Session 2: Microbial Diversity and Evolution
Transcript of Part 2: Interpreting Molecular Fossils of Oxygenic Photosynthesis
00:00:00.00 Hi, and welcome back to iBioSeminars. My name is Dianne Newman, 00:00:04.09 and I am a professor at the California Institute of Technology 00:00:06.29 in the divisions of Biology and Geological and Planetary Sciences. 00:00:10.21 I am also an investigator of the Howard Hughes Medical Institute. 00:00:13.25 This is part three of my three part series in microbial diversity and evolution. 00:00:19.03 And what I would like to give you an example of today 00:00:21.19 is ways that we can through laboratory experiments 00:00:24.27 today gain insight into metabolisms that were evolved in the remote past. 00:00:31.02 And this is a very difficult task and we are just at the beginning of our ability to do it, but I hope through this story 00:00:37.25 I am going to illustrate just how interesting this field of research is and where the important open questions still lie. 00:00:44.08 So the specific example that I am going to tell you about is how we have gained some insight into 00:00:50.12 an important class of biomarkers that are used to record the evolution of oxygenic photosynthesis, 00:00:57.14 one of the most important metabolisms ever to have been invented on the Earth. 00:01:01.06 And today we see examples of oxygenic photosynthesis at work on the globe, 00:01:07.03 as you can see in this beautiful image that is taken from satellite data averaged over three years. 00:01:12.25 And what it is showing you are chlorophyll profiles in the marine oceans. 00:01:17.15 Here you see blooms of phytoplankton that occur in not only the Atlantic, but also 00:01:24.23 here in the Eastern Equatorial Pacific off of South America, and extending far out into the Pacific Ocean. 00:01:35.14 Now these gyres occur when there are blooms of phytoplankton which today 00:01:40.27 are responsible for emitting into the atmosphere most of the oxygen that we breathe. 00:01:47.16 However, these organisms, these marine phytoplankton, are not the ancestors of photosynthesis. 00:01:54.15 They are the ones that are doing it en masse in the world's oceans today, 00:01:57.27 but the process of photosynthesis evolved many billions of years ago on the Earth. 00:02:03.19 And its story is one that we still don't fully understand. 00:02:07.25 One of the most basic questions we don't understand is when did it occur. 00:02:11.09 How did organisms figure out how to use water as a substrate for photosynthesis? 00:02:15.24 One of the things we do know is that this ability to utilize water 00:02:21.27 was evolved through a complex combination of photosystems 00:02:26.16 in different types of primitive photosynthetic organisms, 00:02:30.10 including anoxygenic bacteria like the type shown in these bottles that are called purple bacteria 00:02:36.00 by virtue of the beautiful purple pigments they contain within their membranes, that you can sometimes see 00:02:41.29 when you grow them on certain substrates. 00:02:44.01 And I spoke about these organisms in my introductory talk. 00:02:47.17 And as I said then, at some point the molecular machinery 00:02:51.13 in these organisms was combined with the molecular machinery 00:02:55.12 in other organisms, also of this anoxygenic general class 00:03:00.08 but nevertheless quite distinct organisms. 00:03:02.14 How that happened, we don't know, but an issue right now is that when 00:03:09.00 this did eventually happen, and these beautiful molecular machines came together 00:03:13.24 so as to enable these organisms to utilize water as the energy together of course the real energy source is the sun, 00:03:22.24 but water is the substrate that is then activated by sunlight 00:03:27.06 to a form that is biologically utilizable in the electron transport chain. 00:03:31.19 This is when oxygenic photosynthesis was invented, 00:03:36.02 and the perpetrators of this invention were the cyanobacteria, 00:03:40.07 shown here in this slide of beautiful bubbling tubes 00:03:43.13 where CO2 is being fed into these cultures and these beautiful green organisms 00:03:48.10 with their green chlorophyll pigments are converting that CO2 to 00:03:53.21 biomass coupled to the oxidation of water to molecular oxygen. 00:03:58.16 Now the chloroplast itself as I explained in the first lecture is what evolved into the plastids that today we find 00:04:07.07 in diatoms that you see an example here. These are marine phytoplankton important in the oceans. 00:04:13.11 And of course, terrestrial plants that we are all very well familiar with 00:04:17.22 due to their ability to take water and generate oxygen. 00:04:20.10 But when this group of organisms first arose is a subject of real debate. 00:04:27.29 And it had been thought that there were signs in old rocks, 00:04:33.08 either these banded iron formations, or other geochemical signatures 00:04:36.19 that indicated oxygen likely was rising up in the oceans, being produced by these 00:04:44.26 ancient cyanobacteria, reacting with iron and precipitating it from solution, 00:04:49.11 to form these deposits, and then eventually getting into the atmosphere so that 00:04:53.17 other types of geochemical signatures found in rocks such as mass independent fractionations 00:04:59.17 of sulfur isotopes that we find in different types of sulfur bearing minerals 00:05:05.13 giving a signature that we think is very diagnostic of having oxygen present in the atmosphere. 