Category Archives: journal club

Cyanobacteria Shed

Here’s something new under the sun… cyanobacteria shed in the ocean.


Most people are well aware of the fact that dogs and cats shed, but did you know that oceanic cyanobacteria also shed?* OK, so the cyanobacteria don’t shed hair, and you won’t need a lint brush. In a report in this week’s Science, researchers from the Chisholm lab show that cyanobacteria living in the world’s open oceans are shedding zillions of membrane vesicles. Before you accuse them of pollution, allow me to explain why they are the most important bacteria you’ve never heard of– the Prochlorococci. Don’t worry about being too late to this fountain of knowledge, scientists only discovered them about 25 years ago.

Prochlorococcus species are cyanobacteria that live in the oligotrophic ocean, areas of the world’s open oceans that are nutrient poor when it comes to sources of nitrogen, phosphorous and bioavailable forms of essential metals like iron. These bacteria dominate these otherwise ocean deserts and are possibly the Earth’s most populous species by sheer numbers of individuals. Remember about half of the world’s primary productivity comes from aquatic sources. Prochlorococci are a major contributor to global nutrient cycles and a significant number of the oxygen molecules you are constantly breathing in and out were produced by a Prochlorococcus.

You may be curious as to why scientists have only recently discovered these numerous and critically important photosynthetic organisms. The answer is because of their other trademark- their size.  Measuring in with widths and lengths well under a micrometer, they are the world’s smallest photosynthesizers. Prochlorococci pass through many of the standard filters used for trapping microbes from aquatic samples. They were detected using a cell-sorting technique that could just detect their faint chlorophyll fluorescence signal. These fascinating microbes have been studied in the ocean as well as the laboratory for the last couple of decades, but it is a labor of love. Since Prochlorococci live in such a nutrient-limiting environment, they don’t have much competition. Consequently, these cyanobacteria don’t have to grow very fast, and their small size is another way of conserving their resources. In the laboratory, they don’t grow as fast as the other model cyanobacteria and their extremely small size means scientists have to wait quite a while before significant biomass is achieved for experiments.

So, it is quite a remarkable discovery that something so tiny that has toiled so diligently to fix carbon dioxide into biological molecules would just shed them as tiny vesicles into the ocean. The vesicles released by the Prochlorococci and Synechococci** are small membrane-closed vesicles that contain proteins and even small fragments of DNA and RNA. Surely, the bacteria are not just throwing away precious resources. It’s unlikely that science has stumbled upon the first eating disorder among primary producers, so the obvious question arising from this observation is why the bacteria shed these vesicles.  The authors offer some hypotheses to justify this seemingly wasteful phenomenon.

It could be an example of the cyanobacteria ‘casting their bread upon the waters.’ Preliminary results suggest that- as the saying goes- they find it again. The open ocean environment is a hard place to live, but cyanobacteria are not the only organisms living there. Heterotrophic microbes are also part of the biological community and previous reports have demonstrated that they can stimulate the growth of Prochlorococcus strains. Perhaps the vesicles are a way for the cyanobacteria to feed their beneficial heterotrophic neighbors without putting their cells on the menu.

The vesicles could also serve as bacterial countermeasures against cyanophages, viruses that infect cyanobacteria. For all intents and purposes, the vesicles look like miniature bacteria complete with receptor proteins the phages use to recognize their bacterial victims. The authors were able to observe ‘fooled phages’ that had bound and injected their genetic material into vesicles. In this way, the shed vesicles would be a good defense against these oceanic pathogens. From the phage perspective, an ocean filled with zillions of vesicles, each appearing like a potential host would seem like a lottery they can’t lose. However, the cyanobacteria have ensured the odds are in their favor by shedding many, many more vesicles into the environment relative to the susceptible cells.

Finally, the vesicles could just be the clichéd ‘message in bottles’ of seafaring lore. Even an ocean teeming with trillions and trillions of other Prochlorococci is isolating when reproduction is asexual. For microbes, setting your genetic material adrift on the ocean swells is a hopelessly romantic gesture. That DNA may eventually make its way to another cyanobacterial cell that will find it useful and incorporate the DNA into its genome. It is also possible that the messages are not necessarily genetic. The vesicles could also be used to store and transport specialized metabolites as a defensive or offensive strategy. For now, only the cyanobacteria seem to be able to decipher the messages.


* Hopefully, you weren’t unwittingly gifted any pet cyanobacteria for Christmas.

**larger cyanobacteria than the Prochlorococci, but still not in a position to freely give away hard won nutrients

Fun fact: Dr. Chisholm has also co-authored a couple of children’s books on photosynthesis. Check them out here.


