Sussman Lecture

Like a number of my colleagues around the world, my students and I have been carrying out experiments that attempt to get at one of the major unanswered questions in the life sciences: How do photosynthetic organisms generate oxygen? In this lecture, I’ll focus on three subjects:

First, a brief historical overview will introduce the scientific origins of the problem.

Second, I’ll describe the present state of knowledge concerning photosystem II, which is the name given to the enzyme that traps light energy and uses it to convert water into oxygen. I will spend most of the time discussing the roles of proteins, an area in which there have been a number of novel new observations, but I’ll also spend some time describing research on the ions that cooperate to produce oxygen from water. At the start of this research project, manganese and chloride were known to be required for the oxygen evolution reaction. No other ions were thought to be involved, but as you’ll see, when we began to work on photosystem II’s proteins, we discovered that calcium was also required for in oxygen evolution.

The third part of the talk will be relatively brief, and will expose you to the speculative side of the problem. I have tried to adapt rather complicated material for a general audience, and hopefully I won’t be too technical.

SHOW 2^nd SLIDE

These equations present the global picture of photosynthesis. Photosynthetic reactions are estimated by some scientists to have been in existence for as long as 4 billion years. As shown in these equations, plants, algae and cyanobacteria use light energy, water and carbon dioxide to synthesize constituents of the major food groups as well as oxygen, which is a waste product of photosynthesis. However, nearly all organisms burn up these food groups, consuming oxygen from photosynthesis, and produce the biochemical energy (a cellular chemical called adenosine triphosphate) that is used for synthesis, movement, reproduction and so on. As can be seen at the bottom of these equations, the central component of this series of steps is the ability of photosynthetic reactions to convert solar energy into stable forms of biological energy. A fraction of the approximately 1,000 billion tons of oxygen produced by photosynthesis every year is therefore put to good use in burning up foodstuffs to produce the chemical energy that sustains life on earth. This is the theme that connects light, life and photosynthesis.

Some (but certainly not all) historical milestones in research on oxygen evolution are listed here. Joseph Priestly discovered that plants produce oxygen, and Lavoisier, his contemporary in France seems to have been among the first to study oxygen consumption by living organisms, a process known today as respiration. Those who complain that standards of peer review have softened up in the last couple of hundred years can certainly make a case from the example of Lavoisier.

Martin Kamen and Sam Ruben carried out the first biological isotope experiments as a way to study photosynthesis. They identified the source of oxygen, and confirmed the fate of carbon dioxide. Today, isotopes are widely used in biological research, but these were the very first experiments to use this technique.

Robin Hill carried out some of the first biochemical investigations on photosynthesis. He discovered how to produce oxygen from chlorophyll-containing membranes that he isolated from spinach. This organism has become the "laboratory rat plant" for people who work on photosynthesis. Hill also formulated the current model for photosynthetic electron transfer reactions. He proposed that light is absorbed by enzymes which came to be called photosystems. Hill hypothesized that one photosystem, photosystem II, obtained electrons from water in the light, and transferred them on to the other photosystem, photosystem I, which produced the energy used for carbon dioxide conversion into sugars.

Finally, Bessel Kok and Pierre Joliot discovered the oscillating pattern in oxygen production by photosystem II. I’ll show you more about this experiment in minute.

Here’s a picture of two of Priestley’s experiments with plants and mice. He found that plants could restore the oxygen that had been removed from the atmosphere by burning a candle in a jar, as shown here. Priestley’s method, shown on the right side of this slide, certainly makes a graphic point about the importance of oxygen to life. However, this assay method is no longer a fashionable way to measure photosynthetic activity.

These equations (SLIDE 5) summarize Kamen’s isotope experiments, which were carried out about 150 years after Priestley. The isotope technique showed conclusively that water was the source of oxygen produced by photosynthesis, and that carbon dioxide was converted into sugars.

The results obtained by Kok and Joliot, using a sensitive method for oxygen detection, are shown in (SLIDE 6). This oscillating pattern of oxygen production is caused by illumination of photosystem II with brief pulses of light, like those from a camera flash. Each flash causes an electron to be transferred in photosystem II, and oxygen is produced with a periodic pattern that shows a maximum on every 4^th flash, once the system is up and running. Kok proposed a still-valid model for these oscillations, shown on the next slide.

