Directional Proton Transport

Contemporary cells utilize a variety of complex mechanisms for energy acquisition and transduction. A common motif, however, is the conversion of acquired energy into a proton gradient that is then used to do useful work [27], such as the synthesis of ``high energy'' compounds. The universality of this mechanism suggests that it must have emerged at an early stage of protobiological evolution.

The creation and maintenence of a transmembrane proton gradient requires a system that irreversibly transports protons across the protocellular boundary. There are several possible early sources of protons, including chemical reactions and light. Probably the simplest system which is capable of photo-generating a transmembrane proton gradient consists of polycyclic aromatic hydrocarbons incorporated into liposome membranes [28]. Upon the absorption of light, the chromophore releases protons either to the exterior or the interior of the liposome. Protons in the environment dissipate whereas those inside the liposome accumulate, thereby creating a proton gradient. This system, however, is not directional; it lacks a ``gate-keeper'' mechanism which ensures that protons transferred to the protocell interior are not used to regenerate the protonated chromophore.

A schematic of a simple directional proton transport system is shown in Fig. 2. In this figure, PS refers to a proton source located near the center of the membrane and A₁ and A₂ are a pair of proton acceptors that are part of the gate-keeper complex. The PS initiates the proton transport across the membrane. It could be comprised of a chemical reaction, a chromophore and an ionizable species [29] or, more simply, of a polycyclic aromatic hydrocarbon [28]. The only constraint placed on the PS is that, when protonated, it transfers its proton to A₁. This might not only require that A₁ be in close proximity to PS, but also that the pK_a of A₁ be coupled to the state of PS. The secondary proton acceptor, A₂, is located sufficiently close to the protocell interior that any proton it accepts is quickly released to the aqueous solution. After PS transfers its proton to A₁, the reverse reaction,   might become highly probable. However, all that is needed to create a proton gradient is a nonvanishing probability of irreversible proton transfer from A₁ to A₂. Protons transferred to A₂ would then be injected into the protocellular interior.

 
Figure 2: ``Gate-keeper'' scheme for directional proton transfer. PS is an activated proton source, A₁ and A₂ are proton acceptors that are part of the gate-keeper complex.

After the proton has been tranferred to the protocellular interior, PS must be reconstituted by a proton from the exterior if the proton gradient is to be maintained. Since the proposed proton pump is assymetrical, this can be accomplished if reconstitution from the exterior is faster than from the interior. After deprotonation, the charged PS alters the environment within the gate-keeper complex, slowing down the ``uphill'' proton transfer from the interior. Reconstitution of PS from the exterior would continue until the interior pH of the protocell drops below the pK_a of A₂, at which point the formation of protonated A₂ would begin to drive the back reaction.

Three different mechanisms have been considered for increasing the irreversibility of the transfer of a proton from A₁ to A₂. One mechanism involves a transient water bridge. Transient chains of hydrogen-bonded water molecules have been postulated to account for the anomalously high proton permeability of membranes [30] and for the proton conductivity of the gramicidin channel [31]. Furthermore, ab initio quantum mechanical calculations on the formic acid-water-formate system (see Fig. 3) have revealed that the barrier for proton transfer from a formic acid to a formate ion across a water bridge is only 0.7 kcal/mole [32]. With a pK_a of 3.7, formic acid is a good model for the acidic amino acids glutamate and aspartate which have pK_as in aqueous solution of 3.95 and 4.4, respectively. These two amino acids are good candidates for A₁ and A₂. A transient water bridge between them would then provide an efficient mechanism for irreversibly transferring a proton. The back transfer of the proton is impeded by the free energy required to move the proton uphill towards PS, as well as by the disruption of the water bridge resulting from the hydration of the negatively charged A₁^- moiety. The efficiency of this mechanism will decrease if the pH inside the protocell becomes low enough to cause the spontaneous formation of protonated A₂, driving the back reaction.

 
Figure 3: A sketch of the water-bridged formic acid-water-formate proton transfer system.

In contemporary cells, transmembrane helix bundles are the most common motif for proteins conducting proton transfer. Their simplest precursor might have been a system consisting of a proton donor and an acceptor affixed to a single transmembrane -helix. However, geometrical considerations rule out this arrangement -- all conformations of side chains that bring the proton donor into the proximity of the acceptor, with or without a water bridge, are sterically disallowed.

A somewhat more complex, two-helix system does not suffer from similar steric restrictions. The feasibility of the water-bridge mechanism was tested by constructing a pair of transmembrane helix fragments with sequences conducive to this type of proton transport. At pH 7, only three amino acids could serve as proton acceptors: aspartate (D), glutamate (E) and histidine (H). We used a glutamic acid (E⁰) as the proton donor and glutamate for both A₁ and A₂. We further assumed that the pK_a of the glutamic acid was coupled to a chromophore so that the photo-excited chromophore induced the glutamic acid to ionize. The rest of the helical fragments was constructed from nonpolar leucine (L) and polar, nonionizable glutamine (Q). In the folded fragments, the leucines were in contact with the nonpolar membrane while the glutamines formed a polar core within which the proton transfer could occur. The sequences of the two helix fragments studied are:

The helix fragments were arranged so that the glutamate on fragment 1 (A₁) and the glutamic acid on fragment 2 (proton donor of the PS) formed a bifurcated hydrogen bond. A₁, in turn, is separated from the A₂ glutamate by approximately 5Å. While this distance is too far for direct proton transfer between the two glutamates, it is nearly ideal for a water-bridged transfer.

Another mechanism to prevent back transfer is a conformational shift of A₁ after it is deprotonated. To examine the magnitude of the conformational shift accompanying deprotonation in a simple system, we considered two helix fragments differing only in the protonation state of the first glutamate:

Minimization of each protein resulted in different side chain conformations of the glutamate. The oxygen initially bearing the proton, O, is displaced by 4.8Å after deprotonation. The two relevant torsional angles of the glutamate side chains, NCCC and CCCC, change upon deprotonation from 72 and -176 to 63 and -83, respectively, upon deprotonation. The displacement of the deprotonated carboxyl oxygen is large enough to disrupt a proton transfer chain. When coupled with a transient water bridge, it would ensure the irreversibility of the proton transfer. This coupling of a conformational shift with a hydrogen-bonded chain of proton acceptors to generate irreversibly a transmembrane proton gradient is thought to be used by bacteriorhodopsin [33].

The final gate-keeper mechanism is a generalization of the water-bridged mechanism whereby the water molecule acts as an amphiprotic species -- a species that can both accept and donate protons. A similar gate-keeper mechanism could be constructed from other amphiprotic species. For example, the two ring nitrogen atoms of histidine can both accept protons, although at pH 7 only one of them is protonated. Proton transfer to the second nitrogen, occurs with a pK_a of slightly greater than 7. Proton transfer would proceed from histidine to a secondary acceptor and the back reaction would be prevented by tautomerization of the histidine ring. This mechanism would be very sensitive to the exact nature of the PS since the pK_a's of the two nitrogen atoms are very sensitive to the local environment.

Modern proton pumps are thought to utilize a complex chain of hydrogen-bonded residues as well as internal isomerizations for irreversible proton transport across cell membranes. We have demonstrated that these same mechanisms could have been used by potential precursors of these pumps to facilitate directional proton transport across the protocellular boundary. The structures needed to perform this function are simple and not highly specific and, therefore, are compatible with protocellular conditions.