#and my mosquitoes were LATE in the mail. arrived to late to do any testing… I had to pivot QUICK | Explore Tumblr posts and blogs

kmsherrard · 4 years ago

Text

In praise of roller coaster rides

“...the thousand concurring accidents of such an audacious enterprise….”

-Herman Melville, Moby Dick

Despite what teachers of high school science classes solemnly intone, this business of doing science is the least straightforward endeavor that can possibly be imagined. This was brought home to me in a series of unfortunate events that unfolded this week.

At first, it seemed to be that rare triumph where my simple test of a straightforward prediction actually yielded a clear positive result, instead of the more typical back-to-the-drawing board head-scratcher. If this were a story, that the protagonist was a protein named Diaphanous could serve as a hint that the plot would not prove as solid as one might hope. (Like many genes first discovered in fruit flies, Diaphanous evokes the appearance of animals lacking a functional version of that protein).

The backstory: Lately, my research has been on how stress fibers remodel to accommodate the movements of migrating cells. But as I work on cells in intact tissues, namely the rind of follicular cells that envelops the developing cluster of cells that give rise to a fruit fly egg, I like to consider the natural experiments that unfold in the course of normal development. For example, these follicle cells migrate for a time, going round and round like hamsters running on a wheel, but then they stop and do other things, like flatten out and secrete the eggshell. They still have stress fibers—these are long contractile bundles of a similar composition to muscle, that help attach the cells to the fibrous surface outside them. But these later-stage stress fibers are much stouter and of somewhat different composition.

I had already established that the stress fibers in the migrating cells depend on an unusual partner, amusingly called DAAM, to form. The more typical protein to help build stress fibers is DAAM’s cousin Diaphanous, but I’d done experiments depleting Diaphanous that clearly showed it was not needed in this case. When I depleted DAAM, though, the stress fibers got really wispy. Oddly enough, I’d noticed that in the much later stages, after the cells stopped migrating, had stress fibers unaffected by loss of DAAM.

So the experiment I wanted to do next was to deplete Diaphanous in the later stages. This was not completely straightforward to execute, though, because I had to avoid depleting it too early. I’d already seen that this caused cells to have trouble with their normal round of cell divisions. It’s a common problem in this sort of work that it can be harder to study later processes if you mess things up before they have begun to happen. The solution makes use of the dazzling array of tissue-specific drivers of gene expression that have been invented for fruit flies. They allow you to drive expression of a gene at specific times and places, targeting particular processes you want to study. To keep a gene from being expressed, you can use something called RNAi, which basically makes a cell chop up the instructions for making a protein sent from the DNA so that protein does not get produced.

In short, I needed a driver that acted late in the follicle cells but not early. Our lab did not have such a driver, since we study the earlier stages. But we’d read a paper with some very clever experiments that made use of just such a late driver, one called Cy2. We requested the fly stock from one of the paper’s authors and she promptly mailed it off to us. Fly researchers are awesomely generous. It’s a tradition that goes back to the earliest days of the field over a century ago to share reagents this way.

Chapter the First: Quarantine. The flies arrived and had to be put in quarantine, out of an abundance of caution concerning the possible introduction of mites into our hundreds of lab stocks. In practice, this consists of isolating the vials on the top of the lab refrigerator. All stocks that arrive from elsewhere must be taken through quarantine, save those from the renowned and very reliably mite-free Bloomington stock center. It meant a delay to the start of my planned experiment, until I could obtain 3rd instar larvae and wash them, a rather amusing exercise on which I have previously posted.

So there the flies sat, two healthy vials with clearly written labels: Cy2/(Cyo); Dr/TM6b. This cryptic shorthand conveyed that along with the driver I’d asked for, the flies conveniently included markers on another chromosome, in case I wanted to build more things into the stock. Annoyingly, they were all senescent adults and developing pupal cases—ideal for surviving the mailing process, but the worst possible stage of colony development for obtaining sufficient larvae for my purposes. I would have to wait several weeks for the new generation to produce larvae I could wash.

In pre-covid times, I could have done the cross right away with existing males, dissecting the offspring on a quarantine-use microscope belonging to a neighboring lab. Normally we share a lot of equipment freely in our department. But the physical distancing requirements have temporarily stopped that sort of thing. And we can’t risk getting mites onto the equipment we use for all our normal work.

