Toxins 2015, 7, 1779-1807; doi:10.3390/ oxins7051779
oxins
ISSN 2072-6651
www.mdpi.com/jou nal/ oxins
A icle
Fi s De ec ion o Te odo oxin in G eek Shell ish by
UPLC-MS/MS Po en ially Linked o he P esence o he
Dino lagella e P o ocen um minimum
A is idis Vlamis 1,2, Panagio a Ka ikou 2,*, Ines Rod iguez 1, Ve ónica Rey 1, Ampa o Al onso 1,
Angelos Papazacha iou 2, The is Zacha aki 2, Ana M. Bo ana 1 and Luis M. Bo ana 1
1 Depa men o Pha macology, Ve e ina y School, Uni e si y o San iago de Compos ela,
Lugo 27002, Spain; E-Mails: [email p o ec ed] (A.V.); [email p o ec ed] (I.R.);
[email p o ec ed] (V.R.); [email p o ec ed] (A.A.); [email p o ec ed] (A.M.B.);
[email p o ec ed] (L.M.B.)
2 Na ional Re e ence Labo a o y on Ma ine Bio oxins, Ve e ina y Cen e o Thessaloniki,
Minis y o P oduc i e Recons uc ion, En i onmen and Ene gy, 3A Limnou s ee ,
GR 54627 Thessaloniki, G eece; E-Mails: aop e @gmail.com (A.P.); [email p o ec ed] (T.Z.)
* Au ho o whom co espondence should be add essed; E-Mail: bi[email p o ec ed];
Tel.: +30-2310-552928; Fax: +30-2310-566581.
Academic Edi o : John P. Be y
Recei ed: 5 Ma ch 2015 / Accep ed: 11 May 2015 / Published: 20 May 2015
Abs ac : Du ing o icial shell ish con ol o he p esence o ma ine bio oxins in G eece in
yea 2012, a se ies o unexplained posi i e mouse bioassays (MBA) o lipophilic oxins
wi h ne ous symp oma ology p io o mice dea h was obse ed in mussels om Vis onikos
Bay–Lagos, Rodopi. This a ypical oxici y coincided wi h (a) absence o low le els o
egula ed and some non- egula ed oxins in mussels and (b) he simul aneous p esence o he
po en ially oxic mic oalgal species P o ocen um minimum a le els up o 1.89 × 103 cells/L
in he a ea’s seawa e . Fu he analyses by di e en MBA p o ocols indica ed ha he
unknown oxin was hyd ophilic, whe eas UPLC-MS/MS analyses e ealed he p esence o
e odo oxins (TTXs) a le els up o 222.9 μg/kg. Re iewing o o icial con ol da a om
p e ious yea s (2006–2012) iden i ied a numbe o sample cases wi h a ypical posi i e o
asymp oma ic nega i e MBAs o lipophilic oxins in di e en G eek p oduc ion a eas,
coinciding wi h pe iods o P. minimum blooms. UPLC-MS/MS analysis o e ained sub-samples
om hese cases e ealed ha TTXs we e al eady p esen in G eek shell ish since 2006,
in concen a ions anging be ween 61.0 and 194.7 μg/kg. To ou knowledge, his is he
OPEN ACCESS
Toxins 2015, 7 1780
ea lies epo ed de ec ion o TTXs in Eu opean bi al e shell ish, while i is also he i s
wo k o indica e a possible link be ween p esence o he oxic dino lagella e P. minimum in
seawa e and ha o TTXs in bi al es. Con i med p esence o TTX, a e y hea -s able oxin,
in il e - eeding mollusks o he Medi e anean Sea, e en a lowe le els o hose inducing
symp oma ology o humans, indica es ha his eme ging isk should be se iously aken in o
accoun by he EU o p o ec he heal h o shell ish consume s.
Keywo ds: Aegean Sea; eme ging bio oxins; Medi e anean Sea; cul u ed mussels;
My ilus gallop o incialis; P o ocen um minimum; e odo oxin; oxic episode; UPLC-MS/MS;
ene upin shell ish oxin
1. In oduc ion
Te odo oxin (TTX) is an ex emely po en neu o oxin ha can block sodium channels and hus inhibi
p opaga ion o ac ion po en ials in muscle and ne e cells [1]. Named a e he Te aodon idae pu e
ish amily, TTX is pe haps mos no o ious as he oxin ha causes pu e ish poisoning. TTX can cause
dea h by muscula pa alysis, espi a o y dep ession and ci cula o y ailu e [2]. The minimum le hal and
minimum acu e doses o TTX o human (w . 50 kg) a e es ima ed o be a ound 2 mg and
0.2 mg, espec i ely. Symp oms usually appea wi hin 10–45 min o exposu e, depending upon he
amoun o he oxin inges ed; howe e , some o he epo ed cases we e asymp oma ic un il as much as
3 o 6 h a e exposu e. The ini ial symp om is usually o al pa es hesia, which g adually sp eads o he
ex emi ies and unk. O he ea ly symp oms comp ise as e dis u bance, dizziness, headache,
diapho esis, and pupilla y cons ic ion, po en ially accompanied by gas oin es inal symp oms o
sali a ion, hype sali a ion, nausea, omi ing, hype emesis, hema emesis, hype mo ili y, dia hea and
abdominal pain [3,4].
Te odo oxin is known o occu in a a ie y o ish species and o he o ganisms including a h opods,
echinode ms, an alga, mollusks, wo ms, new s, ogs and oads [5–9]. To da e, howe e , epo ed
occu ences o TTX in bi al e mollusks (clams, cockles, mussels, oys e s, scallops and o he s) a e
e y ew. TTX has been de e mined in New Zealand Paphies aus alis clams [10] and in Japanese
scallops [11]. Simila ly, epo ed occu ences in Eu opean sea ood a e e y limi ed. TTX was i s
de ec ed in 2007, in he cou se o a non- a al human in oxica ion ollowing consump ion o con amina ed
sea snails Cha onia lampas lampas (a gas opod) ha es ed in Spain [12]. Ve y ecen ly, TTX p esence
has been epo ed in bi al e mollusks (mussels and Paci ic oys e s) ha es ed om he sou h coas o
England, along he Channel [13].
Due o he wide TTX dis ibu ion among gene ically-un ela ed animals, combined wi h indi idual,
egional and seasonal a ia ion in TTX le els, he exac o igin o TTX in he ood chain has been long
deba ed, bu s ill he sou ce o he oxin emains p ac ically unknown [5,14]. In con as o he es o
bio oxins ha accumula e in ishe y p oduc s, TTX p oduc ion has no been linked un il now o any
mic oalgal o ganism. Con ibu ion o symbio ic bac e ia has been sugges ed o play a ole in TTX
genesis o ma ine animals [15]. Such symbio ic bac e ia, including Shewanella algae, S. pu e aciens,
Vib io sp., Pseudomonas sp., and Al e omonas e aodonis, a e sugges ed o accumula e in he
Toxins 2015, 7 1781
subcu aneous mucus, o in he in es ine, eleasing he TTX [16,17]. This was la e con i med by he
isola ion o TTX-p oducing bac e ia om di e en TTX-bea ing animals [13,18,19].
P o ocen um minimum (Pa illa d) Schille (p e iously epo ed also as Exu iaella ma iae-lebou iae
Pa ke and Ballan ine and P. ma iae-lebou iae), a phy oplanc onic dino lagella e known o cause ed
ides, is widely dis ibu ed in he seas and oceans o he no he n hemisphe e. P. minimum is gene ally
conside ed ha mless; howe e , episodes o shell ish and human poisoning ha e been so a epo ed in
Japan, Po ugal and No way. In Japan, a mass-poisoning inciden occu ed a Lake Hamana in
Ma ch 1942 a e consump ion o sho -necked clams (Vene upis semidecussa a), causing dea h in 114
ou o 324 pe sons a ec ed [20]. A o al o se en y-one dea hs in he same egion we e la e a ibu ed
o inges ion o oxic oys e s (C assos ea gigas) in Ma ch 1943 and o oxic clams in Ma ch 1949.
Symp oms included hea y li e inju y (nec osis and a y degene a ion), hemo hagic dia hesis wi h
enzy, unconsciousness and coma, and dea h occu ing wi hin 24–48 h o symp oms ini ia ion [21].
