F1 particle – ATP synthase ATPase Sonicate mitochondria (turns inner membrane into vesicles, such that the membrane side that was pointing into mitochondrial matrix now is on the outside.) Observe “lollipop” structures.(A) These can be removed by gentle treatment. (B)The heads are F1 complex (290kD complex, 100Å in diameter) F1 hydrolyses ATP when not bound to membrane pore, Fo. Vesicles containing Fo but not F1 do not make ATP. ATP synthesis is restored when add back F1 to Fo. SDS polyacrylamide electrophoresis (denaturing mw gel, above) shows that F1 particle contains five subunits. After removal of F1, the part left in membrane is Fo. Fo contains 10-14 c subunits, which are integral membrane proteins that insert into membrane. They form a ring. The a subunit is also a membrane protein and it is thought to be conduit for protons. The  subunit associates with the ring and is a rigid elongated subunit that inserts into a hole in the centre of the F1 “head”, which is a hexamer of three  and three  subunits. These have bound nucleotide. The b subunits and  subunits, together with the a subunit, form a “stator”. As the name suggests, this part remains stationary as the other components rotate relative to it. (Rather like the commutator in an electrical motor). ATP synthesis occurs very readily on the enzyme. This is not to say that ATP has a low free energy of formation. Rather, it means that the enzyme binds ADP.Pi and ATP such that they can be easily interconverted, facilitated by stabilising groups on the enzyme. Evidence : if ADP and Pi are added to F1 synthase in presence of H218O, the 18O gets incorporated into Pi. If reaction were heavily in direction of synthesis, this would not happen because hydrolysis would not occur. F1 ATP synthase binds products tightly and the rate limiting step is their release, which also requires most energy input (remember PGK and Myosin). The diagram is a cross section through the structure of F1. In the structure, all three subunits have different conformations. Differences occur between two domains between which ATP is bound (e is empty subunit – no nucleotide; TP is ATP bound). Notice that the long subunit is asymmetric. It has a bulge and interacts differently with the lower domain in the two subunits. Remember that this stalk connects to c ring, which would be in the membrane at the bottom of this figure. This shows conformation of the subunit that binds nucleotide least tightly. This is called the open form and it is shown with bound nucleotide on left and empty on right. There is little change in domains. The other subunits bind nucleotide more tightly and the lower domain is angled upwards to close around ATP site (remember PGK) Binding change mechanism Experiments on the kinetics of ATP production and consideration of the symmetry of the head, which contains three  -subunits, suggested a hypothesis for how ATP might be formed and released in a cyclic and continuous manner. This “binding change” mechanism proposes three states. A “Tight” state in which ATP is very tightly bound and a large activation energy would be needed to release it; an “Open” conformation, in which ATP is weakly bound and could leave readily; a “Loose” conformation in which ADP.Pi are bound but have not yet been converted to ATP. A cyclic mechanism is suggested in which these states interconvert in strict order (the proton gradient biases the rotation of the -subunit to be in only one direction). In A, ATP binds tightly. Subsequently energy input from gradient changes the state of the head so that at C, this ATP molecule is weakly bound and can leave. At same time, ADP.Pi in L site converts to ATP and its site is now tight. Concurrently the old Open site now becomes an L site and can bind ADP.Pi. Note that there are still only one of each of the three conformational states: they are just in different places. So gradient of protons only has to facilitate the switching. Structure suggested that movement of the  subunit would trigger these ordered changes in sequence if it rotated in one direction continuously. Subunits  and  were produced in bacteria and were engineered recombinantly to have hexahistidine tags at protein chain termini. By this means the head can be stuck down on a glass slide coated with Ni2+. Then an actin filament was experimentally attached to the  subunit. This can be seen using a fluorescent drug molecule that binds tightly to actin. The experiment showed that when ATP was added  rotates (remember runs in reverse, as an ATPase, when no proton gradient). Direction was anticlockwise, as shown on the right. When viewed from the head towards the membrane, head runs clockwise, when F1 acts as an ATPase How does the Fo part produces motion? The a subunit is thought to have two channels, or paths, that can conduct protons. Neither half can conduct a proton the whole way through the membrane. ( Left) The c- subunit is formed by two hydrophobic helices that bend back on each other via a hair-pin loops. All amino acids side-chains on the surface are hydrophobic, except for a conserved aspartic acid residue halfway down one helix. Hypothetical model postulates (1) there are two half channels (2) aspartate is protonated when in contact with apolar fatty acid chains (3) protonation or deprotonation can occur only when in contact with the halfchannel in a subunit. The left panel shows that, without further proton addition, the rotor is stalled. It can move neither left nor right because c subunits next channel are negatively charged. If it were to rotate, energy of opposing negative charge next hydrophobic fatty acids would be too high. In middle panel, imagine proton on outside comes down left channel and protonates the aspartate carboxylate group in one of c subunits. Right panel shows this situation. Now ensemble can rotate to left because leftmost of two c subunits apposed to the a subunit is protonoted. The subunit that moves in from right to now contact the a subunit is protonated – but that is energetically favourable. Protons can leave through lower half-channel when the carboxyl side chain of the aspartic acid deprotonates to become aspartate. The sequence of events in one ratchet between c-subunits is shown in the next figure. The net result is that protons flow through the top half-channel, bind Asp 61 in a c-subunit and then rotate the whole way around, until they meet the bottom half channel. Note that the rotor will always run only one way, as long as protons come in from the top and leave by the bottom. Remember that stalk region ( subunit) is attached to the top of the c-ring. Thus this stalk will rotate and, like a cam, trip the sequence of conformational changes in the 33 heads to synthesis ATP. But if the proton concentration was higher at bottom than at top, the rotor could theoretically run in the opposite direction If the head ran in the opposite direction, hydrolysing ATP rather than synthesising it, then the c-ring would also run with it, coupled via the -subunit, such that protons flow would also run in the opposite direction. This mechanism would need a high concentration of ATP and would pump protons against a proton gradient. But this is just another case in which the following equations holds: 0'  G G  RT ln Pr oducts  Reac tants  Although the free energy for the proton transfer is high, it is coupled to ATP hydrolysis, which has a large negative free energy, and so the overall free energy is negative. Such a mechanism could be used to acidify a membrane compartment. Summary F1Fo-ATP synthase is a rotary motor. In mitochondria it conducts protons along their concentration gradient into the inner space and couples this to synthesis of ATP. Protons flow through a half channel in a membrane subunit and protonate an aspartate sidechain in one of a ring of membrane subunits. This rectifies random rotations of the c-subunits so that they rotate in only one direction. The -subunit, which sits on top of the ring of c subunits and is inserted into a ring of  subunits that are outside the membrane, acts like an eccentric cam. It rotates and changes the conformation of three  -subunits in that ring, which bind nucleotide. The -subunit clashes with each -subunit in turn and releases synthesised ATP (the energy requiring and rate-limiting step in ATP synthesis). Formation of ATP on the enzyme does not require much energy, but releasing it does. Rotation of the -subunit changes the conformation of  -subunits in temporal and spatial order (i) to trip synthesis of ATP (ii) to trip its release and (iii) to produce an open comformation ready to bind more ADP.Pi. If ATP concentration is high enough, the motor can run in reverse to pump protons, even against their own concentration gradient, at the expense of ATP hydrolysis.