THE VISUAL PIGMENT CYANIDE EFFECT

— The visual pigment of the Tokay gecko (Gekko gekko) with its in situ absorption maximum at 521 nm has its spectral position at 500 to 505 nm when chloride‐deficient digitonin is used for the extraction. In this case the addition of chloride or bromide to the extract restores the maximum to 521 nm. This property, characteristic of gecko pigments in general, does not occur with any of the rhodopsins that have been tested. Simple salts of cyanide, a pseudohalogenoid with an ionic radius close to those of chloride and bromide and/or its hydrolysis product attacks both this gecko pigment and rhodopsins in the dark. This is seen as a slow thermal loss of photopigment if (sodium) cyanide is present at concentrations above 40 mM for the gecko pigment and 150 mM for the rhodopsins of the midshipman (Porichthys notatus) and of the frog (Rana pipiens). In all cases the loss of the photopigment is accompanied by the appearance of a spectral product with maximum absorption at about 340 nm. Cyanide addition has no effect on the photosensitivity of the native pigments and neither does it alter, as do chloride, bromide and other anions, the spectral absorbance curve. The spectral product at 340 nm also appears when the visual pigments are photolyzed in the presence of cyanide salts below the threshold concentrations given above. Incubation of digitonin‐solubilized all‐ trans‐relina with (sodium) cyanide leads to a reaction product with absorption spectrum similar to that obtained with visual pigments under comparable conditions. We suggest that cyanide addition leads to conformational changes in the visual pigment that results in the release of the retinal chromophore. This is followed by a direct attack of cyanide on the retinal to form the product absorbing in the near ultraviolet. The special point of interest here is that both the gecko 521‐pigment and the rhodopsins are similarly attacked by cyanide addition and form the same product. It is also suggested that this product is an α‐hydroxy acid derivative of retinoic acid, probably α‐hydroxy homoretinoic acid. The opsin site selected by the cyanide action is apparently similar for both classes of visual pigments but must be different from the site in the gecko pigment with which chloride and bromide react. The cyanide‐effect implies that retinal binding and/or Schiffs‐base stability in visual pigments are allosterically controlled. Finally, some practical analytical aspects of this novel effect are discussed.