Experimental Parasitology 127 (2011) 490–499

Contents lists available at ScienceDirect

Experimental Parasitology

journal homepage: www.elsevier .com/locate /yexpr

Effect of garlic and allium-derived products on the growth and metabolismof Spironucleus vortens

Coralie O.M. Millet a,⇑, David Lloyd a, Catrin Williams a, David Williams b, Gareth Evans b,Robert A. Saunders b, Joanne Cable a

a School of Biosciences, Cardiff University, Cardiff CF10 3AX, UKb NEEM Biotech Limited, Unit 1, Willowbrook Technical Units, Llandogo Road, St. Mellons, Cardiff CF3 0EF, UK

a r t i c l e i n f o

Article history:Received 26 May 2010Received in revised form 27 September2010Accepted 11 October 2010Available online 20 November 2010

Keywords:GarlicAllicinAjoeneDithiinsDiplomonadSpironucleus

0014-4894/$ - see front matter � 2010 Elsevier Inc. Adoi:10.1016/j.exppara.2010.10.001

⇑ Corresponding author. Address: Main Building, SUniversity, P.O. Box 915, Cardiff CF10 3AX, UK.

E-mail addresses: MilletCO@gmail.com, milletCO@

a b s t r a c t

Spironucleus is a genus of small, flagellated parasites, many of which can infect a wide range of verte-brates and are a significant problem in aquaculture. Following the ban on the use of metronidazole infood fish due to toxicity problems, no satisfactory chemotherapies for the treatment of spironucleosisare currently available. Using membrane inlet mass spectrometry and automated optical density moni-toring of growth, we investigated in vitro the effect of Allium sativum (garlic), a herbal remedy knownfor its antimicrobial properties, on the growth and metabolism of Spironucleus vortens, a parasite of trop-ical fish and putative agent of hole-in-the-head disease. The allium-derived thiosulfinate compounds alli-cin and ajoene, as well as an ajoene-free mixture of thiosulfinates and vinyl-dithiins were also tested.Whole, freeze-dried garlic and allium-derived compounds had an inhibitory effect on gas metabolism,exponential growth rate and final growth yield of S. vortens in Keister’s modified, TY-I-S33 culture med-ium. Of all the allium-derived compounds tested, the ajoene-free mixture of dithiins and thiosulfinateswas the most effective with a minimum inhibitory concentration (MIC) of 107 lg ml�1 and an inhibitoryconcentration at 50% (IC50%) of 58 lg ml�1. It was followed by ajoene (MIC = 83 lg ml�1, IC50% =56 lg ml�1) and raw garlic (MIC >20 mg ml�1, IC50% = 7.9 mg ml�1); allicin being significantly less potentwith an MIC and IC50% above 160 lg ml�1. All these concentrations are much higher than those reportedto be required for the inhibition of most bacteria, protozoa and fungi previously investigated, indicatingan unusual level of tolerance for allium-derived products in S. vortens. However, chemically synthesizedderivatives of garlic constituents might prove a useful avenue for future research.

� 2010 Elsevier Inc. All rights reserved.

1. Introduction

Aquaculture is the fastest growing animal food production sec-tor, and represented a $63.3 billion global market in 2004 (FAO,2007). Stress and overcrowding of farmed fish render them partic-ularly susceptible to infectious diseases (Wedemeyer, 1997;Pickering, 1998), and pathogens are recognised to be the maincause of financial losses in aquaculture (Meyer, 1991; Subasingheet al., 2000). For instance, in Asia alone, diseases in aquaculturehave been estimated to cause monetary losses of over $3 billionin 1995 (Chua, 1996).

The genus Spironucleus includes several species of parasiticanaerobic protozoan flagellates, some of which infect importantspecies of farmed fish and have a crippling effect on their aqua-culture. The most notorious is Spironucleus salmonicida (Jorgensenand Sterud, 2006), which is responsible for devastating outbreaks

ll rights reserved.

chool of Biosciences, Cardiff

cf.ac.uk (C.O.M. Millet).

in salmon farms (Sterud et al., 1998) and may generate up to100% mortality (Guo and Woo, 2004a,b). Another species, Spiro-nucleus vortens causes significant losses in the ornamental fishindustry. Indeed, although a lineage of S. vortens has been iso-lated from wild ide in Norway (Sterud and Poynton, 2002), thespecies is mostly known to infect tropical fish, and is thesuspected causative agent of hole-in-the-head disease, a verycommon affliction of ornamental cichlids (Paull and Matthews,2001).

Although antimicrobial compounds have been employed heav-ily in aquaculture, chemotherapy of infectious diseases in thisindustry is generally difficult, as the effectiveness of antimicrobi-als tends to be limited in aquatic environments, especially inopen, coastal systems (GESAMP, 1997; Treves-Brown, 1999).Metronidazole, a nitroheterocyclic antibiotic, is the drug of choicein the treatment of a range of anaerobic infections including spi-ronucleosis (Tojo and Santamarina, 1998; Sangmaneedet andSmith, 1999). Although it is currently one of the most widely useddrugs in the world, both in medicine and veterinary practice,metronidazole was shown to be genotoxic, mutagenic and

C.O.M. Millet et al. / Experimental Parasitology 127 (2011) 490–499 491

carcinogenic in rats (Huet et al., 2005; Maher et al., 2008), as wellas mutagenic and cytotoxic in fish (Caves and Ergere Gozukara,2005). This prompted its ban in 1998 from use in aquacultureof European food fish by the Council Regulation 613/98/EEC(L82/14, 1998).

Metronidazole is still employed in the aquarium trade, but itsuse is controversial, as dispersal of this drug on a wide-scale withinoutdoor, ornamental fish farms, typically within food pellets, raisesserious environmental and health concerns. Indeed, this drug isquite water soluble, as well as non-biodegradable, and can there-fore resist water and sewage treatments (Richardson and Bowron,1985), which may result in long-lasting contamination of the aqua-tic environment. This is likely to have a devastating impact on thenatural anaerobic microbial flora (Lanzky and Halling-Sorensen,1997) and may also encourage drug resistance (Chua, 1996; GES-AMP, 1997; Shehadeh and Maclean, 1997; Treves-Brown, 1999).Such severe limitations in the use of metronidazole highlight theneed for alternative Spironucleus treatments, in both food andornamental fish.

Garlic Allium sativum has been heavily employed as an antimi-crobial in medical folklore throughout history (Whitmore andNaidu, 2000). As antibiotic resistance is becoming an increasinglyalarming issue, garlic thus generates an enormous amount ofinterest. It was shown to inhibit many pathogenic Gram-negativeand Gram-positive bacteria, including some vancomycin-resistantenterococci and methicillin resistant Staphylococcus aureus(Ankri and Mirelman, 1999; Ross et al., 2001; Tsao et al., 2003;Iwalokun et al., 2004). Garlic also demonstrates strong antifungalproperties against pathogenic species, such as Candida, Aspergillusand Cryptococcus (Pai and Platt, 1995; Shen et al., 1996; Davis,2005), and proved extremely potent against many human viruses,even destroying HIV-infected cells in vitro (Tsai et al., 1985;Tatarintsev et al., 1992). It also exhibits antimicrobial activityagainst many parasitic protists, including Trypanosoma,Leishmania, Leptomonas, Trichomonas and Entamoeba (Mirelmanet al., 1987; Reuter et al., 1996; Ankri et al., 1997), and was foundeffective in the treatment of giardiasis (Soffar and Mokhtar, 1991;Harris et al., 2000).

