles leads
et al., 2006a). In that study we used flow cytometry
to a significant increase in microbiological water quality. A
combined with viability staining to characterize the loss of
positive impact on health has been documented in several
essential cellular functions in irradiated bacterial cells. The
epidemiological field studies, e.g. in India, where a total of
recorded cellular functions include membrane integrity,
40 % of diarrhoeal diseases and 50 % of severe diarrhoea
membrane potential, efflux pump activity and glucose
episodes were prevented by the use of SODIS (Rose et al.,
uptake activity. We showed that a reproducible sequence of
2006). SODIS water treatment is already used by 2 million
membrane-function breakdown takes place when E. coli is
people and the number is increasing. But despite the fact
irradiated with sunlight or UVA light (Berney et al., 2006a).
that the method works, the exact mechanism of inactiva-
However, it is important to know whether these results
tion of microbial pathogens is not yet known. Therefore, it
translate to enteric pathogens like Salmonella or Shigella,the inactivation of which is the primary goal of SODIS. So
3Present address: Department of Microbiology and Immunology,
University of Otago, PO Box 56, Dunedin 9016, New Zealand.
Reliability of the SODIS method depends not only on the
Abbreviations: DiBAC4(3), bis-(1,3-dibutylbarbituricacid)trimethine oxo-
light dose leading to damage in target cells, but also on
nol; EB, ethidium bromide; 2-NBDG, 2-[N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino]-2-deoxy-D-glucose; PET, poly(ethylene terephthalate); PI,
possible recovery processes in injured cells after irradiation.
So far, no regrowth or recovery of membrane functions in
Solar disinfection of S. typhimurium and Sh. flexneri
injured E. coli cells has been found (Berney et al., 2006a;
Staining procedures. Five fluorescent dyes were used alone or in
Joyce et al., 1996; Oates et al., 2003; Reed, 1997; Wegelin
different combinations: Syto 9 (Invitrogen Molecular Probes),
propidium iodide (PI; Invitrogen), bis-(1,3-dibutylbarbituricacid)-trimethine oxonol (DiBAC4(3); Invitrogen), ethidium bromide (EB;
The present work extends our knowledge about the
Fluka) and 2-[N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino]-2-deoxy-
inactivation mechanism of solar light from the indicator
D-glucose (2-NBDG; Invitrogen). Samples taken from irradiation
bacterium E. coli to two important enteric pathogens,
experiments (sunlight and artificial UVA) were divided into five
Salmonella typhimurium and Shigella flexneri. In addition,
subsamples and immediately stained with two mixtures of fluorescentdyes (Syto 9/PI and Syto 9/EB) and three single fluorescent dyes
we have investigated the ability of these enteric pathogens
[DiBAC4(3), Syto 9 and 2-NBDG]. Samples were incubated in the
to survive and repair damage after solar irradiation.
dark at 37 uC for 5 min (2-NBDG) or at room temperature for10 min [DiBAC4(3)], 15 min (Syto 9/EB), 20 min (Syto 9/PI) and25 min (Syto 9), respectively, before analysis. Prior to flow-cytometric
analysis, samples (~1–56107 cells ml21) were diluted with sterile-filtered bottled water (Evian) to 1 % (v/v) of the initial cell
Bacterial strains. Salmonella enterica serovar Typhimurium ATCC
concentration (~1–56105 cells ml21 final concentration). Stock
14028 (referred to in this paper as Salmonella typhimurium) and
solutions of the dyes were prepared as follows: PI and Syto 9 were
Shigella flexneri ATCC 12022 were used in this study.
used from the LIVE/DEAD BacLight kit (Invitrogen), EB wasprepared in distilled and filtered water at 25 mM, DiBAC4(3) was
Growth media and cultivation conditions. Cells were grown as
prepared in DMSO at 10 mM, and 2-NBDG was dissolved in distilled
described by Berney et al. (2006a) with modifications. In drinking
and filtered water at 5 mM. All stock solutions were stored at 220 uC.