One more thing about those solar-powered sea slugs (aka heterotroph in autotroph’s clothing)…

OK, well really this post is about the algae they eat. I’ve dedicated enough space to an animal on this blog. Regardless of whether or not (and it’s decidedly not) the slugs need the chloroplasts to be photosynthetically active, the kleptoplasts are remarkably stable. At least as far as the light reactions go based on long term chlorophyll fluorescence measurements, they are capable of photosynthetic electron transfer after months of residing within the digestive tract of the slug.

Genome Biol Evol 20132013  5(12) 2540-8, Fig. 1._

How remarkable is that? Remember last time I mentioned that it took hundreds of genes coming from the nuclear genome of a photosynthetic eukaryote to keep their chloroplasts functional? It’s no small feat.

Remember how I told you that photosynthesis researchers had to be fast and cold (and often dark) when it comes to biochemical preparations of chloroplasts and thylakoid membranes? I’ll tell you isolated chloroplasts kept on ice in the dark won’t remain that active for a week much less at ambient ocean temperatures.

I’m sure the slugs still have some yet-to-be-discovered secrets of chloroplast maintenance, but the first one is knowing where to steal their chloroplasts. Acetabularia acetabulum is the food of choice of the chloroplast-thieving sacoglossan slug Elysia timida. Researchers sequenced the chloroplast genome of Acetabularia acetabulum and identified clues as to the longevity of the stolen chloroplasts- the ftsH and tufA genes.

Here’s why that’s important: As far as stability goes, the weak link in the photosynthetic electron transfer chain is the Photosystem II (PSII) enzyme. As part of its normal activity, the core D1 subunit is irreversibly damaged. The damaged protein must be removed and replaced with a newly synthesized version. In photosynthetic organisms, an elaborate PSII damage-repair cycle keeps the light reactions functional. One of the auxiliary factors required for this process is FtsH, a protease involved in the removal of the damaged D1 protein. Normally, the ftsH gene is encoded within the nuclear genome of photosynthetic eukaryotes. Translation elongation factor Tu (encoded by tufA) is essential for the translation of chloroplast proteins. Since the PSII damage-repair cycle and the longevity of chloroplasts require a significant amount of new protein synthesis, the translation elongation factor Tu would be limiting over a period of months if the chloroplasts couldn’t make this protein on their own.

In the case of Acetabularia acetabulum, ftsH and tufA are encoded within the chloroplast genome. Similar trends were found in aquatic algae that were food sources for the sacoglossan slugs. This means the chloroplasts can make this critical factor for themselves, giving them a higher level of autonomy for maintaining their photosynthetic machinery. Elysia timida wouldn’t be so lucky if it tried stealing chloroplasts from the alga Chlamydomonas. Likewise for any plant chloroplast.

Genome Biol Evol 20132013  5(12) 2540-8, Fig. 3._

Additional experiments are necessary to directly evaluate the effect these chloroplast genes have on the long term retention of active kleptoplasts in the sea slugs. I’m sure there’s more to the story of how the stolen chloroplasts are maintained, but the ability to independently make these two essential proteins is a good start.


Reference (open access)

About those solar-powered sea slugs…

Elysia chlorotica,Credit: Patrick Krug Cataloging Diversity in the Sacoglossa LifeDesk via Wikipedia

About those ‘solar powered sea slugs’… a new study has given me reason to change their status from super photosynthesizer to plain heterotroph. It seemed too good to be true: sea slugs that can slurp on algae, digesting all but their chloroplasts which are retained in an active state inside the digestive tract conferring a photosynthetic lifestyle to the sea slug hosts. With such an elaborate mechanism for retaining these photosynthetically active chloroplasts, of course it must be to the nutritional advantage of the sea slugs. The early studies on the sea slugs suggested as much. The community of photosynthesis researchers, including myself, was taken with the idea of heterotrophs becoming photosynthetic. Not that we really wanted to convert the human race to autotrophy, but if some organism had figured out a way to enslave chloroplasts, then we should be researching those animals. Was it possible that slugs knew more than humans about the complexities of chloroplast exploitation?

Photosynthesis in plants and algae involves an elaborate exchange of proteins and signals between the chloroplasts and the cell nucleus for proper function. Hundreds of genes are involved and the system acutely senses the needs of the cell and the primary inputs of photosynthesis (light, water, carbon dioxide). Harnessing the chloroplast for the biochemical benefit of a cell is no small feat. Researchers have been looking for answers as to how the slugs might accomplish this. While there have been some instances of an algal gene present in the sea slugs, these are few and far between.

