A test of the binary chloroplast membrane hypothesis by using a nonpenetrating chemical probe,p-(diazonium)-benzenesulfonic acid

Summary Spinach chloroplasts were exposed to³⁵S-labeledp-(diazonium)-benzenesulfonic acid (DABS), a water soluble compound which does not penetrate lipophilie regions of membranes, and which is highly reactive toward amino acid functionagroups such as -amino, sulfhydryl, histidine, and tyrosine groups. Amino groups inl lipids can also form similar, stable covalent bonds by diazo coupling. Both chloroplast lipids and proteins were labeled with DABS, the total binding being about 1 DABS per 10 chlorophylls, depending on the reaction conditions.After diazo coupling and subsequent digitonin fractionation into photosystems I and II enriched fractions, it was observed that PS-I was more highly labeled than PS-III usually by a factor of 10 to 24 times (on a per chlorophyll basis). After digitonin isolation, however, the PS-II portion bound an amount of DABS similar to the PS-I binding, We interpret these data as consistent with the binary membrane hypothesis (Arntzen. Dilley and Crane (1969),J. Cell Biol. 43:16), which visualizes PS-I on the externa, half of a 90 Å grana membrane, and PS-II occurring on the interior half of thel membrane. The alternative explanation that PS-II and PS-I are arranged as a mosaic, and that the low DABS binding in PS-II is caused by burial of the diazo reactive groups in the interior of the proteins (and only exposed through the denaturing effect of digitonin) is not directly ruled out. However, this alternative is not consistent with the facts that: (a) most of the membrane proteins in PS-I and PS-II are identical in electrophoretic properties and therefore probably have similar overall structures; and (b) digitonin does not lead to appreciable denaturation of proteins, evidenced by the retention of PS-II electron transport activity.