In Bessel Kok’s model, the oxygen evolving reaction operates like a clock driven by light. Discrete states that Kok simply called "S" become more electron deficient as light absorption by chlorophyll withdraws electrons and protons from them. The most electron deficient state (at about 8 o’clock) is called S₄. This state is where photosynthesis makes oxygen. S₄ reacts with two molecules of water and removes four electrons from them. This step releases oxygen and uses the electrons and protons extracted from water to reset the clock to about 11 o’clock, the S₀ state, which starts the process all over again. Oxygen "clocks" rotate a couple of hundred times a second when photosystem II is illuminated. The electrons that are withdrawn from water are finally deposited in the organic molecules that are the constituents of nearly all living organisms on earth, including everyone in this room.

You might have noticed that when the clock is first illuminated, only 3 light flashes are necessary to get the first big burst of oxygen. This bothered Kok, and he got around the problem by proposing correctly that the S₁ state of the clock, at about one o’clock, is the dark-stable state; in other words, when the sun sets, all other S states collapse to S₁ in the dark.

The discovery of oscillations in the oxygen yield from photosystem II provided a useful model for how the reaction system works. The clock moves ahead one electron at a time, and each clock operates independently of other clocks. That is, they all produce oxygen at the same time, so to speak. This means that the clocks are isolated from one another, as would be expected if each clock were a component part of a larger, discrete enzyme system.

A difficulty that confronted researchers working on oxygen evolution at this time was that only fragments of the enzyme system could be isolated from the chlorophyll-containing membranes of plants or algae. None of these fragments were very active in producing oxygen. While I was a post-doctoral fellow at Cornell, I succeeded in isolating photosystem II in a highly active form. When I told some senior colleagues about this discovery, I was not taken seriously. It was made clear to me that junior people couldn’t possibly accomplish what more senior people had failed to do. This was sufficient to turn my efforts in other directions when I set up my lab at Michigan.

Tenure time came and went, and I survived, presumably on account of good behavior with respect to saying nothing about isolating photosystem II. By this time, there was a real need to get on with the biochemistry of photosystem II. A post-tenure sabbatical at Michigan State with my colleague, Prof. Gerry Babcock, provided the opportunity to clean up the photosystem II isolation procedure and to send it off for publication. After relatively mild complaints from one of the reviewers, the paper was accepted.

The asterisk by my name alerted readers to the fact that I was really from the University of Michigan, that I was the senior author, and that any questions or complaints should be directed to me. Instead of complaints, there were several amusing fallouts from the appearance of this paper. The first person to request inside information on how to do the preparation, while the paper was still in press, was one of the senior figures who told me earlier that isolation of photosystem II was impossible. The first author on the paper, Deborah Berthold, who was doing her honors research in Prof. Babcock’s lab, and who helped me with some of the experiments, is now the most widely-cited undergraduate in the field of photosynthesis. The first letters of the last names of the authors (BBY) must have a euphonious ring. To our collective embarrassments "BBY" has come to be used in most scientific journals as an accepted abbreviation for this preparation. A less amusing consequence was that the procedure, when I refined it, was so simple that just about anyone could, and did, use it. This spawned an army of competitors, which has to this day made my life somewhat more lively than I would have liked.

Isolation of photosystem II in its active, intact form proved to be the vehicle by which new properties of the oxygen evolution reaction have been discovered. These discoveries have included a number of unanticipated, surprising and entertaining observations, and this is what I’d like to tell you about.

Here’s a summary of what has been shown to be associated with isolated photosystem II, a list of the gears and springs that make oxygen clock tick. As I mentioned earlier, manganese and chloride were already known to be essential for the oxygen evolution reaction, but calcium was a surprise. Properties of the proteins, many of which serve as scaffolding to bind the various molecules of the enzyme, also caught us off guard. For example, we had no idea that there were any photosystem II proteins that would dissolve in water. Since the enzyme resides in a membrane, we assumed, with good precedent, that photosystem II proteins would require detergents to dissolve them. We finally got around to those detergent experiments, but first, we had to deal with the discovery that things were not what they seemed to be.

One of the first experiments a biochemist does with a newly-isolated enzyme is to determine how many proteins are present in the preparation. This protein analysis employs a technique called gel electrophoresis, shown here. Electrophoresis separates proteins according to their size, from large to small. The proteins are first dissociated from one another in a test tube, so that they will separate in the gel when an electrical field is applied, as shown in the diagram. The dissociation treatment also leaves negative charges on the proteins, so they will migrate in the direction of the positively charged part of the gel, shown at the bottom.