To shorten the waiting time (a frequent concern of fruit fly researchers, especially I would think those of us who work on adult rather than embryonic or larval structures, meaning our crosses must extend to the full 10+ days of development time beyond any stock-building that precedes it), I planned to wash enough larvae to siphon off a number of males for the experimental cross. To that end, I also began “blowing up” the stocks I would obtain the females from; I could virgin them ahead of time and have them all ready to go as soon as their husbands emerged from their pupal cases.

When you’re waiting to wash a quarantine stock, impatient for the experiment to begin, they seem to take longer to develop, much like a watched pot. The stock contained the mutation Tubby, which makes for shorter flies but a longer developmental time, so that was part of it. Also room temperature (on top of the fridge) slows development compared to the flies’ optimal temperature of 25 C (that’s 77 to your Fahrenheiters...and to be honest, most of us American scientists are very compartmentalized in their understanding of Celsius; outside of the lab context we speak it no better than the average U.S. citizen). So far, then, the slowness makes sense both physical and psychological. But why the quarantined flies should always produce their burst of 3rd instar larvae on a weekend day, and on the one weekend day I don’t pop into the lab, is more puzzling. But it is the rule, I have found.

I wasn’t going to let it happen this time. I watched them like a hawk (a mosquito hawk?) and sure enough, it was a Sunday when all the larvae began to wander. Wandering larvae is the other, more romantic name for the 3rd instar of Drosophila melanogaster, because they have at last eaten their fill of the mushy rotten fruit they have been burrowing through, and there is nothing else for them to do but come out into the light and air and begin to claim their inheritance as winged creatures of the sky. First, though, they must choose a spot in which to prepare their new bodies. Here in that lab, they climb around on the clean walls of the vial, above the caramel-colored dollop of food, fat, juicy larvae as big as a good-sized grain of rice, big enough to grasp gently in forceps and take through the three ritual baths, soapy water, ethanol, and salty water, that remove any lurking mites or mite eggs from their surfaces. After being placed in a fresh vial and wicked dry with a twist of Kimwipe (lab Kleenex), they will crawl around a bit more, mingling with their certified-mite-free compatriots. In a few more hours they will settle down, stop moving, and let their skins harden into bark. Inside that bark, they pretty much dissolve themselves, save for a few set-aside clusters of cells. They go on to rebuild their bodies into the adult form, complete with intricate jointed legs and multitudinously-faceted eyes and iridescent, cellophane-like wings over the course of about a week (at room temperature).

I spent several hours washing more larvae than usual to establish a clean stock, wanting to have plenty of extra males to father the experimental crosses. If I’d had access to the quarantine microscope, I could have selected extra male larvae—you can already distinguish males and females at this stage-- but it would not really have saved time. I played the numbers game instead. It was a Sunday afternoon, quietest time of the week in lab, and very peaceful. I took my time and changed the bath solutions often to make sure there wasn’t too much soapy water in the ethanol or too much ethanol in the final rinse. I wanted this all to go smoothly with no delays.

I put the now-lawful vial in the 25C incubator to develop, after carefully copying the genotype from the original handwritten labels: Cy2/(Cyo); Dr/TM6b. Incidentally, there are lots of markers of chromosomes, many going back to the original mutations described by early fly workers such as Calvin Bridges and Alfred Sturtevant. They let you follow with visible traits the invisible genes that you wish to follow through the generations. Various labs have their favorite markers, but some such as Cyo (which makes for curly wings) are ubiquitous, and Dr and TM6b were familiar to me as well. Dr (short for Dropped, I don’t know why) makes the eyes very slitted, and TM6b is a whole set of markers that comprises what is called a balancer chromosome: a chromosome that has been scrambled and rearranged so that even though it still has all its genes, they are in the wrong places. This means that none of the usual recombination between sister chromosomes that occurs when egg and sperm form can happen. The advantage to the researcher is that this keeps genes segregated in predictable places. Otherwise, all those markers would not be reliable indicators letting you keep track of the genes you put in place from one generation to another. TM6b can actually include various different markers, but one of them is Tb, easy to recognize in both the shorter larvae and pupal cases and to some extent discernible in adults as well.