00:05:14.08 That all of this was occurring somewhere in the middle of this time span, and certainly by around 2.4 billion years ago. 00:05:22.12 But when the ability of organisms to first take water arose, before it had this profound global effect, 00:05:30.09 is something that has not been very well constrained. 00:05:33.04 So a dominant approach that has been used to try to date the rise of cyanobacteria in the rock record 00:05:39.05 has revolved around using these organic molecules shown here that are called hopanoids. 00:05:44.04 And these are very complex organic structures that resemble sterols in eukaryotic cells. 00:05:49.19 Now the story I want tell today is a reexamination of these structures as biomarkers 00:05:56.16 to ask whether or not they are appropriate to assigning the rise of this very crucial metabolism that transformed the Earth. 00:06:05.10 So let's talk a minute about what it means to be a biomarker. 00:06:08.24 This is essentially acting as a molecular fossil. 00:06:12.13 That is a compound that is diagnostic of an ancient cell, 00:06:16.24 and a particular cell type if it is to be useful to identify organisms or their metabolisms. 00:06:23.22 So the idea is that a long time ago 00:06:25.24 cells hanging around in the environment, be they in the water column or in sediments 00:06:30.13 eventually through their sedimentation, and then through various diagenetic processes 00:06:37.16 where these sediments were transformed and over time compacted and converted 00:06:42.08 into solid structures that we see today uplifted in mountains and in various areas around the world 00:06:49.27 contain yet remnants of the original molecules from the original cells that once upon a time inhabited these environments. 00:06:58.25 Now these structures are not identical, but they are very, very similar, and so 00:07:04.22 what we are looking at today is what we call the biomarker 00:07:07.10 and this is the molecular fossil of what we infer to have been the original compound. 00:07:12.20 So here you can see something that we call a sedimentary hopane that is the diagenetic... 00:07:18.12 And diagenesis is just simply the process of transforming a sediment over time into a rock. 00:07:24.25 This compound is the diagenetic consequence of transforming the parent compound, 00:07:33.25 which we call the hopanepolyol, which has this core pentacyclic ring 00:07:38.09 here with a tail that has a variety of moieties on various carbon molecules, 00:07:47.06 and this is one where we are just showing hydroxyl compounds at these last positions, but they come in a variety of forms. 00:07:53.20 The bottom line is that some of these chemical decorations are lost, 00:07:57.19 but the backbone carbon skeleton remains. 00:07:59.23 And because it is so structurally similar to the original compound 00:08:03.24 it is very reasonable to infer that we are looking at a fossil of this molecule. 00:08:08.16 Now here is just another way of showing that. This is essentially a cartoon of diagenesis. 00:08:15.25 Where through the water column these cells ultimately sediment down, get incorporated into sediments, 00:08:22.25 and these sediments through geological processes 00:08:25.19 over billions and millions of years of Earth history wind up transforming, 00:08:30.06 and these molecules within them are converted into their most basic form. 00:08:35.27 Now why do people pay attention to these particular molecules? 00:08:40.01 These 2-methylated bacterial hopanepolyols. 00:08:43.12 The reason is because of an important study that came out about a decade ago 00:08:47.20 where it was claimed that 2-methylhopanoids were biomarkers for cyanobacterial 00:08:53.00 oxygenic photosynthesis. And the rationale for this was that at the time this work was done 00:08:58.21 all of the surveys of microbial cells for this particular molecule, 00:09:04.00 and when I say particular I mean the pentacyclic part 00:09:09.15 of the molecule and a methyl group at the second position of the first ring. 00:09:15.21 The other aspects of the molecule are not as crucial to the diagnosis, 00:09:20.27 but this initial part, and particularly the methylation at this second carbon atom 00:09:26.17 really was crucial to the whole story that I am going to tell. 00:09:30.13 So back when this initial study was published the investigators 00:09:34.19 reasonably were able to conclude that these molecules might be biomarkers for cyanobacteria because of the 00:09:41.29 fact that they were unable to detect their production in organisms other than cyanobacteria. 00:09:47.12 This wasn't entirely true because there were a few other strains that even in this initial study were reported 00:09:52.24 to produce these molecules, yet it was argued that in their sedimentary context 00:09:58.03 the most likely progenitors of these molecuels would have been cyanobacteria 00:10:03.02 because they were shallow marine enviroments in which the fossilized compounds were found 00:10:07.27 that were consistent with the growth of a phototrophic organism in this type of a habitat. 00:10:13.19 And also, the argument that the concentration of these molecules in a type that was believed to be 00:10:20.06 the most preservable was by far the highest in modern day cyanobacteria 00:10:24.25 lending credence to the notion that these types of molecules, when seen in the rock record 00:10:29.06 very well may be reflecting an ancient cyanobacterial community. 