water, cells grow very slowly or not at all; therefore, we used
The working concentrations of Syto 9, PI, EB, DiBAC4(3) and 2-
stationary-phase cells, which were shown to be more resistant to
NBDG were 5, 30, 30, 10 and 5 mM, respectively. 2-NBDG was added
SODIS than cells in the exponential growth phase (Berney et al.,
in combination with 2,4-dinitrophenol (final concentration 2 mM)
2006b; Reed, 1997). Luria–Bertani (LB) broth, which was filter-
(Natarajan & Srienc, 2000). At the beginning of each experiment a
sterilized with membrane filters (0.22 mm, Millipore) and diluted to
sample was incubated at 90 uC for 3 min (in a 2 ml Eppendorf tube)
33 % (v/v) of its original strength with ultrapure water, was used for
as a control measurement for inactive bacteria. By comparing the
batch cultivation. Precultures were prepared for each individual batch
staining pattern of heat-inactivated with untreated samples, electronic
experiment from the same cryo-vial, streaking the stock solution onto
gates were set to differentiate negatively and positively stained
Hektoen agar plates (Oxoid) selective for Shigella and Salmonella
species. After 15–18 h of incubation at 37 uC, one colony was picked,loop-inoculated into a 125 ml Erlenmeyer flask containing 20 ml LB
Flow-cytometric measurements. The methods used here have
broth, and incubated at 37 uC on a rotary shaker at 200 r.p.m. At an
been described recently (Berney et al., 2006a, 2007). Flow-cytometric
OD546 between 0.1 and 0.2, an aliquot of the culture appropriate to
measurements were made using a Partec Cyflow space flow cytometer
obtain an initial OD546 of 0.002 was transferred into a 500 ml
with 488 nm excitation from an argon ion laser running at 50 mW or
Erlenmeyer flask containing 50 ml prewarmed LB broth. With this
(for the fluorescent glucose analogue 2-NBDG) 200 mW. Green
procedure, no lag phase was observed. These flasks were then shaken
fluorescence was collected in the FL1 channel (520±20 nm), and red
at 200 r.p.m. on a rotary shaker at 37 uC for approximately 18 h until
fluorescence in the FL3 channel (.590 nm); all data were processed
stationary phase was reached. Stationary phase was confirmed from
with the Flowmax software (Partec), and electronic gating with the
five consecutive OD546 measurements within 1 h.
software was used to separate positive signals from noise. The specificinstrumental gain settings for these measurements were as follows:
Sample preparation and plating. Cells were harvested from batchculture by centrifugation (16 000 g, 3min), washed three times with
FL15490, FL35600, speed 3 (implying an event rate never exceeding
filter-sterilized commercially available bottled water (Evian) and
1000 events s21). All samples were collected as logarithmic (3
decades) signals and were triggered on the green fluorescence channel
546 of approximately 0.01 (corresponding to 1–
56107 cells ml21). To allow the cells to adapt to the mineral water,
(FL1). A routine check of the flow cytometer was performed every day
light exposure of bacterial suspensions was not started until 1 h after
for correct alignment with an FITC bead standard. This ensured
dilution. During exposure, aliquots were withdrawn at different time
accuracy in counting (volumetric counting device) and measured
points and diluted in decimal steps (1021 to 1026) with sterile-filtered
fluorescence intensities. Microscopic observation was performed on
bottled mineral water (Evian). Volumes of 1 ml of appropriate
an Olympus BX50 microscope equipped with filters HQ-F41-007 for
dilutions were withdrawn and mixed with 7 ml liquid tryptic soy agar
PI and EB and HQ-F41-001 for Syto 9, DiBAC4(3) and 2-NBDG (all
(TSA) (Biolife) at 45 uC (pour-plate method). After 20 min, the
solidified agar was covered with another 4 ml liquid TSA (40 uC).
Plates were incubated for 48 h at 37 uC until further analysis. The
Total ATP. For the determination of total ATP, the BacTiter-Glo
standard error of pour plating was always ,10 %.
system (Promega) was used. The BacTiter-Glo buffer was mixed withthe lyophilized BacTiter-Glo substrate and equilibrated at room
Sunlight and UVA exposure. Samples of 10 ml bacterial suspension
temperature. The mixture was stored overnight at room temperature
(see above) were exposed to sunlight or UVA light as described earlier
to ensure that all ATP was hydrolysed (‘burned off’) and the
background signal had decreased. A cell suspension of 100 ml wasmixed in a 2 ml Eppendorf tube with an equal volume of the
Dark storage. Samples of 10 ml bacterial suspension were exposed to
previously prepared BacTiter-Glo reagent (stored on ice). The sample
UVA light (see above). Cellular damages were assessed immediately
was then briefly mixed by once pipetting up and down and put into a
after irradiation and at different time points during dark storage,
water bath at 37 uC for 30 s. The luminescence of the sample was
which was performed at 37 uC for 48 h, holding the cells in the same
measured in a luminometer (model TD-20/20; Turner BioSystems)
medium (sterile-filtered bottled water) as during irradiation. To
immediately after incubation. A calibration curve with dilutions of
exclude the possibility of regrowth, nalidixic acid was included in all
pure rATP (Promega, P1132) was measured for each batch of
the samples at a concentration of 100 mg ml21.
