Ati-Microspcctrophotometry(MSP) shows rhodopsin highly concentrated (about 3.Ommolfl) in rod outer segments (ROS). Calculation of the in uiuo absorption spectrum of human rhodopsin from such data reveals a striking failure to agree with the action spectrum of human rod vision. Agreement is good between the spectral distribution of absorption coefficients and the action spectrum. but the "concentration-broadening" (or "self-screening") introduced by the high end on absorbance at this concentration results in a misfit among the largest in the 93 years comparisons of this kind have been made! To deal with this anomaly, it has been suggested that "conantration-broadening" is inappropriate for rhodopsin in rod vision. This proposal was tested by comparing rod action spectra of 15day-old and adult rats. since the lengths of ROS increase by a factor of about two in maturation. Three lines of cvidena arc inconsistent with it. Although the conundrum remains unexplained. it cannot be dismissed by supposing "self-screening" inappropriate for night vision.
Self-screeningRhodopsin Rod vision Elcctroretinogram Absorption spectrum
INTRODUCMONThe agreement between the absorption (or absorptance) spectrum of the human rod visual pigment (rhodopsin) and the action spectrum of human night vision became the fundamental datum in visual physiology upon its discovery by Koenig (1894). Though never perfect, this elegant convergence of results from molecular physical chemistry and experimental psychology compels the inference that the initial step linking the physical world to its visual perception by the human brain is photochemical. Continued attempts to improve Koenig's approximate agreement followed as understanding deepened. Unfortunately, the agreement in the most recent comparison, estimating the rhodopsin absorption spectrum in situ from spectral analysis of human rod outer segments (ROS) with a microspectrophotometer (MSP) (Bowmaker and Dartnall, 1980), is almost as poor as that in the initial comparison.
TheoryTo be seen, photons must be absorbed. If Z is the incident irradiance [photons of mono-*To whom correspondence should be addressed. chromatic light (of wavenumber, or frequency, m) set-' deg-) and I' the irradiance absorbed, thenwhere c1 is the absorption coefficient of rhodopsin, m, is its wavenumber of peak absorption, c is its concentration and 1 the length of the light path. A plot of Z'(m)Z(q,)/Z'(~)Z(nr) vs m is an absorption spectrum. A plot of Z(m,)/Z(m) vs m for some fixed effect is an action spectrum. Koenig implicitly assumed that at rod threshold, the amount of light absorbed is constant independent of wavelength. This assumption is now a corollary of a general principle of vertebrate photoreception: the principle of univariance (Naka and Rushton, 1966). It is evident from equation (l), assuming univariance, that if the wavenumber dependency of the effect is derived from absorption by rhodopsin alone, the two spectra will be the same.The shape and the width of the absorption spectrum of a pigment varies with ...