00:10:33.15 However, whether or not this was an absolutely valid assumption was debatable for a variety of reasons. 00:10:42.13 One of these reasons is that not all cyanobacteria can make these compounds, 00:10:47.09 so they are certainly not essential for the ability of oxygenic photosynthesis. 00:10:52.06 And so the argument that just because cyanobacteria make them means that they have 00:10:56.13 anything to do with oxygenic photosynthesis 00:10:57.25 is a whole other story in and of itself and is something that really needed to be looked into more for validation. 00:11:06.24 The second reason that this needed to be looked into more is that cyanobacteria 00:11:10.19 are capable of other types of metabolisms beyond oxygenic photosynthesis. 00:11:16.22 Some of them are called facultative oxygenic photosynthesis organisms 00:11:21.14 in that they are capable of using other substrates beyond water 00:11:24.21 in photosynthetic processes such as sulfide, for example. 00:11:28.11 And in addition, they are capable of fermentation. 00:11:30.10 And so just because cyanobacteria make them, doesn't a priori mean that they have anything to do with photosynthesis. 00:11:36.03 Although they might. At this point in the story, nobody knew, and so it merited looking into. 00:11:40.20 And so we got interested in this, and we thought that we would begin our studies 00:11:44.11 by coming up with just some rational criteria for what would constitute a robust biomarker. 00:11:48.29 The first of course is one I am not going to touch on because this really belongs to the province of geologists. 00:11:54.13 And that is that the biomarker must be indigenous to the rocks that it is meant to represent. 00:11:59.26 There's actually great debate about this point, but let's just assume for the sake of argument that when we find these fossils 00:12:06.11 in ancient rocks that they really are as old as we think they are. 00:12:10.15 And we will let the geologists work out indeed whether that's true, 00:12:13.13 but I think that it is fair enough to say that many of these samples 00:12:17.06 now there isn't much of a question. These molecules really are for instance 2.4-2.7 billion years old. 00:12:26.17 So let's just grant that. 00:12:27.14 The second part which now becomes of relevance for biologists is that they must have a unique distribution 00:12:34.05 amongst modern organisms, or they must have a clear evolutionary history 00:12:38.18 where it is obvious from whence they originated. 00:12:41.21 What were the first type of organisms that were able to produce them and/or the third point, 00:12:48.16 which I think is actually the most crucial point, 00:12:51.10 that they have a specific and conserved biological function that's related to the process of interest. 00:12:57.16 And in our case for today's lecture, we are talking about oxygenic photosynthesis. 00:13:01.22 So let's begin now with this second point. 00:13:05.28 Do these molecules, these 2-methylhopanoids have a unique distribution? 00:13:09.19 So to begin to look into this we decided to go to the environment 00:13:14.08 and collect some diverse types of phototrophic organisms to work with. 00:13:17.25 And we began by working with various cyanobacteria 00:13:21.04 that we surveyed for their ability to produce these 2-methyl compounds. 00:13:25.06 As I said, not all of them do, so we picked ones that were abundant producers as our reference strains, 00:13:31.05 and then because we had also been doing other studies in my lab 00:13:34.16 looking at the ability of anoxygenic phototrophs to catalyze the oxidation of iron, 00:13:42.08 under anoxygenic conditions, we decided just for fun we would begin to test 00:13:48.06 whether or not they could produce these compounds as well, 00:13:50.12 thinking that this would be a negative control, because previous reports had shown that related strains were, 00:13:57.17 under the conditions these previous investigators used, not able to generate these compounds. 00:14:03.18 And so here is where we had one of the surprises that occur once in a while in science 00:14:08.19 that are very exciting and at first cause you great distress 00:14:11.12 because you are not sure whether or not it's an artifact. 00:14:14.14 And so this happened when my student Sky Rashby made the surprising discovery 00:14:18.11 that one of our "negative control strains", one that we called Rhodopseudomonas palustris 00:14:23.03 was capable of making these 2-methyl hopanepolyols, that's what this acronym here means, 00:14:28.17 2-MeBHP stands for 2-methylated bacterial hopanepolyol. 00:14:33.06 In concentrations as great as its cyanobacterial counterparts. 00:14:38.02 Now you can be assured that we did all of our proper controls to make sure we weren't contaminating the cultures, 00:14:43.25 and this was a very reproducible, very robust result. 00:14:48.03 And what was really quite significant is that this molecule here, 00:14:52.15 and here you can see liquid chromatographic spectrum, and now a mass spectrum 00:14:58.14 of these peaks, was identical between our cyanobacteria positive controls 00:15:04.07 and what we thought was going to be our negative control. 00:15:06.20 What was really crucial here is that the production of these molecules by our anoxygenic phototroph 00:15:13.24 was conditional upon the manner in which we grew them. 00:15:17.12 And I think this is a very important point that is worth emphasizing. 