BacTiter-Glo buffer. ATP concentration per cell was then calculated
using this calibration curve and the total count measurements (Syto
threefold, the reciprocity law started to fail not only for
the culturability of Sh. flexneri (Fig. 1) but also for all othermeasured viability indicators (data not shown). This clearly
Reproducibility. All field experiments were conducted in three
demonstrates that for these enteric pathogens high
biological replicates on three different days. Irradiation intensity datawere obtained from a weather station located 300 m from the
irradiation intensities in laboratory experiments result in
exposure site (BAFU/NABEL, EMPA Du¨bendorf, Switzerland).
an overestimation of the effect in comparison with the
Sunlight intensity varied with the weather conditions, so the light
same light dose under natural sunlight conditions and that
dose at the sampling points was never exactly the same. Therefore, we
such data are hence of limited use. Therefore, the doses
listed in Table 1 were obtained exclusively from experi-ments performed at light intensities that were in the rangeof natural conditions.
The difference in the fluence needed to achieve a three-logreduction when exposing the cells to different light
intensities was a result of two effects. Firstly, the shape of
The susceptibility of different properties of S. typhimurium
the inactivation curve followed the ‘log-linear with a
and Sh. flexneri to artificial UVA light is shown in Table 1.
shoulder’ model for samples irradiated with natural sunlight
Both organisms were less susceptible to UVA light than the
or artificial UVA of a corresponding intensity. However, the
indicator bacterium E. coli (Berney et al., 2006a). For
shoulder disappeared when cells were irradiated with very
example, when assessed with PI staining, the light dose
high intensities (919 W m22 and 1315 W m22 in Fig. 1).
needed for membrane permeabilization of S. typhimurium
Secondly, the inactivation was three times faster with very
was approximately three times, and for Sh. flexneri
high intensities as compared to natural conditions (slope of
approximately two times, higher than for E. coli.
20.0031 for intensities around 360 W m22 vs 20.0011 with
However, for all three enteric bacteria, the same sequential
inactivation pattern was observed with the measured
In general, the reciprocity law was valid for Sh. flexneri as
long as intensities not exceeding 400 W m22 were applied.
Because of this higher resistance, some of the laboratory
Experiments conducted in this irradiation intensity range
experiments were conducted with much higher irradiation
compared well with natural conditions. Reciprocity for S.
intensities than those normally achieved with natural
typhimurium held over a much wider range (from 50 to
sunlight. Typically, the maximum natural sunlight intens-
700 W m22). However, when very high irradiation
ity at noon on mid-European longitude is about 130 W
intensities (1000 W m22) were applied, the shoulder of
m22 (integrated for the wavelength spectrum 350–
the inactivation curve was shortened (data not shown).
450 nm), whereas in this study, intensities between163 W m22 and 1315 W m22 were used for artificial
UVA exposure to achieve high enough fluences within areasonable time period. We observed that, as soon as
A decrease in culturability of more than 99 % was observed
natural irradiation conditions were exceeded two- to
(Figs 2a and 3a) for S. typhimurium and Sh. flexneri during
Table 1. Comparision of the susceptibility of S. typhimurium and Sh. flexneri to artificial UVA light and sunlight
Figures indicate the approximate light doses (±20 %) in kJ m22 at which .90 % of the cells exhibited the properties indicated. Cells were exposedcontinuously to artificial UVA light with an intensity of 620 W m22 for S. typhimurium and 360 W m22 for Sh. flexneri. In the case of sunlight, cellswere exposed during two consecutive days with a night break of approximately 12 h. Sunlight exposure on the first day reached 2300 kJ m22 for S.
typhimurium and 2500 kJ m22 for Sh. flexneri, respectively. An asterisk (*) indicates that this parameter changed during the night break. Forcomparison, corresponding data for E. coli are shown (Berney et al., 2006a).
Loss of culturability (0.1 % survival) at
.90 % of cells with inactivated efflux pumps at
.90 % of cells unable to take up glucose at
.90 % of cells with depolarized membranes at
.90 % of cells with permeabilized membranes at
Solar disinfection of S. typhimurium and Sh. flexneri
lowest level on the first day of irradiation (Fig. 2b, e).