Here’s a real electrophoresis gel that shows the proteins found in photosystem II. The numbers given next to some of the protein bands in lane 1 give their apparent molecular masses in units called kilodaltons (abbreviated kDa). The largest photosystem II protein has a molecular mass of about 47 kilodaltons, and the smallest, barely visible on the gel, is less than 5 kilodaltons. Lane 2 on this gel reveals the first surprising fact about photosystem II. Proteins of molecular masses of 33, 23 and 17 kilodaltons were removed from the enzyme just by exposing it to alkaline pH. This is a trick biochemists use to separate water soluble proteins from proteins that exist in membranes. The membrane-associated proteins would only be removed from the enzyme if we had been treated with a detergent. So, although photosytem II’s proteins are mainly associated with a membrane (these are the proteins that are shown on lane 2 of the gel), there are three proteins that will dissolve in water if they can be dissociated from the enzyme.

Unfortunately, there’s another result from the experiment in lane 2: removal of the three proteins destroyed oxygen evolution activity and caused all of the manganese to be lost from the sample.

After some head scratching, we found that it was possible to release only the 2 smaller proteins.

Here’s a gel showing the results of this experiment. In this case, manganese was retained by the sample after removal of the 23 and 17 kilodalton proteins, shown in lane 2 of this gel. However, a great deal of activity was still lost. To my dismay, replacing the proteins, as shown in lane three of the gel, did not restore activity. This was not the expected result, based on a lot of experience with other protein systems. Luckily, however, pursuit of this very strange result led finally to the discovery that calcium was required for oxygen evolution activity.

The collected results of a large number of experiments attempting to restore oxygen evolution activity to photosystem II after manipulation of the two proteins are shown in (SLIDE 13). As you can see, this activity loss is not repaired by restoration of the proteins. The ineffectiveness of this strategy suggested that the enzyme might have also lost something relatively small, much smaller than a protein. Finding the "something relatively small" took a bit of time and luck. As this slide shows, a fair amount of activity is restored by the combined additions of calcium and chloride. As I mentioned earlier, chloride was already known to be a required component of the oxygen evolving reaction, and we were using chloride in all of our experiments. It was the calcium result that came as a big surprise. No other biological electron transfer reactions had been shown to require this metal, and calcium itself is unable to exchange electrons with other metals. In other words, calcium is the last metal you’d suspect to be required for oxygen evolution activity, and it’s one of the very last metals we tested as we looked for the missing cofactor. But we weren’t done yet.

Next, it had to be determined if calcium simply mimicked the presence of the polypeptides, or if it was in fact a real cofactor in the reaction. For example, it could be argued that I had failed to put the proteins back in the right place, and calcium was an unnatural substitute for proteins bound to the right place. To resolve this problem, the 23 and 17 kilodalton proteins were removed from photosystem II, and the preparation was divided in half. Proteins were reconstituted to half of the preparation, being careful to avoid calcium and chloride contamination in the process. Finally, ions were added to both preparations, with or without proteins, and photosystem II’s ability to produce oxygen was monitored. Here are the results.

Two findings are shown here. In the first preparation lacking the proteins, activity is restored rapidly by calcium along with chloride (orange bars). As you can also see here, the ions added together do restore activity in the presence of the proteins, but long periods of incubation are required to obtain the effect (blue bars).

This experimental result solved the mystery about the role of these proteins in oxygen evolution, and their relationship to the strange discovery of the calcium requirement. Calcium really is a unique cofactor for oxygen production, and the proteins are needed to trap calcium, and also chloride, in the site where the reaction occurs. Schematic diagrams that explain all of these results are shown on the next slides.

Here’s a diagram (SLIDE 15) of the native oxygen evolving enzyme system in photosystem II, as prepared in its native form from spinach leaves. There’s a good deal of information here, because I’m introducing you to all of the important cofactors of the enzyme. The organic constituents of the enzyme are bound in the green box, which is a representation of the membrane-associated proteins of the enzyme that bind chlorophyll and other organic cofactors.

Light absorbed by the green box drives electrons out of the box, into other enzyme complexes, which direct the electrons to the reactions that convert carbon dioxide to sugars. The electron deficiency in the green "box" is restored by electrons that pass from water, through the manganese atoms, back to an electron-deficient chlorophyll molecule.

