The Smallest Self-sustaining Ecosystem: Part II

IMG_7281

The research process, and a preliminary answer

Preface: “AUGHH, SO BORING!” is probably the first response I expect from this post. I can hear the sound of dozens of laptop screens shutting or browsers navigating away to avoid reading this post. It delves into the daunting world of academic and scientific literature, where the English language is changed into some strange, twisted version of itself, lengthened in sentence and peppered with weird acronyms that are meaningless to all but the most entrenched in the field. But reading research literature is absolutely essential to how we arrive at answers, especially in academic fields like the science. And it’s getting easier to do; academic researchers have realized the importance of making their work accessible to everyone, and have pushed to make the language they use easier to understand.

So if you want to get your feet wet in the academic literature, you want a glimpse into what a lot of grad school is about, or you’re just curious about what researchers in the sciences do for MUCH of their time, this post will tell you.

In the last post I defined our question, which was “What is the smallest self-sustaining ecosystem that we could send to space?”. Now I’ll cover where and how to look for an answer to that question in the scientific literature and give a preliminary answer to our question. By the end of this post, you’ll know how to use scientific literature to answer questions of you own!

Diving into scientific literature

Scientific literature can be daunting because it looks unlike anything else we read, but it is the fount of all scientific information. It is the original source of all the information and data, which is collected by news articles and websites to deliver to you, the reader. As with anything involving interpretation, news articles and websites can leave out important information or get things outright wrong when they report on information from the scientific literature, but the only way for you to know is to go read the original scientific article or research report yourself. So if you want to truly know something, read the scientific literature.

We’ll divide our dive into the scientific literature into three sections: searching the scientific literature for research articles, sorting out which research articles in the literature are useful, and analyzing research articles to answer your question. Doing these three steps successfully is difficult and will be slow, especially if you’re new to the scientific literature. Give yourself time when searching the scientific literature and write your question at the top of your page or next to your computer to keep yourself on track.

Step 1: Searching for research articles that might hold answers

In this step I collect several research articles that might have answers to my question to search through in depth later. The best ways to search the scientific literature are Google Scholar and NCBI’s PubMed. Google Scholar searches research articles and primary literature instead of webpages, but works in a similar way to the standard Google search engine so we won’t cover it. NCBI’s PubMed is also a search engine, but it searches all of the primary literature indexed vetted by the National Institute of Health to create a customizable, highly-specific, extensive output. It is the go-to site when I need to find research articles.

Try PubMed by clicking here and typing in something you want to search.

PubmedSimpleMicrobiome
Search results on PubMed. Each numbered entry in the list is a research article that might hold answers to your question.

Like Google, you’ll see that a search will yield research articles. Clicking on an entry pulls up the PubMed page, which contains the research article information, including the title, author, journal it was published in (which, like magazines have varying degrees of prestige and veracity), and a super-useful summary of the paper called an Abstract that will help you determine if the paper is useful. I prefer to cast a wide net in this step, saving anything that might be useful or interesting to read to answer my question.

PubmedEntrySimpleMicrobiome
An example entry on a research article in PubMed. Your access to full text links will be the rectangles on the upper right. In this case, the full text is available for free.

If you read the Abstract and think this research article has answers, your next goal is to get the full research article, which is easiest through the “Full Text Links” in the upper right of the PubMed entry. This can get…uh…tricky. In the best cases, the article is free either from NCBI or the journal that published it and the box under “Full Text Links” will say “free full text” or “free final version”, which you can click to get to an HTML or PDF version of a research article. If a full version isn’t available for free (you’ll know because the link you clicked on will say ask for $$ to access an article), then there are a few ways you still might be able to get a research article for free. The first is to check the lab website, which can be done by Googling the name of the last author (here, Pace) and the article’s title–some professors will post PDFs of their research articles on their lab website. A second option is Kopernio, which searches a few websites to determine if the article is freely available. If you don’t find a PDF using these methods, you can also contact an author through ResearchGate or via email and ask for a copy of the article; most authors are happy to share. And the last option, one of dubious legality, is to pirate it from the Russian repository Sci-Hub. Whichever way you choose, save the research articles you can as PDFs into a folder to use in step two.

Step 2: Sorting out which research articles matter

Now you have a pile of PDFs in a folder, one (or some) of which may have the answer to your question. But most PDFs won’t and there are more papers in the world than hours you have to read them, so it’s time to start sorting.

FilesSimpleMicrobiome
My file organization system for research. The “not useful” papers usually get deleted, and I go through the “less useful” papers if I can’t find an answer in the papers here. “Round 2” comes from my second round of research (reapeating step 1 again) and I prefer to keep them separate so I know what I have finished reading and need to read next.

I like to sort into three categories: useful, less useful, and not useful. To figure out which papers are useful, I’ll open the file and skim the Abstract, figures, and figure captions — 90% of the time, this is enough for me to figure out if a paper isn’t useful. If I still can’t decide, I’ll read the paper’s Discussion section. If it doesn’t have my answer or any useful related info, it goes into the “not useful” folder. If it doesn’t clearly have my answer but might have some useful info,  it goes into the “less useful” folder. And if I think it’s useful, I leave it in the main folder.