P. minimum (E. ma iae-lebou iae) was a ibu ed as he oxin sou ce, a e es ablishing a ela ionship
be ween shell ish oxici y and seawa e abundance o hese dino lagella es [20,22–24]. A ni ogenous
oxic subs ance, which p oduced symp oms in mice simila o hose obse ed in humans, was isola ed
by Akiba and Ha o i [25] om he mid-gu o he sho -necked clam. This subs ance was named
“ ene upin” and he associa ed synd ome was assigned as Vene upin Shell ish Poisoning (VSP) [21].
Much la e , P. minimum (iden i ied ini ially as Exu iaella bal ica and la e as P o ocen um bal icum)
was inc imina ed o a numbe o human poisoning episodes in Po ugal (Obidos Lagoon) ollowing
shell ish consump ion. The symp oms we e cha ac e ized as esembling hose o pa aly ic shell ish
poisoning (PSP) [26–28]. Repo ed e en s wi h PSP-like symp oma ology in he Zhengjiang and Fujiang
p o inces om 1978 o 1985 we e also a ibu ed o P. minimum, bu de ails a e sca ce [29]. Simila ly,
P. minimum was iden i ied as he cause in a case o poisoning a e mussels consump ion in he
Oslo jo d, No way in 1979, as i had bloomed se e al weeks be o e in he mussel ha es ing a ea. The
symp oms (especially nausea and la e gas oin es inal diso de s) esembled in na u e and de elopmen
o hose o VSP. P elimina y oxin bioassays on mussels sampled om he jo d showed signs o oxici y,
wi h a symp oma ology e y simila o ha o VSP, bu less se e e [30,31]. Toxin p esence has been
occasionally in es iga ed a he same ime wi h a P. minimum bloom. Kimo e al. [32], o ins ance, ailed
o ind any oxici y in Bal ic Sea dino lagella es, whe eas mouse es s e ealed only e y mino symp oms
o oxici y in Black Sea mussels [33]. Red ides caused by P. minimum o E. bal ica (ac ually P. minimum)
ha e also been associa ed wi h a ious e ec s on ma ine o ganisms. Blooms we e ega ded as highly oxic,
since ish and o he ma ine animals died o we e o ced o lee [27,34,35]. In one case, howe e ,
hese dea hs occu ed h ee days a e educ ion o dissol ed oxygen [34]. I should be no ed ha in bo h
o hese cases, u he s udies we e no unde aken o cla i y he sou ce o oxici y.
E ec s o algal blooms on shell ish and aquacul u e ha e been e iewed by Shumway [36], who
sugges ed ha oxic e ec s ha e been o en associa ed wi h P. minimum blooms and ha he oxin
accumula ed in shell ish was able o a ec bo h shell ish ha es ed om bloom wa e and humans
consuming he shell ish. So a , howe e , he only unambiguous link be ween blooms o P. minimum
and shell ish oxici y has been documen ed by G zebyk e al. [37] in a F ench Medi e anean si e. In ha
s udy, a numbe o P. minimum axenic clones, p oducing a wa e -soluble neu o oxic componen ha
apidly killed mice, we e isola ed. P io o ha , mussel oxici y had been de ec ed in he Se e egion on
he F ench Medi e anean coas du ing a la ge bloom o P. minimum and P. micans [37].
Toxins 2015, 7 1782
The symp oma ology obse ed in mouse bioassays indica ed a apid neu ological e ec ; howe e , i was
no possible o p o ide clea e idence ha he dino lagella es we e ac ually he sou ce o his oxici y.
Subsequen ly hough, Dena dou-Quenehe e e al. [38] documen ed he same oxic ac ion ob ained
om shell ish du ing a la ge P. minimum bloom (56 × 106 cells/L), which occu ed in he Salses-Leuca e
lagoon on he F ench Medi e anean coas . Despi e he ac ha his esea ch g oup p o ided a subs an ial
con ibu ion owa ds cla i ica ion o he P. minimum oxin p ope ies, he na u e and s uc u e o he
oxin was ne e iden i ied.
Du ing o icial shell ish con ol o he p esence o ma ine bio oxins in G eece in yea 2012, a se ies
o unexplained posi i e mouse bioassays (MBA) o lipophilic oxins wi h ne ous symp oma ology p io
o mice dea h was obse ed in mussels om Vis onikos Bay–Lagos, Rodopi. This a ypical oxici y coincided
wi h he absence o p esence in ace le els o all known lipophilic, PSP and ASP g oup oxins in mussels,
as well as wi h he p esence o P. minimum a le els up o 1.89 × 103 cells/L in he a ea’s seawa e , whe eas
u he analyses by di e en MBA p o ocols indica ed ha he unknown oxin was o hyd ophilic na u e.
In his con ex , he pu pose o his wo k was: (a) o in es iga e his episode wi h ega d o he na u e o
he oxin(s) causing his a ypical oxici y and (b) o in es iga e oxin p o iles o shell ish om p e ious
yea s (2006–2012) om di e en G eek p oduc ion a eas, collec ed in ime pe iods when P. minimum
was p esen a high abundances.
2. Resul s
2.1. The 2012 P. minimum Episode in Vis onikos Bay–Lagos, Rodopi, G eece
2.1.1. Episode His o y and P elimina y In es iga ion
In la e sp ing-ea ly summe o 2012, a se ies o ou consecu i e mussel samples (1770/2012;
1774/2012; 1786/2012 and 1798/2012) de i ed om he sampling si e Zone 6, Vis onikos Bay–Lagos
( egional uni o Rodopi, G eece; Table 1, Figu e 1A) es ed posi i e in he Yasumo o 1978 (Yas’78)
p o ocol o lipophilic oxins [39], p esen ing an a ypical ne ous symp oma ology p io o mice dea h,
whe eas all ou es ed nega i e o he MBA o PSP oxins and below he limi o de ec ion (0.1 mg/kg)
o he p esence o ASP oxins. Du ing he same ime pe iod, phy oplank on coun s o all known species
associa ed wi h he p oduc ion o egula ed ma ine bio oxin g oups (lipophilic, pa aly ic and amnesic
oxins) we e all ei he unde ec able o p esen a low le els no adequa e o accoun o any shell ish
oxici y. The only mic oalgal species p esen a conside able concen a ions was P o ocen um minimum
(syn. = P. co da um, as he cu en ly axonomically accep ed name; [40]) eaching le els up o
1.89 × 103 cells/L (see Table 2).
Toxins 2015, 7 1783
Table 1. Loca ions o sampling s a ions included in he s udy o each coas al a ea exp essed
by geog aphical coo dina es (longi ude and la i ude).
Coas al A ea Regional Uni Sampling S a ion La i ude (N) Longi ude (E)
Vis onikos Bay–Lagos Rodopi Zone 6 40°58'23.73" 25°07'06.56"
Ke amo i Bay Ka ala Ke amo i 40°52'00.70" 24°40'34.60"
S ymonikos Bay Se es S1—Nea Ke dyllia 40°46'46.49" 23°49'50.00
Chalkidiki X2—Olympiada 40°33'00.60" 23°50'00.00"
The maikos Gul Thessaloniki Kymina 40°30'00.22" 22°41'00.06"
Sa onikos Gul
A iki/Nisson Zone I—Vasilika–Fane omeni 37°59'42.29" 23°27'50.06"
Zone IV—Agios Geo gios 37°58'09.89" 23°25'53.92"
Dy iki A iki/Mega a Ne aki 38°01'30.19" 23°28'52.79"
D epano 37°59'01.02" 23°24'30.09"
Figu e 1. Maps showing loca ions o sampling s a ions: (A) Vis onikos Bay–Lagos (Rodopi);
(B) Ke amo i Bay (Ka ala); (C) S ymonikos Bay (Se es, Chalkidiki); (D) The maikos Gul
(Thessaloniki); and (E) Sa onikos Gul (Dy iki A iki/Mega a, A iki/Nisson). Round ma ks
(•) depic he loca ion o sampling s a ions.
Toxins 2015, 7 1784
Table 2. Sampling da a and analysis esul s o he 2012 P. minimum episode in Vis onikos Bay–Lagos, Rodopi, G eece.
Sampling da a P o ocen um
minimum
(cells/L)
MBA (Yasumo o1978 [39]) MBA (O he ) UPLC-MS/MS
Sample
code
Da e o
collec ion
Regional
Uni Si e Species Resul Symp oma ology P o ocol Resul Symp oma ology Tissue
es ed
TTX
(μg/kg)
4-epi-TTX
(μg/kg)
4,9-anhyd o-
TTX (μg/kg)
Sum o
TTXs
(μg/kg)
1727/2012 7 May
2012 Rodopi Zone 6 Mussels P esence Nega i e
(− − −) No symp oma ology N/A N/A N/A DG 46.2 ND ND 46.2
1770/2012 22 May
2012 Rodopi Zone 6 Mussels 8.26× 102 Posi i e
(+ + +)
3/3 mice dead in
1:00–1:30 h
CRL-MBA-SOP
.5
(DEE ac ion)
Nega i e
(− − −) None
DG 202.9 7.6 <LOQ
(12.4) * 222.9
WF 179.1 <LOQ
(4.2) *
<LOQ
(10.3) * 193.6
1774/2012 28 May
2012 Rodopi Zone 6 Mussels 1.78 × 103 Posi i e
(+ + +)
3/3 mice dead in
1:30–2:00 h
CRL-MBA-SOP
.5
(DEE ac ion)
Nega i e
(− − −)
2/3 mice wi h
mild dia hea DG 185.9 <LOQ
(4.5) *
<LOQ
(15.9) * 206.3
1786/2012 30 May
2012 Rodopi Zone 6 Mussels 1.70 × 103 Posi i e
(+ + +)
3/3 mice dead
(1/3 in 1:15–1:30 h,
1/3 in 1:48 h &
1/3 in 1:51 h)
CRL-MBA-SOP
.5
(DEE ac ion)
Nega i e
(− − −)
1/3 mice wi h
mild dia hea
DG 180.0 <LOQ
(6.8) *
<LOQ
(14.5) * 201.4
CRL-MBA-SOP
.5
(aqueous ac ion)
Posi i e
(+ + +)
3/3 mice dead in
7–10 min
1798/2012 6 Jun
2012 Rodopi Zone 6 Mussels 1.89 × 103 Posi i e
(+ + +)
3/3 mice dead
(1/3 in 1:31 h, 1/3 in
2:30 h & 1/3 in 2:31)
CRL-MBA-SOP
.5
(DEE ac ion)
Nega i e
(− − −) None DG 186.2 15.6
<LOQ
(14.7) * 216.6
1801/2012 11 Jun
2012 Rodopi Zone 6 Mussels 5.65 × 102 Nega i e
(− − −)
2/3 mice wi h
dia hoea
CRL-MBA-SOP
.5
(e he ac ion)
Nega i e
(− − −)
1/3 mice wi h
dia hea DG 56.5 8.8 22.9 88.2
MBA = Mouse bioassay, TTX = e odo oxin, DG = diges i e glands, WF = whole lesh, DEE = die hyle he , N/A = no applicable, ND = no de ec ed. Values wi h an as e isk (*) indica e concen a ions be ween
he calcula ed me hod LOD and LOQ and a e shown as con ibu ions o he sum o TTXs.
Toxins 2015, 7 1785
Acco ding o he documen ed p ocedu es o he G eek Na ional Re e ence Labo a o y on Ma ine
Bio oxins (NRLMB) a ha ime and due o he ac ha Regula ion (EC) 15/2011 [41] was al eady in
place, acco ding o which egula o y moni o ing decisions wi h ega d o lipophilic oxins ha e o be
based on he esul s o he LC-MS/MS me hod (EU-Ha monised SOP-LIPO-LC-MS/MS, Ve sion 4,
Eu opean Re e ence Labo a o y o Ma ine Bio oxins, Vigo, Spain) in case o con lic , samples exhibi ing
a ypical esponses in he ou ine Yas’78 MBA p o ocol, we e as a ule also es ed he ollowing ways.
(a) The ha monized MBA p o ocol o he EU Re e ence Labo a o y on Ma ine Bio oxins
(CRL-MBA-SOP, e sion 5; [42]), which includes a pa i ioning s ep be ween die hyle he (DEE) and
wa e . This pa i ioning s ep is necessa y o con i m he lipophilic na u e o a oxin, in case o a posi i e
esul . In case o a nega i e esul , howe e , i could be indica i e o in e e ences due o he p esence
o hyd ophilic oxins (e.g., PSPs), which a e qui e common and documen ed o he Yasumo o1978
p o ocol; and (b) he UPLC-MS/MS in o de o con i m he p esence o egula ed lipophilic bio oxin
g oups (okadaic acid (OA) g oup comp ising OA, dinophysis oxins (DTXs) and pec eno oxins (PTXs)
oge he , yesso oxin (YTX) g oup and azaspi acids (AZA) g oup), as well as non- egula ed ones, which can
be de e mined by his me hod, such as gymnodimines (GYM), spi olides (SPX) and pinna oxins (PnTXs).
Resul s o he ha monized MBA p o ocol es s (injec ion o DEE ac ion) on he ou posi i e
samples (1770/2012, 1774/2012, 1786/2012 and 1798/2012) a e p esen ed in Table 2. In all ou
samples, a nega i e esul was ob ained, wi h mice su i ing he whole 24-h obse a ion pe iod and
ei he being comple ely asymp oma ic o p esen ing mino clinical symp oms, such as dia hea. In one
o he samples (1786/2012), he aqueous ac ion om he CRL-MBA-SOP p o ocol was also es ed,
p oducing apid dea hs in all h ee injec ed mice (<10 min) wi h se e e symp oma ology om he
ne ous sys em be o e dea h. Analysis by UPLC-MS/MS o he p esence o lipophilic oxins,
on he o he hand, showed only ace amoun s o OA, GYM, SPX and PnTXs (OA: <20 μg/kg,
GYM: <0.2 μg/kg, SPX: <0.2 μg/kg and PnTXs: <0.1 μg/kg), which canno accoun o mouse oxici y.
The samples we e hus designa ed as nega i e o he p esence o egula ed lipophilic oxins and he e o e
sa e o human consump ion, in e ms o egula o y moni o ing. Based on hese esul s, no sani a y
measu es could be aken by he espec i e egional compe en au ho i y. All he abo e indings, could
no jus i y he oxici y obse ed in he Yasumo o 1978 MBA p o ocol, bu also s ongly indica ed ha
he oxic subs ance(s) p oducing he dea hs was mos p obably o a hyd ophilic na u e.
The ac ha a se ies o ou consecu i e samples p esen ed: (a) posi i e MBA esponses in he
Yas’78 p o ocol wi h simila a ypical ne ous symp oma ology and dea h imes; (b) simila esul s in
he p elimina y in es iga ions by he CRL-MBA-SOP p o ocol and by UPLC-MS/MS o lipophilic
oxins; and (c) all coinciding wi h a bloom o a species bibliog aphically documen ed as po en ially
oxic, a leas in a ew, e en limi ed, cases, igge ed ou in e es in u he in es iga ion o he episode.
Fo his eason, me hanolic ex ac s o all samples and whole issue o sample 1786/2012 we e o wa ded
o he Uni e si y o San iago de Compos ela (USC) o c oss-checking and u he es ing. Analysis o
he p esence o lipophilic oxins by UPLC-MS/MS con i med he NRLMB indings, simila ly de ec ing
he p esence o OA, GYM, SPX and PnTXs in ace amoun s. On he o he hand, analysis by HPLC-FLD
o he p esence o PSP oxins in sample 1786/2012 e ealed he p esence o GTX5 a a concen a ion
equi alen o 1.32 μg STX eq/kg, hus no able o accoun o he p esence o oxici y in he MBA.
Finally, analysis by LC-MS/MS o he p esence o e odo oxins by he me hod o Rod iguez e al. [8],
Toxins 2015, 7 1786
esul ed in he de ec ion o TTX analogues in ha sample and hus a ele an UPLC-MS/MS me hod o
he p esence o TTXs was de eloped in NRLMB o be employed o a ull in es iga ion o he episode.
2.1.2. Pe o mance Pa ame e s and Quali y Con ol o he UPLC-MS/MS Me hod o TTXs
The ce i ied TTXs (s anda d con aining TTX), 4-epiTTX and 4,9-anhyd oTTX, oge he wi h
na u ally con amina ed gonad and muscle samples om Lagocephalus scele a us pu e ish used in
o me wo ks [8], p o ided us wi h he e en ion imes (RTs) o TTX and i s analogues. Values p esen ed o
each compound a e he mean RT and ange om minimum o maximum (in b acke s): TTX = 2.73 min
(2.56–2.85 min); 4-epiTTX = 2.53 min (2.43–2.65 min); 4,9-anhyd oTTX = 2.20 min (2.10–2.34 min);
5,6,11- ideoxyTTX = 1.53 min (1.42–1.61 min); 5-deoxyTTX = 2.02 min (1.92–2.10 min);
11-deoxyTTX = 2.24 min (2.13–2.34 min); 11-no TTX-6(R)-ol = 2.26 min (2.15–2.36 min); and
11-no TTX-6(S)-ol = 2.41 min (2.28–2.49 min) (Figu e 2). RTs o TTXs in es ed samples when
compa ed o hose o he e e ence s anda ds and na u ally con amina ed ma e ials we e always wi hin
he ±2.5% ole ance limi allowed o LC-MS me hods in compliance o he equi emen s o EU Decision
2002/657/EEC [43] o each day o analysis.
Figu e 2. UPLC-MS/MS ch oma og ams (TIC) o : (a) na u ally con amina ed gonads ample
om L. scele a us pu e ish; and (b) he ce i ied e odo oxins (TTXs) e e ence s anda d
(TTX = 2000 ng/mL; 4-epiTTX = 49.86 ng/mL; 4,9-anhyd oTTX = 467.4 ng/mL).
Peak scale has been adjus ed in some o he ch oma og ams o allow p esen a ion o
lowe concen a ion peaks.
Toxins 2015, 7 1787
Table 3. Mean TTX eco e y da a o mussel diges i e gland (DG) and whole lesh (WF)
o i ied a i e concen a ions (n = 6).
TTX Concen a ion and Reco e y Tissue TTX Fo i ica ion Le el
100 μg/kg 200 μg/kg 400 μg/kg 800 μg/kg 1200 μg/kg
Mean TTX de ec ed (μg/kg) DG 90.1 211.2 431.5 833.9 1171.5
Mean Reco e y (%) 90.1 105.6 107.9 104.2 97.6
Mean TTX de ec ed (μg/kg) WF 94.3 209.3 384.3 754.6 1211.0
Mean Reco e y (%) 94.3 104.6 96.1 94.3 100.9
Solu ions o TTXs ce i ied e e ence s anda d showed excellen linea i y wi hin he whole ange o
s anda d concen a ions o all h ee analogues con ained in he e e ence s anda d (TTX, 4,9-anhyd oTTX
and 4-epiTTX), wi h R2 alues highe han 0.995 in all analysis days. The limi s o de ec ion (LOD;
S/N > 3) and limi s o quan i ica ion (LOQ; S/N > 10) o he me hod ob ained om he s anda d solu ions
we e: TTX = 0.6 and 2.2 ng/mL; 4-epiTTX = 0.7 and 2.3 ng/mL and 4,9-anhyd oTTX = 1.9 and
6.5 ng/mL, espec i ely, co esponding o TTX = 2.2 and 7.2 μg/kg; 4-epiTTX = 2.3 and 7.6 μg/kg and
4,9-anhyd oTTX = 6.2 and 21.1 μg/kg in shell ish issue, espec i ely.
T ueness o he me hod, assessed h ough TTX eco e y (%) was wi hin he accep able le els
indica ed by he EU Decision 2002/657/EEC [43] in all i e le els es ed and o bo h ypes o issue
(DG o WF) used in he s udy (Table 3).
Figu e 3. TTXs concen a ions (μg/kg) and mouse bioassay (MBA; Yasumo o 1978
p o ocol) esul s compa ed o P o ocen um minimum concen a ions (cells/L) o he 2012
shell ish oxici y episode in Vis onikos Bay–Lagos, Rodopi, G eece.
2.1.3. Applica ion o he UPLC-MS/MS Me hod o he Mussel Samples o he P. minimum
Episode unde In es iga ion
Resul s o he UPLC-MS/MS analyses o he p esence o TTXs wi h ega d o mussel samples
ob ained om he episode in ques ion a e p esen ed in Table 2. The sum o TTXs concen a ions
(TTX, 4-epiTTX and 4,9-anhyd oTTX) exceeded he le el o 200 μg/kg in all ou consecu i e samples
ha es ed posi i e in he Yas’78 p o ocol whe eas he sum o TTXs in he wo MBA nega i e ones
(one sampled p io and one sampled a e he posi i e ones) we e much lowe , wi h he highes
Toxins 2015, 7 1794
Table 6. Con .
Sampling Da a
P o ocen um
minimum
(cells/L)
MBA (Yasumo o1978) UPLC-MS/MS
Yea
Sample
Code
Da e o Collec ion Regional Uni Si e Species
Resul
(24 h)
Symp oma ology
Tissue
Tes ed
TTX
(μg/kg)
4-epi-TTX
(μg/kg)
4,9-anhyd o-
TTX (μg/kg)
Sum o
TTXs
(μg/kg)
2012 1216/2012 22 No embe 2012 Ka ala Ke amo i Mussels
No da a
p o ided. Bloom
epo ed in he
ollowing weeks
Posi i e
(+ − +)
2/3 mice dead in
4:00–20:00 h, 1/3 mice
wi h bad clinical
condi ion
DG 71.0 <LOQ
(3.4) *
<LOQ
(20.7) * 95.0
2012 1224/2012 10 Decembe 2012 Ka ala Ke amo i Mussels 3.13 × 104 Nega i e
(− − +)
1/3 mice dead in
2:00–18:00 h, 2/3 mice
wi h bad clinical
condi ion and mild
dia hea
DG 71.2 <LOQ
(5.2) *
<LOQ
(11.7) * 88.2
2012 1227/2012 17 Decembe 2012 Ka ala Ke amo i Mussels 1.56 × 104 Nega i e
(− − −)
2/3 mice wi h bad
clinical condi ion and
mild dia hea, 1/3 mice
wi h bad clinical
condi ion
DG 91.8 <LOQ
(4.3) *
<LOQ
(13.2) * 109.2
2012 1241/2012 3 Decembe 2012 Thessaloniki Kymina Mussels 174 (in wide
a ea 1.75 × 105)
Nega i e
(− − −) None WF 79.5
<LOQ
(5.6) *
<LOQ
(18.0) * 103.1
MBA = Mouse bioassay, TTX = e odo oxin, DG = diges i e glands, WF = whole lesh, ND = no de ec ed. Values wi h an as e isk (*) indica e concen a ions be ween he
calcula ed me hod LOD and LOQ and a e shown as con ibu ions o he sum o TTXs.
Toxins 2015, 7 1795
Table 7. Sampling da a and analysis esul s o andom ou ine shell ish samples ob ained in June–July 2014.
Sampling Da a MBA (Yasumo o1978) LC-MS/MS
Yea
Sample
Code
Da e o
Collec ion
Regional Uni Si e Species Resul (24 h) Symp oma ology
Tissue
Tes ed
TTX
(μg/kg)
4-epi-TTX
(μg/kg)
4,9-anhyd o-TTX
(μg/kg)
2014 2303/2014 23 June 2014 Chalkidiki X2: Sou h S ymonikos
Gul —Olympiada Mussels Nega i e (− − −) 3/3 mice wi h dia hea HP/WF <LOQ ND ND
2014 272/2014 30 June 2014 Thessaloniki Kymina Mussels Nega i e (− − −) None HP/WF ND ND ND
2014 2313/2014 30 June 2014 Dy iki A iki/
Mega a D epano Mussels Nega i e (− − −) 3/3 mice wi h bad clinical
condi ion and dia hea HP ND ND ND
2014 2315/2014 30 June 2014 Dy iki A iki/
Mega a Ne aki Mussels Nega i e (− − −) None HP/WF ND ND ND
2014 1303/2014 01 July 2014 Ka ala Ke amo i Mussels Nega i e (− − −) None HP/WF <LOQ ND ND
2014 2323/2014 01 July 2014 Se es S1: Nea Ke dyllia Mussels Nega i e (− − −) None HP/WF <LOQ ND ND
2014 283/2014 07 July 2014 Thessaloniki Kymina Mussels Nega i e (− − −) 1/3 mice wi h mild dia hea HP/WF ND ND ND
2014 2332/2014 07 July 2014 Dy iki A iki/
Mega a D epano Mussels Nega i e (− − −) 3/3 mice wi h bad
clinical condi ion HP ND ND ND
2014 2334/2014 07 July 2014 Dy iki A iki/
Mega a Ne aki Mussels Nega i e (− − −) None HP/WF ND ND ND
2014 2337/2014 07 July 2014 Chalkidiki X2: Sou h S ymonikos
Gul —Olympiada Mussels Nega i e (− − −) 1/3 mice wi h bad clinical
condi ion HP <LOQ ND ND
2014 1307/2014 08 July 2014 Ka ala Ke amo i Mussels Nega i e (− − −) None HP ND ND ND
2014 2342/2014 08 July 2014 Se es S1: Nea Ke dyllia Mussels Nega i e (− − −) 1/3 mice wi h dia hea HP ND ND ND
2014 297/2014 14 July 2014 Thessaloniki Kymina Mussels Nega i e (− − −) 1/3 mice wi h dia hea,
2/3 mild dia hea HP ND ND ND
2014 1311/2014 15 July 2014 Ka ala Ke amo i Mussels Nega i e (− − −) 1/3 mice wi h bad clinical
condi ion HP ND ND ND
2014 2360/2014 15 July 2014 Dy iki A iki/
Mega a D epano Venus clams Nega i e (− + −) 1/3 mice dead in 1:05 h,
1/3 dia hea HP/WF ND ND ND
2014 2368/2014 15 July 2014 Se es S1: Nea Ke dyllia Mussels Nega i e (− − −) None HP/WF ND ND ND
Toxins 2015, 7 1796
2.4. In es iga ion o Random Rou ine Shell ish Samples Ob ained in June–July 2014
Diges i e gland and whole lesh samples om a numbe o andomly selec ed samples ob ained in
June and July 2014 om he same a eas (excep Rodopi) whe e p e ious p esence o TTXs was eco ded
(see Sec ions 2.2 and 2.3) we e also analyzed by UPLC-MS/MS o he p esence o TTXs (Table 7).
TTXs we e no de ec able o below he limi o quan i ica ion in hese samples, some o which se ed
as nega i e issue ma e ial o he o i ica ion expe imen s o he s udy.
3. Discussion
3.1. UPLC-MS/MS Me hod Pe o mance
TTXs e e ence s anda d solu ions exhibi ed excellen linea i y in he de eloped UPLC-MS/MS
me hod o he de e mina ion o TTXs o e he whole calib a ion anges o all h ee analogues, wi h
R2 alues highe han 0.995 o all analysis days. Mean TTX eco e y a he i e o i ica ion le els
(100, 200, 400, 800 and 1200 μg/kg) selec ed in he p esen s udy and o bo h es ed ma ices (mussel
DG and WF) was good, anging be ween 90.1% and 107.9%. These alues a e in compliance wi h he
90%–110% ange indica ed as accep able eco e y le els acco ding o he equi emen s o EU Decision
2002/657/EEC [43], when eco e y is assessed by o i ica ion.
3.2. P esence o TTXs in G eek Shell ish
In he p esen s udy, concen a ions o he sum o TTXs in a o al o 38 shell ish samples selec ed o
UPLC-MS/MS analysis anged om 25.0 μg/kg up o a maximum o 222.9 μg/kg. TTX was he mos
abundan o all TTXs and was p esen in all 38 samples, a concen a ions anging om 25.0 o
202.9 μg/kg. The analogues 4-epiTTX and 4,9-anhyd o TTX we e quan i ied in lowe concen a ions,
in 24 and 19 ou o 38 samples, espec i ely. On he o he hand, 5,6,11- ideoxyTTX and he emaining
analogues we e no ound a quan i iable concen a ions, a ac a ibu able o ei he hei absence in ou
samples, as hey p obably esul by me abolism in shell ish [13], o o possible di e ences in me hod
sensi i i y o hese compounds. I should also be no ed ha a numbe o mussel samples nega i e o he
p esence o TTXs we e ound among he ou ine samples o egula o y moni o ing (da a no shown) and
hese we e used o he o i ica ion expe imen s conduc ed in e ms o me hod eco e y assessmen .
The TTXs le els o ou s udy a e subs an ially highe han hose ecen ly epo ed by Tu ne e al. [13] in
bi al e mollusks (blue mussels and paci ic oys e s) ha es ed in 2013–2014 om he sou h coas o
England, along he Channel. This is expec able, aking in o accoun ha TTX is adi ionally occu ing
in wa me opical o sub opical wa e s and ha G eece is loca ed mo e sou he n compa ed o England,
wi h a subs an ially wa me clima e. Fu he mo e, he p esen wo k is mos ly based on analyses o
a ge ed and speci ically selec ed samples, ei he connec ed o he p esence o P. minimum blooms o
de i ed om coas al a eas wi h a p e ious his o y o TTXs p esen in inc eased le els.
Va ious TTXs le els ha e been epo ed in gas opods and o he bi al e species in Eu ope and
in e na ionally. Rod iguez e al. [12] ound e y high TTXs concen a ions (TTX: 315 mg/kg and
5,6,11- ideoxyTTX: 1004 mg/kg) in he diges i e glands o he umpe shell Cha onia lampas lampas
sample ha es ed om he sou h coas o Po ugal and pu chased om Malaga ma ke . Such inc eased
Toxins 2015, 7 1797
TTXs concen a ions in his speci ic sample, howe e , a e no common bu we e expec able due o he ac
ha his sample was implica ed in a ood poisoning inciden . Indeed, Nzoughe e al. [44] epo ed much
lowe TTX le els o 22.4 and 66.6 μg/kg in isce a and muscle, espec i ely, o na u ally con amina ed
umpe shells om he same species, ha es ed om Angei as Coas , Po ugal. Simila ly, he p esence
o TTXs a low le els was also epo ed in wo mo e gas opod species ha es ed om Po uguese coas s:
Gibbula umbilicalis (monodeoxyTTX: 63.81 μg/kg) and Monodon a linea a (TTX: 90 μg/kg and
4-epiTTX: 21 μg/kg) [45]. The only o he epo ed occu ences o TTX in bi al e mollusks a e hose by
McNabb e al. [10],who ound TTX in he New Zealand clam Paphies aus alis a le els up o 800 μg/kg
and by Kodama e al. [11] who epo ed he p esence o TTX in he Japanese scallop Pa inopec en yessoensis.
Ou s udy cons i u es, o ou bes knowledge, he ea lies epo (Ma ch 2006) o TTXs de ec ion in
Eu opean bi al e mollusks, mussels and enus clams, as well as he i s epo o shell ish ha es ed
in he Medi e anean Sea. Al hough he TTXs le els ound in ou samples a e ela i ely low when
compa ed o o he species implica ed in human in oxica ion inciden s, he possibili y o inc ease in hei
concen a ions in he nea u u e canno be excluded. TTXs p esence in bi al es is no ou inely
moni o ed o egula ed in Eu ope o anywhe e else in he wo ld, gi en he absence so a o published
da a demons a ing a isk o TTX in oxica ion om bi al es. In he p esen s udy, TTX p esence was
e ealed due o posi i e esul s and a ypical esponses in he MBA o lipophilic oxins. Un il mid-2011, his
si ua ion would esul in a egula o y closu e o he a ec ed a ea, hus he consume would be p o ec ed
om exposu e o his oxin. This is no anymo e he case, as due o he eplacemen o MBA by
LC-MS/MS as a e e ence me hod o he p esence o lipophilic oxins, such “ alse posi i e” esul s a e
no de ec able, hus exposing he consume s o TTX and analogues oge he wi h all possible oxins no
de ec able by he cu en ly exis ing ou ine es ing me hods.
Taking in o accoun he absence o any o mal egula o y guidance o he p esence o TTX in
shell ish, he maximum concen a ion o 222.9 μg/kg TTX in he p esen s udy, equa es o ca. 28% o
he maximum pe mi ed le el o saxi oxin (STX) equi alen s (800 μg STX equi alen s/kg shell ish
issue), no ing he simila i y in biological ac i i y be ween he wo oxin g oups. The TTXs le el o
222.9 μg/kg would also equa e o a low le el dose o oxin in compa ison o he p oposed minimum
le hal dose (MLD) o TTX o be ween 0.5 o 2 mg [19], bu would be qui e nea o he minimum acu e
dose (0.2 mg) o induc ion o symp oma ology o humans (w . 50 kg) [3]. In he same con ex ,
consump ion o 500 g o shell ish con amina ed wi h 222 μg/kg o TTXs would equa e o he in ake o
ca. 110 μg TTX, i.e., almos 22% o he p oposed MLD i aken as 0.5 mg TTX o a 60 kg human [46].
Such ough calcula ions, o cou se, do no inco po a e any addi ional sa e y ac o s as applied by he
Eu opean Food S anda ds Agency (EFSA) in hei isk assessmen me hods, aking in o accoun
measu emen o oxici y- ela ed unce ain ies [47], and/o he possibili y o high oxin con en a iabili y
in bulk samples o shell ish ac oss ha es ing a eas.
Consequen ly, while he human heal h isk a ising om he speci ic samples analyzed in ou s udy is
a he low, he e is he po en ial o heal h impac s, especially i TTX le els we e signi ican ly highe a
o he imes o in o he a eas associa ed wi h shell ish ha es ing. I is impo an o no e ha while
bac e ial pa hogens may be elimina ed in shell ish p oduc s ollowing e ec i e cooking, TTXs a e hea
s able and will hus no be des oyed in he ood p epa a ion p ocess [13].
Toxins 2015, 7 1798
3.3. Compa ison o Te odo oxin o Vene upin Shell ish Poisoning Toxin
Un o una ely no unp ese ed seawa e samples con aining ali e P. minimum cells we e a ailable in
o de o ob ain cul u es o he speci ic s ain causing he 2012 episode in Rodopi, which would allow an
in-dep h s udy o i s oxin p oduc ion po en ial. Ce ain esul s om he p esen s udy, howe e ,
combined wi h published indings om p e ious wo ks, poin o he suspicion ha Vene upin Shell ish
Poisoning (VSP) oxin associa ed wi h P. minimum blooms could be ela ed o e odo oxin. Rele an
ac s suppo ing his hypo hesis would be: (a) he inc ease and decline o TTXs concen a ions in mussels
o he 2012 episode in Rodopi almos in pa allel wi h inc ease and decline o P. minimum concen a ions
(Table 2 and Figu e 3); (b) he p esence o TTXs in a numbe o samples selec ed acco ding o ins ances
o P. minimum blooms a le els highe han 60 μg/kg and up o a maximum o 194.7 μg/kg (Table 6);
and (c) he p esence o TTXs a much lowe le els (<50 μg/kg) in a coas al a ea wi h a p e ious his o y
o TTXs p esence in inc eased le els, when no blooms o P. minimum we e eco ded (Table 5).
Whe e da a a ailable in he bibliog aphy a e conce ned, impo an ac s include: (a) p e ious
inc imina ion o P. minimum o se e al episodes o human poisoning subsequen o he consump ion o
shell ish, wi h a symp oma ology cha ac e is ic o pa aly ic shell ish poisoning (PSP) [26–29], which is
a he di icul o dis inguish wi h ha o TTX poisoning [19]; (b) mussel oxici y associa ed o he
p esence o P. minimum by a wa e soluble neu o oxic componen , di e en o PSP bu also inhibi ing
calcium channels, ha apidly killed mice [37]; (c) he subsequen isola ion o he same oxic ac ion
om shell ish du ing a la ge P. minimum [38]; and (d) he s ain speci ic and s imula ed by associa ed
bac e ia cha ac e o he oxin p oduc ion by P. minimum [37], also aking in o accoun he gene ally
accep ed bac e ial o igin o e odo oxin. All hese indings a e in ag eemen wi h ou obse a ions wi h
ega d o he hyd ophilic and neu o oxic na u e o he compound causing shell ish oxici y in he 2012
episode in Vis onikos Bay, whe e TTX was de ec ed. Howe e , i would be a he p ema u e o conclude
ha TTX and VSP a e indeed iden ical, as a numbe o expe imen s would be necessa y o e i y his
hypo hesis. Upon occu ence o a u u e P. minimum oxic episode, he e o e, u he esea ch will ha e
o be unde aken, including ob aining seawa e ield samples, es ablishing cul u es o he speci ic s ain
and TTX analysis o he expe imen al ma e ials in o de o elucida e he ac ual mode o P. minimum oxin
p oduc ion and o es ablish he exac , i any, ole o his species ega ding he p esence o TTXs in shell ish.
4. Expe imen al Sec ion
4.1. S anda ds and Reagen s
A ce i ied e e ence s anda d o e odo oxins (CRM-03-TTXs; pu i y >96%), pu i ied om pu e
ish, was ob ained om Labo a o io CIFGA S.A. (Lugo, Spain). The s anda d has a ce i ied e e ence
concen a ion o 25.7 ± 2.1 μg/mL o e odo oxin (TTX) and 3.0 ± 0.4 μg/mL o 4,9-anhyd o
e odo oxin (4,9-anhTTX) and a non-ce i ied e e ence concen a ion o 0.32 ± 0.08 μg/mL o 4-epi
e odo oxin (4-epiTTX). Ce i ied e e ence s anda ds o lipophilic oxins analysis (okadaic acid
(NRC-CRM-OA-c), yesso oxin (NRC-CRM-YTX), pec eno oxin-2 (NRC-CRM-PTX2), azaspi acid-1
(NRC-CRM-AZA1), gymnodimine-A (NRC-CRM-GYM) and 13-desme hyl-spi olide C (NRC-CRM-SPX1)),
we e pu chased om he Na ional Resea ch Council o Canada (Hali ax, Canada). The p e-ce i ied
e e ence s anda d o pinna oxin-G (RM-PnTX-G) was a gene ous gi om Pea ce McCa on (NRC,
Toxins 2015, 7 1799
Hali ax, Canada). Ace ic acid was o analy ical g ade, me hanol o HPLC g ade, ace oni ile and wa e
o LC-MS g ade, whe eas o mic acid and ammonium o ma e we e o mass spec ome y g ade.
Chemicals we e ob ained om Fishe Scien i ic UK (Loughbo ough, UK), Pan eac (Ba celona, Spain),
Fluka (Sigma Ald ich g oup, Schnelldo , Ge many) and Al a Aesa (Ka ls ube, Ge many).
4.2. Sampling and Tes Ma e ials
Samples and espec i e da a on P. minimum coun s and p esence o lipophilic oxins we e de i ed
om he G eek “Na ional P og amme o Moni o ing o Bi al e Molluscs” P oduc ion A eas o he
p esence o Ma ine Bio oxins” du ing he yea s 2006–2014; coo dina ion was p o ided by he Minis y
o P oduc i e Recons uc ion, En i onmen and Ene gy ( o me Minis y o Ru al De elopmen and
Food [48,49]). All samplings we e conduc ed by he ele an P e ec u al Ve e ina y and/o Fishe ies
au ho i ies. Cell coun s o oxic and/o po en ially oxic phy oplank on in seawa e , and speci ically o
he species P. minimum, we e conduc ed by he Labo a o y Uni o Toxic Ma ine Mic oalgae (LUTMM),
Depa men o Biology, A is o le Uni e si y o Thessaloniki (scien i ic coo dina o : G. Nikolaidis (un il
Feb ua y 2010) and M. A senakis (Ma ch 2010– o da e), and da a we e o wa ded o he NRLMB
ini ially by he p e ec u al au ho i ies and la e on by he conduc ing labo a o y, as indica ed by he
moni o ing p og am equi emen s. Mussel (My ilus gallop o incialis) samples and addi ional seawa e
samples, in he case o he main in es iga ed episode o Rodopi we e collec ed om he sampling si e
“Zone 6” in Vis onikos Gul , No h Aegean Sea (Figu e 1). The shell ish samples es ed om p e ious
yea s (mussels: My ilus gallop o incialis and clams: Venus e ucosa) we e ob ained om a ious
sampling si es along he G eek coas s. Random samples om hese la e si es, ob ained in June and
July 2014, we e also es ed in o de o check TTXs backg ound le els in hese a eas and o p o ide
nega i e issue ma e ial o o i ica ion expe imen s (see Figu e 1 and Tables 1, 5–7).
4.3. Seawa e Analyses
Seawa e samples we e analyzed o he p esence o oxic and/o po en ially oxic phy oplank on by
U e möhl’s sedimen a ion me hod [50], using in e ed mic oscopy o iden i ica ion and enume a ion o
he phy oplank on cells. B ie ly, a e ho ough homogeniza ion, sub-samples we e allowed o se le
on 25 mL sedimen a ion chambe s. Cells we e coun ed in an in e ed mic oscope, a 100× and 400×
magni ica ions. I should be no ed ha un il No embe 2012, indi idual concen a ions o P. minimum
cells we e no egula ly epo ed by he es ing labo a o y (LUTMM), as hey we e no pe cei ed as
po en ially oxic in he con ex o he bi al e mollusks’ moni o ing p og am, bu only men ioned as
esponsible o ich hyo oxici y. Fo his eason, addi ional seawa e samples ob ained o in es iga ion
o he 2012 Rodopi episode had o be enume a ed ad hoc du ing he episode de elopmen , in o de o
ob ain a ull da ase wi h ega d o P. minimum p esence o each sample ob ained a ha pe iod.
4.4. Shell ish P epa a ion and Sub-Sampling
Shell ish samples we e washed and shucked upon a i al o he labo a o y; diges i e glands (DG) and
whole lesh (WF) we e sepa a ed and homogenized. All samples we e es ed o he p esence o
lipophilic oxins by mouse bioassay (MBA), acco ding o he p o ocol o Yasumo o e al. (1978) [39]
Toxins 2015, 7 1800
(see Sec ion 4.5), whe eas a numbe o samples we e also u he in es iga ed using he EU-ha monized
MBA p o ocol o he Eu opean Re e ence Labo a o y on Ma ine Bio oxins ( e sion 5; [42]). Acco ding
o he NRLMB labo a o y p ac ice, DG and/o WF sub-samples we e e ained o each es ed sample,
o a leas one yea a e he analysis. Howe e , in he case o samples wi h posi i e esul s o nega i e
esul s wi h se e e symp oma ology, wi h an emphasis being paid on a ypical esponses, sub-samples
a e e ained as a ule o inde ini e ime unde deep eezing condi ions (−70°C), o he pu pose o u he
in es iga ion in due ime.
4.5. Mouse Bioassay Tes s
Mussel DG samples ( ou ine and o i ied wi h TTX a a ious le els) we e es ed o he p esence o
lipophilic oxins by he mouse bioassay (MBA) p o ocol o Yasumo o e al. (1978). B ie ly, a 20 g
po ion o DG was ex ac ed h ice wi h 50 mL ace one each ime and il e ed h ough a cellulose il e .
The combined oxin ex ac was e apo a ed o d yness and esuspended in 1% Tween-60 o a inal
olume o 4 mL (5 g DG/mL). Each one o h ee mice (Albino Swiss, 18–20 g body weigh ) was injec ed
in ape i oneally wi h 1 mL o his solu ion.
Whe e u he in es iga ion was equi ed, mussel whole lesh samples we e also es ed o he
p esence o lipophilic oxins using he ha monized EU-MBA p o ocol ( e sion 5; [42]). An aliquo o
100 g o whole lesh issue homogena e was weighed in o 500 mL plas ic cen i uge con aine s (Sigma
Labo zen i ugen GmbH, Os e ode am Ha z, Ge many), o which ace one (300 mL) was added.
The mix u e was homogenized o 2 min using an Ul a u ax (15 mm sha , 10,000 pm; IKA,
S au en, Ge many) and cen i uged a 1200× g o 10 min a 4 °C (Cen i uge Sigma 4K15C; Sigma
Labo zen i ugen GmbH, Os e ode am Ha z, Ge many). The supe na an was ans e ed h ough a il e
pape in o a labeled 1 L ound-bo om lask, while he issue esidue was e-ex ac ed in he same way
wi h 200 mL o ace one. The combined 500 mL il a e was o a y e apo a ed unde acuum a
42 °C ± 2 °C (Ro a apo R-200, Buchi, Flawil, Swi ze land), un il comple e emo al o ace one. The
aqueous esidue o he ound-bo om lask was ans e ed o a 100 mL glass cylinde and olume was
adjus ed o 100 mL by addi ion o de-ionized wa e , while 100 mL o die hyl e he we e used o inse
o he esidue o he ound bo om lask. Bo h aqueous and die hyl e he ex ac s we e ans e ed o a
500 mL sepa a o y unnel. The mix u e was agi a ed and ollowing sepa a ion o he aqueous and die hyl
e he phases, he aqueous laye was ans e ed back in o he ound-bo om lask. The wa e was
e-ex ac ed wice wi h 100 mL o die hyl e he and sepa a ed in he sepa a o y unnel yielding ca.
300 mL o die hyl e he ex ac , which was backwashed wice wi h 20 mL de-ionized wa e . The e he
phase was collec ed in ano he ound-bo om lask and o a y e apo a ed o d yness unde acuum a
42 °C ± 2 °C. The d y esidue was esuspended in 1% Tween-60 o a inal olume o 4 mL
(25 g whole lesh/mL). When necessa y, he esuspended ex ac was u he homogenized using an
Ul a u ax (8 mm sha ) p io o injec ion. Each one o h ee mice (Albino Swiss, 18–20 g body weigh )
was injec ed in ape i oneally wi h 1 mL o his solu ion.
In bo h p o ocols, he c i e ion o oxici y es ablished by he EU Regula ion 2074/2005 (2005) was
employed, which is he dea h o wo ou o h ee mice wi hin 24 h o injec ion wi h an ex ac equi alen
o 5 g o DG o 25 g o whole lesh issue. This cons i u es a posi i e esul o he p esence o he
lipophilic oxins men ioned in he abo e Regula ion. The mice we e allowed labo a o y eed and wa e
Toxins 2015, 7 1801
ad libi um h oughou he obse a ion pe iod. All animal manipula ions we e pe o med in acco dance
wi h he EU Di ec i e 86/609/EEC (1986) and he EU Recommenda ion 2007/526/EC (2007),
unde o icial license om he P e ec u al Ve e ina y Se ice o Thessaloniki, G eece (licensed acili ies:
o conduc ion o MBAs no. EL54BIO07 and o b eeding mice no. EL54BIO08).
4.6. UPLC-MS/MS Analysis (Ul a-Pe o mance Liquid Ch oma og aphy Coupled o Tandem
Mass Spec ome y)
4.6.1. Sample T ea men
Aliquo s o 2 g o sample issue (DG o WF) we e ex ac ed in 6 mL o 1% ace ic acid in me hanol
by o ex mixing o 5 min (Mul i Reax, Heidolph Ins umen s, Schwabach, Ge many), hen u he
homogenized o 1 min using an Ul a u ax (18 mm sha , 24,000 pm; IKA, S au en, Ge many) and
ul asonic wa e ba h (10 min, 40 kHz) (T910DH, Elma T anssonic Digi als, Singen, Ge many). The
ubes con aining he homogenized ex ac s we e igh ly closed wi h pa a ilm® (3M) o p e en sol en
e apo a ion and we e placed in a wa e ba h o 10 min a 100°C. A e cooling o oom empe a u e,
olume was adjus ed o 7.5 mL, whe e necessa y, wi h he me hanolic ace ic acid solu ion and ex ac s
we e cen i uged a 3000× g o 10 min a 4°C (4K15C, Sigma Labo zen i ugen GmbH, Os e ode am
Ha z, Ge many). The supe na an s we e il e ed h ough 0.20 μm me hanol compa ible nylon sy inge il e s
(Ti an 3, The mo Scien i ic, Nash ille TN, USA) and we e placed in ials o UPLC-MS/MS analysis.
4.6.2. UPLC-MS/MS Me hod
An Acqui y UPLC sys em coupled o a TQD mass spec ome e (Wa e s, Manches e , UK) equipped
wi h a Z-Sp ay ESI sou ce was used o sample analysis. The UPLC sys em was equipped wi h an
Acqui y UPLC BEH Amide column (2.1 × 150 mm, 1.7 μm) and an Acqui y UPLC BEH VanGua d
P e-Column (2.1 × 5 mm, 1.7 μm, 130 Å), wi h column o en main ained a 25 °C. The mobile phase
consis ed o 100% wa e in channel A and 95% ace oni ile (ACN) in channel B, bo h con aining 2 mM
ammonium o ma e and 50 mM o mic acid. An isoc a ic elu ion was used wi h a low a e o 0.4 mL/min
a a a io o 35% A and 65% B o a o al o 5 min and he injec ion olume was 10 μL (pa ial loop wi h
needle o e ill mode). The mass spec ome e ope a ed in elec osp ay posi i e mode wi h he capilla y
ol age se a 2.8 kV. The TQD mass spec ome e ope a ed wi h he ollowing op imized sou ce-dependen
pa ame e s (ESI sou ce): capilla y ol age 2.8 kV, cone ol age 40 V, desol a ion empe a u e 350 °C,
desol a ion gas low 800 L/h N2, cone gas low 50 L/h N2, sou ce empe a u e 120 °C, collision gas low
0.10 mL/min A gon. The cone ol age and collision ene gy we e op imized o TTX by s anda d
in usion. MassLynx 4.1 wi h QuanLynx so wa e (Wa e s, Manches e , UK) was used o da a p ocessing.
The mass spec ome e ope a ed in MRM, de ec ing in posi i e mode, analyzing a leas wo p oduc
ions pe compound: one o quan i ica ion ( he mos abundan ) and ano he o con i ma ion. The
ansi ions employed we e: TTX and 4-epiTTX (m/z 320 > 302/162/60), 4,9-anhTTX (m/z 302 > 256/162),
5,6,11- ideoxyTTX (272 > 254/162), 11-deoxyTTX and 5-deoxyTTX (304 > 286/176) and 11-no TTX-
6(S)-ol and 11-no TTX-6(R)-ol (290>272/162) and we e selec ed acco ding o Yo su-Yamashi a e al. [51],
Rod iguez e al. [8] and McNabb e al. [10] ollowing op imiza ion, whe e possible, by in usion o
e e ence s anda ds.
Toxins 2015, 7 1802
Fo he calib a ion cu e, se ial dilu ions om he ce i ied TTXs s ock s anda d we e pe o med
using 1% ace ic acid in me hanol o p o ide eigh calib a ion s anda ds. The concen a ion anges we e:
TTX: 15.63–2000 ng/mL, 4-epiTTX: 0.39–49.86 ng/mL and 4-9-anhTTX: 3.65–467.4 ng/mL. These
compounds we e quan i ied using hei peak a eas o calcula e amoun s and using he ele an cu e
ob ained om he s anda d, while he emaining analogues we e quan i ied using he TTX cu e and
assuming an equal mola esponse ac o . TTXs concen a ions in samples we e exp essed in μg/kg. The
calib a ion ange was chosen in o de o include concen a ions o TTX cu en ly ound in a ious aqua ic
species [7,8,12,13,44,45,52–54] and also o encompass he EU egula o y limi (800 μg/kg) o PSP oxins
se by Regula ion (EC) no 853/2004 [55]. To o e come he challenge o he lack o s anda ds o some
o he TTX analogues, gonad and muscle samples o a na u ally-con amina ed Lagocephalus scele a us
pu e ish used in o me wo ks [8] we e injec ed in he UPLC-MS/MS, his way de e mining he
espec i e e en ion imes (RT) (Figu e 2).
4.6.3. UPLC-MS/MS Me hod Quali y Con ol
Nega i e mussel DG and WF issues (i.e., ee om TTX) we e o i ied a i e concen a ions,
i.e.,100, 200, 400, 800 and 1200 μg/kg, equi alen o, espec i ely, 0.125, 0.25, 0.5, 1 and 1.5 imes he
egula o y limi o PSP oxins o he pu pose o e alua ing he me hod pe o mance. Fo i ied ma e ials
we e ex ac ed ollowing he ex ac ion p ocedu e desc ibed in Sec ion 4.6.1 wi h six independen
eplica es o each concen a ion. The mean me hod eco e y o each concen a ion and issue was
calcula ed. Linea i y o calib a ion cu es, as well as limi o de ec ion (LOD) and limi o quan i ica ion
(LOQ) based on signal o noise a ios o all h ee analogues included in he ce i ied TTXs s anda d
we e also assessed.
5. Conclusions
The p esen wo k, o ou knowledge, cons i u es he ea lies epo (Ma ch 2006) o he p esence o
TTXs in Eu opean bi al e mollusks, mussels and enus clams, as well as he i s epo o shell ish
ha es ed in he Medi e anean Sea. Taking in o accoun he e idence p esen ed o TTXs occu ence in
Eu opean bi al e mollusks, and he adi ional occu ence o hese oxins in wa m opical wa e s,
an impo an ques ion is whe he his is linked o clima e change. TTXs ha e un il ecen ly been assumed
no o occu in bi al e mollusks, pa icula ly in empe a e wa e s. Since hese bi al e species a e la gely
consumed, he p obabili y o human heal h haza ds seems o be inc easing. The e o e, despi e he low
concen a ions de ec ed, anging om 25.0 o 222.9 μg/kg, i is e iden ha he p esence o TTX and
analogues should be moni o ed in all species ha can po en ially accumula e he oxins and can be used as
human ood, especially in combina ion wi h moni o ing o P. minimum blooms.
In he p esen wo k, p elimina y indica ions we e ound o he exis ence o a possible link be ween
P. minimum blooms and p esence o TTXs in shell ish. Al hough epo ed associa ion be ween
P. minimum blooms and human oxici y a e a e, u he esea ch is equi ed, when ano he episode o
P. minimum combined wi h shell ish oxici y e-occu s in he u u e, in o de o elucida e he mode o
P. minimum oxin p oduc ion and o es ablish he exac , i any, ole o his species ega ding he p esence
o TTXs in shell ish.
Toxins 2015, 7 1803
Acknowledgmen s
The esea ch leading o USC esul s has ecei ed unding om he ollowing FEDER co unded-g an s.
F om CDTI and Technological Funds, suppo ed by Minis e io de Economía y Compe i i idad,
AGL2012-40185-CO2-01 and Conselle ía de Cul u a, Educación e O denación Uni e si a ia,
GRC2013-016, and h ough Axencia Galega de Inno ación, Spain, ITC-20133020 SINTOX. In addi ion
om he Eu opean Union’s Se en h F amewo k P og amme managed by REA—Resea ch Execu i e
Agency (FP7/2007–2013) unde g an ag eemen 315285 CIGUATOOLS and 312184 PHARMASEA.
Inés Rod iguez is suppo ed by a ellowship om Subp og ama de Fo mación de Pe sonal In es igado
(AGL2012-40185-CO2-01), Spain.
In dep h in es iga ion o he oxic episodes leading o he esul s and publica ion o he p esen wo k
was unde aken by he G eek Na ional Re e ence Labo a o y o Ma ine Bio oxins (NRLMB) o ul ill
he equi emen s o EU Regula ion 178/2002/EC (A icles 6 and 7) ega ding isk analysis and
communica ion and scien i ic in o ma ion needed o isk assessmen and EU Regula ion 882/2004/EC
(a icle 7) wi h ega d o anspa ency and in o ma ion o he public. Collabo a ion o all he s a o he
NRLMB is g ea ly app ecia ed. Thanks a e also exp essed o all he G eek egional e e ina y se ices
o hei con ibu ion o he shell ish samplings and o p o ision o he seawa e analyses esul s o he
p esence o po en ially oxic mic oalgae. The use o cell coun s’ da a wi hin he pe iod 2006–2009 and
2012 ega ding P. minimum p esence in seawa e , de i ed om he G eek “Na ional P og amme o
Moni o ing o Bi al e Molluscs’ P oduc ion A eas o he p esence o Ma ine Bio oxins” and conduc ed
by he Labo a o y Uni o Toxic Ma ine Mic oalgae (LUTMM), Depa men o Biology, A is o le
Uni e si y o Thessaloniki (scien i ic coo dina o : G. Nikolaidis (un il Feb ua y 2010) and M. A senakis
(Ma ch 2010– o da e)), as well as he es ic ions o LUTMM ega ding he use o da a p oduced wi hin
2013–2015 due o con ac e ms and ISO 17025 equi emen s a e acknowledged. Thanks a e also
exp essed o he anonymous e iewe s o he manusc ip o all he imp o emen s in oduced acco ding
o hei p oduc i e commen s.
Au ho Con ibu ions
A is idis Vlamis conduc ed he seawa e analyses o e-enume a ion o he p esence o P. minimum
o he 2012 Rodopi episode in he addi ional seawa e samples aken. He was esponsible o all
p epa a ions o s anda ds and shell ish samples and he ele an analyses by UPLC-MS/MS o he
p esence o TTXs. He handled UPLC-MS/MS and mic oalgae da a p ocessing and in e p e a ion and
also con ibu ed o expe imen al design, UPLC-MS/MS me hod de elopmen , he manusc ip
p epa a ion and comple ion. Panagio a Ka ikou was esponsible o he expe imen al design, con ibu ed
o UPLC-MS/MS me hod de elopmen and manusc ip p epa a ion and handled he submission.
Angelos Papazacha iou pa icipa ed in UPLC-MS/MS me hod de elopmen and was esponsible
oge he wi h The is Zacha aki and Panagio a Ka ikou o he conduc ion and da a collec ion o all MBA
analyses. Inés Rod iguez, Ve ónica Rey and Ampa o Al onso we e in cha ge o USC analysis
LC-MS/MS and HPLC-FLD). Luis M. Bo ana was co- esponsible o he expe imen al design and
o e all supe ision o he esea ch wo k included in his manusc ip .