Interestingly, a more than twofold increase in ATP
concentration was observed initially before it rapidly
decreased at approximately 1000 kJ m22. A similar increase
in cellular activity with increasing exposure in the initial
phase was observed also for the uptake of the fluorescent
For Sh. flexneri a similar general pattern was observed, but
with some distinct differences in the magnitude of the
effects. Culturability dropped over more than 3 additional
log-units during the night break (Fig. 3a). After the night
break, 80 % of the cells had lost their membrane potential
and 20 % were even permeabilized (Fig. 3c, d). About 40 %
of the cells were unable to take up glucose (Fig. 3f). As in S.
typhimurium, ATP concentration and efflux pump activity
in Sh. flexneri had already reached final levels on the first
day. Comparable to S. typhimurium, a slight increase inglucose uptake activity was observed during irradiation onthe first day (Fig. 3d).
Fig. 1. Deviations from light dose reciprocity appearing withdifferent intensities of artificial UVA light irradiation in comparison
The overall fluence rate resulting in membrane permeabi-
to long, low-intensity irradiation with sunlight, shown for the
lization in .90 % of the cells (8000 vs 4500 kJ m22 in S.
culturablity of Sh. flexneri. Bacterial cells were harvested from the
typhimurium, and 6000 vs 4500 kJ m22 in Sh. flexneri) and
stationary phase of an LB batch culture, washed three times and
loss of membrane potential in .90 % of the cells (6000 vs
diluted in bottled mineral water. C.f.u. were measured by pour
2300 kJ m22 in S. typhimurium, and 4500 vs 2500 kJ m22
plating and sensitivity was recorded as c.f.u./(c.f.u. at time zero).
in Sh. flexneri) was clearly reduced with discontinuous
Horizontal dashed lines indicate the detection limits. Lines
represent modelled inactivation curves with the program GInaFIT(Geeraerd et al., 2005). Empty diamonds (e) represent averagedcontrols. (a) Artificial UVA light was applied at the following
Dark storage of S. typhimurium after irradiation
intensities: ¾, 163 W m”2; $, 332 W m”2; #, 370 W m”2; &,
For a better understanding of the overnight processes in
710 W m”2; +, 732 W m”2; h, 786 W m”2; -, 802 W m”2; m,
the outdoor experiments, irradiation and subsequent
919 W m”2 and g, 1315 W m”2. (b) Sunlight irradiation on three
storage in the dark of S. typhimurium was investigated in
different days in biologically independent triplicates: ¾, $ and #
the laboratory (Fig. 4). This was done because of the higher
on 23 August 2006; &, + and h on 31 August 2006; -, m and g
resistance of S. typhimurium resulting in only 2-log
reduction during one full day of exposure. Bacterial cellsuspensions were irradiated with a dose of 1500 kJ m22,
one day of solar irradiation (8 h; 2300 and 2500 kJ m22,
corresponding to half a day of sunlight or one full day of
respectively). Efflux pump activity and ATP concentration
exposure under overcast conditions. Culturability, efflux
had reached their lowest level by the end of the day (Figs
pump activity, membrane potential, membrane permeab-
2b, e and 3b, e). In contrast, inactivation of glucose uptake
ility and ATP content per cell were measured immediately
ability, loss of membrane potential and loss of membrane
after exposure and for a period of up to 48 h at eight
integrity were not observed with S. typhimurium and Sh.
subsequent time points during storage in the dark. This
flexneri during one day of sunlight irradiation (Figs 2f, c, d
aspect was investigated in more detail because it is possible
and 3f, c, d). Therefore, we decided to continue exposure
that a small part of the cells might survive the irradiation
and that these survivors might start growing on either lysedcells or assimilable organic carbon in the water. For
Interestingly, S. typhimurium and Sh. flexneri lost cellular
example, it has been shown that some pathogenic bacteria
activity during dark storage at 37 uC overnight. This is
are able to grow on natural substrates in bulk water (Vital
clearly shown for S. typhimurium, where an additional one-
et al., 2007, 2008). To exclude this possibility and to make
log reduction in c.f.u. was observed after the night break
sure that only the cells originally exposed to light were
(Fig. 2a). Accordingly, 60 % of the cells lost their
analysed, we added nalidixic acid to inhibit cell division.
membrane potential during the night break and a slight
The concentration of nalidixic acid required to keep S.
reduction in membrane integrity was also observed
typhimurium from dividing was determined in a minimal
(Fig. 2b). Furthermore, the percentage of cells able to take
inhibitory concentration experiment.
up glucose increased by 40 % on the first day and wasfollowed by a loss of 80 % during the night break (Fig. 2f).
After the bacteria had received a ‘half-day’ UVA dose, c.f.u.
ATP concentration and efflux pump activity reached their
had decreased by approximately 1 log-unit (Fig. 4a).
Fig. 2. Viability parameters of stationary-phase cells of S. typhimurium exposed tosunlight. (a) Culturability [log(c.f.u. ml”1)] ($),with the dashed line indicating the detectionlimit, and log(total cell concentration ml”1) (.).