At this point, scientists had to reconsider the hypothesis that the stolen chloroplasts were conferring some kind of autotrophy to the slugs. A recent report in Proceedings of the Royal Society B addresses this assumption. Previous work had established that the presence of kleptoplasts offered an advantage to the slugs during starvation studies, but was it because they were photosynthetically active? Christa and co-authors were able to make more carefully-controlled weight measurements during starvation experiments of different kinds of kleptoplast-containing sea slugs. They compared weight loss of sea slugs starved (1) in the light, (2) in the dark, and (3) in the light in the presence of an inhibitor of photosynthesis. It turns out that there was no difference in the rate of weight loss among these conditions. This means that it doesn’t matter whether the chloroplasts are photosynthetically active or not. Consequently, the slugs are not gaining any autotrophic advantage, but are likely just holding onto them as a food reserve.

The sea slugs and the algal chloroplasts still have an odd relationship that begs for a deeper meaning. The overwhelming majority of the Earth’s biosphere depends on heterotrophs eating photosynthetic organisms, but I can’t think of another example of another heterotroph that saves organelles in specialized structures within its digestive tract (biochemically active or not). It seems like an over-engineered way to stash a midnight snack.

As an omnivore myself, I can’t be too condemning of these sea slugs for feasting on algae. However, given this new information, Elysia chlorotica is the herbivore equivalent of Buffalo Bill* saving parts of its victims as an ornate garment. This revelation definitely gives a creepy vibe to heterotrophy. Are the ‘kleptoplasts’ decorative? Maybe we are on the brink of photosynthetic fashion, remember these guys?

Why chloroplasts and why not some other organelle like mitochondria? Do the sea slugs use some elaborate form of communication based on chlorophyll fluorescence signals? Elysia chlorotica may not need the stolen chloroplasts to be photosynthetically active, but they still need something from them. Researchers have shown that the slugs cannot continue through development past a certain point without ingesting the algae and simultaneously stealing the chloroplasts. For their diet and development to be so tightly linked, there must be some intimate connection between these two species (or species and organelle) that scientists haven’t identified yet. There may be something special about chloroplasts,**but for Elysia chlorotica it isn’t photosynthesis.

This is one strange relationship that doesn’t fit nicely into our typical interspecies interaction categories. It’s not a symbiosis. I would call either organism a parasite. It’s more than an herbivore and its lunch. One thing that is safe to say about it, Elysia chlorotica and its stolen chloroplasts will keep researchers busy for quite some time into the future.


* I’m talking about this guy from the Silence of the Lambs and not this guy from American Wild West history.

**Of course there’s something special about chloroplasts, haven’t you been reading the blog? But I guess Elysia and I will have to agree to disagree on the importance of photosynthesis.

References and Links:

NAL1 Spikes Back

Mature rice panicle against blue sky. Part of ...

Mature rice panicle against blue sky. Part of the image collection of the International Rice Research Institute (IRRI) . (Photo credit: Wikipedia)

Earlier this fall, I highlighted a research paper on a gene responsible for increased yields in rice. Variants of this previously characterized gene played a significant role in increased photosynthesis rates and ultimately higher grain production. In this week’s PNAS, a separate group reports on the importance of a NAL1 variant for increased rice yields.*



Briefly, I will go through the experimental highlights of the new paper…

Using different genetic lines of rice, Fujita and co-authors were able to isolate a trait conferring increased grain yield. They designated the genetic element controlling the trait ‘SPIKE**’ for increased spikelet number. When this element was mapped to a specific location within the rice genome, the researchers found that it corresponded to the NAL1 gene that has been previously characterized. The SPIKE allele has three key amino acid changes in the NAL1 gene. The authors perform a number of physiological assays on rice plants containing the following types of NAL1: the SPIKE allele, SPIKE overexpression construct, SPIKE silencing construct and NAL1 (wild-type version for typical rice cultivars). The SPIKE allele was correlated with increase in total spikelet number. The authors were also able to show that the presence of the SPIKE allele had a positive impact on grain yields in certain rice varieties.

Characterization of yield-related traits of a NIL for SPIKE. (A) Plant morphologies. (B) Panicle structures. (C) Flag leaves. (D) Cross-sections of panicle neck. (Scale bars: A, 20 cm; B, 10 cm; C, 5 cm; D, 500 µm.)

Characterization of yield-related traits of a NIL for SPIKE. (A) Plant morphologies. (B) Panicle structures. (C) Flag leaves. (D) Cross-sections of panicle neck. (Scale bars: A, 20 cm; B, 10 cm; C, 5 cm; D, 500 µm.)

