Siliceous and Delicious

After last night’s very disappointing CLS microscopy session, this morning’s DSA meeting with Beau was just the elixir I needed. He got me back on track, thinking positively about the months ahead, and perhaps more importantly, also about the past twelve months. Helpful as ever, Beau provided his notes on our conversation:

Confocal work: Ben reported that the inverse staining option wasn’t really returning positive results, and showed that the best pictures so far have come from the standard method. Sadly, he got news from the lab tech that with his departure at the end of the summer, the machines will probably be disappeared to various locations. This puts a hard 6 month deadline (at best) on the confocal work.

Presentation: Ben didn’t get a chance to work on this, due to various EPS 8 commitments, but sounded like he knew exactly what should be going where, and has plenty of sexy pictures to add. He needs to do some 3-D reconstructions of some of the newer confocal pics, but otherwise it’s all under control in that department.

Reflections from B: It’s quite a blow to hear that the machines will be gone at the end of the summer. I entirely appreciate the feeling that, despite all the hard work you’ve been putting in, circumstances appear to be conspiring against you in ways that are absolutely not under your control. But what is under your control is how you make productive use of the time ahead. Since you’ve assured me that it’s possible to get the measurements done within 6 months (possible or necessary, I guess both may be the same at this point!), you can now focus on pulling out all the stops towards getting this project whacked on the head during that period. Getting this done will not only leave you with a thesis chapter ready to go, but also prove a tremendous morale boost, which will help the other work you’re doing. The more you work on the machines, the better you’ll get at taking the pictures, and the better the dataset you’ve gathered will become. So hang in there, focus on the prize, and make sure you’ve got a clear path laid out for all the work during the summer.

For next week: Ben will run through his presentation, and for bonus points, also talk about potential milestones for the confocal work during the summer.

Spent the afternoon grading my part of the EPS 8 labs, then went home.

Came in this morning to find the epoxy in my latest attempt at slide-making had indeed cured, but that there were also bubbles. Just as many as when heat-curing. Rats! Spent a fair bit of time trawling the interwebs, trying to find any information I could about how to add stuff to Epo-Tek 301 epoxy without making bubbles. I could find almost nothing of any use, although I am beginning to suspect that it’s the water in which the fluorescein salt is dissolved that’s the source of the problem. Perhaps trying to use fluorescein in methanol will lead to happier results?

I futzed about in the lab for a bit and tried to prepare a slide using epoxy mixed with fluorescein in methanol, but I saw bubbles forming immediately upon placing the cover slip on the slide. I thought, why not try the lead weights lying around next to the scales? I’d always been concerned that they would just crush the fossils between coverslip and slide, but given the desperation of the moment I figured I’d try it anyway. The result: I glued the lead weight to the coverslip. And there were bubbles in the epoxy anyway. Woohoo.

In the meantime, the epoxy had started to harden a little. I thought I’d try making one more slide with that epoxy, on the off-chance that waiting time had something to do with it. It didn’t seem to make a difference. Putting a weight on it made the epoxy squeeze out all over the place, but there were still bubbles nonetheless.

Then another incredibly frustrating time on the CLSM. Evangelos informs me that, in all likelihood, the facility will be dismantled after his departure. Fuck.

Spent the morning trying, at long last, an old idea on the CLSM front: making the epoxy fluoresce (rather than the fossils). It’s an exciting idea, because—it it were to work—it would actually yield images of the whole skeletal volume, rather than just an image of the surface covering the skeletal volume. However, as with everything I’ve tried, there appear to be challenges.

First, neither straight fluorescein nor the fluorescein sodium salt will dissolve directly in either of the epoxy components. So, I tried adding to the epoxy fluorescein sodium salt (which appears to be far more soluble in general) that had first been dissolved in a small volume of either ethanol or DI. The resulting epoxy is wonderfully fluorescent (when illuminated, it glows bright green like the radioactive brick in the Simpons titles sequence) but seems to form bubbles no matter what. I tried lowering the curing temperature from 80˚ to 60˚C, and this prevented bubble formation in a blank trial slide—but bubbles still formed when I attempted to prepare a slide with fossils. Under the microscope, it appeared that each individual bubble was nucleated on (or at least closely associated with) a fossil, making matters worse.

I had to stop after a couple of hours to turn my attention to an even more pressing task: writing the next EPS 8 lab. Jc and I sat down and together wrote the first part, on the Signor-Lipps effect, fairly quickly. The rest went substantially less quickly. We got only part of the way through the P/T extinction section before the day was over and it was time to head out for my squash game at 6 pm.

After squash I returned to work, having scheduled time at the CLSM to see if my inverse-staining method worked. It was a frustrating exercise—for the first part of the evening, I seemed to be able to get some images, though the matrix fluorescence was far from even throughout the whole background. Rather it seemed that there the fluorescence was bright in the center of the field of view and decayed to almost no fluorescence at the edges. The non-fluorescent areas at the center of the image—constituting the frustule structure I’m trying to image—were poorly defined and not nearly as crisp as the images I’ve been able to obtain by staining the frustule. What’s more, focusing to the bottom of the preparation—where the fossil sits on the coverslip—brings the plane of focus to a point where it appears not to intersect any epoxy, as the fluorescence disappeared altogether, leaving blackness and consequently no image. The most frustrating part ensued when I tried to rectify the situation by changing the filter set-up, after which the image disappeared altogether and all I seemed to be able to get was either a black screen (no light reaching the detectors) or a fully saturated screen (the laser directly hitting the detector). I wasn’t able to make the image reappear after that.