00:15:22.21 And that is, oftentimes in these interdisciplinary fields 00:15:26.07 they begin with a group of investigators who are excellent at a particular type of measurement 00:15:31.12 but not necessarily excellent at all, because who is? You know we are all specialists. We are all good at doing what we do. 00:15:39.20 And so the pioneering scientists who began this work were outstanding organic geochemists, 00:15:45.06 but were not people who routinely grew microorganisms in their own laboratories 00:15:49.00 so they were dependent upon getting strains from their colleagues 00:15:52.27 who would send them microbial samples grown up in whatever random growth condition 00:15:57.04 they happened to be doing at the time that they shipped the strain. 00:16:00.00 However, today, when we grow them in our laboratory, being microbiologists, 00:16:04.17 we are aware of the metabolic versatility of these organisms and are able to grow them under a variety of conditions. 00:16:10.04 And because of this we were able to find that it was under certain conditions 00:16:15.05 that these organisms made the molecule, and had we only looked at them in a more narrow window, 00:16:21.09 we would have missed their production entirely. 00:16:23.09 So there is an important lesson here, and this nice cartoon 00:16:26.15 of this filamentous cyanobacterium that is supposed to represent the cyanobacterium Nostoc punctiforme 00:16:32.17 that I'll show you later in the talk 00:16:34.06 is saying, "Do you make 2-methylhopanoids?" to our favorite model anoxygenic phototroph, 00:16:39.19 Rhodopseudomonas palustris, and she is answering, "Yes, I do." 00:16:42.10 This very nice cartoon was made by Sky, who is a very talented artist. 00:16:46.16 Now an actual image of this purple bacterium is shown here 00:16:51.25 in a transmission electron micrograph, taken by my postdoc Ryan Hunter, 00:16:55.18 that he overlaid upon sediments from the Sippewissett Salt Marsh in Woods Hole, Massachusetts. 00:17:01.21 And the reason we did this is because as I've been saying, 00:17:05.00 the geological record ultimately is emanating from sediments such as these, 00:17:10.03 where today we find purple bacteria and other types of phototrophs growing. 00:17:13.23 And one can infer that these types of environments also were good homes 00:17:19.11 for these organisms back billions of years ago. 00:17:22.05 And so what we are looking at when we are looking at the fossils 00:17:25.11 are ancient remnants of these types of habitats. 00:17:28.06 So what we thought we would begin to do would be to study Rhodopseudomonas palustris 00:17:33.01 to ask whether or not we could gain insight into the molecular biological basis for the production of these compounds, 00:17:40.21 and have this help us parse whether or not they were accurate biomarkers 00:17:44.09 for cyanobacteria or, and/or, oxygenic photosynthesis. 00:17:49.00 Now R. palustris is a terrific organism to use for many reasons. It grows rapidly. 00:17:55.02 It is metabolically versatile. It's genome is sequenced. 00:17:57.17 And it is amenable to genetic manipulation. And so it's a system primed for discovery 00:18:03.19 in terms of the biosynthesis and the cellular function of 2-methyl BHPs. 00:18:08.01 So one of the first things that we wanted to answer was whether or not 00:18:11.14 there was a specific enzyme responsible for methylating hopanoids 00:18:15.05 at the C2 position because given the fact that these molecules were produced very conditionally 00:18:22.14 upon specific quirks of how one were to grow them, what would be desirable 00:18:29.08 in terms of asking the question which organisms writ large are able to make these molecules 00:18:33.25 would be to bypass this growth requirement and simply be able to go to an organism's genome 00:18:39.00 to see if there were some gene that were diagnostic, that could predict the ability to make these molecules. 00:18:44.17 So Paula Welander, a postdoc in my laboratory, 00:18:47.07 sat down with the genome of Rhodopseudomonas palustris 00:18:50.01 and through some very clever bioinformatic work identified a whole part of the chromosome that encoded genes 00:18:57.03 that she suspected might have something to do with hopanoid biosynthesis. 00:19:00.13 And that's because in the middle of this cluster there is a gene annotated as shc that had previously been shown 00:19:07.19 to encode an enzyme responsible for the first step in hopanoid biosynthesis, 00:19:13.11 the cyclization of squalene into the first pentacyclic ring. 00:19:20.16 Now what we wanted was the enzyme responsible for a later part in the biosynthetic pathway 00:19:28.27 and that is the methylation at C2 because a wide variety of bacteria are capable of making hopanoids in general, 00:19:35.03 but what had been thought, as I told you, to be the diagnostic feature was this methyl group at the second position. 00:19:42.03 So if we could find a methylase that was specifically responsible for putting this methyl group on that position 00:19:49.02 we might be in a position to ask whether or not different organisms beyond cyanobacteria and this one freak 00:19:56.08 discovery we made, were capable of also making 2-methylhopanoids. 