(b) EB-positive cells, (c) DiBAC4(3)-positive
cells, (d) PI-positive cells. Values were calcu-
lated as percentage of total cell concentration.
(e) Average ATP concentration per cell. (f) 2-
NBDG-positive cells (able to take up glucose)calculated as percentage of total cell concen-tration. Light dose on day 1, 2300 kJ m”2; day2, 1200 kJ m”2 (overcast conditions). Thenight break is indicated by a dotted line in eachgraph. In all graphs, unirradiated control
samples are displayed as empty symbols.
Error bars represent standard deviations from
three biologically independent experiments.
During subsequent dark storage, c.f.u. decreased by 5 log-
unirradiated control cells were also hampered at the
units over 24 h. Efflux pump activity was lost completely
beginning of the experiment, when about 50 % of the
just after irradiation and was not regained (Fig. 4b). The
population showed no efflux pump activity, probably as an
Fig. 3. Viability parameters of stationary-phase cells of Sh. flexneri exposed to sunlight.
(a) Culturability [log(c.f.u. ml”1)] ($), with thedashed line indicating the detection limit, andlog(total cell concentration ml”1) (.). (b) EB-positive cells, (c) DiBAC4(3)-positive cells, (d)
PI-positive cells. Values were calculated as
percentage of total cell concentration. (e)
Average ATP concentration per cell. (f) 2-
NBDG-positive cells (able to take up glucose)
calculated as percentage of total cell concen-tration. Light dose on day 1, 2500 kJ m”2; day2, 1300 kJ m”2 (overcast conditions). Thenight break is indicated by a dotted line in eachgraph. In all graphs, non-irradiated control
samples are displayed as empty symbols.
Error bars represent standard deviations from
three biologically independent experiments.
Solar disinfection of S. typhimurium and Sh. flexneri
Fig. 4. Dark storage of S. typhimurium afterexposure to a dose of 1500 kJ m”2 artificial
UVA light, applied with an irradiation intensityof 205 W m”2 for 120 min. To exclude the
possibility of regrowth, nalidixic acid (100 mgml”1)
Culturability [log(c.f.u. ml”1)] ($), with thedashed line indicating the detection limit, andlog (total cell concentration ml”1) (.). (b) EB-positive cells, (c) DiBAC4(3)-positive cells, (d)PI-positive cells. Values were calculated as
percentage of total cell concentration. (e)
Average ATP concentration per cell. (f) 2-
NBDG-positive cells (able to take up glucose),
calculated as percentage of total cell concen-
tration ($, numbers on left side of graph), andgeometric mean of green fluorescence of the2-NBDG-positive cell population (., numberson right side of graph and dashed line for
detection limit). In all graphs, non-irradiated
effect of nalidixic acid. Efflux pump activity was later
functions during continuous artificial UVA exposure was
regained in the control sample. Membrane potential was
similar in both tested organisms and corroborates our
lost in only a very small part of the cell population during
results for E. coli (Berney et al., 2006a). The similarity in
irradiation, but a relevant fraction of the cells (about 60 %)
inactivation pattern suggests that the molecular mechan-
lost their membrane potential in the following 24 h of dark
isms involved in the inactivation and killing of bacterial
storage (Fig. 4c). Integrity of the cell membranes did not
pathogens due to solar irradiation are similar or even
change significantly during dark storage: only a slight
identical in the three enteric bacteria. Therefore, E. coli can
increase in the percentage of permeabilized cells was
be considered a good model organism for such investi-
observed (Fig. 4d). ATP content of the cells was reduced to
gations. However, as suggested in an earlier study (Berney
10 % compared to the control just after the treatment and
et al., 2006b), both Sh. flexneri and S. typhimurium were
levelled off to ,5 % during the first 10 h after the
found to be more resistant than E. coli.
treatment. Also the fraction of cells able to take up glucosedecreased from about 90 % just after the treatment to zero
during the first 24 h. The population that was still able totake up glucose during dark storage became increasingly
In experiments with Sh. flexneri and S. typhimurium, we
less fluorescent, which indicates that the uptake rate of
observed two modes of exposure that caused deviation
fluorescently labelled glucose decreased with time.
from the reciprocity law: firstly, exposure to either veryhigh or very low irradiation intensity, and secondly, split
We never observed a regain of culturability in irradiated S.
exposure by a pause in irradiation of up to 14 h.
typhimurium or Sh. flexneri cells after 24 h or 48 h dark
Interestingly, light dose reciprocity was valid for E. coli
storage after irradiation with different light doses in
(Berney et al., 2006b) and S. typhimurium (this study)
laboratory or field experiments (data not shown).
within a much broader intensity range (50–700 W m22)than for Sh. flexneri, which already showed distinctdeviation when light intensities were two- to threefold
higher than those occurring under natural conditions.
