Russell's Blog

New. Improved. Stays crunchy in milk.

These are not the microbes you are looking for

Posted by Russell on October 04, 2011 at 6:45 p.m.
A few months ago, I tweeted, "I've been working in a microbiology lab for two years, and just realized we don't actually have a microscope. Huh."

Jack Gilbert and some other people proceeded to give me grief for what I intended as an interesting observation about the current state of the art in microbiology. So, I decided to remedy the situation. Evidently, we do have a microscope, I just didn't know where it was.

Here are some cool things I found by randomly poking around in some of my samples from Borax Lake. This first thing I found is probably some kind of diatom from in the sediment of the little hot spring just north of Borax Lake. I'm not looking for diatoms, but it looks really, really cool.

Here they are at 100x magnification.

This is somewhat less cool-looking, but is probably what I'm actually looking for. In the little bubble of water surrounding the granule in the center, there were a couple little rods hopping around. No clue what they are, they're there, doing what they do.

3D printing update

Posted by Russell on October 04, 2011 at 12:19 a.m.
I've been working a bit on the software that generates my 96-well dilution plate. I have a new version that cuts the plastic use by about 80% and print time by about the same. Also, it now prints with the wells upside-down on the build platform, which should help cut down contamination during the printing process.

Things to do :

  • I'm going to try cutting the plastic use even more by adding a skirt around the plate (like a normal titer plate), and adjusting the outer height of each well.
  • Add a fill-line to each well.
  • Raise the well edges a little more, and add drain-holes between wells to prevent spillage between wells and to make filling easier.
  • Add embossed row and column labels.
  • Add an embossed text area for user notations (e.g., for which sample group is this plate calibrated).
Hmm. I might pull this off yet.

Also, if you are interested in this stuff, the UC Davis Biomedical Engineering made me instructor of a variable unit class (graded P/NP) called "Research internship in robotics for the laboratory" for Winter 2012. Sign up for BIM192, sec 2 (the CRN is 24791).

Borax Lake : Sample collection and processing

Posted by Russell on October 01, 2011 at 1:57 a.m.
I've admired Rosie Redfield's lab-notebook-as-blog from the moment I started reading it, and I've been looking for an excuse to steal pilfer abscond with adapt her idea. However, most of the work I've been doing up to this point has been computational, and try as I might, I can't make myself keep a lab notebook for programming. That's what things like GitHub are for. I've started actually doing some of the laboratory and field work for my thesis project, and so I finally have an excuse to do some open lab and field notebook blogging.

Rosie might be amused to that I'm starting off with some work I'm doing at an arsenic-heavy lake, although Borax Lake is known more for boron than arsenic. If I bugger up my assays, I just hope she'll get on my case in the comments before I submit anything for publication.

I will write more about this as I go along, but my goal for my thesis project is to try to get an idea about the modality of microbial migration. Specifically, I want to know if microbial taxa, when they colonize a new environment, arrive individually or as an existing consortium. I hope to find out by reconstructing population structures from metagenomic samples from widely dispersed but ecologically similar environments.

I learned about Borax Lake from Robert Mariner of the US Geological Survey, who was kind enough to respond to my emails and patiently discuss his survey results over the course of several lengthy telephone calls. He also volunteered a lot of useful information, such as which sites have rattlesnakes and where they are likely to be found, and helped enormously in search and selection of sampling sites. Without this help, I probably would have had to give up on this project as I originally imagined it.

Borax Lake was one of many thousands of sites across the American West surveyed over a 40 year long project USGS project led by Ivan Barnes and Robert Mariner to study the chemistry and isotopic composition of mineral springs. Extensive analysis of Borax Lake water was conducted by in August of 1972 by John Rapp, and again in July of 1991.

Dr. Mariner pointed out that Borax Lake is administered by the Nature Conservancy, and a little bit of Googleing and emailing got me in touch with Jay Kerby, the Southeast Oregon Project Manager for the Nature Conservancy. Jay was very helpful, and walked me through the process of obtaining sampling permits for my project.

Before I talk about Borax Lake, I need to say that it is absolutely essential that you obtain explicit, written permission before collecting samples. As scientists, we've got to get this stuff right if we want to avoid stuff like this. The fact that some researchers did not (for whatever reason) obtain permission to use the cells they used to make important discoveries, or did not cooperate in good faith with the originators of those cell lines, has made it much more difficult for me to do my own research. Kary Mullis, if you're reading this, thanks for PCR (really), but...

Anyway, I'm not sure if Borax Lake itself is going to be a good candidate for my project (it has very unique chemistry), but it is surrounded by ephemeral pools of brine that may be good analogs to coastal salt ponds. You could think think of this as island biogeography, but inverted; I'm looking for islands of ocean isolated by oceans of land.

The lake itself has a very peculiar mineralized ledge a few inches above the shore. The water has been precipitating an extremely hard material for a very long time. I tried to collect a small amount of it to examine in the lab, and discovered that it is as hard as concrete. Even with the aid of a hammer, I couldn't dislodge any small pieces. I didn't want to damage the ledge itself by taking a larger piece, so I left without any samples of the precipitated material. Borax Lake is sitting atop a thirty foot high pedestal of this stuff.

The Nature Conservancy has been working on some plans to make the site more accessible, but I don't imagine it will get many visitors. It is way off the beaten path. Next time I visit, I'm going to bring a truck. The lake is off of a very lonely state road, up several miles on unpaved, unmarked fire roads, followed by a few miles of ATV tracks. A horse would probably the the ideal way of getting there, but my trusty little Toyota still managed.

