Mechanism of polarized light perception.

As background for a report on our current selective adaptation experiments in decapod crustaceans, the various facts and hypotheses generally relevant to intraretinal sensitivity to polarized light in arthropods as well as cephalopods have been marshaled. On the basis of this review, the following working hypotheses have been made. 1) One ommatidium in the compound eye is the functional unit in image perception but contains in its component retinular cells subunits which can work independently in detecting other visual parameters, such as polarization. 2) Single retinular cells do respond differentially to light polarized in various planes. 3) Light sensitivity, including e-vector detection, is localized in the rhab domeres, which comprise closely packed arrays of microvilli protruding axially from retinular cells; the dichroism of the photopigment molecules, which are contained within the microvilli, provides the molecular basis of e-vector detection. 4) The visual pigment molecules have their major dichroic axis aligned predominantly parallel to the long axis of the microvillus containing them; typically all microvilli in a single rhab domere are closely parallel to one another, thus comprising at the cellular level a unit dichroic analyzer with maximum optical density to photons vibrating in the direction parallel to these microvillous protrusions. 5) In most decapod crustaceans, in cephalopods, and in some insects the microvilli in all rhabdomeres of a retinula are oriented in only two directions, perpendicular. to each other. Therefore, e-vector perception must depend at the retinular level on a two channel system consisting of a pair of dichroic analyzers with their major transmitting axes fixed at a 90 degrees angle determined by the two directions of microvillus orientation. Our new results on selective adaptation in the eye of Cardisoma provide direct experimental evidence for such a two-channel analyzer in which the pair of functional units have their maximum sensitivity to polarization in the same retinal directions as the rhab dom microvilli observed in electron micrographs. In turn, these directions correspond with the vertical and horizontal axes of the animal's normal spatial orientation. In e-vector detection the seven retinular cells of a single decapod ommatidium thus form two operational subgroups of four and three cells, respectively (39). The correspondence of the electrophysiological evidence for a dual polarization analyzer with the perpendicular directions shown by the microvilli in a single rhabdom strengthens the idea that one ommatidium is enough for detecting e-vector orientation. On this evidence we may conclude that the model developed above for a two-channel polarization analyzer effectively accounts for the relevant spectrophotometric, fine-structural, electrophysiological, and behavioral data currently available for a considerable number of arthropods and cephalopods.