00:19:59.12 So to do that Paula looked at this genomic region, and she identified a gene over here 00:20:07.26 that seemed to have attributes that might be consistent with its product being 00:20:13.17 capable of catalyzing this methylation reaction. 00:20:16.21 And so to test that she made a clean in frame deletion of this gene, 00:20:22.10 and then she asked whether or not the mutant 00:20:24.19 strain that she called delta 4269 was capable of making methylated hopanoids in comparison to the wildtype. 00:20:32.22 And so here what you are looking at is a liquid chromatogram showing peaks of elution 00:20:38.03 of various types of hopanoid structures. 00:20:40.19 These over here are called...this one is diploptene, and this one 00:20:45.19 has a methyl group at the 2 position, so it is called 2-methyldiploptene 00:20:49.14 and this is a different type of hopanoid, with a slightly different R group on the end. 00:20:57.07 And again here, this peak that elutes first is the 2-methylated version of the molecule. 00:21:02.06 And as you can see when you compare it to the mutant that Paula made, 00:21:05.15 there is no peak here in the foreground showing that the 2-methylated versions 00:21:11.09 of these hopanoids are not being made in this mutant background. 00:21:14.04 Now when she complements and adds back to this mutant strain the wildtype 00:21:18.21 gene, then again she sees, indeed, just where she should, the peaks 00:21:23.29 that indicate the production of the methylated versions of those hopanoids. 00:21:27.22 So this was a very nice demonstration that indeed this methylase was responsible for 00:21:32.24 producing these methylated hopanoids. 00:21:36.02 Now in collaboration with another postdoc in the lab, Maureen Coleman, 00:21:40.27 Paula and Maureen went on to show that this methylase gene that they dubbed hpnP, 00:21:46.22 was much more broadly distributed than we had suspected. 00:21:50.28 And so what matters here is not the individual names of these organisms, 00:21:55.29 by any means, but simply the fact that what you are looking at is a complicated tree 00:22:02.04 that's drawn using the 16S ribosomal DNA sequence for a whole variety of organisms. 00:22:09.04 And what you are looking at now is in blue those organisms on this tree that are capable of producing hopanoids in general. 00:22:18.06 And there are many, as I said. 00:22:20.06 But the really key piece is that here, which is this outer ring 00:22:24.12 in which you can see in three different groups, in the cyanobacteria, in the acidobacteria, 00:22:32.17 and down here in the proteobacteria, evidence for the hpnP gene from the genome. 00:22:39.20 And this evidence is based on a very stringent cutoff where we are asking for a very high degree of similarity 00:22:47.05 in sequence identity between putative hpnP sequences 00:22:52.03 and our sequence of hpnP that we have shown is required for the methylation at the C2 position. 00:22:59.11 Now the other thing to note is that in every case that these sequences have been found in the genome 00:23:11.06 where the measurements of the actual ability to make the compound have been made, we have found 00:23:16.18 a 100 percent correspondence. And inversely, we have never found an instance 00:23:23.05 for an organism that is capable of making 2-methylated BHPs where they don't contain this gene. 00:23:29.03 So while of course it is possible that in the future another type of methylase will be discovered, 00:23:35.17 right now there is a 100% correspondence between position of this gene in the genome 00:23:40.26 and the ability to make the compound as measured directly. 00:23:44.28 So it appears as though it is a fairly robust indicator. 00:23:48.03 All right so, there are a few things to note here, as I said before, not all cyanobacteria make this. 00:23:54.06 So a priori, it is certainly not required for oxygenic photosynthesis. 00:23:57.00 The question is whether...and number two, more than cyanobacteria make it. It is not just a case of one random organism 00:24:05.02 being a flier. No. We find that there are many other types of bacteria here in this proteome, 00:24:12.02 alpha-proteobacteria family, that are capable of producing these compounds. 00:24:16.00 Some of these actually have even been shown in that very original paper to produce the methylated hopanoids. 00:24:22.04 It just had been claimed that they weren't producing them in abundance that was relevant for diagenesis. 00:24:27.03 And so that deserves a bit of re-examination. 00:24:29.27 But here is this whole other group, the acidobacteria, that hadn't previously been appreciated to make these molecules. 00:24:38.25 And I should say that this is just the tip of the iceberg, because these are only the genomes 00:24:42.08 of microbes that currently have been sequenced, which represent a minute fraction of the microbial diversity 00:24:48.01 in nature, and so now we have the tools where we can go out into the environment and ask more broadly, 00:24:54.01 where do we see these sequences? What type of organisms are likely containing them? 00:24:58.22 And we'll get a much better picture of the phylogenetic diversity of the organisms that make them. 00:25:03.00 However, to get back to the central point in terms of whether or not this is a robust biomarker, 00:25:09.26 even though today we see that these molecules are distributed amongst many organisms, 00:25:15.14 they still might be a valid biomarker of cyanobacteria 00:25:17.28 provided we can have some good evidence suggesting 00:25:23.05 that their root evolutionarily was within the cyanobacterial group. 