This is one of the hotsprings just north of Borax Lake. The first record I have of it is from May 1957 by D.E. White of the USGS. It was visited again in June 1973 by Robert Mariner, and again in September 1976 by Robert Mariner and Bill Evans. The next visit was in July 1991 by Robert Mariner. I measured a surface temperature of 65°C. To my surprise, I saw a couple of Borax Lake chubb swimming around near the cooler (but not much cooler) periphery.

I took four kinds of samples : Unprocessed water samples in 500ml bottles, unprocessed sediment samples in 50ml conical tubes, processed water samples for environmental DNA in Sterivex filters, and processed sediment samples for environmental DNA using Zymo's Xpedition Soil/Fecal miniprep kits. I divided the unprocessed samples between the freezer and the 37° room, and I'll save my notes on the filtered water samples for another article.

One of the unusual things about the Xpedition miniprep kit is that the first spin column is not a DNA binding column; it's more like a crap-catcher. So, you are supposed to keep the flow-through, not discard it as you would with a DNA binding column. John got a little ahead of himself, and discarded the flow-through from four columns before he realized the protocol was different from, well, just about all of the other DNA extraction mini-preps on the market. Fortunately, I collected many extra samples. Also, when I split the work between John and myself, I split up the samples into evens and odds, so that neither of us would be working on all of one group of replicates.

This led to an important lesson : Do not discard the lysis tubes after you've removed the supernatant. It occurred to me that the wreckage of beads, muck and buffer at the bottom of the spent tubes was probably full of DNA, so I added 500 μL of molecular-grade water, vortexed them, and put them back into the centrifuge at 10,000g for a minute, and spun the supernatant through the orange-capped columns. Two of the four yielded plenty of DNA. I'd probably have gotten more if I'd used lysis buffer instead of water, and the bead-beater instead of the vortexer.

I'm still not totally sure what rationale to apply for the last step. The Xpedition miniprep lets you elute the DNA with anywhere from 10 to 100 μL of buffer. If you use less elution buffer, you get less total DNA, but the DNA you get will be at higher concentration. Elute with more, and you get more DNA, but at lower concentration. The actual amount of DNA can vary over four orders of magnitude, and so guessing right is very helpful. But... impossible. I decided to elute in 50 μL, and that seems to have worked OK for my purposes.

I then measured the DNA concentration in a Qubit fluorometer with Invitrogen's Quant-iT high sensitivity assay for dsDNA. Because this requires me to go one-by-one, this is not how I would like to quantify my samples in the future. But, for thirty seven samples, it was easy enough.

Sample μg/mL Source Description
1 0.723 Don Edwards Wildlife Refuge Salt crystals from site A23
2 0.582 Don Edwards Wildlife Refuge Salt crystals from site A23
3 0.531 Don Edwards Wildlife Refuge Salt crystals from site A23
4 0.824 Don Edwards Wildlife Refuge Salt crystals from site A23
5 0.209 Don Edwards Wildlife Refuge Salt crystals from site A23
6 27.8 Don Edwards Wildlife Refuge Mat community from site A23
7 - Don Edwards Wildlife Refuge Mat community from site A23 (field processing failed)
8 2.57 Borax Lake Sediment (poor collection)
9 44.9 Borax Lake Sediment
10 44.9 Borax Lake Sediment
11 25.1 Borax Lake Sediment
12 47.1 Borax Lake Sediment
13 48.2 Borax Lake Sediment
14 41.2 Borax Lake Sediment
15 41.2 Borax Lake Sediment
16 40.9 Borax Lake Sediment
17 26.1 Borax Lake Sediment
18 31.6 Borax Lake Sediment
19 - Borax Lake Sediment (lost during extraction)
20 81.6 Borax Lake Sediment
21 35.6 Borax Lake Sediment
22 35.9 Borax Lake Sediment
23 - Borax Lake Mat community from hot spring
24 - Borax Lake Mat community from hot spring
25 3.13 Borax Lake Mat community from hot spring
26 44.0 Borax Lake Mat community from hot spring
27 - Borax Lake Mat community from hot spring (salvaged sample)
28 - Borax Lake Mat community from hot spring
29 2.76 Borax Lake Mat community from hot spring (salvaged sample)
30 - Borax Lake Mat community from hot spring
31 - Borax Lake Mat community from hot spring (salvaged sample)
32 - Borax Lake Mat community from hot spring
33 0.72 Borax Lake Mineralized mat community from hot spring
34 22.8 Borax Lake Mineralized mat community from hot spring
35 100 Borax Lake Mineralized mat community from hot spring
36 29.2 Borax Lake Mineralized mat community from hot spring
37 39.2 Borax Lake Mineralized mat community from hot spring (salvaged sample)

For the 19 samples that had DNA concentrations above about 20 μg/mL, I ran a gel to check the size distribution. It looks like the Zymo miniprep performed about as well as they claimed; most of the fragments seem to be between 5 and 10 kilobases, with a fair amount of DNA in fragments larger than 10 kilobases.

I only need about a picogram of input DNA for each transposase tagmentation library, and I only need fragments bigger than about 3 kilobases. So, this process exceeds my absurdly modest requirements by a lot.

I should mention that Anna-Louise Reysenbach graciously lent me a pH probe to use in the field after mine turned out to be dead as a doornail. Issac Wagner, a postdoc in her lab, spent a couple of hours helping me get their field probe calibrated with my meter. Unfortunately, their probe turned out to be in only somewhat better condition than mine, and Anna-Louise asked that I leave it in Portland rather than risk taking bad data with it. I drove directly from Borax Lake to the UC Davis Genome Center in about seven hours, and immediately took the chemical measurements on our benchtop pH meter. It didn't work out, but I still greatly appreciate the help from Anna-Louise and Issac! (Also, thanks goes to my little sister Anna, who has to take Portland's MAX over to Portland State to return the ailing pH probe.)