Susceptibility of S. typhimurium and Sh. flexnerito solar light
High irradiation intensity reduces the light doserequired for inactivation
For the first time, cellular functions in S. typhimurium andSh. flexneri during solar and UVA exposure were followed
Light dose reciprocity has been observed in a majority of
not only by plating but also with viability staining and ATP
biological and medical applications, while so called
measurements. The pattern of sequential loss of cellular
‘reciprocity law failures’ have been mostly observed in
experiments conducted at either very low or very high
radiant fluxes (Martin et al., 2003). In the case of Sh.
The data presented in this study strongly suggest that
flexneri irradiated with UVA light we observed that the
SODIS inactivates S. typhimurium and Sh. flexneri by
shoulder of the inactivation curve became less pronounced
inhibiting the respiratory chain. The ATP content per cell
or was even eliminated with high irradiation intensities.
decreased rapidly in E. coli (Berney et al., 2006a) and Sh.
The existence of a shoulder was interpreted earlier by either
flexneri (this study) upon irradiation with sunlight, while
the presence of repair mechanisms that are able to slow
in S. typhimurium an initial increase was observed before a
down the light effect and/or damage to more than one
sharp sustained decline. We propose that after an initial
target within the cell (Harm, 1980; Sommer et al., 2001). If
activation of energy metabolism, which is reflected in
a shoulder is eliminated, as seen in our experiments for
increased glucose uptake and ATP level, respiration stops
high-intensity irradiation of Sh. flexneri and also S.
and the remaining ATP is either consumed by various
typhimurium, repair mechanisms might be too slow or
recovery processes (Kobayashi et al., 2005), or, more likely,
damaged themselves. Elimination of the shoulder results ina reduction of the light dose required for inactivation.
through the maintenance of the membrane potential viathe F1F0-ATPase. Consistent with this proposal is theobserved increase in glucose uptake activity, which
Split exposure reduces the light dose required for
provides ATP via substrate-level phosphorylation to fuel
the proton-pumping ATPase to maintain the membrane
We found that S. typhimurium and Sh. flexneri are more
potential at a critical level even in the absence of a
susceptible to the same light dose when exposed to sunlight
functioning electron-transport chain. Since the medium
over two days (with a break overnight) than with continuous
used in this study (bottled mineral water) contains only
artificial UVA light. This finding was reproduced in the
low levels of assimilable organic carbon, the cells will
laboratory with discontinuous irradiation on two consec-
eventually die from ATP exhaustion and loss of the
utive days (data not shown). This is most likely because the
membrane potential. In fact, it has been proposed earlier
bacterial cells are irreversibly damaged and consequently they
that components of the respiratory chain like menaqui-
continue to lose viability even when irradiation is stopped.
nones and dehydrogenases could be inactivated by UVAlight (Jagger, 1981).
This is in line with the observation that exposure of cells froma human carcinoma cell line to short intervals of UVA lightwas more cytotoxic than continuous UVA irradiation
(Merwald et al., 2005). However, if the time between
In our experiments, injured bacterial cells irradiated with
exposures exceeded 2 h, the cells were able to recover and,
sunlight or UVA light were never observed to be able to
therefore, were less susceptible than with one single dose of
regrow. This corroborates the work of other authors (Joyce
UVA. Similar results were reported for Saccharomyces
et al., 1996; Oates et al., 2003; Reed, 1997; Wegelin et al., 1994).
cervisiae and E. coli irradiated with UVC light (Dzidic etal., 1986; Harm, 1968; Salaj-Smic et al., 1985; Sommer et al.,
The lack of regrowth in cells irradiated with polychromatic
1996) and human dermal fibroblasts irradiated with UVA
UV light in water disinfection processes seems to be the
light (tanning bed radiation) (Hoerter et al., 2008).
main advantage compared to monochromatic UVC light
Therefore, our data suggest either that the repair mechanisms
and therefore is of great interest to the water disinfection
of S. typhimurium and Sh. flexneri were already inactivated
community (Kalisvaart, 2001, 2004; Oguma et al., 2002;
after 8 h of continuous irradiation, or that the lack of
nutrients in the suspension did not allow the induction of an
It has been shown that non-cultivable cells of S.
