The mass spectrum of penta acetyl gitoxin is reported and discussed. This cardenolide derivative has recently been the subject of extensive metabolic study. Fragmentation processes are described which account for all of the major ions. High resolution measurements, metastable peaks and mass shifts, for the propionyl derivative, are reported to establish the validity of these proposals.
A brief indication of the potential value of this technique, when combined with deuteroacetylation, to metabolic studies in this field is given, along with comments on the spectra of related compounds.
CARDENOLIDES (cardiac glycosides) have long been of interest because of their pharmaceutical p ~t e n c y . ~. ~. ~
Recent studies of the metabolites of acetylated carden-olides4 to have resulted in ambiguous conclusions because there has been no technique for locating acetyl groups within these complex molecules. We wish to illustrate the potential value of mass spectrometry to these studies.
The mass spectrum of penta acetyl gitoxin (IV), Fig. 1, can be rationalised if one considers two major fission processes. The first of these corresponds to a fragmentation already well documented in monosaccharide s p e ~t r a . ~ The initially formed ion radical, localised on the cyclic ether oxygen of one sugar,8 initiates cleavage of the adjacent glycosidic bond to generate A type ions. (These are lettered in accordance with Kotechov's nomen~lature.~) It is of interest that other important fission processes found in monosaccharide spectra, for example C,-C!, cleavage (E type ions) or ring cleavage, have been suppressed in this trisaccharide.
Thus the peak at m/e 215 corresponds to ion A,, containing the sugar S, (m/e 21 5.0927; Cl,Hl,O, from high resolution measurements). Two other ions m/e 155.071 (C,H,,O,; A2) and m/e 95. 049 (C,H,O;A : ) are related to A, by loss of one and two molecules respectively of acetic acid. Metastable peaks at m* 118-5 (m/e 215 + m/e 155) and m* 58-3 (m/e 155 + m/e 95) were found supporting this conclusion.
Location of the ion radical site within the S, ring should afford an A2 ion at mle 387. No ion corresponding to this was found, but its related ion A,l (m/e 327.143; ClGH2,0,) was prominent. In the case of fragmentation of sugar S, both anticipated peaks, A, at m/e 559 and A; at m/e 499, were found. High resolution confirmed the C24H,,011 composition (m/e 499-2196) of the latter but also revealed an isobaric component presumably arising by breakdown of the aglycone (m/e 499.2708 ; C,,H,,O,). No further evidence for fragmentation of the aglycone rings was found in the spectrum of compound IV.
The second major fission process again involved cleavage of the saccharide chain, but with charge retention on the aglycone. This, termed T cleavage by us, can be represented by two possible schemes. The first being simple cleavage of the C,-0 bond, with charge retention on the steroid A ring. Alternatively one can consider a tertiary carbonium ion, formed by loss of OH' from C14, undergoing sequential 9. 5