U VD isinfectionofGiardia lamblia
Cysts inW ater
K A R L
G A E T A N F A U B E R T ,§
W I L L I A M
M A R KD . S O B S E Y³
Department of Civil and Environmental Engineering,
Duke University, Durham, North Carolina 27708,
Department of Environmental Sciences and Engineering,
University of North Carolina at Chapel Hill, Chapel Hill,
North Carolina 27599-7400, Institute of Parasitology,
McGill University, Ste. Anne-de-Bellevue,
Quebec, Canada H9X 3V9, and Trojan Technologies Inc.,
London, Ontario, Canada
G . L I N D E N , *, ²
G W Y - A M S H I N ,³
C A I R N S ,⊥A N D
The human and animal pathogen Giardia lamblia is a
waterborne risk to public health because the cysts are
ubiquitous and persistent in water and wastewater, not
completely removed by physical-chemical treatment
processes, andrelatively resistant tochemical disinfection.
Given the recently recognized efficacy of UV irradiation
against Cryptosporidiumparvumoocysts, the inactivation
temperature by near monochromatic (254 nm) UV
irradiation fromlow-pressure mercury vapor lamps was
determined using a ªcollimated beamº exposure system.
Reductionof G. lamblia infectivity forgerbils was very rapid
and extensive, reaching a detection limit of >4 log
within a dose of 10 J M-2. The ability of UV-irradiated G.
lambliacysts torepairUV-induceddamagefollowing typical
drinking water and wastewater doses of 160 and 400
J M-2was also investigated using experimental protocols
typical for bacterial and eucaryotic DNA repair under
both light and dark conditions. The infectivity reduction of
G. lamblia cysts at these UV doses remained unchanged
evidence of either light or dark repair of DNA damage
caused by LP UV irradiation of cysts was observed at the
UV doses tested. We conclude that UV disinfection at
practical doses achieves appreciable (much greater than
4 log) inactivation of G. lamblia cysts in water with no
evidence of DNA repair leading to infectivity reactivation.
Giardia lamblia is oneof themost important health-related
waterborneoutbreaksoverthelastseveral decades, many of
which had met regulatory standards for turbidity and
coliforms (1). Cysts of G. lamblia are consistently present at
high concentrations (102-104/L) in wastewater (2), highly
to conventional water and wastewater treatment processes,
ventional filtration system achieves some physical removal
of G. lamblia at practical doses and contact times (4-6). On
the basis of the in vitro viability assays of excystation and
vital dyestaining, Riceand Hoff (7) reported that G. lamblia
cystsarealso very resistant to UV irradiation. Craik et al. (8),
using an in vivo mouse animal infectivity assay, reported
water and wastewater treatment plants worldwide now
practicing UV disinfection, employ monochromatic low-
pressure (LP) UV lamps for irradiation, not polychromatic
MP UV lamps.Furthermore, G.murisisnot ahuman patho-
gen, and significant physiological differences exist between
G. lamblia and G. muris. Therefore, the extent of UV
inactivation ofthehuman pathogen G.lamblia asmeasured
by an animal infectivity assay has not been reported.
In addition to a need for evaluating the effect of UV irra-
to repair UV-mediated DNA damage and restore infectivity
also requires investigation. Knowledge of DNA repair ca-
pability will aid in proper design of UV disinfection systems
ability to repair DNA lesions caused by either exogenous
(UV light, ionizing radiation, and environmental chemicals)
orendogenous(oxidativedamageand structural instability)
mechanisms (9). It is well-known that some of the health-
related microorganisms have one or more of these repair
pathways (i.e., photoreactivation), such as most of the
indicator bacteria (10-12) and some pathogenic bacteria
Cryptosporidium parvum did not exhibit restoration of cell
culture infectivity following specific light and dark repair
protocols(15).Itisnotknown whetherG.lamblia cystshave
DNA repair pathways capable of reversing DNA damage
caused by UV irradiation and thus are able to restore
The objectives of this research were to (1) determine the
LP mercury UV lamps and (2) determine the kinetics and
extentofDNA repairofUV-irradiatedG.lamblia cystsunder
the conditions of dark and light repair typically employed
for bacterial, eucaryotic, and mammalian cell DNA repair
M aterials andM ethods
Parasites. G. lamblia (CH-3) cysts were purchased from
from experimentally infected Mongolian gerbilswerescreened
toremovelargedebris,then mixedwithzinc sulfatesolution
(ZnSO4, 1.2specific gravity), and centrifuged at1500rpm for
5min.Cystsrecovered in thesupernatant werewashed with
distilled water, resuspended in deionized buffer water
solution containing antibiotics (Gentamycin at 1 mg/mL),
and stored at 4 °C.