00:25:28.11 So to get at that question of what is the root of this enzyme, 00:25:32.07 where was it first invented, Maureen Coleman decided to look at the phylogenetic history of the enzyme family. 00:25:39.08 So she made a tree where she was able to show, and now this is an unrooted tree, 00:25:44.03 the different branches of the organisms that we know can make this. 00:25:48.27 All right, so these are two different groups of proteobacteria. Here we have the cyanobacteria. 00:25:52.20 And here we have the acidobacteria. 00:25:55.07 And what we wanted to know was which of these groups was the first to invent this enzyme. 00:26:01.01 Now here is where it became complicated, and also where it became very important to do a rigorous 00:26:06.21 statistical analysis to align these sequences in various ways, use different algorithms, 00:26:12.18 to infer the most parsimonious evolutionary path to their divergence. 00:26:19.01 And when Maureen did this, quite literally by doing an in silico experiment, by varying a variety of parameters, 00:26:26.24 none of which was clear what was the absolute correct assumption to make, 00:26:32.00 but all of which would affect the topology of the trees that ultimately would result. 00:26:36.29 What she found was that the root, at least at present, is ambiguous, 00:26:41.19 that statistically there was as good support for these three different topologies, 00:26:47.14 that led to very different answers. 00:26:49.26 The first where the ancestral root of this methylase would be in the cyanobacteria. 00:26:54.01 The second where it was quite simply unresolved, 00:26:58.08 where there was no clear branching pattern, 00:27:00.27 of which subset would have been the earliest one to make this enzyme, 00:27:06.11 or the opposite result where the ancestor would be inferred to be an alpha-proteobacterium. 00:27:12.18 We hope that in the future as more sequences are gained we will be able to, 00:27:17.21 with greater confidence, resolve between these three options, but what we can confidently say 00:27:21.28 now is that we cannot with confidence interpret that cyanobacteria were even the evolutionary inventors of this. 00:27:30.11 And so the case remains open whether or not in the future they will be proven to be so, 00:27:35.25 but at present it is invalid to argue that there is an absolute certainty 00:27:41.12 that these molecules are biomarkers even for the cyanobacteria themselves, 00:27:46.03 saying nothing about oxygenic photosynthesis. 00:27:48.05 So let's cut to the chase and talk about oxygenic photosynthesis 00:27:51.02 because that is really what motivated this all along. 00:27:54.18 Could we find a biomarker that would give us insight into the process of oxygenic photosynthesis? 00:27:59.15 And of course, just because an organism makes a molecule, doesn't necessarily 00:28:02.28 mean that it has anything to do with a particular function. 00:28:06.04 And so what we wanted to ask was whether or not there was a connection between 00:28:10.22 the ability to produce this molecule and photosynthesis directly. 00:28:15.05 So, as I told you, hopanoids resemble sterols. 00:28:20.16 And if you take a look at the biosynthetic pathway 00:28:25.00 for cholesterol over here and compare it to bacteria hopanoids over here, 00:28:32.13 you see that they go through very similar structures. 00:28:36.22 Of course the rings are different between them, but they are reminiscent. 00:28:42.15 And so it would be reasonable when thinking about the universal functions for these molecules 00:28:48.01 that we would turn to sterols for some inspiration. 00:28:50.26 And this is a very fascinating field in cell biology in eukaryotes, 00:28:56.06 and I am sure there are some interesting commentaries on this in other iBioSeminar series that you can refer to. 00:29:01.13 Very briefly what I will just say now is that it has been demonstrated 00:29:05.10 that sterols in eukaryotic membranes have roles in both membrane fluidity and integrity. 00:29:10.12 They can play roles in membrane curvature. They are also important 00:29:16.05 in creating what are called lipid rafts, that involve the recruitment 00:29:20.13 of specific lipids and proteins to particular places within the membrane. 00:29:24.06 All of these are very active areas of research in eukaryotic cell biology, 00:29:28.01 and one can reasonably ask whether or not in bacterial and in archeal, 00:29:33.13 or even microbial- eukaryo systems these hopanoid molecules 00:29:39.02 that resemble sterols might play similar functions. 00:29:41.26 And one can go further to ask whether or not in any way these functions are related specifically to oxygenic photosynthesis. 00:29:50.01 So, one thing that we can do to begin to address this is simply to figure out in different types of microbial cells 00:29:58.16 where hopanoids reside. And the point here is that it is not obvious, 00:30:05.05 because microbial membranes are actually very complex, 00:30:08.07 particularly in phototrophs where we have these 00:30:10.17 beautiful inner cytoplasmic membrane lamellar invaginations. 00:30:15.11 We can have other types of invaginations 00:30:17.06 that are either vesicular or tubular, there are many different types of 00:30:20.20 internal membrane structures that can form, 00:30:22.22 and as I said this is a wonderful opportunity for future research in microbial cell biology. 