One final observation, and another strike against the inverse-imaging idea, was that the stained epoxy seemed to bleach quite rapidly. After just a couple of minutes of scanning a specimen (in the process of trying to get the z-stack set up right, which incidentally is also a total pain in the backside with the Olympus software), the fluorescence started to wane, and when I moved the stage by a small distance, the previous edge of the scanning field was visible as a bright field of fluorescence, showing starkly by contrast how much the scanned area had bleached.

In summary: as much as I’d fallen in love with the idea over the past few days, I’m no longer convinced the inverse-staining method is the solution to my problems. I will try one more time, with Evangelos’ help, and then decide.

Found a website that lists excitation and emission wavelengths for a bunch of fluorescent dyes, to figure out what lasers I’ll need to test the dyes I started experimenting with yesterday on the CLSM. Evangelos has been evasive as seems to be the norm for CNS employees—whilst being very generous with assurances of his willingness to help and his certainty that we can make things work out on the Olympus CLSM as a back-up, he was decidedly stingy in releasing any information about when the Zeiss scope would be coming back online (“the Zeiss will be down for a while”).

Anyway. Best not to dwell on the barriers, but work on those paths that show some horizon. The relevant wavelengths:

• Fluorescein: 494, 518 nm
• Calcofluor: 440, 500-520
• Ruthenium red: couldn’t find any information on ruthenium red as a fluorescent dye. As far as I can tell, it’s used as a histological stain and as an inhibitor of Ca uptake and release in mitochondria. Not as a fluorescent dye. Did Jacques make a mistake here?

Spent the rest of the morning making slides of these—and then, in a fit of frustration over the mess and sheer dirt in the lab, cleared and scrubbed most of the counter space of the wet lab to make the space a little more conducive to doing research. It felt good to clean, and the place looks much better, so I was inspired to move ahead and try again to weigh my radiolarian samples from Dave. This, again, was not so successful. I had bought some glass petri dishes yesterday and washed them, but in drying them they had accumulated a large amount of lint from the paper towel I used—and under the picking scope it became clear just how big the lint fibers are compared to the radiolarians (they are orders of magnitude larger). I will have to thoroughly clean the microscope stage and petri dish and get it as dust- and lint-free as I can before I start my next weighing attempt. Another problem with the newer method I tried today (which involved stroking the brush on the dry petri dish to remove picked radiolarians from the brush to the destination petri dish) is that I never know exactly where I’ve already deposited radiolarians. Thus, I risk picking up previously deposited radiolarians when I’m trying to put down new ones. A possible solution (now that I’m going to be evaporating away the water before weighing) is to fill the destination petri dish with water and then swirl the brush in one end of that to release the radiolaria, but then tilt the dish slightly to encourage the rads to settle on the other end of the petri dish.

At 3 pm I was due to meet Evangelos to have him show me the Olympus CLSM, but when I walked over there I could not find him anywhere. I did bump into Tanja Bosak, though, who asked how things were going for me—I gave the honest answer, which always leaves me feeling about as low as I can get, and although she wished me well, this was about all I could take for the day, so I rustled up a group of fellow EPSers and made a beeline for the Berryline.

A good day for getting the lab cleaned. A good day for being sunny and warm.

Searched a little more for recipes online for the other fluorescent stains I got from Jacques. Ruthenium red allegedly dissolves in DI water. I made preparations of four different stains:

1. Fluorescein in methanol (this, as I have previously found out, doesn’t dissolve particularly well).
2. Fluorescein sodium salt in DI water (this dissolves much, much better—it makes a thick, deep red solution).
3. Ruthenium red in DI water (a few grains turn the water deep purple, but the grains don’t dissolve fully, even with several minutes of vortexing, leaving some small granules in the solution and thus presumably to problems further down the line).
4. Calcofluor in DI water with some KOH (mixed about 30-70, 1 pellet dissolved in about 100 ml of water, giving a solution with pH ~12). Makes a pale yellow solution, though there seem to be undissolved specks in it (hard to tell).

For each dye, prepared two samples of the coarse fraction of ODP/Zoe 27#, put one sample of each on the hotplate at 60˚ for two hours, and then filtered the samples down and washed the filters into small vials with DI water (the same procedure as before). Unfortunately ran out of time at this point and had to leave the samples in the vial with the DI water. This makes me somewhat nervous as most of these dyes are water-soluble so will presumably come right off the diatoms again. Maybe if I quickly prepare slides from them..? Yeah. Did it. And it feels good to at least have tried something today. We’ll see tomorrow how it works out.

It’s been quite a while, but now that the insanity of EPS 8 has subsided, and I’m no longer sick as a dog, it’s time to return to the CLSM project. It’s been well over a month since I last worked on this project, so just to review briefly where I have arrived: I have had the most success (see images posted on Jan. 27th) working with the supernatant of a saturated fluorescein solution with the suspended excess fluorescein particles settled out at the bottom of the vial. The sample suspended in this solution was filtered through an 0.6 µm, 24 mm wide polycarbonate filter disc that actually fits the filtering equipment I bought with Nick Tosca’s help last year. I rinse the sample briefly with DI water and then remove the filter disc into a glass vial; I use a methanol rinse only for cleaning the fritted disc after the filter has been removed.

Unfortunately, I realized too late that I was actually expected to be present at this afternoon’s midterm exam review for EPS 8, so rather than actually translating my desire to continue work on the CLSM project, I instead sat and prepared for the midterm review.

After the midterm review, had just under an hour to get cracking. But the lab was out of purple nitrile gloves, so I had to abort mission and vow to continue tomorrow morning.