LP UV IrradiationSystem andRadiometry.Aªcollimated
beamº bench scale UV apparatus consisted of two 15-W LP
mercury vapor germicidal lamps emitting nearly monochro-
matic UV radiation at 253.7 nm that was directed through
a circular opening to provide incident radiation normal to
*Corresponding author phone: (919) 660-5196; fax: (919) 660-
5219; e-mail: email@example.com.
³University of North Carolina at Chapel Hill.
⊥Trojan Technologies Inc.
Environ. Sci. Technol. 2002, 36, 2519-2522
Petri dish.UV irradiance(W M-2) at 253.7nm wasmeasured
with a radiometer and UV 254 detector (International Light
IL1400,Newburyport,MA) thathadbeen factory-calibrated,
traceableto National InstituteofStandardsand Technology
UV Dose Determination. The measured incident irra-
diance at the surface of the test liquid (approximately 0.6 W
M-2) was corrected for any nonhomogeneity of irradiation
across the surface area of the Petri dish to provide a value
for the average incident irradiance. The average irradiance
in the mixed suspension was determined mathematically
depth, accounting for UV absorbance of the G. lamblia test
were computed as the product of average irradiance and
time (in seconds), and required exposure times were calcu-
lated by dividing the desired UV dose by the average UV
irradiance. Note that units for UV dose are presented as J
M-2where 10 J M-2) 1 mJ cm-2. The UV absorbance of the
G. lamblia test suspension ranged from 0.25 to 0.46 cm-1at
UV Disinfection Experiments.Ineachexperiment,5-mL
volumes of phosphate buffered saline (PBS) containing
mL in 60 × 15 mm cell culture Petri dishes were irradiated
using the bench scale UV ªcollimated beamº system while
(23-25 °C). After predetermined exposure times necessary
from the UV irradiation system and diluted serially 10-fold
for subsequent infectivity assays.
DNA Repair Experiments. Experimental samples of G.
as described previously and then wrapped with aluminum
foil immediately after UV exposure. Onesamplewas kept at
4 °C as an experimental control. Replicate samples were
transferred to 25 °C and 37 °C incubators. One dish at each
temperature was illuminated by a 15 W fluorescent lamp at
a distance of 25 cm with slow stirring (light repair) and the
other was only stirred while kept wrapped with aluminum
foil (dark repair). Conditions were 25 °C for 4 h and 37 °C
for 2 h, as commonly used in DNA repair assays of UV-
age (14, 11, 17-19). After incubation, the samples were im-
mediately serially diluted and dosed into animals for
in 8-10-week old female Mongolian gerbils (Meriones
Canada (St. Constant, PQ). At least 10 days before experi-
mental infection, the animals were treated once with a
solution (20 mg/gerbil) of metrodinazole (Flagyl; Rhone
Poulenc, Montreal, PQ) which was administered by gavage.
This treatment ensured that the gerbils were free from all
previous intestinal infections (including Giardia), as dem-
onstrated by three consecutive examinations of feces over
a period of 7 days. Following each experimental treatment,
gerbils were dosed with G. lamblia cysts by gavage. Feces
werecollected daily starting 3daysafterdosing until day 25.
