Turning an opinion inside-out: Rees and Eisenberg's commentary (Proteins 2000;38:121-122) on "Are membrane proteins 'inside-out' proteins?" (Proteins 1999;36:135-143).

In any scientific discussion it is paramount to recognize what is agreed upon and what are the areas of contention. We are happy to conclude that Rees and Eisenberg in their recent commentary agree with us that membrane proteins are not “inside-out” proteins. They also concede that their original article, describing the use of the hydrophobic moments to model and understand membrane protein structure, is conceptually incorrect, because it is based on the “inside-out” nature of membrane proteins as its foundation. The disagreement is based upon whether Rees and Eisenberg’s 1989 Science article entitled “Hydrophobic organisation of membrane proteins” is also a proponent of the “inside-out” nature of membrane proteins. Rees and Eisenberg, now claim that their intentions in their 1989 Science article were to discredit the “inside-out” nature of membrane proteins. We could not disagree more, in fact Rees and Eisenberg’s 1989 Science article is one of the strongest and most influential proponents of the “insideout” nature of membrane proteins as we readily demonstrate below. The original concept for the “inside-out” model for membrane proteins was introduced by Engelman and Zaccai, after a neutron diffraction study of bacteriorhodopsin. In referring to bacteriorhodopsin it is stated that: “ . . . the protein is inside-out compared with normal distribution of polar and non-polar amino-acids found in soluble proteins.” Although this model may have been useful to describe some of the unusual properties of this archaean protein, it is clearly of limited applicability to the structures solved since. For the currently published high-resolution structures, hydrophobic organization is a poor indicator of the orientation of transmembrane helices. Rees and Eisenberg’s 1989 Science article describes the study of the hydrophobic organization of Rhodobacter sphaeroides PRC and is highly supportive of the “insideout” nature of membrane proteins. Indeed, one only needs to look at the first two sentences of the abstract in order to come to this conclusion: “Membrane-exposed residues are more hydrophobic than buried interior residues in the transmembrane regions of the photosynthetic reaction center from Rhodobacter sphaeroides. This hydrophobic organisation is opposite to that of water-soluble proteins.” If the residues in the exterior are more hydrophobic than the residues in the core, and this situation is reversed from water soluble proteins (which are “outside-out”), how can PRC not be “inside-out” according to Rees and Eisenberg? To reiterate, since water soluble proteins are “outside-out” and the hydrophobic organization of the PRC is opposite to that, then PRC cannot be “outside-out” as well. The above quotation from Rees and Eisenberg’s PRC study unambiguously specifies a direction for the TM helix amphipathicity, whereby the more polar residues are oriented towards the core of the protein. Polarity is a relative concept, in which one should compare the polarity of a solute with the polarity of its solvent. Water is less polar than brine, TM helices are less polar than water, and decane is less polar than TM helices. Rees and Eisenberg’s analysis indicated that, compared to the lipid solvent, PRC has a polar core. The magnitude of the core hydrophobicity may be comparable to the cores of aqueous domains, but given that our interests were concerned with the residues within the plane of the lipid bilayer, the PRC article does indicate that the core was more polar than the exterior of the protein and the lipid environment. In our analysis a different conclusion was reached: The polarities of the core and exterior of mem-