appropriate repair response. It remains to be determined
typhimurium produced by UVA irradiation do not retain
whether or not the cells are able to recover when cell damage
infectivity for mice (Smith et al., 2000). Our study now
is less severe (shorter irradiation period) or when an
indicates that this is most likely due to irreversible damage
appropriate amount of nutrients is available in the water.
occurring during exposure to sunlight. Bacterial cells that
Discontinuous UVC exposure of E. coli was shown to
are impaired in glucose uptake and oxidative phosphoryla-
induce the SOS response, which increases DNA repair
tion may not be able to regrow, because uptake of nutrients
activity (Dzidic et al., 1986; Salaj-Smic et al., 1985).
and the maintenance of a membrane potential are regarded
However, UVA probably causes more complex damage to
as prerequisites for survival and replication.
the cells (several targets may be affected by UVA light, ascompared to UVC, where predominantly DNA damage is
observed) (Jagger, 1981). Hence, this is likely to require amore complex repair machinery. In an earlier microarray
SODIS and artificial UVA light kill enteric bacteria most
study we showed that in E. coli both DNA-repair genes and
likely by inactivation of the respiratory chain and
genes involved in oxidative stress response are induced
subsequent exhaustion of ATP. Our results show that even
upon irradiation with sublethal UVA light intensities
the resistant strain of S. typhimurium, which appeared to
suffer only minor damage after half a day of sunlight, was
Solar disinfection of S. typhimurium and Sh. flexneri
actually damaged to an extent such that regain of viability
was not detected. In fact, our results suggest that it is even
Motomatsu, A., Honjoh, K. & Iio, M. (2005). Identification of factors
favourable to store the treated water overnight before
involved in recovery of heat-injured Salmonella Enteritidis. J FoodProt 68, 932–941.
consumption, because injured cells will die from ATPexhaustion.
Martin, J. W., Chin, J. W. & Nguyen, T. (2003). Reciprocity lawexperiments in polymeric photodegradation: a critical review. ProgOrg Coat 47, 292–311.
McGuigan, K. G., Joyce, T. M., Conroy, R. M., Gillespie, J. B. &Elmore-Meegan, M. (1998). Solar disinfection of drinking water
This project was financially supported by the Velux Foundation
contained in transparent plastic bottles: characterizing the bacterial
(project number 346). We thank Margarete Bucheli, Martin Wegelin,
inactivation process. J Appl Microbiol 84, 1138–1148.
Regula Meierhofer, Silvio Canonica and Gregory M. Cook for
Honigsmann, H. & Trautinger, F. (2005). UVA-induced oxidativedamage and cytotoxicity depend on the mode of exposure. JPhotochem Photobiol B 79, 197–207.
Natarajan, A. & Srienc, F. (2000). Glucose uptake rates of single E. colicells grown in glucose-limited chemostat cultures. J Microbiol Methods
Acra, A., Karahagopian, Y., Raffoul, Z. & Dajani, R. (1980).
Disinfection of oral rehydration solutions by sunlight. Lancet 316,1257–1258.
Oates, P. M., Shanahan, P. & Polz, M. F. (2003). Solar disinfection(SODIS): simulation of solar radiation for global assessment and
Berney, M., Weilenmann, H.-U. & Egli, T. (2006a). Flow-cytometric
application for point-of-use water treatment in Haiti. Water Res 37,
study of vital cellular functions in Escherichia coli during solar
disinfection (SODIS). Microbiology 152, 1719–1729.
Oguma, K., Katayama, H. & Ohgaki, S. (2002). Photoreactivation of
Berney, M., Weilenmann, H.-U., Ihssen, J., Bassin, C. & Egli, T.
Escherichia coli after low- or medium-pressure UV disinfection
(2006b). Specific growth rate determines the sensitivity of Escherichia
determined by an endonuclease sensitive site assay. Appl Environ
coli to thermal, UVA and solar disinfection. Appl Environ Microbiol
Reed, R. H. (1997). Solar inactivation of faecal bacteria in water: the
Berney, M., Weilenmann, H. U. & Egli, T. (2006c). Gene expression of
critical role of oxygen. Lett Appl Microbiol 24, 276–280.
Escherichia coli in continuous culture during adaptation to artificialsunlight. Environ Microbiol 8, 1635–1647.
Rose, A., Roy, S., Abraham, V., Holmgren, G., George, K., Balraj, V.,Abraham, S., Muliyil, J., Joseph, A. & Kang, G. (2006). Solar
Berney, M., Hammes, F., Bosshard, F., Weilenmann, H. U. & Egli, T.
disinfection of water for diarrhoeal prevention in southern India.
(2007). Assessment and interpretation of bacterial viability by using
the LIVE/DEAD BacLight Kit in combination with flow cytometry.