This period of collection was chosen because it represents
the latent, the acute, and the elimination phase of Giardia
infection in gerbils (20). The total number of cysts released
previously (21). The presence of trophozoites in the small
intestine was determined at day 15 after infection in one
gerbil of each group and at the end of the experiment in all
remaining gerbils after they were sacrificed in a CO2 gas
chamber.Thesmall intestinewasslitlongitudinally, divided
on a shaking plate (100 cycles/min), the supernatant was
of trophozoites was determined by examining the sediment
(MPN) as described in the next section.
values based on the presence or absence of cysts and
trophozoitesin individual animalsthathad been inoculated
Results for the numbers of animals positive or negative for
infection out ofthetotal numberofanimalsinoculated with
a given sample dilution (cyst concentration) were used to
estimate the MPN using the Thomas equation. Because the
number of cysts was used for the MPN calculation, the unit
of the MPN is the number of infectious cysts/total number
each experiment) rather than MPN/mL.
Data Presentation. For each test, G. lamblia infectivity
in control samples was computed as an MPN and taken as
N0, the initial concentration. The average infectivity con-
centrations of UV-irradiated samples were similarly com-
puted as MPNs and taken as Nd. The proportions of initial
infectious G. lamblia remaining at each dose (d) were
computed by dividing the concentration at each UV dose
(Nd) by the initial concentration (N0). These values for
infectivity reduction were then transformed to log (Nd/N0)
and plotted against their corresponding UV doses (semilog
plots) for linear regression analysis.
values, and the MPN log reductions of each of the 6 UV
disinfection experiments (trials) performed are shown in
Table 1. In the first and second trials, approximately equal
samples and high UV doses up to 1500 J M-2were tested
because only moderate levels of inactivation of G. lamblia
cysts infectivity by LP UV were expected. As a result, the
detection limit of the gerbil infectivity assay in these trials
was only about 2.5-2.6 log; therefore, the reductions were
already beyond the detection limits, even at the lowest UV
doses tested (20 and 50 J M-2). To better determine the
amount and kinetics of inactivation, both lower numbers of
cysts in untreated control samples (for determining gerbil
infectivity titer endpoint) and lower UV doses (to give less
infectivity reductions) were used in the subsequent trials.
The cysts used in these experiments were highly infectious
such that doses of only 5-10 cysts werecapableof infecting
gerbils, and infectivity reductions of up to ∼5 log could be
followed. Despite this greater limit for detection of UV
infectivity reduction, it was not until an extremely low UV
dose of only 5 J M-2was delivered that some infectious G.
lamblia cysts remained in the UV-irradiated samples.
Thereduction kineticsofG.lamblia infectivityforseveral
different doses of LP UV irradiation applied to cyst suspen-
sions in PBS at pH 7.3 and room temperature are shown in
of the poor detection limits for infectivity reduction in the
first two trials. The reduction of G. lamblia cyst infectivity
for gerbils was extensive at very low doses and reached >4
log within a dose of 10 J M-2in these trials. Inactivation
kineticsappearedtobeapproximately first-orderthrough at
least 4 log (99.99%) reduction, with no ªtailingº (retardant
or decreasing rate) inactivation kinetics. This result is in
contrast to the findings of a previous study reporting a
ªplateauº of inactivation at about 2.5 log reduction for UV-
between the two studies is due to different Giardia species
state of the cysts, the UV dosimetry conditions, the suscep-
tibility and variability of the animals to the infection, the
difference in the suspension waters used, or other factors.