Now, for the water soluble proteins. The experiments on the preceding slides indicate that photosystem II uses the 23 and 17 kilodalton proteins to form part of the structure that traps calcium and chloride near manganese in the oxygen evolving site. In this way, the ions are concentrated by photosystem II. Release of the 23 and 17 kilodalton proteins in the laboratory opens this site to the external environment, and calcium and chloride are lost.

In the laboratory, loss of the ions is repaired if we add them in high concentrations to the protein-depleted enzyme, where they can find their sites in photosystem II very quickly (SLIDE 16)(the orange bars in the graph I showed you). In nature, such ion concentrations are never available, and hence the need for proteins to form a container which traps them.

Lastly, if the proteins are removed and then replaced without including calcium and chloride, the enzyme is inactive, and oxygen evolution activity is recovered only slowly (SLIDE 17)(the blue bars in the graph I just showed you), as calcium and chloride diffuse slowly to their sites of action inside the barrier created by these water soluble proteins. I’ll come back to why calcium and chloride are necessary for the oxygen evolution reaction later.

What about the 33 kDa water soluble protein?

The third member of the trio of water soluble proteins is also the largest, at 33 kilodaltons. If this protein is removed along with the 23 and 17 kilodalton proteins, calcium and chloride are lost, but manganese also begins to fall out of the enzyme. This causes a destruction of activity that cannot easily be repaired. Also, when the 33 kilodalton protein is removed, the oxygen evolution reaction is damaged. The oxygen clock makes a couple of turns and shuts down, even if adequate amounts of calcium and chloride are supplied.

Because of its wide-ranging effects on manganese activity in photosystem II, the 33 kilodalton protein has also been called Manganese Stabilizing Protein. We’ve been trying to learn more about how this protein carries out its functions by changing its amino acid composition, using techniques called bacterial overexpression and site-directed mutagenesis. We’ve received a good deal of assistance from Prof. Eran Pichersky in doing these experiments.

Here’s a scheme for how bacterial overexpression is done. Reading from the top of the slide, the gene for Manganese Stabilizing Protein from spinach is inserted into a circle of DNA that will infect the bacterium E. coli. Next the bacteria is tricked into synthesizing Manganese Stabilizing Protein. The techniques of molecular biology can be used to introduce amino acid changes at various sites in the DNA that code for the protein, so that the bacterium will produce mutant forms of Manganese Stabilizing Protein.

In the next step, the native spinach protein is removed, and replaced with a bacterially-produced protein (SLIDE 20). Then, the properties of oxygen evolution activity are examined with the new protein in place. In these experiments, calcium and chloride are added to make up for the absence of the 23 and 17 kilodalton proteins in this particular form of the enzyme. The experiment works! A green box from spinach, and a Manganese Stabilizing Protein from a bacterium will produce oxygen evolution activity that cannot be distinguished from the native enzyme system.

Here’s an amino acid sequence map of Manganese Stabilizing Protein. The bold-face letters identify a number of the amino acids where mutations have been made. The point of this slide is to illustrate the power of molecular biology-a large number of manipulations over the whole protein sequence can be carried out in a relatively short time, and changes can be made in amino acids that would not be possible using the usual techniques of biochemistry. I’ll give you a general summary of results, but first I’ll show you just one experiment to illustrate the kinds of data that have been obtained.

This slide shows an experiment in which we’ve systematically removed one amino acid at a time from the back end of Manganese Stabilizing Protein. This is done by inserting what is called a "stop" code into DNA which replaces the code for an amino acid and eliminates it from the protein. The first couple of mutant proteins have lost the amino acids glutamine and glutamate. As you can see, they aren’t very interesting. However, when the amino acid leucine is removed in addition to glutamine and glutamate, things start to happen. This mutant protein is unable to restore oxygen evolution activity, although it will bind to photosystem II. In addition, its structure in solution is drastically modified. This mutant’s size is increased, and the data collected so far indicate that it’s lost what is called "secondary structure", coils and ripples in the protein backbone that are caused by amino acids interacting with one another. The situation is actually even more dramatic when four amino acids are removed. I’ll return to what this change in the Manganese Stabilizing Protein is telling us in just a minute.

A brief summary of results from overexpression and mutation experiments on Manganese Stabilizing Protein are shown below. Without mutations, the bacterial version of the protein works just fine. By mutation, amino acids have been identified that are important for maintaining the protein’s structure. Other amino acids control its ability to bind to photosystem II, and there are also amino acids that are important in promoting the electron transfer activity of the manganese atoms. A current project is devoted to analyzing these mutant proteins in an effort to distinguish between defects that can be traced to structural changes in Manganese Stabilizing Protein, and defects caused by amino acid changes that directly affect electron transfer activity of manganese itself.