What I’d really like to go through in this post is a comparison of the two papers…

  1. By my reading the papers use very similar methodology geared toward uncovering genetic elements in rice that increase yields. I will admit that some of the finer points of rice genetics may elude me, but basically both groups have used rice breeding in various cultivar backgrounds to tease apart the complex trait of grain yield.
  2. It should be noted that the earlier Scientific Reports paper happened upon the NAL1 gene because they were assessing photosynthetic properties. Remember, their working name for the allele they found was GPS (GREEN FOR PHOTOSYNTHESIS). The PNAS paper focused solely on physiological parameters based on yield. This highlights how photosynthesis and yield are connected in crop plants. Of course, given my bias for photosynthesis, you can probably guess which group I’m giving bonus points to.
  3. The natural variants of NAL1 found in both articles hinge on differences in the two main varieties of rice Oryza sativa indica vs. Oryza sativa japonica. You can guess the origin of each based on their names, but that just refers to the stock genetic material. Both types of rice are grown all over Asia. The Scientific Reports study used cultivars of the two varieties that show the greatest photosynthetic difference. Their work showed that the natural variant of the NAL1 gene they found was key to increased yields in the indica variety, but the gene did not appreciably increase yields in the japonica genetic background. The PNAS study emphasizes the importance of the SPIKE allele to the yields of indica cultivars, which is consistent with the earlier report.
  4. I’m still left wondering what NAL1 actually does and what kinds of alleles these are giving these advantageous agronomic traits. I have to admit I got a little lost in the physiological markers used for judging the different rice plants as well as in the genetic nomenclature for the different NAL1 alleles used in both papers. It is still not clear to me whether GPS and SPIKE represent loss of function or gain of function alleles. I’m leaning toward a decrease in expression or function relative to the NAL1 versions found in other rice lines. Don’t get me wrong. I do appreciation the enormous amount of genetics and plant physiology work represented in both of these papers, but the biochemist in me is screaming, “So, what does the NAL1 protein do anyway? What is it not doing or doing differently with the GPS and SPIKE mutations?”
  5. The PNAS paper does not cite the work of the Scientific Reports paper. For those scientists in the audience, the significance of this is immediately apparent. For everyone else, here’s the big deal. When it comes to scientific publishing, there’s no prize for second place. Usually, for such similar articles, the second one just never sees the light of day. The work isn’t novel any more. Now, there are numerous instances when competing groups publish very nearly at the same time (same journal issue) or close enough that both manuscripts would have been in peer review at the same time (not at the same journal). However, when they are several months apart***, the authors of the second article generally have to answer questions about the first one. As in- how is this new work different? more important? better? an extension of the published work? Upon initial submission of the PNAS manuscript, the Scientific Reports paper wasn’t even accepted, but surely Fujita and co-authors would have noticed the Scientific Reports article during the manuscript revision phase. Of course, maybe not, but I really think someone (author or reviewer) should have noticed the earlier work.
  6. I don’t consider these two papers to be completely redundant. I’m sure there is someone out there that will enlighten me to the nuances of rice genetic techniques. Nevertheless, I think that instances such as this strengthen the scientific enterprise. Increasing agricultural yields is a high stakes field, and it is good to know that the science behind novel breeding markers is sound and validated. Given the fact that some reports suggest that an absurd amount of scientific studies are irreproducible, this kind of redundancy or overlap is welcome.

So, there are lessons here for both rice specifically and science in general.



* Check it out. It’s Open Access!

** Geneticists always get to come up with snarky names for the genes or alleles of genes they work on.

*** I realize that the time between actual publication dates for hard copy journals may not always tell the complete story with respect to scientific timelines. Pro-tip: All manuscripts have footnote information on when the manuscript was initially submitted, finally accepted, and published (online and/or) hard copy. For the case of these two papers, the Science Reports article was accepted a week after the PNAS paper was submitted for review. And yes, I’m sure these are different groups. If there was ANY overlap, they definitely would have cited the earlier work. Rule #2 of academic publishing is ‘cite yourself or no one else will.’ (Rule #1 is, of course, Publish or Perish.)

The Dark Side Comes to Life in Fluorescent Colors

Something new under the sun… carboxysome assembly in real time.

Today’s post features research out of the Kerfeld Lab from a close colleague of mine, Jeff Cameron.* After doing his thesis work on redox homeostasis in cyanobacteria, he moved on to a postdoc on the dark side of photosynthesis.** His new paper just out in Cell today sheds new (fluorescent) lights on a longstanding question in cyanobacterial carbon fixation.

Cameron et al. Cell Volume 155, Issue 5 2013 1131 – 1140

Schematic for cyanobacterial cell structure showing membrane systems and carboxysomes.

Schematic for cyanobacterial cell structure showing membrane systems and carboxysomes.