Chapter the Second: Cross Purposes. Fast forward two weeks (you can—I sadly could not—this being November of 2020, I would certainly have appreciated the distraction). So I waited, none too patiently, for the new adults to emerge. Meanwhile, I tended the stocks I would virgin for females: two different RNAi lines for Diaphanous and one, a control, for its cousin DAAM which I already knew was not required for the later-stage stress fibers. I built up a collection of ladies in waiting, captured shortly after their eclosion and isolated in vials away from all male contact, so I could be sure their offspring would be the genotype I wanted. [A note about the term ‘eclosion’: one might be tempted to call the emergence of the adults from their pupal cases ‘hatching’, but that term is reserved for the larvae coming out their eggshell. You only hatch once, even in the doubled lifestyle of these metamorphosing beasties.]

Finally the washed flies began to eclose. All my usable Cy2 flies were in that one vial. I briefly knocked them out with carbon dioxide gas, used a fine paintbrush to separate the males, and added 3 males each to the three bevvies of expectant females. There were still a few males left, enough to establish the new stock of Cy2 for future use.

At last, more than a month after conceiving it, I’d begun the experimental cross. It would be two more weeks before I had the flies to dissect and the beginnings of an answer. Fly work involves a lot of waiting, and to cope with that we tend to have a lot of irons in the fire. All that juggling can be rather distracting. Sometimes, depending on how other experiments have gone in the interim, I’ve unfortunately moved on from the original urgency of a question by the time the flies are ready to examine. It’s a hazard of the work.

Though I did not yet realize it, I’d made two mistakes. First, I should have looked a bit more carefully at those Cy flies. Second, I should have done the proper control. Sure, crossing them to the DAAM flies was a pretty good control, but there was an even stricter one, that tested whether the driver stock alone had any effect (it should not, but you like to be sure). I should have crossed the Cy2 flies to what we call wild-type, a stock called w1118 that has white eyes, incidentally [link] the first fly mutant ever identified and the foundation of fly genetics.

I hadn’t wanted to use up any more of my precious males, and figured I could always do that control later, if the experiment turned out promising. A lot of us cut corners that way, and it isn’t necessarily less efficient. But sometimes it snarls you up and wastes your time instead of saving it, and makes you go through all sorts of contortions trying to make sense of your data with less information than you should have had.

Chapter the Third: The Experiment. I waited out that two weeks, pursuing other work and trying not to pay too much attention to the news. I wore my mask and stayed in touch with my loved ones over zoom and the like. I hung up bird feeders to entertain my cats and my family alike. I went on long walks by the lake. Time passed. At last the grand day arrived: my experimental flies had begun to eclose. I gassed them and tapped them out of the CO2 pad. Now here was a wrinkle I’d shoved to the back of my mind: those extra markers that I didn’t need, the Dr and TM6b. In a clean experiment I’d have gotten rid of them, but that would have required another couple generations. I’d wanted a quick provisional answer, in order to decide whether it was worth the time and trouble to do the more careful version of the experiment. So: would I dissect the TM6b-carrying flies, or the Dr-carrying flies? It had to be one or the other. The balancer chromosome carries a number of mutations so it would be more likely to do something weird to the cells I was interested in. Not that that was very likely, but I might as well be careful. Dr it was then: that only affected the eyes, as far as I knew. What were the chances it would mess up my experiment on stress fibers in follicle cells?

But none of the flies had Dr eyes. That was odd. I looked closer. Half of them sure looked like Tb flies, shorter and a bit chubbier, though you never want to depend on your ability to discern that marker in adults. The others, the longer ones? They did have some oddly short hairs on their dorsal thorax (around the back of the lower neck, if you want to be anthropomorphic about it), much shorter than the clipped ones you see with the marker Stubble. It kind of reminded me of a marker I’d seen once or twice. Well, that must be what these were; maybe the label had been written wrong.

Impatient to get the experiment done, I swept the short-haired flies into a fresh vial with a bit of yeast. The yeast was to encourage egg production (they’re called fruit flies or vinegar flies, but it’s really the yeast on the rotting fruit that they’re after). I added a few males which were there for the same end. You could say the way to a fine set of ovaries is through both the heart and the stomach. Two more days to go before the dissection. For good measure I put some plain-vanilla w1118 flies on yeast to serve as extra controls.