Because of the unusual size increase caused by removal of amino acids from Manganese Stabilizing Protein, we suspected that it might have something in common with a number of other proteins distributed through many organisms, including humans. To see if this was so, we heated Manganese Stabilizing Protein, and found that its activity was not damaged. This is a strange result; the proteins of most organisms are not stable when heated. In fact, some of you have no doubt observed, and eaten, the disgusting result that is obtained when you heat spinach leaves. The meaning of Manganese Stabilizing Protein’s heat resistance is presented with the other conclusions, on the next slide.

A number of results from several laboratories, including ours, indicate that the protein is essential for maintaining the ability of manganese to receive and pass along electrons originating from water. Such activity corresponds with the ability of the oxygen evolving site to bind calcium and chloride in the light. In other words, efficient functioning of the electron deficient S-states, such as S₂, S₃ and S₄, requires Manganese Stabilizing Protein. So, in addition to forming a wall of the inorganic ion "box", Manganese Stabilizing Protein is also likely to have a more direct role in positioning calcium and chloride near manganese atoms.

Now, for the second item on this slide, the explanation of the strange heat stability of Manganese Stabilizing Protein. At the risk of sounding like a broken record, such a result was totally unexpected. However, when this odd behavior is combined with all of its other properties, Manganese Stabilizing Protein is seen to belong to a family of very interesting proteins that are involved in a wide range of biological processes, a few of which I’ve listed here. These proteins only possess a limited amount of organized structure when they’re dissolved in water. This explains their resistance to heating. The current hypothesis is that a disorganized structure enables such proteins to be flexible and to match up with other proteins in the larger biological complexes of which they are members. The resulting protein-protein interactions are most often beneficial, for example in protein synthesis or oxygen evolution. No all interactions are beneficial, however, as in the case of amyloid plaque formation in Alzheimer’s patients.

This schematic diagram (25^th SLIDE), which you’ve already seen, gives the present view of Manganese Stabilizing Protein and its smaller water-soluble partners. These important proteins form the a structure that traps the essential ion cofactors, manganese, calcium and chloride, that are the catalysts of water oxidation. This structure also prevents their loss from the enzyme under a range of conditions. Without these proteins, photosynthetic oxygen evolution could not occur.

Up to now, I’ve ignored the green box part of photosystem II, the proteins that are closely associated with one another and that interact very strongly with lipids in a membrane system. These proteins are typically very hard to work with, because they won’t dissolve in water, and can only be separated from one another using detergents. With some effort, it proved possible to remove proteins of the "green box" that aren’t essential for oxygen evolution activity. In doing so, the simplest protein unit capable of oxygen evolution could be defined....................

What is shown here is the two-step dissociation of hydrophobic membrane proteins from photosystem II. We have to give up the 23 and 17 kilodalton proteins in these experiments, but the calcium/chloride trick can be used to assay activity. The lane on this gel that is labeled 3 reveals the protein content of the purest preparation we have been able to obtain which retains the oxygen evolving reaction. There are 6-7 membrane-associated proteins in the highly purified enzyme, along with the 33 kilodalton Manganese Stabilizing Protein. So, here is the smallest protein unit capable of oxygen evolution, the heart of photosynthetic electron transfer, and therefore the heart of life on earth.

Here’s a schematic representation, a cartoon, of a low-resolution view of how the proteins of photosytem II are arranged in a membrane (SLIDE 27). The heart of the system is here along with the 33, 23 and 17 kilodalton proteins. The general location of manganese is shown. I’d like to conclude my talk by telling you about how we think manganese and calcium and chloride are organized with respect to one another. Finally, I’ll speculate on how these ions participate in the oxygen evolving reaction.

A very brief summary of what is known about the ions that are needed to catalyze oxygen evolution is shown below. The data here are from a number of research groups, including results from biochemical experiments done in my laboratory as part of interdisciplinary spectroscopic studies carried out with the groups of Jim Penner-Hahn in the Chemistry Department here, and Hans van Gorkom’s group at Leiden University in Holland. Prof. Penner-Hahn is an expert in the technique of X-ray absorption spectroscopy. He uses a linear accelerator to obtain information on the oxidation states, and structures, of the ions in photosystem II. This is big-time science! Prof. van Gorkom’s group has devised techniques for analyzing the various S-states of photosystem II. Our joint results, along with those from other research groups, are summarized here.