Many studies have attempted to breach the basis of garlic’santimicrobial properties, however, the question is unexpectedlychallenging, as the plant has a complex chemistry and containshundreds of compounds (Whitmore and Naidu, 2000), some ofwhich are extremely unstable (Freeman and Kodera, 1995;Fujisawa et al., 2008a,b) and may act synergistically (Chunget al., 2007). Diallyl thiosulfinate or allicin, an oxygenated sulphurcompound extracted from steamed garlic, was originally assumedto be the main antimicrobial in garlic (Cavallito and Bailey, 1944).However, later studies demonstrated allicin-free garlic extractsretain activity against microbes (Fujisawa et al., 2008a,b) andseveral other compounds in garlic seem to also act as potentantimicrobials. Amongst these, the sulphonated compoundvinyl-dithiins and ajoene, a product of the mild oxidation ofallicin, have attracted the most interest, as they seem to generatea powerful, broad antimicrobial action, at concentrationssometimes lower than that of allicin (Yoshida et al., 1987; Biereret al., 1995; Naganawa et al., 1996; Whitmore and Naidu, 2000;Ledezma et al., 2002).

Garlic, allicin and other allium-derived compounds have proveneffective against several anaerobic protozoan flagellates, howevertheir usefulness in the treatment of S. vortens has not been inves-tigated and susceptibility of the organism to these compounds isunknown. In this study, membrane inlet mass spectrometry andautomated optical density monitoring was used to investigatethe effect of whole garlic, allicin, ajoene and dithiin on the gasmetabolism and growth of S. vortens in vitro, in comparison tometronidazole.

2. Materials and methods

2.1. Organisms and cultures

Spironucleus vortens, ATCC 50386, was obtained from ProfessorJ. Kulda, Charles University, Prague. The pathogen was first isolatedfrom a lip lesion in a freshwater angelfish in Florida in 1991(Poynton et al., 1995). Trophozoites were cultured axenically at20 ± 0.5 �C in 15 ml screw-capped Falcon tubes on Keister’smodified TYI-S-33 medium containing (per liter): casein digest(BBL), 20 g; yeast extract (Oxoid), 10 g; glucose (Sigma), 10 g;bovine bile (local slaughter house), 1 ml; NaCl (Merck), 2 g;L-cysteine. HCl (Sigma), 2 g; ascorbic acid sodium salt (Fluka),0.2 g; K2HPO4 (Merck), 1 g; KH2PO4 (Merck), 0.6 g; ferric ammo-nium citrate (Sigma), 22.8 mg, heat-inactivated newborn calfserum (Difco), 100 ml. The pH was adjusted to 6.8 with NaOH priorto filter sterilization (0.22 lm pore size). Subculturing was per-formed routinely at 72 h intervals. Cells were counted on a FuchsRosenthal haemocytometer slide, using 0.4% (v/v) Trypan blue asa viability indicator. Motility was also used to assess viability. Cul-tures were monitored routinely to confirm axenic growth. Expo-nentially growing cultures were harvested by centrifugation at800g for 2 min at room temperature in a bench centrifuge (MSEminor), washed twice in phosphate buffer (0.1 M, pH 6.8) andresuspended in PBS (0.1 M, pH 6.8) containing 10 mM glucose.

2.2. Inhibitors

Freeze-dried powder of whole Chinese garlic was obtained fromCultech Ltd. and kept in a sealed bag at �30 �C. HPLC analysis ofthis garlic by NEEM Biotech Ltd. revealed that the powder con-tained 0.223 (w/w)% allicin, but no detectable amounts of ajoene.Stock solutions (20 mg ml�1) were prepared by dissolving thepowder in PBS (0.1 M, pH 6.8) or culture medium, vortex mixingfor 10 min and centrifuging at 3000g for 20 min. Solutions werethen filtered through 0.2 lm porosity syringe filters, diluted as re-quired in culture medium and used immediately. Metronidazole(Sigma) was dissolved in acidified buffer (final concentration50 mM), slowly brought back to pH 6.8 with NaOH, filteredthrough a 0.2 lm porosity syringe filter, diluted as required andused immediately.

Allicin was produced by NEEM Biotech Ltd. according to the USpatent 7179632 and HPLC analysis established its purity at 99% (w/w). Ajoene and dithiin were also prepared by NEEM Biotech Ltd.according to the methods described by Block et al. (1986). Ajoenewas produced by mild oxidation of a solution of pure allicin.Briefly, a solution of allicin was refluxed as a 10% solution in 3:2acetone/water for 4 h, diluted with methanol and repeatedly ex-tracted with pentane to remove non-polar materials and poly-sulphinates: the mixture was then extracted with methylenechloride and concentrated. The resulting oily extract of ajoenewas a mixture of diastereoisomers with a ratio of 3:1 E:Z. andwas established to be 75% (w/w) pure by HPLC. HPLC revealed thatthe remaining 25% (w/w) contained 12 breakdown products of alli-cin, including dithiins. The respective proportion of these com-pounds was not determined. Dithiins were concentrated fromraw garlic as described below. Chopped garlic pieces were soakedin methanol for 3 days. The concentrate was suspended in waterand extracted with diethyl ether. The diethyl ether extract wasconcentrated, and the residue was stored at 25 �C for 4 days as a10% solution in methanol, suspended in water and repeatedly ex-tracted with hexane to separate the less polar compounds fromthe more polar ones. The aqueous methanolic layer was then ex-tracted with methylene chloride. Concentration of the hexaneand methylene chloride extracts separately afforded yellow oils.

492 C.O.M. Millet et al. / Experimental Parasitology 127 (2011) 490–499

The major non-polar components were thereby identified as diallyldisulfide, diallyl trisulfide, diallyl tetrasulfide, allyl methyl trisul-fide, 2-vinyl-4H-1,3-dithiin, 3-vinyl-4H-1,2-dithiin and allicin.Concentration of dithiins in the resulting oil was 32% (w/w) byHPLC. Allicin, ajoene and dithiin were diluted in DMSO as requiredand used immediately.