On the appointed day, I got out my fiercely pointed #55 forceps and began the dissection. I nearly messed up by dissecting the early stages by habit—the technique to do so destroys most of the older egg chambers—but luckily remembered what I was about it time, and switched to the method to optimize acquisition of undamaged later stages. I fixed for 15 minutes in 4% paraformaldehyde, rinsed three times in phosphate-buffered saline solution with Triton-X detergent, and added a stain that would label the filamentous actin, the principle component of stress fibers among many other cellular structures. I put it in the lab fridge (the one where no food is allowed!) to stain overnight. The next morning, early, I came in and rinsed off the stain and made slides. Then I went to the womb-like room where one of my favorite workhouse microscope lives, the renowned Nikon 800 laser scanning confocal microscope. I did the necessary 2020 ritual wipe-down of all surfaces with 70% ethanol, and fired her up.

And oh, it was beautiful. I was so disciplined; I began with the controls to set up the correct laser intensity and gain at which to collect all the images, so the brighter ones would not be out of the range of measurable brightness and everything could be properly quantified. But it was already clear from the what I saw on the computer screen as I centered examples, focused, and took images that the experimental egg chambers had strongly reduced stress fibers. I took lots of pictures, happy that for once my experiment had gone as planned and given me a clear answer.

Also, can I just say how much I love the stain Oregon Green phalloidin? The name itself is lovely: as a native of the Pacific northwest I find it so evocative: the green of deep cushiony moss and ferns and forests of hemlock and douglas firs; and phalloidin itself is a stain derived from mushrooms with which those forests are rife. (Phalloidin, now there’s a scary toxin: it binds so tightly to filamentous actin that it stops your heart. Unlike a lot of other toxins, it doesn’t make you nauseated, so you absorb it until it’s too late for any antidote. But that’s why it’s such a good stain. You just have to wear gloves, or wash your hands after pipetting it. And we all wash our hands so often nowadays it makes no never mind.) There’s red phalloidin, and far-red phalloidin, and even ultraviolet phalloidin (but most microscopes don’t have the right filter sets to light it up very well): but green phalloidin is the king as far as I’m concerned. So bright, and a short enough wavelength (only 488 nanometers, vs. 566 or 647) that it shows up structures the more finely. You can definitely see the difference: it’s sharp as can be.

So, I had the preliminary results I had hoped for: the Diaphanous flies had reduced stress fibers. It doesn’t actually happen to me all that often, that I get a clear answer, either what I predicted or the opposite which is almost as good in science. At least that’s progress, an increase in understanding. No, usually I stumble over these head-scratchers of outcomes. Interesting results, but interesting in a complicated way that require a lot more work to make sense of, if you ever do. It’s partly down to most of my experiments involving imaging with a microscope: you get a lot of unexpected information that way, if you keep your eyes open. But it’s also that I seem to be attracted to the sort of problem that does not yield neat answers—the way some people are attracted to overly hairy guys on motorcycles who are a bit too into mild-altering substances and petty crime. I think I’m the one to straighten them out, but usually I’m the one who gets burned. But this time I had prevailed!

This was just a start; of course I needed to replicate, do some more dissections, get more numbers, reach levels of statistical unassailibility. In particular, I didn’t have as many clear examples of the DAAM control as I needed. Also, I’d do the proper control, and maybe even un-double-balance that Cy2 stock to get rid of the pesky extra markers.

Chapter the Fourth: The morning after. Yeah, and now I’d better take the time to figure out what is going on with that marker that is not Dr. Because, unlikely as it was, wouldn’t it be a shame if it were somehow affecting my results? Worst-case scenario—because that’s how we self-questioning scientists have to operate, ever since the dawn of time or at least the Enlightenment—worst-case scenario, then, is this marker, whatever it is, is the thing responsible for the reduction in stress fibers. Oh, but that’s very unlikely, I tell myself. Besides, the DAAM controls didn’t have reduced stress fibers.

I looked at the original handwritten label, still on the vial of flies on top of the fridge in quarantine. Maybe that D might actually be a P. What was Pr? I’d never heard of it.

I went to the master compendium of fruit fly genetics, FlyBase.org, and looked up Pr. Purple, an eye color gene on the first chromosome. I was looking for a gene on the third chromosome, so that couldn’t be it. I tried a different approach: I DuckDuckWent (DuckDuckGoed doesn’t sound right; if you haven’t heard of it, it’s a more private alternative to Google) images of Drosophila markers. There was that classic poster I’ve seen hanging in various labs, of the most common markers. And there was that marker I’d been reminded of, with the very short hairs. Sn it was called. Could that be my marker? It would have to be some pretty bad handwriting, to make an S look like a D; r to n is easier to imagine.