I’ve mentioned before that of the oxygenic photosynthetic organisms, cyanobacteria have an elaborate system for concentrating CO2 at the site of Rubsico. This prevents Rubisco from getting confused and using O2 instead, a costly mistake for the cells. Sure, these cells have an array of transporters as well as carbonic anhydrases to interconvert CO2 and bicarbonate. But the real heart of the cyanobacterial Carbon concentrating mechanism (CCM) is the carboxysome- an intracellular microcompartment where the cells corral all of their Rubsico. This structure allows the cells to handily funnel all of their acquired CO2 so that it can be fixed by Rubisco. No, I didn’t say it’s an organelle. Cyanobacteria are prokaryotes; while they may break the rules by having an internal thylakoid membrane system, they do not have true organelles in the sense that eukaryotes do.

So what are they exactly? They are protein-based structures, in which an array of Rubsico is encapsulated by hexameric and pentameric shell proteins. Carboxysomes have been well-studied from a structural standpoint and the major protein components have been identified.

Schematic model of the α-carboxysome assembly containing RuBisCO small (dark green) and large (green) subunits and carbonic anhydrase (red). The shell is composed of hexamers (blue), pseudohexamers (light blue, magenta, and light green), and pentamers (yellow) From Kinney et al 2011 Photosynthesis Research

However, all of this beautiful structure work doesn’t provide any information on the dynamics of these components as the cells must construct them.

Cameron and co-authors describe an experimental system that allowed them to follow the assembly of carboxysomes from scratch. This system also allowed them to eliminate carboxysome subunits one at a time and monitor the effects on assembly. From all of these results, they were able to piece together a model for carboxysome biogenesis, something that wasn’t possible from looking at static pictures from single mutants.

Graphical Abstract Cameron et al 2013 Cell Biogenesis of a Bacterial Organelle: The Carboxysome Assembly Pathway

Graphical Abstract
Cameron et al 2013 Cell
Biogenesis of a Bacterial Organelle: The Carboxysome Assembly Pathway

Here’s how the method breaks down:

Observations: Mutants of Synechococcus elongatus PCC 7942 that lack the carboxysome structural proteins cannot confine their Rubisco to small areas and require higher than air-levels of CO2 to grow photosynthetically. Individual mutants of the various carboxysome subunits have variable effects on carboxysome structure, but these are not useful for understanding how the carboxysome subunits come together in a temporal sequence during assembly.

Hypothesis: If a system could be generated to synchronize carboxysome assembly (going from no carboxysomes present to fully assembled carboxysomes), then researchers could probe the effects of the presence/absence of structural components on assembly. This information will allow researchers to deduce the order of assembly events.

Experiment: A synchronized carboxysome assembly system was generated by starting from a mutant of Synechococcus elongatus PCC 7942 lacking all carboxysome structural subunits (no carboxysomes). These structural subunits were added back to this mutant background (as a complete set and as with individual missing subunits) under the control of an inducible promoter. These various cell lines were also engineered to express a fluorescent version of Rubisco were produced to serve as a marker for carboxysome assembly (diffuse fluorescence = no assembly; tight fluorescent spots = assembly).

Results: The control cell line (fluorescent Rubisco + complete set of structural carboxysome subunits under inducible control) showed a diffuse localization of Rubisco throughout the cells prior to the induction of carboxysome genes. After the addition of the inducer chemical, the fluorescence from Rubisco focuses into small green spots indicative of carboxysome assembly. The cases where less-than-the-complete-set of carboxysome subunits were added back to the original mutant cell line, the induction yielded variable results for carboxysome assembly. Some did not assemble any carboxysomes (diffuse Rubisco fluorescence) while others showed intermediate patterns of Rubisco fluorescence (abnormal or partially assembled carboxysomes).

Conclusions: The authors deduced a new model for carboxysome assembly. First Rubisco proteins and the subunit CcmM aggregate together to for pro-carboxysomes (PCs). CcmN is then added to the PCs and begins to associate with the outer shell proteins CcmK2 and CcmO. Finally, when the CcmL protein is incorporated, the carboxysome completely enclosed by its protein shell and released from the PC.

Think Ahead: Carboxysomes are a prototype for other bacterial microcompartments. Other bacteria use similar protein-based microcompartments for specialized metabolism. This system and the insights on the timing sequence of its assembly can applied to these other systems. In any case, these experiments lay the groundwork for engineering similar microcompartments for medical or biotech purposes. For instance, plants might be coerced to make carboxysomes for more efficient carbon fixation. Alternatively, it may be possible to create novel microcompartments in other bacteria for the purposes of efficiently making useful metabolites.

Check out the link below for the fluorescence microscopy video of carboxysome biogenesis. Sorry, there’s a paywall for 12 months.

UPDATE: There is a free version of Supplemental Video #2. Thanks JGI!

This is after all carboxysome genes have been turned on and you can watch them move and even see some bud off after awhile. (I still prefer the other video where you can watch them form from nothing.) Also, in the current edition of the video caption, they credit James Cameron instead of Jeff Cameron. Sure, James does great cinematic work- just not this one. (It should be fixed on Monday.)

