A crucial result of electron flow from water to NADP⁺ is the formation of a proton gradient. Such a gradient can be maintained because the thylakoid membrane is impermeable to protons. As discussed in Chapter 21, energy inherent in the proton gradient, called the proton-motive force (Δp), is described as the sum of two components: a charge gradient and a chemical gradient. The dual nature of Δp is nicely illustrated in chloroplasts. In chloroplasts, nearly all of Δp arises from the proton gradient, whereas, in mitochondria, the contribution from the membrane potential is larger. The reason for this difference is that the thylakoid membrane, though impermeable to most ions, including H⁺, is quite permeable to Cl^– and Mg²⁺ acquired from soil. The light-induced transfer of H⁺ into the thylakoid lumen is accompanied by the transfer of either Cl⁻ in the same direction or Mg²⁺ (1 Mg²⁺ per 2 H⁺) in the opposite direction. Consequently, electrical neutrality is maintained and no membrane potential is generated.

The proton-motive force generated by the light reactions is converted into ATP by the ATP synthase of chloroplasts, also called the CF₁-CF₀ complex (C stands for chloroplast and F for factor). CF₁-CF₀ ATP synthase closely resembles the F₁-F₀ complex of mitochondria. CF₀ is analogous to F₀ in mitochondria and conducts protons across the thylakoid membrane, whereas CF₁, which is analogous to F₁, catalyzes the formation of ATP from ADP and P_i.

CF₀ is embedded in the thylakoid membrane. It consists of four different polypeptide chains known as I (17 kDa), II (16.5 kDa), III (8 kDa), and IV (27 kDa) with an estimated stoichiometry of 1:2:(10–14):1. Subunits I and II correspond to the b mitochondrial subunits, whereas subunits III and IV correspond to subunits c and a, respectively, of the mitochondrial F₀ subunit. CF₁, the site of ATP synthesis, has a subunit composition α₃β₃γδε. The β subunits contain the catalytic sites, similar to the F₁ subunit of mitochondrial ATP synthase. Protons flow out of the thylakoid lumen through ATP synthase into the stroma. Because CF₁ is on the stromal surface of the thylakoid membrane, the newly synthesized ATP is released directly into the stroma (Figure 22.21). Recall that NADPH formed through the action of photosystem I and ferredoxin-NADP⁺ reductase also is released into the stroma. Thus, ATP and NADPH, the products of the light reactions of photosynthesis, are appropriately positioned for the subsequent dark reactions, in which CO₂ is converted into carbohydrate.

The activity of the ATP synthase is sensitive to the redox conditions in the chloroplast. For maximal activity, a specific disulfide bond in the γ subunit must be reduced to two cysteines. The reductant is reduced thioredoxin, which is formed from ferredoxin generated in photosystem I, by ferredoxin-thioredoxin reductase, an iron-sulfur containing enzyme:

Conformational changes in the ε subunit also contribute to synthase regulation. The ε subunit appears to exist in two conformations. One conformation inhibits ATP hydrolysis by the synthase, while the other, which is generated by an increase in the proton-motive force, allows ATP synthesis and facilitates the reduction of the disulfide bond in the γ subunit. Thus, synthase activity is maximal when biosynthetic reducing power and a proton gradient are available. We will see in Chapter 23 that redox regulation is also important in photosynthetic carbon metabolism.

On occasion, when the ratio of NADPH to NADP⁺ is very high, NADP⁺ may not be available to accept electrons from ferredoxin. In these circumstances, specific large protein complexes allow cyclic electron flow that powers ATP synthesis, a process called cyclic photophosphorylation. In this situation, the electrons in reduced ferredoxin can be transferred back to the cytochrome b₆ f complex rather than to NADP⁺. Each electron then flows back through the cytochrome b₆ f complex to reduce plastocyanin, which can then be reoxidized by P700* to complete a cycle (Figure 22.22). The net outcome of this cyclic flow of electrons is the pumping of protons by the cytochrome b₆ f complex. The resulting proton gradient then drives the synthesis of ATP. In cyclic photophosphorylation, ATP is generated independently of the formation of NADPH. Photosystem II does not participate in cyclic photophosphorylation, and so O₂ is not formed from H₂O.

We can now estimate the overall stoichiometry for the light reactions. The absorption of 4 photons by photosystem II generates one molecule of O₂ and releases 4 protons into the thylakoid lumen. The two molecules of plastoquinol are oxidized by the Q cycle of the cytochrome b₆ f complex to release 8 protons into the lumen. The absorption of 4 additional photons by photosystem I generates four molecules of reduced plastocyanin, which subsequently reduce four molecules of ferredoxin. The four molecules of reduced ferredoxin generate two molecules of NADPH. Thus, the overall reaction is

The 12 protons released in the lumen can then flow through ATP synthase. Let us assume that there are 12 subunit III components in CF₀. Twelve protons must pass through CF₀ to complete one full rotation of CF₁. Just as in mitochondrial ATP synthase, a single rotation generates three molecules of ATP. Given the ratio of three molecules of ATP to 12 protons, the overall reaction is

Thus, 8 photons are required to yield three molecules of ATP (2.7 photons/ATP).

Cyclic photophosphorylation is a somewhat more productive way to synthesize ATP than noncyclic photophosphorylation (the Z scheme). The absorption of 4 photons by photosystem I leads to the release of 8 protons into the lumen by the cytochrome b₆ f system. These protons flow through ATP synthase to yield two molecules of ATP. Thus, every 2 photons absorbed yield one molecule of ATP. No NADPH is produced.

The complexity of photosynthesis, seen already in the elaborate interplay of multiple components, extends even to the placement of the components in the thylakoid membranes. The thylakoid membranes of most plants are differentiated into stacked (appressed) and unstacked (nonappressed) regions. Stacking increases the amount of thylakoid membrane in a given chloroplast volume. Both regions surround a common internal thylakoid lumen, but only unstacked regions make direct contact with the chloroplast stroma. Stacked and unstacked regions differ in the nature of their photosynthetic assemblies (Figure 22.23). Photosystem I and ATP synthase are located almost exclusively in unstacked regions, whereas photosystem II is present mostly in stacked regions. The cytochrome b₆ f complex is found in both regions. Plastoquinone and plastocyanin are the mobile carriers of electrons between assemblies located in different regions of the thylakoid membrane. A common internal thylakoid lumen enables protons liberated by photosystem II in stacked membranes to be utilized by ATP synthase molecules that are located far away in unstacked membranes.

What is the functional significance of this spatial differentiation of the thylakoid membrane system? The positioning of photosystem I in the unstacked membranes gives it direct access to the stroma for the reduction of NADP⁺. ATP synthase, too, is located in the unstacked region to provide space for its large CF₁ globule and so that it can bind stromal ADP and phosphate and release its ATP product into the stroma. The raw materials for CO₂ reduction and carbohydrate synthesis are thus accessible to the stromal enzymes. The tight quarters of the appressed region pose no problem for photosystem II, which interacts with two molecules found in the thylakoid lumen—a small polar electron donor (H₂O) and a highly lipid soluble electron carrier (plastoquinone).