I went back to FlyBase and looked up Sn. It was the gene Singed. Like if you got to close to the outdoor fire pit on the patio (a way to safely hang out with your friends outdoors even during the Chicago winter), and singed your eyebrows most of the way off (and no, I haven’t done that yet). Also on the first chromosome, though. But look here, this is interesting: Singed is an actin-bundling protein. I read further down the page that summarized the work of dozens or hundreds of researchers over the decades. Yes, it was expressed in the ovaries, and yes, it was known to affect stress fibers. That would be worrying if it were my marker. Lucky it’s not.

I wasn’t getting anywhere. I tried yet another method, going to the webpage for the Bloomington stock center. It’s very well organized, and they have a page showing the details of all the balancer stocks they keep. There ought to be a clue here, for any marker that a researcher could assume another lab would recognize. I go down the list to the TM6b stocks, and find it. Pri, aka Pr, for Prickly. Causes short thoracic bristles. That’s my guy.

Back on FlyBase, I learn that Prickly is one of the classic mutants discovered in the early days of fly research. And this is weird: it has not been annotated. That is, nobody has figured out what gene it is a mutation of, let alone what biological processes it participates in or what tissues it’s expressed in (this matters because if it’s not active in the follicle cells, my experiment would still be valid). They could; it’s a straightforward enough task given that the whole genome is sequenced, but apparently it’s not one that anyone’s found worthwhile. So all we know is it makes very short, deformed bristles that look to me a lot like those of Sn.

Okay, now I am getting worried. What are the chances that this is NOT a protein that affects something like actin bundling and therefore messes up stress fibers? Maybe I had only seen what I wanted to see with the DAAM control. That’s a hazard of doing science, because it’s a hazard of being human. That’s why controls are so important. I consider my experiment in this new and harsher light. Maybe the Diaphanous results are just a phantom of wish fulfillment, summoned by this Prickly hitchhiker I’d never meant to take along for the ride.

I’d already begun the proper control that would answer this question, but meanwhile, while I wait for those flies to emerge, is there anything else I can do? Maybe I should dissect those formerly scorned Tubby flies; at least they lack Prickly. But according to the list at Bloomington, that particular stock has a number of other mutations on its TM6b chromosome, including one called Bri. Bri is a twin of Pri in more ways than one: it also causes very short bristles, and is also unannotated so we have no idea what protein it makes or when or where it acts in the body. Without asking the researchers who sent me the flies, I had no way of knowing if Bri was in there or not.

It would be a bit awkward quizzing them about their flies. We all tend to overdo the shorthand in labeling our stocks, and don’t always remember all the extra mutations lurking there. It’s tripped me up before, when I uncovered interacting mutations I hadn’t known to worry about until they unhinged my crosses. Don’t get me started on vermillian eye color: it’s a real bear. Either way, I’d have to check the controls and unbalance the stock to have a real answer, so probably better not to pester them.

I can’t resist having a quick peek at the TM6b flies though; I’ll be dissecting them tomorrow and should know by Sunday or Monday if the Diaphanous results are evaporating or not...that is, if Bri or something else is not further muddying the waters. A positive result would be definitive; a negative one will require further research. Well, either one will require further research, but one will be more cheerful and the other more like putting nails in a coffin of my hopes one more time. And that, my friends, is what it’s like to do science. (At least I get to see more Oregon green on the confocal, though).

Epilogue. What lessons can we draw from this (mis)adventure, this stomach-churning roller coaster ride of thrills and doubts that is my life in science?

1. Do the proper controls from the beginning. (Although that would have cut out the thrills as well as the doubts, so to be honest, I’m not totally on board with this one).

2. Take the time to look at the flies you are about to cross, and make sure they have the markers you expect. Harder, probably unrealistically hard, is to make sure they don’t have the markers you don’t expect. That would require a Rumsfeldian level of perceiving unknowns unknowns.

3. Remember the limitations of shorthand for conveying a genotype, which like the face we present to the world is invariably far more complex than there is room enough and time to write out.

4. Murphy’s law reigns supreme in this world of ours. What were the chances that the unwanted marker I’d thought I could ignore for a first-pass experiment would turn out to be a different marker I’d never heard of that might affect stress fibers in my cells? Still, it made for a good story, which I haven’t come across in all this interminable slog of an Autumn.

#drosophila #science #research #lab work

1 note · View note