Jeff Cameron, the James Cameron of carboxysome cinema.



*Jeff and I were graduate students together in Himadri Pakrasi’s lab at WashU. I think he still hears my voice in his head when he does biochemical purifications. “Fast and Cold! Jeff. Fast. And. Cold.”

**Yes, yes, I know. Light-independent. Whatever. ‘Dark’ makes for catchier writing and storytelling.


New: D1 Processing (in plants!)

Our last journal club was a throwback article of mine on D1 processing and PSII assembly in cyanobacteria. Just this week, a new report in PNAS describes what happens in plants when the D1 protein isn’t processed. Che and co-authors still found that D1 processing is essential for PSII assembly, but there are some new details that reveal differences between the cyanobacterial and plant systems. Check out my previous post for more background info.

C-terminal processing of reaction center protein D1 is essential for the function and assembly of photosystem II in Arabidopsis* Che et al PNAS 110: 16247-16252

Here’s how the science breaks down:

Observations: Processing of the precursor D1 (pD1) protein in cyanobacteria is critical for PSII assembly and photosynthetic growth in cyanobacteria. This processing step does not occur in mutants of the C-terminal processing protease (CtpA), leaving only the pD1 protein in the thylakoid membranes. Cyanobacterial PSII complexes containing only the pD1 protein have specific problems with the assembly of the lumenal side of the complex. There are some differences in PSII between cyanobacteria and plants, so what are the consequences of the absence of pD1 processing in plants?

Hypothesis: Based on the genome sequence of Arabidopsis, there are a few genes that could encode C-terminal processing proteases. The Arabidopsis CtpA (AtCtpA) protein (responsible for pD1 processing) will be found in the thylakoid lumen and mutants (atctpa) lacking this protein will accumulate the pD1 protein and not be able to grow photosynthetically. The extent to which PSII complexes are assembled in the atctpa mutants may be the same or different relative to the previously described cyanobacterial mutants.

Experiment: Identify the AtCtpA gene based on sequence signatures that predict lumenal targeting. Confirm that the AtCtpA candidate is present in the thylakoid lumen and characterize mutants for PSII assembly and function.

Results: Che and co-workers were able to identify AtCtpA and show it was found in the thylakoid lumen. There were two mutants atctpa-1 and atctpa-2 available for the gene they identified. The atctpa-1 mutant did not accumulate and AtCtpA protein, but the atctpa-2 mutant accumulated ~2% of the amount of AtCtpA found in wild-type plants. The atctpa-1 plants could only be grown in the presence of sucrose and had defects in chloroplast development. The atctpa-2 mutant could grow without sucrose, but was sensitive to high light. The PSII complexes in these mutants were analyzed by native gel electrophoresis, which separates the protein complexes in a mixture. Using this technique the authors found that the mutants had fewer PSII dimers and PSII-LHCII complexes, which are the forms of the fully assembled and functional complexes in plants. Using this technique, they did not see significant differences in the assembly state of the lumenal extrinsic proteins.

Conclusions: Processing of the pD1 protein is an essential step in PSII assembly in plants. In the absence of AtCtpA (as in the atctpa-1 mutant), plants cannot assemble PSII complexes sufficient to support photosynthetic growth. Only a small amount of AtCtpA is necessary to support some photosynthetic growth (as in the atctpa-2 mutant), but small amounts are insufficient to keep up with the pD1 processing demands plants face under high light conditions where replacement of the D1 protein must occur more frequently. The main difference in the PSII assembly state of pD1-containing complexes vs. normal was that the pD1-containing complexes did not assemble into the larger dimer forms with the LHCII antenna attached. This means that pD1 processing is required for attachment of the antenna, which is something we didn’t know before and didn’t really expect. However, this makes sense because it would be problematic for plant cells to attach a light-harvesting antenna to a partially assembled complex that would otherwise be sensitive to light.

Think Ahead: Future studies will focus on exactly how the presence of the pD1 protein prevents attachment of the LHCII light-harvesting antenna complex. The authors did not have any data on the presence of the Psb27 protein in their pD1-containing PSII complexes. They reported on just the PSII components known to be part of the assembled complex. I would hypothesize that this extrinsic protein accumulates on pD1-containing PSII in plants as well and should be detectable using a specific antibody.