If you happen to find the answer to your question while skimming a paper, that’s awesome! It does happen, especially if you’re looking for a specific fact or statistic. But for more complex questions, you’ll probably need another step.

Step 3: Finding your answer by analysis and synthesis

In this step, you’ll read the articles you’ve found in depth to piece together the information you need to answer your question. This step takes the longest – while the last two may have taken hours, reading through all the papers you think are relevant and synthesizing an answer can take days, or if the question is really big or you’re really new to this process, weeks.

Since this is the long haul, it’s best to start by getting organized. Pick a method of keeping notes, be it on paper or digitally, and stick with it. If you have more than five or six papers, you’ll likely want to keep track of your notes by paper. I use Mendeley to organize my papers, and OneNote to keep track of all my notes on them, including quotes and excepts from papers I’ve read.

PapersNotes
A sample from my OneNote research notes to answer this question.

Now, down to the actual reading. It will take time for you to read a research paper, especially if you’re new at it. Be patient with yourself and start with small steps, maybe reading only one paper a day. If you’re also new to the field of your question, then doubly so– you’ll be learning new words and vocabulary as well as how to read a paper. Look up definitions, write down acronyms on a sheet of paper next to you, and take notes somewhere on things relevant to your question. Try to rephrase the findings of a paper in your own words, which tests how well you understand it. Be patient and steadfast and the papers will grow easier to read.

The last part of this step is synthesis, using what you’ve learned from the research and your brain to make an answer. It’s difficult and will also take practice. Try to answer your question after each paper you read; write it down, read it, and see if there are still missing pieces in your answer. This isn’t a research paper, so your answer need not and should not be a whole essay unless that is what you need to answer your question. A sentence, three sentences, or a paragraph is better if that’s all it will take.

The true test of whether you’ve answered your question is to give the answer to someone else. Take your question and answer to a teacher or friend or family member. Tell them the question and explain the answer to see if you can help them understand. If you’re successful, congratulations! You just taught someone something they might not know. If you’re not or they disagree with you, try to be humble and ask for advice. If they disagree, ask where they get their answer and go look it up. If they ask questions, you can return to the research and dig for more answers. Like an adventurer, keep following the threading trail until you find the answer. It might be something no one has ever answered before.

A preliminary answer to our question, or one is the loneliest number

Phew! The above was an essay I hadn’t originally intended to write, but I started it and it seemed to matter so much in a day and age where there seem to be so many questions that need answers. Thank you if you read through all of that.

I promised a preliminary answer this week and an answer you shall receive!  An early answer to “What is the smallest self-sustaining ecosystem that we could send to space?” is….drumroll please….

ONE (Chivian et al., Science, 2008). Yes, a ecosystem of one was discovered deep in the Earth’s crust. It was found  in a mine shaft in South Africa at 2.8 km deep, or around 933 floors if you take a floor as about 3 meters tall. If you were taking a standard elevator down to this depth, you’d be on waiting in it for a little over two and a half hours.

SimplestMicrobiomePaperChivian2008
The Chivian et al. 2008 paper describing the discovery of D. audaxviator.

The lonely microbe’s name is Delsulforudis audaxviator (we’ll call it D. audaxviator), which as far as I can tell with Google means “the bold traveler of the sulfur lineage.” And it’s capable of living alone because it carries a complete toolkit of genes that help it do everything it needs to survive – helping it ‘eat’ carbon dioxide, carbon monoxide, and formate, ‘drink’ the nitrogen it needs from out of the air as nitrogen gas, and ‘breathe’ sulfate compounds as we breathe oxygen. While we breathe out carbon dioxide, it breathes out hydrogen sulfide gas, which smells of rotten eggs. It’s a one-microbe jack-of-all-trades, do-it-yourself word down there, and D. audaxviator is well-prepared.

So now that we have an answer, are we done? Uh, no. This is a bad answer to our question because it doesn’t meet the rules we set out! Look again at the description of D. audaxviator above. What do you notice? Well, it makes hydrogen sulfide gas like we make carbon dioxide. That’s a waste product. Where is that waste product going? If it just hung around, it would build up. And where are carbon dioxide, carbon monoxide, format, nitrogen gas, and sulfates coming from? If D. audaxviator uses them all up, does it starve? Does it suffocate? This systems fails to meet our criteria of self-sustaining.

If you read the source paper (linked here), you’ll find two things of note. One, many of the things D. audaxviator needs to survive are made from chemical reactions in the earth and somehow either make it to D. audaxviator or D. audaxviator finds them. This same method might be used to sweep away any waste products made, so D. audaxviator doesn’t’ have to worry about them. Unfortunately, we don’t have the luxury of shunting off waste like this while in space. We won’t have the comfort of earth.