00:30:26.28 But getting back to the focus here, the question is, and in this cell, 00:30:31.29 what I am showing you is a thin section of my favorite organism Rhodopseudomonas palustris, 00:30:36.06 where do the hopanoids reside? Are they in the outer membrane or are they in the cytoplasmic membrane? 00:30:40.27 Are they in this membrane called the inner cytoplasmic membrane, 00:30:46.14 which is where the photosynthetic machinery is housed? 00:30:48.27 And the same question can be asked of cyanobacteria, where you see these inner cytoplasmic membranes, 00:30:55.04 which also host the photosynthetic machinery in cyanobacteria. 00:30:59.20 And this is an organism that isn't even photosynthetic, 00:31:02.17 but does indeed produce 2-methylated hopanoids 00:31:06.13 and as you can see, it has complex membrane systems as well. 00:31:09.10 So where are the hopanoids, in which of these membrane systems? 00:31:12.03 And just for the remainder of this time, I am only going to focus on showing you work 00:31:16.07 we've done using a representative cyanobacterium called Nostoc punctiforme. 00:31:20.28 So Nostoc punctiforme is a terrific microbial system for studying cellular differentiation. 00:31:26.24 And the reason for this is that there are three states it can exist in, as the vegetative growing cells, 00:31:32.16 some of which can differentiate to form specialize structures called heterocysts, 00:31:36.27 and this is where nitrogen fixation occurs in this organism. 00:31:39.09 But these organisms are also capable of creating a spore like state, 00:31:45.10 that is known as an akinete, and this is a very stress resistant cell type 00:31:49.26 that forms under stressful conditions that is not metabolically active. It is metabolically quiescent. 00:31:55.13 So Dave Doughty, a postdoc in my laboratory decided to take Nostoc 00:32:01.11 as a beginning entry way into understanding 00:32:04.25 where hopanoids would partition in cyanobacteria. 00:32:07.28 And what he decided to focus on were the different membranes that I just described: 00:32:12.14 the outer membrane, the inner membrane, and the inner cytoplasmic membrane 00:32:17.15 in the vegetative, heterocyst, and the akinete cell types. 00:32:21.05 And what you can see when you do these thin sections of the organisms 00:32:24.16 and look at them under electron microscopy, 00:32:26.24 is that the cell types change quite dramatically. 00:32:30.02 For example, there are very few, what he is labeling here as thylakoids, 00:32:34.14 this is just borrowing a term from eukaryotic cell biology 00:32:37.20 but what this is also really talking about here is the inner cytoplasmic membrane 00:32:42.15 that these cyanobacteria have that houses the photosynthetic apparati. 00:32:47.28 They have very few when they are in the akinete state, which makes sense, because these are resting cell types. 00:32:53.04 The heterocysts have a much thickened cell envelope, and these are the cells in which nitrogen fixation occurs. 00:33:01.15 And this is just a normal view of a vegetative cell, 00:33:04.12 where you have what Dave calls a thylakoid membrane, and what I have been calling the inner cytoplasmic membrane. 00:33:10.04 As well as an outer membrane and an inner membrane in the overall cell envelope. 00:33:15.27 Now what we wanted to ask was depending upon the way in which the cell was grown, 00:33:22.10 would we find that hopanoid content and also hopanoid localization to these three membranes change? 00:33:28.13 And the answer was dramatically yes. 00:33:31.04 And so what Dave showed was when he was able to grow them as vegetative cells, 00:33:36.14 as heterocysts. Isolate the heterocysts, and then grow them so that they went into the akinete cell type 00:33:43.24 and then fractionate each of those three different cell types with respect to the cytoplasmic membrane, 00:33:48.06 the thylakoid membrane, and outer membrane, 00:33:51.20 that for these cell types the concentration of hopanoids, micrograms of hopanoids per milligram of total lipid, 00:33:58.20 varied quite dramatically. 00:34:01.01 And what I'll have you note here is that for the cytoplasmic membrane and the thylakoid membrane 00:34:05.08 the scale is much smaller than for the outer membrane. 00:34:09.24 Here this is only going up to ten, and here it goes up to sixty. 00:34:14.11 And so while indeed we do find some of these hopanoids in akinetes 00:34:20.08 and in vegetative cells being generated in the thylakoid membrane 00:34:25.15 by far the greatest concentrations are in the outer membrane of the akinetes, which are the resting cell types. 00:34:33.15 So as I told you, akinetes can be very resistant. They are survival structures. 00:34:40.08 But what I want to add to that description is the fact that 00:34:43.06 they can develop when cells are grown completely in the absence of light. 00:34:48.14 And so while in this experiment Dave did he'd originally grown them up photosynthetically, 00:34:53.02 and then subjected them to starvation conditions so that they went into the akinete type. 00:34:57.04 They do not need to ever go through the photosynthetic phase in order to become akinetes. 00:35:02.20 Therefore, since the majority of the methylated hopanoids in this organism 00:35:08.04 are localized to the outer membrane of the akinetes, one can infer that 00:35:13.19 the most likely source of 2-methylated hopanoid deposition into sediments, at least from this organism, 00:35:19.17 would be from cells where photosynthesis is not occurring. 