2.3. Membrane inlet mass spectrometry

The effect of various concentrations of garlic, allicin and metro-nidazole on gas metabolism by S. vortens was investigated bymembrane inlet mass spectrometry by measuring levels of dis-solved O2, CO2 and H2 in continuously stirred cell suspensions oflog-phase trophozoites. Full methodology is described in Lloydet al. (1992). Briefly, dissolved gases were measured using a Halseries quadrupole mass spectrometer (Hiden Analytical) fittedwith a 10 cm, 1.6 mm dia. stainless steel inlet probe covered by agas-permeable polymer membrane (1.56 mm outside dia.,0.5 mm internal dia., 5 cm length) sealed into a 1 cm length ofquartz-glass tube. The inlet orifice was 100 lm dia., and the mem-brane used was silicone rubber catheter tube. The probe was usedto measure concentrations of dissolved gases within a 5 ml cellsuspension maintained at 20 �C in a closed Perspex vessel withmagnetic stirring at 200 rpm. Mass to charge ratios (m/z) used tomeasure the concentrations of H2, ethanol, O2 and CO2 were 2,31, 32 and 44, respectively. Solubilities of H2 and O2 at 20 �C were810 and 1390 lM, respectively (Wilhelm et al., 1977). CO2 calibra-tions employed additions of 3.3 mM NaHCO3. At 20 �C and pH 6.9,total free CO2 (as opposed to HCO3

� and CO32� which are not de-

tected by the mass spectrometer) was 25% (Wilhelm et al., 1977).Ethanol is known to give a signal on the m/z = 2 channel, at whichH2 is normally detected. To avoid non-specific detection of ethanolon this channel, both ethanol and H2 calibrations were performedat m/z = 2. Ethanol calibrations employed stepwise additions of2 mM C2H5OH. Under the conditions employed here, addition of6 mM ethanol did not cause any increase on the m/z = 2 channel,thus ensuring that this channel is specific for the detection of H2.Log-phase trophozoites were harvested by centrifugation as de-scribed above, washed thoroughly in PBS (�3), resuspended in5 ml PBS with 10 mM added glucose, and introduced into the reac-tion vessel. Inhibitors were added into the reaction vessel at thestart of each experiment using a Hamilton syringe. Experimentswere replicated 3–6 times. In order to determine whether a givenconcentration of an inhibitor had an significant effect on gas pro-duction, the data obtained from all replicates of individual concen-trations tested was pooled and one-way analysis of variance(ANOVA), as well as Fisher a priori analysis were used to detect sta-tistically significant differences between individual concentrationsat a 95% confidence level (Minitab 13).

2.4. Growth in Bioscreen C

Bioscreen C (Labsystems, Finland), an automated optical den-sity-monitoring system, was used to monitor growth of S. vortensin 100 well honeycomb plates, in the presence of garlic, garlic de-rived compounds and metronidazole. Inhibitors were diluted inbuffer, culture medium or DMSO (see above), added to 340 ll ofKeister’s modified TY-I-S33 culture medium, and 10 ll of homoge-nised log-phased cultures and introduced into triplicate wells. Inorder to avoid optical density differences between wells in theBioscreen experiments, garlic solutions were dissolved in culturemedium and all dilutions of allicin, ajoene, dithiin and metronida-zole were prepared so that the same volume of buffer or DMSO(0.35 ll) was introduced into all wells. For garlic and metronidazoleexperiments, triplicate control wells were filled with 340 ll ofmedium and 10 ll inoculate. As allicin, ajoene and dithiin solutions

contained DMSO, separate controls containing the same amount ofDMSO were set up, and three control wells were filled with 339.7 llculture medium, 0.34 ll DMSO and 10 ll inoculate. Turbiditywithin wells was measured by OD on a vertical light path, using awideband filter (420–580 nm) every 20 min for 120 h. Heat transferfluid circulation maintained a constant temperature of 25 �C, thetemperature reported as optimal for the growth of this organism(see Sangmaneedet and Smith, 2000). Plates were shaken at lowamplitude for 5 s before each reading. OD measurements werelogged, plotted against time, and exponential growth rates (l), dou-bling times (Td) and final yields were extrapolated from growthcurves by applying the following formula:

l ¼ 2:303ðlog10ODt2 � log10ODt1Þ=ðt2 � t1Þ

Td ¼ ln 2=l

yield ¼ ODtf � ODt0

With ODt0 the optical density at the start of the experiment, andODtf the optical density after 100 h of growth.

In order to determine whether a given concentration of aninhibitor had an effect on exponential growth rates or final growthyield, the data obtained from all replicates of individual concentra-tions tested was pooled and one-way analysis of variance (ANO-VA), as well as Fisher a priori analysis were used to detectstatistically significant differences between individual concentra-tions at a 95% confidence level.

The lowest concentration of the compound to completely inhi-bit growth, or minimum inhibitory concentration (MIC) and theconcentration at which the growth was reduced by 50% (inhibitoryconcentration at 50% or IC50%) were derived from fitted curves ofplots of total yield against inhibitor concentrations. Complete inhi-bition of growth by a compound was achieved when growth ratewas null.

3. Results

Of all the allium-derived compounds tested, the ajoene-free mix-ture of dithiins and thiosulfinates was the most effective, followedby ajoene, freeze-dried Chinese garlic and allicin (see Table 1).

3.1. Effect of freeze-dried Chinese garlic, allicin and metronidazole onthe gas metabolism of S. vortens

When introduced into air saturated buffer, S. vortens trophozo-ites rapidly consume O2 and produce CO2. Once microaerobic condi-tions are reached, H2 started being produced (see Millet et al., 2010).Freeze-dried Chinese garlic significantly inhibited O2 consumption,CO2 production and H2 production of S. vortens at concentrations of0.6 mg ml�1 and above (one-way ANOVA and Fisher a priori analy-ses of pooled data at a 95% confidence level; Fig. 1A). Allicin concen-trations of up to 50 lg ml�1 did not affect gas metabolismsignificantly. At 100 lg ml�1, allicin significantly reduced rates ofO2 consumption and H2 production but CO2 production was unaf-fected (Fig. 1B). Metronidazole inhibited H2 production at concen-trations of 85.6 lg ml�1 and above, and CO2 production atconcentrations of 171.2 lg ml�1 and above. O2 consumption wasnot affected by up to 342.4 lg ml�1 metronidazole (Fig. 1C).

3.2. Effect of freeze-dried Chinese garlic, allium-derived compoundsand metronidazole on the growth rates and total yields of S. vortens inTY-I-S33 culture medium

In the absence of an inhibitor, S. vortens exhibited a biphasicgrowth curve (see Fig. 2), and growth could therefore be characterised

Table 1Comparison of the action of metronidazole, garlic and allium-derived compounds on the growth of Spironucleus vortens. MIC = minimum inhibitory concentration,IC50% = concentration at which the growth was reduced by 50%.

Inhibitor MIC(lg ml�1)

IC50%

(lg ml�1)Significantly reduces growth ratefrom (lg ml�1)

Significantly reduces final growth yieldfrom (lg ml�1)

Freeze dried garlic >20,000 7900 1000 10,000Allicin >160 >160 10 40Ajoene (75%) and allicin (25%) 107 58 10 80Dithiins (32%) and ajoene-free mixture of

thiosulfinates83 56 5 40

Metronidazole <8.56 <8.56 <8.56 <8.56

Fig. 1. Influence of (A) freeze-dried powder of whole Chinese garlic, (B) allicin and (C) metronidazole on gas metabolism of Spironucleus vortens. Whole garlic concentrationsare indicated on the lower x axis of (A) and allicin content of that whole garlic on the upper x axis of (A). Washed organisms were resuspended in air-saturated PBS containingthe inhibitor and 10 mM glucose, and introduced into a closed reaction vessel at 20 �C. The mass spectrometer was programmed to scan mass spectra repeatedly, collecting128 spectra, before presenting data quasi-continuously at m/z values 2, 32 and 44, corresponding to peak values for H2, O2 and CO2, respectively. Rates of O2 consumption, CO2

production and H2 production were derived by linear regression analyses, and lines were fitted between data points using Microsoft Excel. Experiments were replicated 3–6times at each concentration tested and standard errors are plotted around the means.

