Dinoflagellate Gymnodinium catenatum as the source of paralytic shellfish toxins in Tasmanian shellfish

Toxicon, Vol. 2S, No. I0, pp. 1105-1111, 1987. Printed in Great Britain.

G041-0101/87 $3.(I0"+ .~) © 1987 Pergamon Jourmds Ltd.

DINOFLAGELLATE GYMNODINIUM CA TENA TUM AS THE SOURCE OF PARALYTIC SHELLFISH TOXINS IN TASMANIAN SHELLFISH YASUKATSU OSHIMA, t MAMIKO HASEGAWA, t TAKESHI YASUMOTO, ~ GUSTAAF HALLEGRAEFF2 a n d SUSAN BLACKBURN2 1Department of Food Chemistry, Tohoku University, Tsutsumidori-Amamiya, Sendal, Miyagi 980, Japan, and 2Divisionof Fisheries Research, CSIRO Marine Laboratories, G.P.O. Box 1538, Hobart, Tasmania 7001, Australia (Accepted for publication 8 April 1987)

Y. OSmMA, M. HASEGAWA,T. YASUMOTO,G. HALLEGRAEFFand S. BLACKBURN.Dinoflagellate Gymnodinium catenatum as the source of paralytic shellfish toxins in Tasmanian sheUf'mh. Toxicon 25, 1 1 0 5 - l l l l , 1987. ~ Paralytic shellfish toxins in both cultured cells and natural phytoplankton blooms of the dinoflagellate Oymnodinium catenatum from inshore Tasmanian waters (Australia) were analyzed by high performance liquid chromatography, thin layer chromatography and electrophoresis techniques. The dinoflagellate toxins were dominated by low potency sulfocarbamoyl saxitoxin derivatives (98-99 mole~o in total), including gonyautoxin VIII (C2) and its epimer (CI) and suifocarbamoyl gonyautoxins 1 and IV (C3 and C4). Mussels and oysters contaminated by the dinoflagellate showed similar toxins, but contained larger proportions of C3 (40- 57 mole~e) and more potent carbamate toxins (7 - 23 mole% total).

INTRODUCTION

Gymnodinium catenatum Graham is a tittle known unarmoured, chain-forming dinoflagetiate. When it was first described from the Gulf of California (GRAHAM, 1943) and later from Argentinian waters (BALECH, 1964), this plankton species was not known to be toxic. Since then this organism has been circumstantially linked with paralytic shellfish poisonings (PSP) in Mexico (MOREY-GAINES, 1982; MEE et al., 1986) and Spain (ESTSADA et al., 1984; FRAOA and SANCHEZ, 1985). From a red tide event in Mexico in 1979, which caused three human deaths, MEE et al. (1986) confirmed the presence of a neurotoxin in natural dinoflagellate material using mouse bioassay. However, the precise chemical composition of the toxin produced by this dinoflagellate is still unknown. In the period J a n u a r y - June 1986, dense blooms (104 - 10~ cells/liter) of G. catenatum occurred in Australia in inshore southern Tasmanian waters. Subsequent monitoring for toxins in commercial shellfish by mouse bioassays showed that mussels, oysters and scallops from areas affected by the dinoflagellate blooms were highly contaminated with paralytic shellfish toxins (up to 8000 ~g saxitoxin equivalent/100 g shellfish). Two human poisonings with typical PSP symptoms progressing from numbness of the tips, fingertips and toes to mild respiratory problems were also reported (HALLEGRAEFFand SUMNER, 1986). This incident prompted the present detailed chemical study of the toxin produced 1105

Y. OSHIMA et al.

1106

b y G. c a t e n a t u m . T o x i n s in n a t u r a l b l o o m o r g a n i s m s , as well as in a r t i f i c i a l l y c u l t u r e d dinofiagellates, were analyzed by high performance liquid chromatography (HPLC) techniques which have been specially developed for paralytic shellfish toxins. The toxin compositions of Tasmanian mussels and oysters contaminated by the dinoflagellate were c h a r a c t e r i z e d in o r d e r t o assess p o t e n t i a l h e a l t h risks f o r h u m a n s h e l l f i s h c o n s u m e r s . MATERIALS AND METHODS

Dinoflagellates The dinoflagellate Gymnodinium catenatum was isolated from the Derwent Estuary, Tasmania, on 9 January 1986. The non-clonal algal cultures were grown in stationary 1500 ml Erlenmeyer flasks at 20°C under 108 ¢,tEinstein/sec/m2 white fluorescent light and 12 hr:12 hr light:dark cycles. The culture medium consisted of filtered Derwent estuary seawater (34%0 salinity) with nutrients added at 1/5 dilution of medium G of LOEeLICtl and SMITtl (1968). Cultures grown at full strength seawater contained predominantly single cells and only short ( 4 - 5 cells) chains (culture 1), whereas cultures grown at lower salinities (23-26%0 ) showed predominantly chain-forming (up to 16 cells) dinoflasellate (culture 2). Natural dinoflagellate blooms were collected with a vertical 37/an mesh free-fall plankton net from the Derwent estuary on 18 April and from the adjacent Huon estuary on 5 June 1986 (see HALLI~GRAEFFand SUMNER, 1986, for sample locations), In both field samples G. catenatum comprised about 80°70 of the total plankton hiomass, with the dinoflagellates Ceratiumfurca, C. fusus and C. tripos present as contaminants. All samples were frozen at - 20°C immediately after collection, air freighted to Japan under dry ice and kept at - 2 0 ° C until further analysis. SheiO'tsh One sample (20-40 animals) of wild, intertidal mussels (Mytilus edulis plamdatug) was collected from Port Cygnet (Huon estuary) on 24 February and further samples of longline cultured muuels were collected from Killala Bay on 28 April and from Deep Bay on 5 JuRe 1986 (the latter are commercial muuel farms located in the Huon ettuary). Another sample of wild, intertidal oysters (Crassostrea g~as) was aim collected from Port Cygnet on 24 February. The Port Cygnet samples represent the uncooked left-overs of a shellfish meal which caused unambiguous PSP symptoms in two adult humans who had consumed three - four dozen raw oysters and boiled mussels in conjunction with alcohol. All shellfish specimens were frozen at - 20°C and air freighted to Japan under dry ice. High performance liquid chromatographic analysis of toxins Dinofla~llate cells were suspended in 0.1 N acetic acid, homogenized for 5 min with a Nihonseiki US-50 sonicator and cmttrifoged for 5 rain at 6000 8. Shellfish toxins were extracted according to the method of the Association of Official Analytical Chemists for mouse bioassays (WILLIAMS, 1984). HPLC toxin analyses were carried out as described by OSH.~h~ et ai. (198,0 and SULLIVANet ai. (1985) with some modifications. Three chromatographic systetm were used to separate the following three toxin categories: (I) 8onyautoxin VIII and its ll-epimer (GTX, and e4~TX6; Fig. 1) were analysed on Develmil C8 column (4.6 x 250 mm; Nomura Chemical) truing a mobile pha~ of I mM tetrabutylammonium phosphate 0PIC A; Waters) adjusted to pH 6.5 with acetic acid; (2) gnnyautoxins I - VI (GTX, - GTXd Fig. 1) were analyzed on a Deveiosil C8 column using a

R1 /O R4

"~

H

H R I N ' ~ " ~ " N X_.~ ~,,

H ~//~'N~ ~"

2

~......OHOH

R2

R3

gonyautoxin 1 gonyautoxin I1 gonyautoxin I11 gonyautoxin IV gonyautoxin V gonyautoxin VI gonyautoxin VIII (CI) epi-gonyautoxin Vlll (C2) C3 C4 saxitoxin neosaxitoxin decarbamoylsaxitoxin

OH H H OH H OH H H OH OH H OH H

R2 H H OSO~OSO~ H H OSO/H H OSO/H H H

FIG. 1. STRUCTURESOF PARALYTICSHELLFISHTOXINS.

R3 {)SO/ OSOjH H H H H OSOjOSOT H H H H

R4 CONH2 CONH2 CONH2 CONH2 CONHSO~ CONHSChCONHSO/ CONHSO/" CONHSOjCONHSO~CONH2 CONH2 H

PSP Toxins in Gymnodinium catenatum

110"/

mobile phase consisting of 1 mM l-heptanesulfonic acid (PIC BT; Waters) and 10 mM acetic acid adjusted to pH 7.2 with tris(hydroxymethyl)aminomethane; (3) saxitoxin (STX), neosaxitoxin (neoSTX) and decarbamoylsaxitoxin (dcSTX) were resolved on a Develosil ODS column using a mobile phase consisting of 1 mM l-hexanesulfonic acid (PIC B6; Waters), 20 mM acetic acid and 5 mM phosphoric acid adjusted to pH 7.5 with NH,OH. All mobile phase flow rates were 0.8 ml/min. A solution of 7 mM periodic acid and 50 mM sodium phosphate adjusted to pH 12 with NaOH was introduced to the eluate at a flow rate of 0.8 ml/min and the resulting mixture heated while passing through Teflon tubing (0.5 rnm internal diameter, 10 m) immersed in a water bath (75°C). The reaction mixture was acidified with 0.5 N acetic acid and the resulting fluorescent derivatives were monitored at 410 nm wavelength with a Shimadzu RF-530 fluoromonitor using 340 nm as excitation wavelength. Samples (5 - 10 ~) of the extracts were injected into the HPLC system either directly or after being heated for 5 rain with an equal volume of 0.4 N HCI in a boiling water bath (compare Figs 2 and 3). Standard toxins from Protogonyaulax tamarensis (OsHIMA et al., 1984) and Pyrodinium bahamense var. compressa (HARABA et al., 1982) were used as references.

Thin-layer chromatographic and electrophoretic analysis o f C3 and C4 toxins Extracts from Deep Bay mussels were treated on a charcoal column and the toxins eluted with 1070acetic acid in 20°70 aqueous ethanol were chromatographed on a Bio-Gel P-2 column (2.0 x 90 era; Bio-Rad) using 0.05 N acetic acid as eluant. Fractions of 10 ml were collected and analyzed by HPLC as described above. After lyophilization, the fractions were analyzed on Silicagel 60 thin layer plates (Merck) using a solvent system of pyridine/ethyl acetate/acetic acid/water (15:7:3:6) and analyzed by electrophoresis on cellulose acetate strips (Separax; Fuji Film Co.) in 83 mM Tris - HCI buffer (pH 8.7) for 20 rain at 0.6 raA/cm. Toxins were viewed under u.v. light (365 nm) after spraying with 1070hydrogen peroxide and heating for 5 rain at 120°C (HAnADA et al., 1982). Mouse bioassay The lethal potency of the dinoflageilate and shellfish samples was measured at Tohoku University by the standard mouse bioassay 0VILLtAMS, 1984) using ddY strain, male mice. With this technique, I mouse unit (MU) is defined as the amount of toxin to kill a mouse weighing 20 g in 15 rain.

RESULTS

Both cultured and wild cells of the dinoflagellate G. catenatum showed high lethal potencies (300- 1300 MU/g wet weight) in the standard mouse test. Figure 2 illustrates H P L C chromatograms of the toxins in the Huon Estuary bloom material, with GTXs and epiGTXs being dominant (Fig. 2a) and trace amounts (< 1 mole070) of GTX4, GTXs (Fig. 2b) and dcSTX (Fig. 2c) also present. Following acid treatment of the same extract, additional GTX~, GTX2 and GTX3 peaks appeared on the chromatogram (Fig. 3b) and GTX+ was now the dominant toxin. GTX3 and GTX2 apparently derived from GTXg and epiGTXs, respectively, through hydrolysis of the side chain sulfonate. The formation of GTX~ and GTX4 is indicative of the presence of N-sulfocarbamoyl derivatives of GTX~ and GTX+ (C3 and C4, respectively; HALL and REICHARDT, 1984) in the original extract. These toxins could not be identified by HPLC, due to the absence of the standard toxins, and their concentration were therefore estimated from the amount of GTXI and GTX4 generated through acid treatment. Further confirmation on the occurrence of C3 and C4 toxins were obtained through TLC and electrophoretic analysis. Two components eluted at 110- 140070 and 130- 150070 bed volume from a Bio-Gel P-2 column showed greenish yellow fluorescent spots at Rl 0.54 and 0.50 on TLC. They migrated to the cathode on electrophoresis, showing a mobility relative to saxitoxin of - 0 . 4 5 and -0.50, respectively. All these data are comparable to those values reported for C3 and C4 in previous studies (HALL and REICHARDT, 1985; NOGUCHI et al., 1982). In H P L C using tetrabutylammonium as counter ion C3 and C4 toxins were detected at retention times of 19 and 28 rain, respectively, equal to that of two unidentified peaks as shown in Fig. 2a. Toxin compositions (in mole070) of the Tasmanian dinoflagellate and shellfish samples analysed are summarized in Table 1. The Derwent and Huon Estuary dinoflagellate

1108

Y. OSHIMA et al. (a]

(b)

[c)

(C4) epiGTXB

GTXe epiGTX e

GTXe GTXs (C3)

I_

GTX4 I

A___J L_

[

f

J

I

0

I0

20

30

L,.

[

J

I

t

0

I0

20

0

I0

•

. 1 20

Min

FIG. 2. HPLC CHXOMATOGP.AMSOFTOXINSFIOMA NATtmALGymnodiaium catenatum ULOOMIN HUON~rUAtV, T,U M ~ .

chr~

~

are:

00~

C8~

~ phasel ~ ~ral~>",~ma.

ium phosphate (pH 6,5); (b) Devak~l C8 ~ u m m , : ~ ~ I m M l.h~canmURonk add/ 10 mM acetic acid (.oH 7.2); (c) ~ ODS column, m o b k plume I mM l ~ f o n i c acid/20 mM acetic acid/$ taM phmphork acid (pH 7.5). Teodm weTemonitored by fluorescem~ at 410 tun following heating with 7 mM periodic Itdd/50 mM sodium pho~hate.

(b)

(o)

(c) GTX4

GTXr-4

GTX r-4

STX 0cSTX

GTX z GTX3 I

iit t 0

J Io

J 20

....L 3O

1

I

0

10

[ 20

I O

STX

t IO

i 20

Min

Fro. 3. HPLC C ~ L q O M A ~

OFTOX~S mmM A N & ~ ~

The same ¢0~ract as ~

in F~. 2 ~

HCl. ~

~

~

~

~

~

O y m ~

¢'m~.n

m.OOM

.

~ o m

~ at 100°C for 5 rain in 0.2 N as in l~.g. 2.

1109

P S P Toxins in Oymnodinium catenatum TABLE 1. TOXIN COMPosrrloNS OF ~

TASMAN~ o t ~ *

Dinoflagellate

Specimens

Culture 1'

Lethality ( M U / g )

1300

Oonyautoxin 1 Gonyautoxin I Gonyautoxin 11 Gonyautoxin IV Gonyautoxin V Gonyautoxin VI Gonyautoxin VIII Epigonyautoxin VIII C3* CA~ Saxitoxin Neouxitoxin Deearbamoylmxitoxin

. Tr --0.3 . 10.3 52.2 6.3 30.5 -. 0.4

Culture 2*

.

Port Cygnet

Kilinia Bay

Deep Bay

Killala Bay

NT

300

21

170

260

43

3.1 0.3 0.2 4.1 1.8

0.5 --3.6 1.5 . 1.9 7.0 57.1 22.1 0.1 . 6.1

1.8 --5.2 2.7

0.7 0.2 0.1 4.0 3.1

3.9 14.2 43.7 12.2 Tr

3.5 6.5 39.6 40.2 --.

16.2

2.1

.

Tr Tr -Tr .

. 11.1 36.4 18.8 33.8 --

.

. 0.3

Oysters

Huon estuary

910 .

Mussels

Derwent estuary

-w -0.5 . 1.2 6.3 31.3 59.8 0.2 . 0.8

-_ 0.8 0.8 . 1.6 9.4 17.9 68.4 -. 1.2

.

. 4.3 12.9 45.6 22.3 0.3

.

. 5.3

*All results (except for lethality) are s h o w n in m o l e ~ . N T not tested; - - not detected; Tr less t h a n 0.1%. *G. catenatum (Derwent estuary isolate) grown in a full strength seawater m e d i u m (salinity 34°/,0 ). *The same isolate grown in a lower salinity m e d i u m ( 2 3 - 2 6 ° / ~ ). ' C o n t e n t s were estimated from the a m o u n t o f gonyautoxin I a n d gonyautoxin IV generated by acid treatment.

materials, collected from separate bloom events (2 months apart) and distinct localities (50 km apart), showed similar toxin compositions with C3 (18 - 31e/s) and C4 (60- 68e/.) being dominant. In contrast, the cultured dinoflagellates contained higher proportions of GTXs (10- 11e/0)and epiGTXa (36-52e/s), most likely due to physiological conditions (e.g. nutrient stress) in culture being different from those in the natural environment. Lethal potencies of the different shellfish samples varied significantly (21- 260 MU/g shellfish meat), but no apparent differences were observed in the toxin compositions of shellfish with different toxin levels. Compared to the dinofiagellate bloom material, shellfish showed higher proportions of C3 than of C4 toxin, which is most likely the result of epimerization during bioaccumnlation. Mussels contained two to three times more C3 (44- 57e/e) than C4, whereas oysters contained about equal amounts of C3 and C4 (each 40e/e). Several carbamate toxins (GTXI, GTX2 and GTX3) which were absent from the dinoflagellate cells were also detected in the shellfish. The proportion of dcSTX was higher in shellfish, especially mussels ( 5 - 16e/e), than in the dinoflagellate cells. DISCUSSION

The present study provides the first evidence for the toxigenicity of the dinofiagellate G. catenatum, based on HPLC detection of saxitoxin derivatives in unialgal dinoflagellate cultures. The toxin prof'des of the Tasmanian dinoflagellate and shellfish samples are characterized by unusually high proportions of sulfocarbamoyl toxins, which comprised 9 8 - 9 9 % and 77-93e/s, respectively, of total toxins. Predominance of sulfocarbamoyl toxins has also been reported from some strains of the dinoflagellates Protozonyaulax tamarensis and P. catenella (HALL and R B I ¢ ~ , 1984; OSHIMAand YASUMOTO, 1985; CEMBELLAand TAYLOR, 1985), but has not been reported previously from shellfish. The taxonomic affinities of G. catenatum are not clear. To date, the only other dinoflagellate taxa known to produce paralytic shelifish toxins are the chainforming Pyrodinium bahamense var. compressa and species from Protogonyaulax ( T A Y L O R ,

1110

Y. OSHIMA et aL

1985). Accordingly, MOPEY-GAINES (1982) suggested that G. c a t e n a t u m may be a generic strain o f the armoured dinoflagellate P r o t o g o n y a u l a x cateneila which has lost its ability to produce cellulose cell wall plates. This is supported by similarities in cell morphology, chain formation, details o f the pellicular layer and apical pores and toxin chemistry (present work), but other features, such as cyst morphology of the two taxa, are distinct (unpublished data). While net toxicity o f the dinoflagellate P r o t o g o n y a u l a x may vary significantly with culture conditions, toxin composition appears to be a conservative property of a particular clone (BOYER et al., 1985). In contrast, in the present work the toxin composition o f G. c a t e n a t u m varied between laboratory cultures and field populations and the factors controlfing toxin production by this dinoflagellate will be the subject o f further studies. The ultimate toxicity o f shellfish to humans does not ordy depend on the abundance and toxic potency o f the dinoflagellates being f'dtered, but also on the chemical transformations o f the various toxins, either by shellfish themselves or during food storage, food processing and food digestion by human consumers. The predominance o f sulfocarbamoyl toxins in Tasmanian sh~_lf'mh raises questions as to the applicability o f the standard mouse test in assessing potential health risks for human seafood consumers. Firstly, the lethal potencies o f the dominant toxins C3 and C4 are not well known. Secondly, while the sulfonate toxins generally have a low lethal potency, they are easily transformed under mild acidic conditions to the corresponding carbamate toxins, with a concominant several fold increase in toxicity, ~ on the toxin components present (HALL and REICHARDT, 1984). While these conversions are easily a~,omplished in vitro in the laboratory, we still do not know (except for our experimental studies on GTXs; H A R A D A et aL, 1984) under what conditions these conversions may also occur in v i v o in the human stomach. As currently employed, the i.p. mouse bioassay may thus substantially underestimate the potential human oral potency o f samples containing predominantly sulfocarbamoyl saxitoxin analogs. REFERENCES BAL~,

E. (1964) El plancton de Mar de Plata durante el periodo 1961 - 1962. Boln Inst. Biol. mar. Univ. nGc.

B. Airas 4, 1.

Box,n , G. L., SULLIVAN,J. J., A~V,eS,SON, R. J., I'D~mSON, P. J. and TAYLOR,F. J. R. (1985) Toxin production in three isolates of Protofonyaulax sp. In: Toxic Dinoflagellates, p. 281 ( A N ~ , D. M., WH[T1B,A. W. and BADEN,D. G., Eds). New York: ~ . CEMBELLA,A. D. and TAYLOR,F. J. R. (1985) Biochemical variability within the Protogonyaulax tamarensis/ calenetle speciescomplex. In: Toxic Dinoflngellates, p. 55 (ANDERSON,D. M., WHrr~, A. W, and BADL~N,D. G., Eds). New York: Elsevier. ~A, M., SANCtleY.,F. J. and FRAC3A,S. (1984) Gymnodinium catenatum (Graham) en Ins rias Gallegas (N.O. de Espana). I n ~ c i d n p ~ q . 41, 31. F~C,A, S. and SANCHeZ,F. J. (1985) Toxic and ~ y toxic dinoflageUatesfound in ~ Riw (NW Spain). In: Toxic Dinoflafel~tes, p, 51 (ANDnSON, D. M., WHrrE, A. W' and BADEN,D. G., Eds). New York: Elsevier. GgAHAM,H. W. (1943)Gymmodinium cateltatum, a new dinoflagellatefrom the Gulf of California. Trans. Am. m/crox. Sac. 62, 259. HALL,S. and RIBICHAitlYr,P. B. (1984)Cryptic paralytic shellfish toxins. In: Seqfood Toxins, p. 113 (RAOELIS, E. P., F..d.). W ~ n , DC: American ~ Society. HALL~31tAe~,G. and SUMNR, C. (1986) Toxic plaakum btooms affect thelff'mhfmms. Aust. Fish. 4$, 15. HAgADA,T., OSmMA,Y., KAMIYA,H. ~ Y ~ , T. (1982)Confirmation of parlflytk sheitfith toxins in the dinoflqellate Pyrod/nhun babamen~ vat. compras~ and bivalvesin Pslau. Bull. Jim. Soc. ~ e m . Fish. 48, 821. HAgADA,T., OSmMA,Y. and YASUMOTO,T. (1984)A~teument of potential activation of gonyautoxin V in the

PSP Toxins in Gymnodinium catenatum

1111

stomach of mice and rats. Toxicon 22, 476. LOEBLICH, A. R. and SMITH, V. E. (1968) Chloroplast pigments of the marine dinoflagellate Gyrodinium resplendens. Lipids 3, 5. MEE, L. D., ESPINOSA,M. and DIAZ, G. (1986) Paralytic shellfish poisoning with a Gymnodinium catenatum red tide on the Pacific coast of Mexico. Mar. envir. Res. 19, 77. MO~Y-GAXNES, G. (1982) Gymnodinium catenatum Graham (Dinophyceae): morphology and affinities with armoured forms. Phycologia 21, 154. NO6UCHI, T., ONOUE,Y., MARUYAMA,J., HASHIMOTO,K., Nlsmo, S. and IKEDA,T. (1983) The new paralytic shellfish poisons from Protogonyaulax catenella. Bull. Jpn. Soc. scient. Fish. 49, 1931. OSHIMA,Y. and YASUMOTO,T. (1985) Content of carbamoyl N-sulfated saxitoxin analogues in Protogonyaulax tamarensis and Protogonyaulax catenella from Japanese coastal waters. Bull. mar. Sci. 37, 773. OSHIMA, Y., MACHIDA, M., SASAKI, K., TAMAOKI, Y. and YASUMOTO, T. (1984) Liquid chromatographic-fluorometric analysis of paralytic shellfish toxins. Agric. biol. Chem. 48, 1707. SULLIVAN,J. J., JONAS-DAvIES,J. and KENTALA,L. L. (1985) The determination of PSP toxins by HPLC and autoanalyzer. In: Toxic Dinoflagellates, p. 275 (ANDERSON,D. M., WHITE, A. W. and BADEN, D. G., Eds). New York: Elsevier. TAYLOR, F, J. R. (1985) The taxonomy and relationships of red tide flagellates. In: Toxic Dinoflagellates, p. 11 (ANDERSON, D. M., WHITE, A. W. and BADEN, D. G., Eds). New York: Elsevier. WILLIAMS, S. (1984) Paralytic shellfish poison. In: Official Methods of Analysis, p. 344 (WILLIAMS,S., Ed.). Arligton: Association of Official Analytical Chemists.