Appl Environ Microbiol 73, 3283–3290.
Salaj-Smic, E., Dzidic, S. & Trgovcevic, Z. (1985). The effect of a splitUV dose on survival, division delay and mutagenesis in Escherichia
Berney, M., Vital, M., Hu¨lshoff, I., Weilenmann, H.-U., Egli, T. &
Hammes, F. (2008). Rapid, cultivation-independent assessment ofmicrobial viability in drinking water. Water Res 42, 4010–4018.
Smith, R. J., Kehoe, S. C., McGuigan, K. G. & Barer, M. R. (2000).
Effects of simulated solar disinfection of water on infectivity of
Dzidic, S., Salaj-Smic, E. & Trgovcevic, Z. (1986). The relationship
Salmonella typhimurium. Lett Appl Microbiol 31, 284–288.
between survival and mutagenesis in Escherichia coli after fractionatedultraviolet irradiation. Mutat Res 173, 89–91.
Sommer, R., Haider, T., Cabaj, A., Heidenreich, E. & Kundi, M.
(1996). Increased inactivation of Saccharomyces cerevisiae by protrac-
Geeraerd, A. H., Valdramidis, V. P. & Van Impe, J. F. (2005). GInaFiT,
tion of UV irradiation. Appl Environ Microbiol 62, 1977–1983.
a freeware tool to assess non-log-linear microbial survivor curves. IntJ Food Microbiol 102, 95–105.
Sommer, R., Pribil, W., Appelt, S., Gehringer, P., Eschweiler, H.,Leth, H., Cabaj, A. & Haider, T. (2001). Inactivation of bacteriophages
Harm, W. (1968). Effects of dose fractionation on ultraviolet survivalof Escherichia coli. Photochem Photobiol 7, 73–86.
in water by means of non-ionizing (UV-253.7 nm) and ionizing(gamma) radiation: a comparative approach. Water Res 35, 3109–
Biological Effects of Ultraviolet Radiation.
Cambridge: Cambridge University Press.
Vital, M., Fuchslin, H. P., Hammes, F. & Egli, T. (2007). Growth of
Hoerter, J. D., Ward, C. S., Bale, K. D., Gizachew, A. N., Graham, R.,
Vibrio cholerae O1 Ogawa Eltor in freshwater. Microbiology 153, 1993–
Reynolds, J., Ward, M. E., Choi, C., Kagabo, J. L. & other authors
(2008). Effect of UVA fluence rate on indicators of oxidative stress inhuman dermal fibroblasts. Int J Biol Sci 4, 63–70.
Vital, M., Hammes, F. & Egli, T. (2008). Escherichia coli O157 can growin natural freshwater at low carbon concentrations. Environ Microbiol
Jagger, J. (1981). Near-UV radiation effects on microorganisms.
Wegelin, M., Canonica, S., Mechsner, K., Fleischmann, T., Pesaro, F.
Joyce, T. M., McGuigan, K. G., Elmore-Meegan, M. & Conroy, R. M.
& Metzler, A. (1994). Solar water disinfection: scope of the process
(1996). Inactivation of fecal bacteria in drinking water by solar
and analysis of radiation experiments Aqua 43, 154–169.
heating. Appl Environ Microbiol 62, 399–402.
Zimmer-Thomas, J. L., Slawson, R. M. & Huck, P. M. (2007). A
Kalisvaart, B. F. (2001). Photobiological effects of polychromatic
comparison of DNA repair and survival of Escherichia coli O157 : H7
medium pressure UV lamps. Water Sci Technol 43, 191–197.
following exposure to both low- and medium-pressure UV irra-
Kalisvaart, B. F. (2004). Re-use of wastewater: preventing the recovery
diation. J Water Health 5, 407–415.
of pathogens by using medium-pressure UV lamp technology. WaterSci Technol 50, 337–344.
Are Farmer’s Markets right for “We’re small scale producers. We have a small breeding herd, so we’re there from the birth of the piglets, right through to the end product. I pack the meat at the butcher’s, and I make the sausages myself, and then we take the products to the farmers’ markets. So every step of the way to purchase their fresh food and other products we’
PRODUCT DATA SHEET 9.7-inch TabletPC for Android MEDIALINE mPAD® i973A 9.7-inch Tablet PC IPS wide viewing angle panel thin design TFT LCD with LED backlit support Android® 4.0 external USB 3G dongle support The new MEDIALINE mPAD® i973A tabletPC has a large screen of 9.7-inch ( size as iPAD® ). Powerfull graphics ( 2D/3D GPU Mali-400 ), user friendly and