As illustrated in Table 2, as compared to various health-
sensitive to LP UV irradiation than other health-related
waterborne enteric microbes. One possible explanation for
this increased sensitivity, which was also proposed for C.
parvum oocysts (15), is that compared to enteric bacteria
overall sizeand genomesize, thusproviding alargernucleic
acid ªtargetº for UV photons. This target theory may also be
thereason greaterinactivation isachievedatlowerdosesfor
G. lamblia cysts than for C. parvum oocysts, because the
The results of dark and light DNA repair experiments for
UV-irradiated G. lamblia cysts dosed at 160 and 400 J M-2
are presented in Table 3. The reductions of G. lamblia cyst
exposure of the UV-irradiated cysts to either light or dark
TABLE1. Raw Data for Anim al Infectivity Testing andCom putedInactivationfor U VDisinfectionof G. lamblia Cysts
4 10 8/8
aCyst Dose is the number of Giardia cysts dosed into each animal.bInfectivity response is the number of animals with living stages of Giardia
present/total number of animals infected.cControl A ) stirred; control B ) not stirred.
FIGURE 1. Inactivation of G. lamblia by LP UV irradiation in phosphate-buffered saline (pH 7.3) at roomtemperature for trials 3-6. All
data are ªgreater thanº values.
TABLE2. Com parative U VInactivationof Som e H ealth-Related
Bacteria, Viruses, andProtozoans
of either light or dark repair of DNA damage caused by LP
UV irradiation of G. lamblia cysts at these doses causing
high (> 4.2 log) infectivity reductions.
kinetics and extent of inactivation of G. lamblia under UV
assayscoupled with theextremesensitivity oftheGiardia to
UV irradiation, many of the data points were scored as
ªgreater thanº values. Thus, although it was clear that low
extent of inactivation was able to be determined. To attain
these parameters in future studies, it will be necessary to
either use lower UV doses such that complete inactivation
of cysts does not occur, use higher titers for cyst dosing to
increasethelikelihood that an infectiouscystswill bedosed
into ahost animal, orto utilizeasourceofGiardia cyststhat
at low viable cyst doses in control samples.
In conclusion, UV irradiation appearstobevery effective
for inactivation of thehuman pathogen G. lamblia in water,
based on the gerbil animal model for giardiasis. The results
of this study also indicate no phenotypic evidence of either
light or dark repair of UV-damaged DNA in G. lamblia cysts
irradiated with 160 or 400 J M-2of LP UV and exposed to
typical conditions employed for bacterial, eucaryotic, or
mammalian cell DNA repair. The UV doses tested in the
reactivation experiments were typical of the UV doses used
in drinking water and wastewater treatment practices. For
example,thelowdoseof160J M-2istheminimumUV design
municipally treated drinkingwaterdelivered tohouseholds,
systems treating nonmunicipal drinking water and at the
municipal drinking water treatment plant to disinfect water
disinfected drinking water or to inactivate a full range of
pathogens including bacteria and viruses at the municipal
treatment plant will provide adequate disinfection and also
protect against DNA repair and reactivation of G. lamblia
and be activated under different environmental conditions
to lower UV doses may potentially exhibit DNA repair
capability because of the lesser levels of UV-induced DNA
damage.Therefore,althoughUV dosesmuchlowerthan 160
only these higher UV doses have been experimentally
demonstrated here to protect against reactivation of cyst
typic studies are recommended to further investigate the
presence of DNA repair activities in G. lamblia cysts under
a variety of UV irradiation and subsequent DNA repair
shown to bevery sensitiveto low dosesofLP UV irradiation,
properly designed and operated LP UV disinfection systems
provide evidence of G. lamblia control at current UV design
doses that are compatible with the intended use of UV to
inactivate a broader range of target organisms.
This research was supported by funds from Trojan Tech-
nologies Inc., London, Ontario, Canada.
(1) Craun,G.F.WaterborneDiseasesintheUnitedStates.CRC Press:
Boca Raton, FL, 1986; p 295.
(2) Sykora, J. L.; Sorber, C. A.; Jakubowski, W.; Casson, L. W.;
Gavaghan, P. D.; Sapiro, M. A.; Schott, M. J. Water Sci. Technol.
1991, 24, 187-192.