The second thing to note in the paper is that while researchers found more than 99.9% of the genetic info from the mine sample comes from D. audaxviator, that is not 100%. There could be some few other microbes living down there, taking the waste products of D. audaxviator as food and turning them into something else. Or maybe they’re taking other compounds and turning them into the compounds D. audaxviator needs to survive. It’s hard to tell for now, and we may not get an answer anytime soon. The researchers couldn’t clearly identify what these other microbes were because there was so much D. audaxviator down there. It’s the majority, and finding the other microbes would be like playing a game of Where’s Waldo, except imagine your book is now more than 2 km deep in the earth and the only chance you get to look at it is through a grainy photograph that’s been fed into a shredder and then put back together again. It might be a while before we get an answer.

So we found an answer to the question, but it’s not a very good one. Let’s follow the trail of research using Step 3 to find a better answer.

Followthetrail

Engineering Life to Work as Expected

What I wrote below highlights some of the risks of engineering living things, the goal of a field called “synthetic biology” . Is anything unclear? Email me! Ask questions! Your advice will make these posts better.
20160618_115442.jpg
Microbes growing in lab
We’re at an amazing point in scientific history where we know enough about how living things work to begin tinkering with them. There have been thousands of studies in the last century looking at modifications of living things. At first, the tools we had were crude and the questions we asked were “what happens if make a bunch of mutations in a living thing? What kinds of properties does the living thing show? And what changed in the a living thing’s genes, in its DNA, that caused this change?” As our tools became more precise, so too did our questions. We began to ask what happened when we took out one gene, or added another. “What properties appear when we do this?” we asked. These two ages of research brought is much of what we know about how a genome, a collection of a living thing’s genes (stored in the DNA), work together to create life.
Now, we’re taking this knowledge and beginning to modify genomes with the goal of creating living things that exhibit specific traits, in a field dubbed “synthetic biology”. Want an apple that doesn’t brown when sliced? We’ve got you covered. How about a microbe that produces an extremely expensive anti-malaria drug, making it more accessible? Yep, we have that too. These advances are the products of the last three decades, where concerted effort and millions of dollars bore fruit to create life with an engineered function. But with the ability to do create living things with desired properties, a new problem arose: how do we ensure a living thing we create works the way we intended?
IMG_2510
Sea life grows in a plastic cup washed up during a storm; we may one day engineer life to break down plastic in the oceans
To understand this problem, we have to look at the biological purpose of life. Every living creature has a sole directive: to make more of itself, as many copies of itself as possible. From single-celled bacteria to the complex mass of tissues and organs we call animals, all living things strive to give rise to more of themselves. This directive manifests everywhere around us: the battle for self-preservation, consumption and conflict over resources, the care shown to offspring. It also leads to extreme scenarios, such as when a male spider accepts his death for a chance to mate with a female; in this case, the chance to create more life overrides even the powerful instinct of survival. At every level, in every living thing, this rule acts to reward those that succeed at making more of themselves, as these offspring give rise to more offspring. The winners that are better at following the directive keep winning. The losers disappear from the menagerie of life.
This is where our aspirations of engineering functions into life meets reality. Living things aren’t interested in what we want them to do and our engineered functions are forever secondary to the sole directive of life. Even worse, the changes we make to create these engineered functions are often contrary to this sole directive: they make the living thing less capable of making more of itself. The result is that once we’ve created a living thing with a specific function, such as production of biofuels, it begins trying to undo what we’ve done. The offspring of our living things should have the same engineered functions, but those that manage to undo what we’ve done are better at making more of themselves. In as little as one generation, the function we engineered can disappear. The system is never stable.
IMG_0185
A tree overgrows a wall in Hong Kong. Life has a tendency to escape any bounds put on it.
This is a serious problem because if we engineer a living thing and it doesn’t work as expected, what will it do? In the best case scenario, the living thing loses its function and we’ve got to find a way to get it working again. There are some straightforward ways to do this that I won’t cover here, but basically this scenario just costs some time and money. More worrying is a worse case scenario: where the function we engineered into a living thing to exist within certain limits or parameters, but it breaks free. This is the “Jurassic Park” scenario, where something meant to be contained and safe gets loose and wreaks havoc. And because the function of limiting a living thing to certain parameters is both incredibly useful and directly contrary to life’s sole directive, it’s less a matter of if and more a matter of when this happens. As Jeff Goldblum’s character Ian Malcom says, “life finds a way.”
For those of you now panicking, calling research institutions demanding they stop their work, hang on a moment. You’re right to be worried, and I want you to know that we researchers are right there with you. An overwhelming majority of us are driven to research because we want to do good, and the scenario in which our research creates something harmful is a nightmare. That concern is what drove  a temporary moratorium of research on mutant varieties of the avian flu that could prove more dangerous. It’s what’s behind a push to keep the locations of certain extremely rare animals secret, to prevent poachers from decimating their populations. And for those of us making forays into engineering life, it’s why we’re already thinking about how the functions and safety mechanisms we build could fail, and how we can prevent that. These are the problems that keep us awake at night.
In my case, it’s also the subject of my PhD thesis. I’ll explain that in the next installment.