SUMMARY OF IONS Mn, Ca, Cl THE ROLES OF INORGANIC IONS IN OXYGEN EVOLUTION MANGANESE (Mn) à The dark stable S1 state of photosystem II, the 4 manganese atoms have oxidation states of +4/+4/+3/+3 à A manganese atom is oxidized on every S-state advancement with the possible exception of S3 à S4 à The photosystem II-associated manganese cluster is comprised of two individual dimers of manganese atoms, one all +4 oxidation state, one all +3 Calcium (Ca2+) à Only Sr2+ replaces Ca2+ in activation of oxygen evolution; lanthanides are inhibitory, suggesting that Ca2+ may bind the water molecules that are converted to oxygen à Ca2+ imparts stability to binding of the manganese atoms Chloride (Cl-) à Chloride is required for the S-state transitions S2 à S3, and S3 à S4à S0 à Chloride is required for reduction of S1to lower S-states (S0, S-1, S-2) à Chloride is likely to bind to manganese atoms during the S- sate cycle, and to act as a regulator of the ability of these atoms to donate or receive electrons in the course of oxygen evolution chemistry

First, manganese. There’s a growing consensus that in the dark-stable S₁ state, manganese ions in photosystem II exist as manganese 3+ or manganese 4+, rather than as the electron rich form of the metal, manganese 2+. There is also general agreement that a manganese atom loses an electron on most S-state advancements. It’s likely that these electrons come from the manganese 3+ atoms, since they will give up an electron more readily than will a manganese 4+ atom, which is more electron deficient. Structural results from X-ray absorption spectroscopy experiments suggest that the manganese atoms are organized into 2 clusters, of 2 manganese each.

Second, calcium. This ion is replaced functionally only by Sr²⁺. We’ve tried other metals, but none of them will restore oxygen evolution. Because of this observation, it seems probable by analogy with other protein systems that calcium may be the site where water is bound before its converted into oxygen. We have also been able to show that calcium imparts stability to the binding of the manganese atoms. This is consistent with what is known about the structural role of calcium in other protein systems.

Finally, chloride. Understanding the role of chloride has been a real challenge. In joint experiments with Hans van Gorkom, we could show that chloride is needed to advance the oxygen clock S-states from S₂ on through the step where oxygen is produced and released. We’ve also been able to show that it’s necessary if the clock is to be reversed, or pushed backwards. All of the probing experiments we’ve been able to do suggest that chloride must be bound to manganese, but there is still no absolute proof of this.

The next slide assembles these data and speculations about the ions of photosystem II into a structural picture of the site where oxygen is produced as it might exist in the dark-stable S₁ state.

Here are the two pairs of manganese atoms I mentioned previously. It is believed that these manganese atoms are connected by oxygen atoms that are not the source of oxygen released into the atmosphere. Of course, there are other bonds to the manganese atoms, but they’re not shown here because I’m not very good at using PowerPoint. We have an excellent estimate of the distances that separate the ions, which is given on this slide in Angstroms (the A’s with circles above them; one angstrom is about 1-10 billionth of a meter). One pair of manganese atoms, shown in yellow as manganese 4+ atoms, are probably as electron deficient as possible. This pair of metals probably help to preserve the structure of the oxygen evolution site. The manganese 3+ pair shown in red are therefore the manganese atoms that are involved in reactions that lead to oxygen production from water. We know that Calcium is further from either pair of manganese atoms than the manganese atoms are from one another, and we strongly suspect that chloride is bound to one of the manganese 3+ atoms to facilitate its activity in transferring electrons on to the organic cofactors of the membrane-associated proteins.

We can now take this model, mix it with the oxygen clock (SLIDE 30) and generate a speculative sequence of events to describe the oxygen evolving reaction itself. This will present a hypothetical picture of how manganese atoms are involved in first transferring electrons out of the clock, and finally, how they are involved in withdrawing electrons from water. In what I’ll show you, calcium will be used to bind and transfer water to a manganese atom, and chloride will regulate the path of electron transfer by tagging a single manganese atom that will pass electrons on to the chlorophyll-catalyzed light reactions. Remember that what you will see is in some respects speculative.