(3) LeChevallier, M. W.; Norton, W. D. J. AWWA 1995, 87, 54-68.
(4) Gibson, C. J., III; Hass, C. N.; Rose, J. B. Parasitology 1998, 117
(5) Sobsey, M. D. Water Sci. Technol. 1989, 21, 179-195.
(6) Clark,R.M.;Hurst,C.J.;Regli,S.In SafetyofWaterDisinfection:
Balancing Chemical and Microbial Risks; Craun, G. F. Ed.;
International Life Sciences Press: Washington, DC, 1993; pp
(7) Rice, E. W.; Hoff, J. C. Appl. Environ. Microbiol. 1981, 42, 546-
(8) Craik, S. A.; Finch, G. R.; Bolton, J. R.; Belosevic, M. Water Res.
2000, 34, 4325.
(9) Friedberg, E. C.; Walker, G. C.; Siede, W. DNA Repair and
Mutagenesis; ASM Press: Washington, DC, 1995.
(10) Whitby, G. E.; Palmateer, G.; Cook, W. G.; Maarschalkerweerd,
J.; Huber, D.; Flood, K. J. Water Pollut. Control Fed. 1984, 56,
(11) Harris, G. D.; Adams, V. D.; Sorensen, D. L.; Curtis, M. S. Water
Res. 1987, 21, 687-692.
(12) Lindenauer, K. G.; Darby, J. L. Water Res. 1994, 28, 805.
(13) U.S. EPA Design Manual: Municipal Wastewater Disinfection;
EPA/625/1-81/021; U.S. Environmental Protection Agency:
Cincinnati, OH, 1986.
(14) Das, G.; Kaveri, S.; Das. J. 1981. Biochim. Biophys. Acta. 1981,
(15) Shin, G.; Linden, K. G.; Arrowood, M. J.; Sobsey, M. D. Appl.
Environ. Microbiol. 2001, 67, 3029.
(16) Morowitz, H. J. Science. 1950, 111, 229.
(17) Patrick, M. H.; Haynes, R. H. Radiat. Res. 1964, 23, 564-579.
(18) Resnick, M. A.; Setlow, J. K. J. Bacteriol. 1972, 109, 979-986.
(19) Sutherland, B. M.; Cimino, J. S.; Delihas, N.; Shih, A. G.; Oliver,
R. P. Cancer Res. 1980, 40, 1934-1939.
(20) Belosevic, M.; Faubert, G. M.; MacLean, J. D.; Law, C.; Croll, N.
A. J. Infect. Dis. 1983, 147, 222-226.
(21) Belosevic, M.; Faubert, G. M. Exp. Parasitol. 1983, 56, 93-100.
C. M.; Qualls, R. G.; Johnson, J. D. Appl. Environ. Microbiol.
1985, 49, 1361-1365.
(23) Wilson, B. R.; Roessler, P. F.; Van Dellen, E.; Abbaszadegan, M.;
Gerba, C. P. Proceedings AWWA, Water Quality Technology
Conference, Toronto, Canada, Nov 15-19, 1992, 219-235.
(24) Meng, Q. S.; Gerba, C. P. Water Res. 1996, 30, 2665-2668.
(25) Snicer, G. A.; Malley, J. P.; Margolin, A. B.; Hogan, S. P. UV
Inactivation of Viruses in Natural Waters; AWWA Research
Foundation & AWWA: Denver, CO, 1998.
(26) Nieuwstad,T.J.;Havalaar,A.H.J.Environ.Sci.Health 1994,29,
Received for review October 4, 2001. Revised manuscript
received March 25, 2002. Accepted March 26, 2002.
TABLE3. Effect of Dark andLight Repair Conditions on
Infectivity of G. lamblia Cysts Exposedtoa DeliveredDose
of 160 and400 J M-2of LP U V
log G. lamblia reduction
37 °C, 2 h
25 °C, 4 h
37 °C, 2 h
25 °C, 4 h