This study emphasizes the need to use different model organisms to provide information about a seemingly universal process in photosynthetic organisms. My work in cyanobacteria gave certain kinds of information and the work in Arabidopsis present in this report gave slightly different information. The sources of these differences come from a couple of possible places. The first obvious reason is that there could be inherent differences in the way the process works in cyanobacteria vs. plants. I’ve outlined in the My Research section that the lumenal portions of the PSII enzyme have different protein compositions in cyanobacteria vs. plants. The C-terminal extension of the pD1 protein is also different in cyanobacteria (16 amino acid extension) vs. plants (9 amino acid extension). Plants and cyanobacteria also have very different light-harvesting antenna systems. Plants have the LHCII antenna complex, which is comprise of integral membrane proteins that bind chlorophyll pigments, while cyanobacteria have phycobilisome complexes, which are comprised of soluble, extrinsically-associated bilin pigment-containing proteins attached to the cytoplasmic face of the complex. Thus, any effects (however unexpected) on the antenna system relationship with PSII would be expected to be quite different between these two types of photosynthetic organisms.

The second reason is that the experimental techniques available in these two model organisms are different and come with their own technical limitations. Cyanobacteria allow for precise genetic manipulation and tagging of certain proteins, allowing me to selectively purify PSII complexes containing a histidine-tagged CP47 protein. These researchers didn’t have that option with Arabidopsis and had to rely on native gel electrophoresis to separate the PSII complexes from the other protein complexes in the thylakoids of their mutants. This different experimental method can provide different information with respect to protein composition of the complexes. For example, the extrinsic lumenal proteins largely migrated as free proteins in both wild type and their atctpa-1 mutant. This may be because the electrostatic interactions that hold these proteins to the complex do not survive in an electric field, resulting in their release during electrophoresis. This makes it difficult to say how strongly the PsbO, PsbP and PsbQ proteins are attached to their pD1-containing complexes. However, they are able to detect a small amount of PsbO associated with these complexes, which is more than I was able to find in the cyanobacterial equivalent.

Finally, I just have to say ‘Bravo!’ to the researchers for taking on this kind of biochemical work in Arabidopsis mutants. If you read the footnotes on my post, I complained about how difficult it was to grow my cyanobacterial mutant and isolate such small amounts of its PSII. I cannot communicate to you how much more difficult the biochemistry must have been in this study. Arabidopsis is a great model plant for genetic reasons, but even wild-type plants are on the small side for providing adequate material for biochemical work. The atctpa-1 mutants were much smaller than wild-type and had to be maintained on sucrose-containing medium. I shudder to think of how many of those small plants went into the one lane of the native gel figure in the paper (and consequently how many lab-hours that required).

Find out more information on some of the techniques used in this paper on the Experiments in Photosynthesis Research page.


* It’s open access!

Throwback Thursday: Roose and Pakrasi 2004

Yesterday I gave an overview of the research projects I’ve worked on. So in that spirit, I give you a ‘Throwback Thursday’ Journal Club article from my own publication record:

Roose, J.L. and Pakrasi, H.B. (2004) Evidence that D1 processing is required for manganese binding and extrinsic protein assembly into Photosystem II.  Journal of Biological Chemistry. 279: 45417 – 45422.

This was the first paper I published that was all my own work. We were interested in piecing together the necessary steps for Photosystem II assembly with respect to the lumenal side of the complex where the water-splitting reaction occurs. There are many specific steps in assembling the functional PSII complex, but we were interested in three general things: pD1 processing, manganese cluster binding and extrinsic protein association. What is the order of these steps? Are some of these steps absolutely contingent on others for progression along the PSII assembly pathway? Can some of these steps occur only partially?

The D1 protein is the subunit of PSII found at the center of the complex, which binds many of the necessary cofactors for electron transfer. It is first made in a precursor form (called pD1) that must be cleaved by a specific protease (called CtpA). In cyanobacteria, the C-terminal 16 amino acids of pD1 (on the lumenal side of the thylakoid membrane) are removed to yield the mature D1 protein.

The manganese cluster of PSII is the chemical machinery for the water-splitting reaction. It is comprised of four manganese atoms, one calcium ion and one chloride ion and largely held in place by residues of the D1 protein. Notably, the C-terminal residue of the mature D1 protein is responsible for coordinating at least one of the manganese atoms.

The C-terminus of the D1 protein and the manganese cluster are all located on the lumenal side of the complex. Also in this vicinity, three extrinsic proteins (PsbO, PsbU, PsbV) were known to bind to the complex.

At the time I was working on this project some of the lower-resolution PSII crystal structures were coming out and it was clear from those static pictures that the D1 C-terminus, the manganese cluster and the extrinsic proteins were all tangled together and their assembly (and disassembly for the repair cycle) would all be interrelated.

The spatial relationships among D1, the Mn cluster, PsbO, U, and V

The spatial relationships among D1, the Mn cluster, PsbO, U, and V

Here’s the breakdown of how we sorted out some of the details:

Observations: The ∆ctpA mutant, which lacks the D1 processing protease CtpA, contains only the pD1 form and this mutant does not accumulate functional PSII complexes. However, we didn’t know exactly what state of assembly they were in with respect to how much of the manganese cluster was assembled and how many of the extrinsic proteins were associated. It had been previously shown that ∆ctpA cells contained about half as much manganese per chlorophyll as control strains, suggesting at most only two of the four manganese atoms could be assembled into PSII complexes when the D1 protein has not been cleaved.