Starting from S₁, chlorophyll photoreactions withdraw an electron and release a proton from a manganese 3+ atom. The resulting manganese 4+ atom binds a hydroxide molecule created by loss of a proton from a water molecule (SLIDE 31). An electron is transferred internally between manganese atoms, so S₂ now has one bound water molecule, and a manganese atom has lost an electron. This moves the clock from 1 to 4 o’clock.

Now, a chlorophyll photoreaction removes another electron from the manganese atom, and another water molecule moves from calcium and binds to the newly formed manganese 4+ atom to give the S₃ state. It’s now 6 o’clock, oxygen time (SLIDE 32). At this point, we’ve hit a sort of chemical dead end. I already said that it’s difficult to remove electrons from manganese 4+, but the clock is one electron short of moving to 8, which will give the S₄ state where oxygen can be produced.

Because of the chemical dead end, the final step is the most speculative of all of the reactions I’ve shown you. An electron is removed from S₃, possibly by a transient oxidation of manganese 4+ to 5+, or by withdrawal of an electron or hydrogen atom from a bound water. What does seem reasonable in terms of work that’s been done by Prof. Pecoraro’s group here at Michigan, is that an intermediate peroxide molecule forms. This is shown in the S₄ diagram as the pair of oxygen atoms with a bond between them. This peroxide molecule can surrender electrons to the manganese 3+ and 4+ atoms, moving the clock to 11, and releasing oxygen (SLIDE 33).

The last step is a photoreaction that removes an electron from S₀ to restore the enzyme to the S₁ state, back where we started from (SLIDE 34).

I emphasize that a number of models like this one exist, all of which are based on realistic possibilities for reactions between water and manganese atoms. In fact, right now there are probably more models than there are data to support any of them. One of the obvious challenges confronting researchers working in photosynthesis today is to sort out the exact molecular details of the water oxidizing reaction.

What I’ve shown you this afternoon is that an interesting and complicated enzyme, photosystem II, has really emerged from the shadows, that we now have a much clearer picture of how its parts fit together, and how it works so plants can make oxygen. This enzyme system has proven to be frustrating to work with, and at the very same time fascinating as it has revealed its secrets. As I’ve told you, we now know that photosystem II is made up two types of proteins, those that are water soluble and those that are membrane-associated. Three ions, manganese, calcium and chloride form the site where oxygen is produced, and chlorophyll and some other organic cofactors, form the part of the enzyme that traps light and transfers electrons on to other parts of the photosynthetic apparatus.

The picture of how the components of photosystem II cooperate with one another is constantly improving, but my past experience with this enzyme convinces me that there are a number of interesting surprises lying ahead, in the future. No other biological system traps light, makes biochemical energy and unloads oxygen as a waste product. So it’s not surprising in hindsight that photosystem II has proven to be so unique in its properties.

We have come a long way in satisfying one of the aphorisms of my post-doctoral mentor, Efraim Racker, who was a biochemist who led the way in attacking, disassembling, and reconstructing complicated enzyme systems.

" DON'T THINK CLEAN THOUGHTS WITH DIRTY ENZYMES" Efraim Racker

What Ef meant by this saying, was that you should get an enzyme as pure as possible and understand its activity before you start cooking up models and speculations about how it works. Ef also used to go around saying "Troubles are good for you". I swear, I could never figure out what he meant by that, until I started taking apart photosystem II.

Ef got one-upped by the Swedish biochemist Lars Ernster

" DON'T THINK DIRTY THOUGHTS WITH CLEAN ENZYMES" Lars Ernster

Right now, this saying represents another challenge facing investigators working on photosystem II. It’s a useful saying to keep in mind when the urge arises to invent models, rather than to generate more data.

In concluding I’d like to acknowledge the people over the past two decades who have done the work that I’ve been talking about today, post-docs, graduate students, and undergraduates who have been wonderful colleagues. People without asterisks are not unemployed, by the way. They’re still working on the project, or in graduate or medical school at another institution. I’ve had two excellent technicians, whose names are not shown here. I’d like to mention Dimitris Demetrious and Jeanine Ross. I’d also like to thank NSF and USDA for supporting this research from its inception.

Last, but certainly not least, I’d like to acknowledge a group of faculty colleagues here at Michigan and elsewhere with whom I’ve collaborated on a number of the experiments that I’ve discussed. Research on complex systems like photosystem II requires a wide range of knowledge, I’ve been extremely fortunate to have these individuals as colleagues, and as friends.