FISHERY BULLETIN: VOL. 69, NO, 2 



Figure 3. — EPR lipid signals in a hydroperoxide-treated 

 protein and in a sacroplasmic protein. A. Rockfish myo- 

 fibrillar protein upon removal from the freeze-drier 

 (lower trace). The same material 40 min after incor- 

 porating a small amount of a hydropero.xide mi.xture pre- 

 pared from a marine oil (upper trace). B. Freeze-dried 

 rockfish sacroplasmic protein after storage in air at 

 room temperature for 1 day. Arrows denote g ^ 2. 



observed for dilute solutions of low molecular 

 weight radicals. Without exception, all powdered 

 materials of proteinaceous nature, which ex- 

 hibited any type of signal at all, gave a single 

 absorption line in the "free-spin" or g = 2 re- 

 gion. The g = 2 signal is exemplified in Figure 

 la (for solvent-extracted myofibrillar protein), 

 or in Figure lb for human serum albumin. I will 

 have more to say about the g — 2 signal; how- 

 ever, let us concern ourselves for the moment 

 with other resonances which are seen in those 

 samples containing oxidizable lipid in addition 

 to protein. 



Freeze-dried Pacific cod, silver salmon, rock- 

 fish, and other marine fish, though devoid of the 

 g = 2 signal initially (before lipid oxidation has 

 taken jilace) soon give rise to two resonances 

 when samples are exposed to air — the central 

 g = 2 resonance and, downfield (to the left) 

 from the central resonance, an area of EPR ac- 

 tivity which I have designated as the "lipid 

 signal" region (Figures 2 and 3). Unlike tissue 

 samples, many single proteins considered to be 

 quite pure exhibit a g = 2 resonance only. 



When, however, a thin film of oxidizable lipid 

 is deposited on such materials and the mixture 

 is exposed to air, in addition to the central res- 

 onance, a lipid signal is also observed (Figure 

 1 ) . A preliminary study of lipid signals in var- 

 ious models is to be found in the recent work 

 of the author (1970). Although it has not been 

 possible to measure the g-value with the neces- 

 sary precision needed to fingerprint the radical 

 completely, the available data suggest a radical 

 of the peroxy type. This is further illustrated 

 by the two traces of Figure 3a. No indications 

 of hyperfine splitting (hfs) are also consistent 

 with a radical of this nature. 



CHARGE TRANSFER IN TISSUES 



In the present study, it is of particular inter- 

 est to find that in carefully handled freeze-dried 

 tissue samples, there often occurs after the lipid 

 signal reaches a maximum, an abrupt increase 

 in the g = 2 region. In those "native" samples 

 containing a complement of cellular lipid, this 

 may indicate a charge migration (a strong 

 D-»A — donor-to-acceptor interaction; strong 

 charge-transfer process) between a cellular con- 

 stituent acting as a donor and a peroxy radical 

 acceptor. Such a process is also consistent with 

 the observation that it is at this point in time 

 that the lii)id signal begins to decay. This draws 

 our attention to the likelihood that once radical 

 content has increased to some critical concentra- 

 tion, overlap of wave functions between a rad- 

 ical acceptor and a donor is sufficient to allow 

 reactions to proceed. The abrupt change in the 

 g = 2 region is illustrated by the spectra of 

 Figure 3. Figure 3b for sacroplasmic i)rotein 

 under air for 1 day is to be compared with the 

 lower trace of Figure 3a for the same material 

 immediately on removal fi'om the freeze-dryer. 



Another point in favor of a mechanism of 

 this type is the fact that only proteins are really 

 effective as matrices for the formation as well 

 as for the decay of radicals. Powdered glass, 

 quartz wool, and amino acids are essentially 

 without effect when used as substrates for thin 

 films of reactants. Although there ai-e many 

 unanswered questions concerning the mechanism 



374 



