BChl a-protein from P. aestuarii

As indicated in Sections 3 and 4, exciton analysis of spectra is a potentially useful way of interpreting the spectra, understanding (B)Chl-(B)Chl interactions, and possibly verifying structural hypotheses. One also sees that it is not without controversial aspects, although as long as structural information is incomplete one is still free to assume structural features that suit the exciton analysis and thereby minimize controversy. The situation is quite different when detailed knowledge of structure is available, i.e., in practice when a structural model based on X-ray diffraction at or near atomic resolution has been established. This is the situation to be considered in the remainder of this chapter. 

Historically the first X-ray structure [43-45] to undergo exciton analysis was that of the water-soluble BChl a-protein from the green photosynthetic bacterium Prosthecochloris aestuarii. The analysis [16] raised questions, and controversies, that remain unresolved after a decade. It is reviewed again here to emphasize these difficulties, to correct some misconceptions in the literature [4,46,47] regarding possible sources of the difficulties, and to discuss more recent developments. Exciton analysis of photosynthetic pigment-protein complexes is iipt likely to become a truly useful procedure until it produces results that agree with all relevant spectra of this particular complex. 

Standard exciton analyses (Ref. 16 also unpublished calculations by R.E. Fenna and independently by L,L. Shipman) based on the atomic coordinates published by Fenna et al. [44] produce calculated absorption and CD spectra with multiline features that resemble observed spectra in multiplicity and splitting energies. However, the observed overall spectral red-shift is not reproduced theoretically, nor is the pattern of intensity borrowing in the Qy absorption spectrum or the magnitudes and signs of Qy CD bands. These are the aiionialies that continue to plague all exciton-analytical efforts in photosynthesis. 

In their 1978 analysis, Pearlstein and Hemenger [16] presented two sets of theoretical results, one of which is remembered and the other not. The former comes from a nonstandard approach in which each Qy dipole is assumed to be rotated 90° in the plane of its BChl macrocycle relative to the direction assigned [27] by molecular orbital theory. This single assumption virtually solves all of the anomalies just noted, except for the overall red-shift. However, this striking finding has not led anywhere, because so far no physical basis for such a rotation of electronic transition moments has been proven. Nonetheless, there may be a kernel of truth in the transition-moment-rotation hypothesis (see below). 

The forgotten theoretical results of the 1978 paper are contained in a discussion 

---