C.O.M. Millet et al. / Experimental Parasitology 127 (2011) 490–499 493

by the following parameters: exponential growth rate in the first logphase (l1), exponential growth rate in the second log phase (l2), andthe total yield after 100 h of growth (see Millet et al., in press). Addi-tion of inhibitors affects growth rates and yield, as exemplified inFig. 2, which shows the effect of selected ajoene concentrations onS. vortens growth curves. The effects of freeze-dried Chinese garlic, al-lium-derived inhibitors and metronidazole on the exponentialgrowth rates (l1 and l2) of S. vortens are presented in Fig. 3. MICsand IC50% were derived from the influence of inhibitors on total yieldafter 100 h of growth, as shown in Fig. 4. Both growth rates were sig-

nificantly reduced in the presence of garlic by 1 mg ml�1 (see Fig. 3A)and total yield was lowered by 10 mg ml�1 garlic (one-way ANOVAand Fisher a priori analyses, see Fig. 4A). The lowest concentrationof garlic to completely inhibit S. vortens growth (MIC) was not deter-mined in the current study, as it was above the highest concentrationtested (20 mg ml�1, see Fig. 4A). An IC50% of 7.9 mg ml�1 is derivedfrom the fitted curve of total yields (Fig. 4A).

Exponential growth rates were significantly decreased at allicinconcentrations of 10 lg ml�1 and above (see Fig. 3B), and totalyield was lowered from 40 lg ml�1 (see Fig. 4B). Fig. 4B shows that

Fig. 2. Representative example of the output of automated optical density monitoring (Bioscreen C) of Spironucleus vortens cultures upon addition of inhibitors. In TY-I-S33culture medium, S. vortens exhibits a biphasic growth; the first (l1) and second (l2) exponential growth rates are indicated on the control growth curve. In order to illustratehow addition of inhibiting compounds affects growth curves, growth rates of S. vortens upon addition of increasing ajoene concentrations are plotted. Triplicate samples wereanalysed for six ajoene concentrations ranging from 0 to 120 lg ml�1, however in the interest of clarity, only a single replicate for three selected ajoene concentrations testedis plotted.

494 C.O.M. Millet et al. / Experimental Parasitology 127 (2011) 490–499

both MIC and IC50% were above the highest concentration tested(160 lg ml�1).

The crude extract of ajoene (75% ajoene, 25% allicin and allicinbreakdown products) significantly reduced l1 from 5 lg ml�1, whichcorresponds to 3.75 lg ml�1 of pure ajoene. l2 was lowered at con-centrations of 10 lg ml�1 of crude oil (equivalent to 7.50 lg ml�1 ofpure ajoene) and above (see Fig. 3C), and total yield was significantlyreduced from 80 lg ml�1 (equivalent to 60 lg ml�1 of pure ajoene,see Fig. 4C). An MIC of 107 lg ml�1 of crude ajoene extract was de-rived from the fitted curve of total yields (see Fig. 4C). This is equiva-lent to 80 lg ml�1 of pure ajoene. The IC50% of crude ajoene extractderived from the fitted curve of total yields was 58 lg ml�1

(Fig. 4C). This is equivalent to 43.5 lg ml�1 of pure ajoene.The crude extract of dithiins (32% dithiins, 68% mixed thiosulf-

inates) significantly reduced both exponential growth rates from5 lg ml�1 (equivalent to 1.60 lg ml�1 of pure dithiins, seeFig. 3D), and lowered total yield at concentrations of 40 lg ml�1

and above (equivalent to 12.8 lg ml�1 of pure dithiins, seeFig. 4D). An MIC of 83 lg ml�1 of crude dithiins extract was derivedfrom the fitted curve of total yields (Fig. 4D). This is equivalent to27 lg ml�1 of pure dithiin. The IC50% of crude dithiin extract de-rived from the fitted curve of Fig. 4D was 56 lg ml�1, which isequivalent to 18 lg ml�1 of pure dithiin.

Exponential growth rates (see Fig. 3E) and total yields (seeFig. 4E) of S. vortens were significantly reduced upon addition of8.56 lg ml�1 metronidazole and above. Fig. 4E shows that bothMIC and IC50% for metronidazole were under 8.56 lg ml�1, the low-est concentration tested.

4. Discussion

This study shows that garlic and allium-derived compoundshave an inhibitory effect on S. vortens. Of the all allium-derivedcompounds tested, the ajoene-free mixture of dithiins andthiosulfinates was the most effective, followed by ajoene, rawgarlic and allicin, however the concentrations of garlic and

allium-derived compounds required to inhibit growth of the para-sites are much higher than that of metronidazole (see Table 1).

Garlic slows gas metabolism and growth of S. vortens at concen-trations above 500 lg ml�1. However, extremely high concentra-tions (10 mg ml�1 and above) are required to significantly affectthe final growth yield after 80 h, suggesting that the lower concen-trations of the compound only have a transient, static effect on S.vortens. IC50% was 7.85 mg ml�1, but the MIC could not be deter-mined, as the highest concentration tested (20 mg ml�1) did notcompletely inhibit growth after 80 h. Much lower concentrationsof garlic have been shown to completely inhibit growth of thedistantly related diplomonad parasite, Giardia duodenalis(MIC = 0.30 mg ml�1; Harris et al., 2000, 2001) and administrationof twice daily doses of 1 mg ml�1 aqueous garlic for 72 h wasshown to cure giardiasis and completely eliminate cysts fromstools in humans (Soffar and Mokhtar, 1991). Although there is asomehow large genetic distance between Giardia and Spironucleus(Jorgensen and Sterud, 2007), such a strong difference in suscepti-bility to garlic remains surprising. Variation in experimental designfor MIC determination may also have contributed to the stark dif-ference in MIC values between Spironucleus and Giardia, as MIC val-ues in Giardia were based on cell density after 24 h of growth(Harris et al., 2000, 2001), while in the current study, they werebased on optical density of cultures after 100 h of growth. As someof the many volatile compounds in garlic are known to be quiteunstable in aqueous solutions (Freeman and Kodera, 1995;Fujisawa et al., 2008a), the length of growth incubation may indeedhave had a strong influence on the values for MIC. Nonetheless,these results show that the use of a single dose of concentratedaqueous garlic would not be efficient in the treatment of spironu-cleosis, and the effect of multiple doses over several subculturesshould be investigated to assess potential usefulness of this plantin managing the parasite.

