That's certainly a neat trick. But, since this is a logarithmic
effect, pre-specifying an end point would mean that you're sacrificing
accuracy, right? Normally, say you have an infinitely long gel, you
have your DNA running down it in different columns, and as the hours
roll by the distances increase nonlinearly. So if you cut it short
after one minute, that will hardly be the same result as a million
years later. So, can anyone confirm here my suspicions about
invitrogen's "just use a comb near the bottom where you'd like to
Since every gel has a finite length though, you might as well just use
these wells at the end, plus a webcam to track when precisely each
column of stained DNA hits. Some friends and I were talking about this
(or, rather, we're talking about this in the background at the
moment)- some extra ideas would be:
(1) four-directional control of DNA
(2) webcam for visual tracking of stained DNA (we can call it a
DNAlogger, just throw an arduino or PIC at it, right?)
(3) some way of using the refractive index of DNA to your advantage so
as to not have to have a nasty stain on your dna
For protein gel electrophoresis + using the refractive index, see:
The instantaneous monitoring of polyacrylamide gels during electrophoresis.
Real-Time Monitoring of Polyacrylamide Gel Electrophoresis by Schlieren Optics
("A band containing as little as 0.3 µg of a protein could be detected.")
DIY phase contrast (expensive)
Phase contrast set-ups using the Zernicke method can be an expensive
option for the amateur, but home made phase contrast although
potentially cheap, at first seems a daunting prospect. Etching phase
plates accurately using hydrofluoric acid would be technically
difficult and hazardous. In 1953, Prof. Wilska, in a letter to Nature,
described the use of phase plates using a sooted pattern on glass.
This can either be directly on the back lens of the objective, or on a
glass plate, a cover slip for example, mounted close to the objective
focal plane. There is a useful Discussion in The Microscope journal
of 1953 describing the technique. This link also gives access to
transcripts of the correspondence in Nature.
I found that my Watson 30x and 40x parachromats had a ledge behind the
lens in about the right place to mount a phase plate as described in
an article on DIY Phase Contrast on the Quecket Microscopical Club
website, and having seen other correspondence on this technique, but
using a smoked stripe directly on the lens (although this also
referred to the use of a separate plate), decided to have a go. I
tried using a circular cover slip, applying soot from a candle and and
scraping it off to give a stripe, but found this difficult to do in a
controlled fashion. I then tried masking the coverslip with adhesive
tape before sooting, and this worked quite well, although one needs to
select a tape that leaves minimal adhesive residue once removed. The
substage stop was made by masking a transparent disc with PVC tape to
leave a stripe that aligned just inside the phase plate stripe when
the objective back-lens was viewed using a pinhole eyepiece.
I do make phase plates (and modulation plates) and that article was a
little ahead of what I intended to write about, and got sidelined from
(excuse the contorted syntax, but I'm sure you know what I mean!) My
method is not original - it's based on a method that was shown to me
by my old friend and mentor, the late Leslie Martin. