II was subjected to descending paper chroma- 

 tography (table 6). 



5. Three methods have been devised for de- 

 tecting the toxin. These are: a toxicity test 

 under a standard set of conditions; a spectro- 

 scopic test using the absorbance of a CCI4 

 solution at 270 m^ (e obs. = 790, e calc. = 1900)'; 

 and the optical activity test that detects certain 

 centers or units of asymmetry. 



6. Substance II is optically active, which in- 

 dicates a center ox' molecular unit of asymmetry 

 (Morrison and Boyd, 1967). An example of a 

 center of asymmetry would be an asymmetric 

 carbon atom as in 1-alanine; an example of a 

 molecular unit of asymmetry would be tris 

 (ethylenediamine) chromium (III) ion. The 

 specific optical rotation of substance II is +68 

 (maximum, table 6) which may be compared 

 with +128 for the poison of Gymnodinium 

 catenella (Schantz et al., 1966). The former 

 value was obtained at 546 m/i, and the latter 

 was probably obtained at 589 m^i. 



STRUCTURAL FEATURES 



A complete structural determination of the 

 toxin (s) was beyond the scope of the present 

 study, but several structural features became 

 apparent. These are valid, of course, for the 

 toxin at the described state of purification. 



First, the spectrum of substance I evidently 

 contains a carbonyl group, = C = ; that of sub- 

 stance II does not (fig. 3). Both substances 

 have infrared spectra that are consistent with 

 organophosphates R-P0(0R)2, where R is an 

 oxygenated hydrocarbon moiety. 



Second, the toxin molecule contains some 

 asymmetric feature, either an asymmetric car- 

 bon atom (s) or a larger asymmetric unit. 



Third, some structural features are indicated 

 by the mass spectrum (Beynon, 1960; Biemann, 

 1962) , though the spectrum must be interpreted 

 with some care because of the volatility of sub- 

 stance II. For example, the maximum formula 

 weight of a stable fragment is 508. Any frag- 

 ments having greater formula weights are too 

 volatile or too short-lived to be detected. It is 

 also possible that the substance readily decom- 

 poses into three fragments with similar 



'The observed value is based on a molecular weight of 650; the 

 calculated value is based on an empirical formula weight of 1545. 



formulas ; thus, the value of 508 corresponds to 

 about one-third of the calculated empirical 

 formula weight. 



Finally, it may be hypothesized that a toxin 

 portion of molecular weight 650 decomposes 

 into two fragments with m/e values of about 

 508 and 147. The latter fragment might be a 

 phosphate ester OP (OR) OR. 



Peak assignments, which appear to be reason- 

 able on the basis of available data (table 7), 

 include the following: 



1. The peak at m/e = 31 probably corre- 

 sponds to phosphorus-31. 



2. The peak at m/e = 63 is attributable to 

 a POo unit, which should come from the 

 R'-PO( OR) 2 structure. 



3. The P(OR)(OR') fragment might be 

 assigned to the peaks at m/e = 138- 

 143. If R and R' is CH2CH3, a m/e 

 value of 137 would be expected ; if R is 

 CH = CH2 and R' is CH2CH = CH2 m/e 

 value of 142 would be expected. The 

 second possibility is consistent with 

 the ultraviolet absorption spectrum. 

 Cleavage of the vinyl group would ac- 

 count for peaks at about 27 (CH = 

 CH2) and 121 [PO(OCH = CH2)OCH2]. 

 Further cleavage of a methylene frag- 

 ment would account for a peak at about 

 107. 



ACKNOWLEDGMENTS 



J. R. Linton provided helpful discussion and 

 assistance in the study of acetylcholinesterase 

 activity ; J. A. Zoltewicz, University of Florida, 

 helped us obtain the mass spectrogram of sub- 

 stance II ; S. M. Ray and W. B. Wilson, Texas 

 A&M University, provided initial cultures and 

 advice; the BCF (Bureau of Commercial Fish- 

 eries) gave financial support; and the Public 

 Health Service gave a Research Career Award 

 (to DFM, 1 KO 4 GM 425691) from the Na- 

 tional Institute of General Medical Sciences. We 

 were helped by personnel of the BCF Biological 

 Laboratory, St. Petersburg Beach, including 

 James E. Sykes, Director, who with J. K. Mc- 

 Nulty contributed helpful advice and criticisms ; 

 J. H. Finucane, who helped collect bloom sam- 

 ples; and Dorothy Hartswick, who helped pre- 

 pare this manuscript for publication. Personnel 



442 



U.S. FISH AND WILDLIFE SERVICE 