Allicin, believed to be the main antimicrobial in garlic (Cavallitoand Bailey, 1944; Whitmore and Naidu, 2000; Hunter et al., 2005),showed surprisingly little inhibitory effect on gas metabolism of S.vortens, and only inhibited O2 consumption and H2 production at

C.O.M. Millet et al. / Experimental Parasitology 127 (2011) 490–499 495

100 lg ml�1. Even at this high concentration, CO2 production wasnot significantly reduced. Growth rates were reduced from con-trols at concentrations equal or exceeding 10 lg ml�1. However,as with garlic, much higher concentrations (40 lg ml�1 and above)were required to significantly reduce the final yield, which sug-

Fig. 3. Influence of (A) freeze-dried powder of whole Chinese garlic, (B) allicin, (C) ajoenevortens. Exponential growth rates (l1 and l2) during the biphasic growth of S. vortensaddition of inhibitors. Whole garlic concentrations are indicated on the lower x axis oconcentrations from the ajoene (B) and dithiin (C) extracts are plotted on the upper x axderived by linear regression analyses. Experiments were replicated three times at each

gests that at low concentrations, the action of the compound is alsomainly transient and microbiostatic, rather than truly microbiocid-al. MIC and IC50% could not be determined, as the highest concen-tration of allicin tested (160 lg ml�1) did not completely inhibitgrowth or even reduce it by 50%. Again, this finding is rather sur-

, (D) dithiins and (E) metronidazole on the exponential growth rates of Spironucleusin Keister’s modified TY-I-S33 culture medium were recorded by Bioscreen C uponf (A) and allicin content of that whole garlic on the upper x axis of (A). Crude oilis and pure compound content on the lower x axis. Exponential growth rates were

concentration tested and standard errors are plotted around the means.

496 C.O.M. Millet et al. / Experimental Parasitology 127 (2011) 490–499

prising, as most bacteria investigated for sensitivity to allicin haveIC50% ranging from 3 to 15 lg ml�1 (Gupta and Ravishankar, 2005),and antifungal activity is generally displayed between 0.3 and32 lg ml�1 (Ankri et al., 1997; Yamada and Azuma, 1997; Davis,2005; Fry et al., 2005). So far, only mucoid strains of some bacteriawere found to have IC50% above 100 lg ml�1 (Uchida et al., 1975;Ruddock et al., 2005), and such high level resistance is probablydue to the physical protection conferred by the mucus S. vortens.ATCC was isolated from a diseased angelfish in 1991 (Poynton

Fig. 4. Influence of (A) freeze-dried powder of whole Chinese garlic, (B) allicin, (C) ajoenein Keister’s modified TY-I-S33 culture medium. Whole garlic concentrations are indicatedof (A). Crude oil concentrations from the ajoene (B) and dithiin (C) extracts are plotted onafter 100 h (recorded by Bioscreen C) is plotted against inhibitor concentration, and a fi(MIC) and the concentration at which 50% of growth is inhibited (IC50%). Experiments weraround the means.

et al., 1995), and maintained in culture thereafter. Although theauthors used antibiotics on the cultures only sporadically, priorto its arrival in the laboratory, this strain was routinely subculturedwith amikacin (250 lg ml�1) and penicillin G (1000 U ml�1), whichmay have contributed to select for organisms with high toleranceto antimicrobials. Indeed, although these drugs are not designedto target eukaryotic cells, they are far from innocuous (Cunha,2001; Van Bambeke et al., 2003), and may therefore promote theappearance or over-expression of traits that could interfere with

, (D) dithiins and (E) metronidazole on the total growth yield of Spironucleus vortenson the lower x axis of (A) and allicin content of that whole garlic on the upper x axisthe upper x axis and pure compound content on the lower x axis. Total yield by OD

tted curve is used to derive the lowest concentration to completely inhibit growthe replicated three times at each concentration tested and standard errors are plotted

C.O.M. Millet et al. / Experimental Parasitology 127 (2011) 490–499 497

the action of allium-derived compounds. For instance decreasedpermeability of the membrane or increased efficiency of generalefflux mechanisms could lead to a lowered efficiency of allium-derived compounds.

It should be noted that 7.85 mg of garlic (the IC50% for this com-pound), contains 18 lg of pure allicin, however, the IC50% of pureallicin was higher than 160 lg ml�1. Moreover, the concentrationof allicin ranged from 0 to 100 lg ml�1 in the mass spectrometricanalysis of gas metabolism, and from 0 to 160 lg ml�1 in the auto-mated optical density monitoring of growth. These allicin concen-trations are much higher than that contained in the whole garlicpowder experiments (0–13.38 lg ml�1 for the mass spectrometry,and 0–44.60 lg ml�1 for the growth monitoring, see Figs. 1 and 3),and yet they had considerably less inhibitory effect on gas metab-olism, growth rates and final yields. This suggests that despite thelong-standing belief that allicin is the main antimicrobial com-pound in garlic, the observed inhibitory properties of garlic on S.vortens are not due to allicin, or that allicin is only efficient whenacting synergistically with other garlic components.

Ajoene crude oil (75% pure, with 25% allicin and allicin break-down products) was more efficient than allicin, as it reducedgrowth rates of S. vortens from 5 to 10 lg ml�1, equivalent to3.75–7.5 lg ml�1 of pure ajoene oil. As for whole garlic and allicin,significant inhibition of growth yield after 80 h only occurred atmuch higher concentrations (80 lg ml�1 of crude oil, which corre-sponds to 60 lg ml�1 of pure ajoene). The MIC and IC50% wererespectively 107 and 58 lg ml�1 of crude ajoene extract. This cor-responds to 80 and 44 lg ml�1 of pure ajoene (Fig. 4C) These valuesare much lower than that of either whole garlic or allicin, andunless extraordinarily powerful synergistic effects are at play, theallicin content of that crude oil (25%), could hardly be deemedresponsible for this stronger antimicrobial effect, as the muchhigher concentrations of pure allicin tested (Figs. 3B and 4B) gen-erated far less inhibition in S. vortens. However, if ajoene appearsmore efficient at inhibiting growth of S. vortens than either allicinor whole garlic, the values obtained for MIC of the compound arestill considerably higher than that found in other parasitic protozoaand yeasts; 2.34 lg ml�1 for Leishmania (see Ledezma et al., 2002)and 5.5–13 lg ml�1 for Candida, Schizosaccharomyces and Saccharo-myces (see Yoshida et al., 1987; Naganawa et al., 1996). Rather,they fall in range with the values obtained for the most resistantof Gram-positive bacteria (4–136 lg ml�1) and Gram-negativebacteria (116–152 lg ml�1, see Naganawa et al., 1996; Ohtaet al., 1999; Whitmore and Naidu, 2000).

The dithiin extract used in the current study was concentratedfrom chopped garlic, and its purity was only 32%. The remainingfraction was a mixture of diallyl disulfide, diallyl trisulfide, diallyltetrasulfide, allyl methyl trisulfide and allicin, but individual com-pounds were not quantified. As most of these compounds havebeen found to have some antimicrobial activity (Avato et al.,2000; O’Gara et al., 2000; Whitmore and Naidu, 2000; Tsao andYin, 2001; Tsao et al., 2007), it is difficult to assess the actual anti-microbial effect of dithiin here, but the mixture was slightly moreefficient at inhibiting growth of S. vortens than the crude extract ofajoene. Growth rates of the organism were inhibited from5 lg ml�1 of crude oil, and final yield was reduced significantlyat a concentration of 40 lg ml�1 of the crude oil and above. MICand IC50% were 83 and 56 lg ml�1, respectively, which is slightlylower than what is obtained for the crude ajoene extract. Thissmall difference in antimicrobial activity may be attributed tothe higher potency of dithiins, but may also be explained by thesynergistic effect of the many antibacterial compounds present inthe mixtures. As the antimicrobial activity of dithiins documentedso far were lower than that of ajoene (MIC >100 lg ml�1 vs. 15–20 lg ml�1 for ajoene against Helicobacter pylori; see Ohta et al.,1999), the synergistic hypothesis seems more likely. However fur-

ther work, using better defined and purer extracts, is required toaddress this question.

In comparison to allium-derived compounds, metronidazolecompletely inhibited growth rates and final growth yield ofS. vortens at very low concentrations (MIC and IC50% being bothwell below 8.56 lg ml�1). These concentrations are much lowerthan the recommended peak serum concentrations for thetreatment of Giardiasis in humans (40 lg ml�1 when administeredorally and 18–25 lg ml�1 by IV (Madanagopalan et al., 1975; Free-man et al., 1997), which confirms the susceptibility of S. vortens tothis drug (Sangmaneedet and Smith, 1999). Strangely, the muchstronger inhibitory effect of metronidazole on growth rates andfinal growth yield of S. vortens is not reflected in its inhibition ofH2 and CO2 production, which is in the range of allicin exposure.This may be due to the delayed action of the drug, which requiresactivation by the parasites, before exhibiting antimicrobial proper-ties. The absence of inhibition of O2 consumption by metronidazolehas already been described in Trichomonas vaginalis (Lloyd andPedersen, 1985; Yarlett et al., 1986; Ellis et al., 1992) and Giadiaintestinalis (Gillin and Reiner, 1982; Sousa and Poiares-Da-Silva,1999), and is probably due to the fact that the drug can only beactivated under anaerobic conditions.

In conclusion, although this study established that all allium-derived compounds have an inhibitory effect on S. vortens, the po-tency of these compounds against this parasite was much lowerthan for most of other protozoa, fungi and bacteria investigatedso far. In sharp contrast to metronidazole, which had biocidal ef-fects at very low concentrations, allium-derived compounds slo-wed the growth of S. vortens at low concentrations, but onlyinhibited total growth yield at much higher doses. This was partic-ularly true of whole garlic, the concentration of which that was re-quired to lower final growth yield being 10 times that needed toslow growth rates (versus four times for allicin and eight timesfor ajoene and dithiins). Interestingly, MICs of ajoene and dithiinswere less than twice as high as their IC50%, while for whole garlic,the MIC value was more than 2.5 times higher than the IC50%. Manyaquarists believe that addition of 1–2% garlic in the feed cures avariety of parasitic infections, however, to date, few conclusivestudies have been conducted in vivo. In light of our investigations,it appears that such garlic concentrations would have little antimi-crobial effect on S. vortens trophozoites within the intestine, espe-cially as the acid conditions in the stomach are likely to destroymany of the active compounds in garlic (Freeman and Kodera,1995; Fujisawa et al., 2008a,b). Higher doses of garlic, or purifiedthiosulfinates may prove more effective, however, an assessmentof the toxicity of such doses for the fish is required to establishthe practical usefulness of these compounds. It seems likely thatrepeated treatments regimes would be required to eliminate theparasites. Alternatively, some garlic constituents could be used aslead compounds in the chemical synthesis of agents with increasedpotency.

Acknowledgments

The authors are deeply indebted to Prof. Jaroslav Kulda for hisexpert advice and guidance in the culture of the organism. Thanksalso to Dr. Sue Plummer of Cultech Biospecialities, Baglan, PortTalbot for supplying freeze-dried garlic powder. Thanks to Dr.Eshwar Mahenthiralingam for allowing the use of his Bioscreen Cautomated optical density reader and to Drs. Jonathan Wood andVictoria Gray for their advice on Bioscreen data analysis. This workwas funded by the William E. Morgan Scholarship to COMM, anEPSRC (EP/H501118/1) studentship to C.F.W. with CASE Partner,Neem Biotech Ltd. and a Natural Environment Research Council,UK, Research Fellowship to J.C. (NER/J/S/2002/00706).

498 C.O.M. Millet et al. / Experimental Parasitology 127 (2011) 490–499

References

Ankri, S., Miron, T., Rabinkov, A., Wilchek, M., Mirelman, D., 1997. Allicin from garlicstrongly inhibits cysteine proteinases and cytopathic effects of Entamoebahistolytica. Antimicrobial Agents and Chemotherapy 10, 2286–2288.

Ankri, S., Mirelman, D., 1999. Antimicrobial properties of allicin from garlic.Microbes and Infection 2, 125–129.

Avato, P., Tursil, E., Vitali, C., Miccolis, V., Candido, V., 2000. Allylsulfide constituentsof garlic volatile oil as antimicrobial agents. Phytomedicine 3, 239–243.

Bierer, D.E., Dener, J.M., Dubenko, L.G., Gerber, R.E., Litvak, J., Peterli, S., Peterli-Roth,P., Truong, T.V., Mao, G., Bauer, B.E., 1995. Novel 1,2-dithiins: synthesis,molecular modeling studies, and antifungal activity. Journal of MedicinalChemistry 14, 2628–2648.

Block, E., Ahmad, S., Catalfamo, J.L., Jain, M.K., Apitz-Castro, R., 1986. Antithrombicorganosulfur compounds from garlic: structural, mechanistic and syntheticstudies. Journal of the American Chemical Society 108, 7045–7055.

Cavallito, C.J., Bailey, J.H., 1944. Allicin, the antibacterial principle of Alliumsativarum. I. Isolation, physical properties and antibacterial action. Journal ofthe American Chemical Society 66, 1950–1951.

Caves, T., Ergere Gozukara, S., 2005. Genotoxicity evaluation of metronidazole usingthe piscine micronucleus test by acridine orange fluorescent staining.Environmental Toxicology and Pharmacology 19, 107.

Chua, F.H.C., 1996. Aquaculture health management in Singapore: current statusand future directions. In: FAO Fisheries Technical Paper. FAO Publications,Rome, pp. 115–126.

Chung, I., Kwon, S.H., Shim, S.T., Kyung, K.H., 2007. Synergistic antiyeast activity ofgarlic oil and allyl alcohol derived from alliin in garlic. Journal of Food Science 9,M437–M440.

Cunha, B.A., 2001. Antibiotic side effects. Medical Clinics of North America 1, 149–185.

Davis, S.R., 2005. An overview of the antifungal properties of allicin and itsbreakdown products – the possibility of a safe and effective antifungalprophylactic. Mycoses 2, 95–100.

Ellis, J.E., Cole, D., Lloyd, D., 1992. Influence of oxygen on the fermentativemetabolism of metronidazole-sensitive and resistant strains of Trichomonasvaginalis. Molecular and Biochemical Parasitology 1, 79–88.

Food and Agriculture Organization of the United Nations (FAO), 2007. The State ofWorld Fisheries and Aquaculture 2006. FAO Publications, Rome.

Freeman, C.D., Klutman, N.E., Lamp, K.C., 1997. Metronidazole. A therapeutic reviewand update. Drugs 5, 679–708.

Freeman, F., Kodera, Y., 1995. Garlic chemistry: stability of S-(2-propenyl)2-propene-1-sulfinothionate (allicin) in blood, solvents and simulatedphysiological fluids. Journal of Agricultural and Food Chemistry 43, 2332–2338.

Fry, F.H., Okarter, N., Baynton-Smith, C., Kershaw, M.J., Talbot, N.J., Jacob, C., 2005.Use of a substrate/alliinase combination to generate antifungal activity in situ.Journal of Agricultural and Food Chemistry 3, 574–580.

Fujisawa, H., Suma, K., Origuchi, K., Kumagai, H., Seki, T., Ariga, T., 2008a. Biologicaland chemical stability of garlic-derived allicin. Journal of Agricultural and FoodChemistry 11, 4229–4235.

Fujisawa, H., Suma, K., Origuchi, K., Seki, T., Ariga, T., 2008b. Thermostability ofallicin determined by chemical and biological assays. Bioscience, Biotechnology,and Biochemistry 11, 2877–2883.

Group of Experts on the Scientific Aspects of Marine Environmental Protection(GESAMP), 1997. Towards a Safe and Effective Use of Chemicals in CoastalAquaculture. Reports and Studies 65. FAO Publications, Rome.

Gillin, F.D., Reiner, D.S., 1982. Effects of oxygen tension and reducing agents onsensitivity of Giardia lamblia to metronidazole in vitro. BiochemicalPharmacology 22, 3694–3697.

Guo, F.C., Woo, P.T., 2004a. Experimental infections of Atlantic salmon Salmo salarwith Spironucleus barkhanus. Diseases of Aquatic Organisms 1–2, 59–66.

Guo, F.C., Woo, P.T., 2004b. Detection and quantification of Spironucleus barkhanusin experimentally infected Atlantic salmon Salmo salar. Diseases of AquaticOrganisms 1–2, 175–178.

Gupta, S., Ravishankar, S., 2005. A comparison of the antimicrobial activity of garlic,ginger, carrot, and turmeric pastes against Escherichia coli O157:H7 inlaboratory buffer and ground beef. Foodborne Pathogens and Diseases 4, 330–340.

Harris, J.C., Plummer, S., Turner, M.P., Lloyd, D., 2000. The microaerophilic flagellateGiardia duodenalis: Allium sativum (garlic) is an effective antigiardial.Microbiology, 3119–3127.

Harris, J.C., Plummer, S., Lloyd, D., 2001. Antigiardial drugs. Applied Microbiologyand Biotechnology 5–6, 614–619.

Huet, A., Mortier, L., Daeseleire, E., Fodey, T., Elliott, C., Delahaut, P., 2005.Development and validation of new analysis methods for the detection ofresidues of contaminants in biological matrices. Analytica Chimica Acta 534,157.

Hunter, R., Caira, M., Stellenboom, N., 2005. Thiolsulfinate allicin from garlic:inspiration for a new antimicrobial agent. Annals of the New York Academy ofSciences, 234–241.

Iwalokun, B.A., Ogunledun, A., Ogbolu, D.O., Bamiro, S.B., Jimi-Omojola, J., 2004. Invitro antimicrobial properties of aqueous garlic extract against multidrug-resistant bacteria and Candida species from Nigeria. Journal of Medicinal Food 3,327–333.

Jorgensen, A., Sterud, E., 2006. The marine pathogenic genotype of Spironucleusbarkhanus from farmed salmonids redescribed as Spironucleus salmonicida n. sp.Journal of Eukaryotic Microbiology 6, 531–541.

Jorgensen, A., Sterud, E., 2007. Phylogeny of Spironucleus (eopharyngia:diplomonadida: hexamitinae). Protist 2, 247–254.

L82/14, C.R.E., 1998. Veterinary drug residues: regulatory issues.Lanzky, P.F., Halling-Sorensen, B., 1997. The toxic effect of the antibiotic

metronidazole on aquatic organisms. Chemosphere 11, 2553–2561.Ledezma, E., Jorquera, A., Bendezu, H., Vivas, J., Perez, G., 2002. Antiproliferative and

leishmanicidal effect of ajoene on various Leishmania species: ultrastructuralstudy. Parasitology Research 8, 748–753.

Lloyd, D., Pedersen, J.Z., 1985. Metronidazole radical anion generation in vivo inTrichomonas vaginalis: oxygen quenching is enhanced in a drug-resistant strain.Journal of General Microbiology 1, 87–92.

Lloyd, D., Ellis, J.E., Hillman, K., Williams, A.G., 1992. Membrane inlet massspectrometry: probing the rumen ecosystem. Society of Applied BacteriologySymposium Series, 155S–163S.

Madanagopalan, N., Rao, U.P., Somasundaram, A., Lakshmipathi, T., 1975. Acorrelative study of duodenal aspirate and faeces examination in giardiasis,before and after treatment with metronidazole. Current Medical ResearchOpinion 2, 99–103.

Maher, H.M., Youssef, R.M., Khalil, R.H., El-Bahr, S.M., 2008. Simultaneousmultiresidue determination of metronidazole and spiramycin in fish muscleusing high performance liquid chromatography with UV detection. Journal ofChromatography B 2, 175–181.

Meyer, F.P., 1991. Aquaculture disease and health management. Journal of AnimalScience 69, 4201–4208.

Millet, C.O.M., Cable, J., Lloyd, D., 2010. The dipolomonad fish parasite Spironucleusvortens produces hydrogen. Journal of Eukaryotic Microbiology 57, 400–404.

Millet, C.O.M., Lloyd, D., Williams, C., Cable, J., 2010. In vitro culture of thediplomonad fish parasite Spironucleus vortens reveals unusually fast doublingtime and atypical biphasic growth. Journal of Fish Diseases. doi:10.1111/j.1365-2761.2010.01213.x.

Mirelman, D., Monheit, D., Varon, S., 1987. Inhibition of growth of Entamoebahistolytica by allicin, the active principle of garlic extract (Allium sativum).Journal of Infectious Diseases 1, 243–244.

Naganawa, R., Iwata, N., Ishikawa, K., Fukuda, H., Fujino, T., Suzuki, A., 1996.Inhibition of microbial growth by ajoene, a sulfur-containing compoundderived from garlic. Applied Environmental Microbiology 11, 4238–4242.

O’Gara, E.A., Hill, D.J., Maslin, D.J., 2000. Activities of garlic oil, garlic powder, andtheir diallyl constituents against Helicobacter pylori. Applied EnvironmentalMicrobiology 5, 2269–2273.

Ohta, R., Yamada, Y., Kanedo, H., Ishikawa, K., Fukuda, H., Fujino, T., Suzuki, A., 1999.In vitro inhibition of the growth of Helicobacter pylori by oil-macerated garlicconstituents. Antimicrobial Agents and Chemotherapy 7, 1811–1812.

Pai, S.T., Platt, M.W., 1995. Antifungal effects of Allium sativum (garlic) extractagainst the Aspergillus species involved in otomycosis. Letters in AppliedMicrobiology 1, 14–18.

Paull, G.C., Matthews, R.A., 2001. Spironucleus vortens, a possible cause of hole-in-the-head disease in cichlids. Diseases of Aquatic Organisms 3, 197–202.

Pickering, A.D., 1998. Stress responses of farmed fish. In: Black, K.D., Pickering, A.D.,(Eds.), Biology of Farmed Fish. Sheffield Academic Press, Sheffield, pp. 222–255.

Poynton, S.L., Fraser, W., Francis-Floyd, R., Rutledge, P., Reed, P., Nerad, T.A., 1995.Spironucleus vortens n-sp. from the fresh-water angelfish pterophyllum scalare –morphology and culture. Journal of Eukaryotic Microbiology 6, 731–742.

Reuter, H.D., Koch, H.P., Lawson, L.D., 1996. Therapeutic effects and applications ofgarlic and its preparations. In: Koch, H.P., Lawson, L.D., (Eds.), Garlic: TheScience and Therapeutic Application of Allium sativum L. and Related Species.Williams and Wilkins, Baltimore, pp. 135–213.

Richardson, M.L., Bowron, J.M., 1985. The fate of pharmaceutical chemicals in theaquatic environment. Journal of Pharmacy and Pharmacology 1, 1–12.

Ross, Z.M., O’Gara, E.A., Hill, D.J., Sleightholme, H.V., Maslin, D.J., 2001. Antimicrobialproperties of garlic oil against human enteric bacteria: evaluation ofmethodologies and comparisons with garlic oil sulfides and garlic powder.Applied Environmental Microbiology 1, 475–480.

Ruddock, P.S., Liao, M., Foster, B.C., Lawson, L., Arnason, J.T., Dillon, J.A., 2005. Garlicnatural health products exhibit variable constituent levels and antimicrobialactivity against Neisseria gonorrhoeae, Staphylococcus aureus and Enterococcusfaecalis. Phytotherapy Research 4, 327–334.

Sangmaneedet, S., Smith, S.A., 1999. Efficacy of various chemotherapeutic agents onthe growth of Spironucleus vortens, an intestinal parasite of the freshwaterangelfish. Diseases of Aquatic Organisms 1, 47–52.

Sangmaneedet, S., Smith, S.A., 2000. In vitro studies on optimal requirements for thegrowth of Spironucleus vortens, an intestinal parasite of the freshwater angelfish.Diseases of Aquatic Organisms 39, 135–141.

Shehadeh, Z., Maclean, J., 1997. Review on the state of the world aquaculture. FAOFisheries Circular 886, 1–163.

Shen, J., Davis, L.E., Wallace, J.M., Cai, Y., Lawson, L.D., 1996. Enhanced diallyltrisulfide has in vitro synergy with amphotericin B against Cryptococcusneoformans. Planta Medica 5, 415–418.

Soffar, S.A., Mokhtar, G.M., 1991. Evaluation of the antiparasitic effect of aqueousgarlic (Allium sativum) extract in hymenolepiasis nana and giardiasis. Journal ofEgyptian Society for Parasitology 2, 497–502.

C.O.M. Millet et al. / Experimental Parasitology 127 (2011) 490–499 499

Sousa, M.C., Poiares-Da-Silva, J., 1999. A new method for assessing metronidazolesusceptibility of Giardia lamblia trophozoites. Antimicrobial Agents andChemotherapy 12, 2939–2942.

Sterud, E., Mo, T.A., Poppe, T.T., 1998. Systemic spironucleosis in sea-farmedAtlantic salmon Salmo salar, caused by Spironucleus barkhanus transmittedfrom feral Arctic char Salvelinus alpinus? Diseases of Aquatic Organisms 1,63–66.

Sterud, E., Poynton, S.L., 2002. Spironucleus vortens (Diplomonadida) in the Ide,Leuciscus idus (L.) (Cyprinidae): a warm water hexamitid flagellate found innorthern Europe. Journal of Eukaryotic Microbiology 2, 137–145.

Subasinghe, R.P., Bondad-Reantaso, M.G., McGladdery, S.E., 2000. Aquaculturedevelopment, health and wealth. In: Technical Proceedings of the Conference onAquaculture in the Third Millennium. NACA, Bangkok and FAO, Rome, pp. 167–191.

Tatarintsev, A.V., Vrzheshch, P.V., Schegolev, A.A., Yershov, D.E., Turgiev, A.S.,Varfolomeyev, S.D., Kornilayeva, G.V., Makarova, T.V., Karamov, E.V., 1992.Ajoene antagonizes integrin-dependent processes in HIV-infected T-lymphoblasts. AIDS 10, 1215–1217.

Tojo, J.L., Santamarina, M.T., 1998. Oral pharmacological treatments for parasiticdiseases of rainbow trout Oncorhynchus mykiss. I: Hexamita salmonis. Diseases ofAquatic Organisms 1, 51–56.

Treves-Brown, K.M., 1999. Availability of medicine for fish. The Journal of the FishVeterinary Society 4, 40–58.

Tsai, Y., Cole, L.L., Davis, L.E., Lockwood, S.J., Simmons, V., Wild, G.C., 1985. Antiviralproperties of garlic: in vitro effects on influenza B, herpes simplex and coxsackieviruses. Planta Medica 5, 460–461.

Tsao, S.M., Yin, M.C., 2001. In vitro antimicrobial activity of four diallyl sulphidesoccurring naturally in garlic and Chinese leek oils. Journal of MedicalMicrobiology 7, 646–649.

Tsao, S.M., Hsu, C.C., Yin, M.C., 2003. Garlic extract and two diallyl sulphides inhibitmethicillin-resistant Staphylococcus aureus infection in BALB/cA mice. Journal ofAntimicrobials Chemotherapy 6, 974–980.

Tsao, S.M., Liu, W.H., Yin, M.C., 2007. Two diallyl sulphides derived from garlicinhibit meticillin-resistant Staphylococcus aureus infection in diabetic mice.Journal of Medical Microbiology (Pt. 6), 803–808.

Uchida, Y., Takahashi, T., Sato, N., 1975. The characteristics of the antibacterialactivity of garlic (author’s transl.). Japanese Journal of Antibiotics 4, 638–642.

Van Bambeke, F., Michot, J.M., Tulkens, P.M., 2003. Antibiotic efflux pumps ineukaryotic cells: occurrence and impact on antibiotic cellular pharmacokinetics,pharmacodynamics and toxicodynamics. Journal of AntimicrobialsChemotherapy 5, 1067–1077.

Wedemeyer, G.A., 1997. Effects of rearing conditions on the health andphysiological quality of fish in intensive culture. In: Iwama, G.K., Pickering,A.D., Sumpter, J.P., Schreck, C.B., (Eds.), Fish Stress and Health in Aquaculture,Society for Experimental Biology, Seminar Series 62. Cambridge UniversityPress, Cambridge, pp. 35–71.

Wilhelm, E., Battino, R., Wilcock, R.J., 1977. Low-pressure solubility of gases inliquid water. Chemical Reviews 2, 219–259.

Whitmore, B.B., Naidu, A.S., 2000. Thiosulfinates. In: Naidu, A.S. (Ed.), Natural FoodAntimicrobial Systems. CRS press, New York, pp. 349–380.

Yamada, Y., Azuma, K., 1997. Evaluation of the in vitro antifungal activity of garlic.Antimicrobial Agents and Chemotherapy 11, 743–749.

Yarlett, N., Yarlett, N.C., Lloyd, D., 1986. Ferredoxin-dependent reduction ofnitroimidazole derivatives in drug-resistant and susceptible strains ofTrichomonas vaginalis. Biochemical Pharmacology 10, 1703–1708.

Yoshida, S., Kasuga, S., Hayashi, N., Ushiroguchi, T., Matsuura, H., Nakagawa, S.,1987. Antifungal activity of ajoene derived from garlic. Applied EnvironmentalMicrobiology 3, 615–617.