00:35:24.11 And because of this we can go back and interpret the rock record and ask the question, 00:35:32.06 when we find both these molecular fossils of 2-methylated hopanoids 00:35:36.17 as well as these morphological fossils that paleontologists who are very clever and able to look at 00:35:44.04 these images and actually tell you when it is an akinete versus some other type of cell 00:35:49.03 and date them to very old times in the rock record. 00:35:54.00 Here we are talking about billions of years, so 1.5, 1.6, and 2.1 billion year old samples 00:36:00.12 where the akinete cell types as diagnosed by these paleontologists 00:36:05.13 are known to have been produced, we can ask whether or not 00:36:11.04 the 2-methylated hopanoid record that we find at similar ages 00:36:16.11 might have been produced through the shedding of the akinete envelope 00:36:19.12 into the sediments and may be reflecting a stress response 00:36:22.23 as opposed to anything to do with oxygenic photosynthesis per se. 00:36:26.27 That said, there is still more work to be done in examining whether there is some type of indirect connection 00:36:34.26 with oxygenic photosynthesis. There may be. 00:36:37.13 And we have a lot more to discover, so this is a very exciting field to be in 00:36:40.22 and will remain a very fruitful area of research for many years. 00:36:44.15 So to summarize what I hope I have been able to show you through an example 00:36:49.00 is how we can re-examine molecules that have been used as biomarkers for ancient metabolisms, 00:36:55.16 and ascertain whether or not these are valid biomarkers. 00:36:58.24 In this particular case, we have shown that these compounds 00:37:01.23 2-methylated bacterial hopanepolyols, or hopanes, as they are called in the rock record, 00:37:07.00 are likely to have been generated by many diverse bacteria. 00:37:11.02 And we can not yet say which type of organism invented the ability to make them. 00:37:15.19 So the jury is still out whether they can be used as biomarkers of cyanobacteria. 00:37:19.01 That said, even if cyanobacteria did invent them first, 00:37:24.29 it is not at all clear that they have anything to do with oxygenic photosynthesis 00:37:28.15 and in fact, all of the available evidence really would most reasonably lead you to conclude otherwise. 00:37:33.20 And that is because cellular localization studies show that they are mainly in membranes 00:37:40.28 that do not have anything to do with active cells, 00:37:43.27 nor do they necessarily have to reflect going through a photosynthetic process. 00:37:48.28 And in experiments which I don't have time to show you here, 00:37:51.21 we've been able in Rhodopseudomonas palustris to actually knockout these methylated compounds, 00:37:57.13 and find that even for the process of anoxygenic photosynthesis 00:38:01.04 they are not required and so, it doesn't appear that they play a direct role in metabolism. 00:38:07.08 But what it does seem to be important is that they are involved in stress response, 00:38:11.06 and likely this is manifesting through effects in membrane permeability and membrane fluidity. 00:38:17.18 Potentially there are other roles, and we await future discoveries to eliminate the full menu of options 00:38:25.06 that these molecules are serving in terms of their biological function in modern cells. 00:38:29.26 All right, so to summarize what I hope I have been able to show you from this story 00:38:33.23 is that we can use studies of modern bacteria to gain insight into the genesis of ancient metabolisms. 00:38:40.06 And while there is still a lot more work to do, we have a very exciting 00:38:44.25 future ahead of us where we have these remarkably geostable 00:38:48.14 compounds that hang around in rocks that are incredibly old, going back over two billion years, 00:38:53.21 and we need to be able to interpret them rigorously. And so in the case of these 2-methylated bacterial hopanes, 00:39:01.09 it appears that they really are no longer the best indicators of oxygenic photosynthesis, 00:39:06.10 however, they might be good indicators of microbial stress responses and other things. 00:39:12.23 And so in the future we hope to understand what their biological functions 00:39:17.15 are in both cyanobacteria and the other organisms that make them, 00:39:22.07 and be able to utilize these molecules as biosignatures of wonderful inventions in the history of cell biology. 00:39:28.23 So with this, I would like to acknowledge the people in my lab 00:39:31.29 that I have been referring to who have done this work. 00:39:34.08 It's been a collaboration over many years, and it began at Caltech with the student Sky Rashby, 00:39:40.05 co-advised by my colleague Alex Sessions, and in collaboration with my colleague Roger Summons at MIT. 00:39:44.18 And then continued by two postdocs, primarily Paula Welander and Dave Doughty. 00:39:52.04 Paula doing the work in Rhodopseudomonas palustris, and Dave with Nostoc, 00:39:56.08 and now continuing with Rhodopseudomonas, helped out by phylogenetic work with Maureen Coleman 00:40:02.02 and microscopy by Ryan Hunter. 00:40:03.26 And the Howard Hughes Medical Institute, NASA, and the NSF have contributed to supporting our research.