Hypothesis: In the absence of D1 processing (only pD1 is present), PSII complexes will not be able to fully assemble their manganese clusters. Maybe, only one or two of the manganese atoms will be bound to the complex. The extrinsic proteins may or may not be able to stably associate with complexes containing only the pD1 protein.

Experiment: Isolate pD1-containing PSII complexes and analyze them for manganese content and protein composition.* In the cyanobacterium Synechocystis sp. PCC 6803, the ∆ctpA mutation was combined with another mutation called HT3, in which the large PSII membrane protein CP47 contains a histidine-tag. This histidine-tag acts as a specific hook for purifying PSII complexes for subsequent analysis.

Results: While we obtained similar results for manganese content at the level of membranes in ∆ctpAHT3, no manganese atoms were detectable in at the level of PSII complexes in this mutant. We were able to observe some chlorophyll fluorescence changes upon addition of excess manganese, but never any water-splitting activity. Analysis of the protein components of the ∆ctpAHT3 PSII complexes revealed some key differences relative to the control as well. The PsbO, PsbU and PsbV proteins (all lumenal extrinsic proteins**) were absent from ∆ctpAHT3 PSII. These proteins were found to accumulate within the thylakoid lumen, but were not associated with pD1-containing PSII complexes using a partial membrane solubilization procedure. One protein, Psb27 (also a lumenal extrinsic protein)***, was found to be more abundant in the ∆ctpAHT3 PSII complexes relative to the control.

PSII Comparison

PSII Comparison

Conclusions: Our data showed that processing of the pD1 protein is required for manganese cluster assembly in PSII. When only pD1 is present, no manganese atoms are stably associated with PSII, although at least one manganese cluster can access its site to cause the changes in the chlorophyll fluorescence transients we observed. None of the extrinsic proteins (PsbO, PsbU and PsbV) can bind to PSII in the presence of pD1, but these proteins exist as stable soluble proteins within the lumenal space of the thylakoid membranes. The pD1-containing PSII complexes contained the core membrane protein components, but no manganese cluster and none of the extrinsic proteins associated with functional PSII complexes. Altogether this means that the pD1 processing event is an early event in PSII assembly that is absolutely required for manganese cluster assembly and stable association of the extrinsic proteins.

Order of PSII assembly events for the lumenal side of the complex

Order of PSII assembly events for the lumenal side of the complex

Think Ahead: The Psb27 protein is more abundant on ∆ctpAHT3 PSII, which represent partially-assembled PSII complexes, so maybe this protein only transiently associates with PSII assembly intermediates. It may be a factor required for assembly, but not part of the final complex. The structural data on PSII is great, but we need to remember that PSII exists as a dynamic population with complexes constantly cycling through various states of assembly. Future work must be geared toward developing better tools to analyze the diverse forms of this enzyme in order to piece together the details of the assembly and repair cycles.

Here is the link to the full article:


* It sounds too elegant and straightforward, right? That does NOT mean it was technically simple. I started this project as a bright-eyed first semester graduate student rotating in the Pakrasi lab. I tackled it with enthusiasm because no one told me how nearly impossible it would be to obtain the ∆ctpAHT3 PSII complexes I needed for analysis. It took me the better part of 3 years (yeah, yeah, rotations, classes and teaching assignments simultaneously) to optimize the procedures I needed. The ∆ctpAHT3 strain (any strain containing only pD1) was an absolute chore to work with- it grew slowly, was light sensitive, didn’t have much PSII. I was constantly tending to tens of liters and tens of liters-worth of cells. I rejoiced to the point of tears when I recovered even ten micrograms of its PSII (typical yields from much less starting volume can be hundreds of micrograms), and No, there is none of that sample left in a freezer anywhere.

** Notably, the PsbQ protein, another lumenal extrinsic protein which had only recently been identified in cyanobacterial PSII preparations, was also missing from the ∆ctpAHT3 PSII complexes. This protein would be the focus of some of my later investigations as well.

*** No one cared much about the small (11 kDa) Psb27 protein before this result. It had been known as a protein component of PSII, but no one knew what it did. It didn’t even get a letter for its name; PSII researchers had just run out of letters in the alphabet and decided to name that one ‘27’. After seeing that gel result, I had hope that Psb27 would be doing something interesting in PSII assembly. This was a hypothesis that would plague me for the rest of my thesis work. As we Southerners like to say, “It stuck in my craw.” Stay tuned for the exciting conclusion (or just read ahead from the links on the My Research page).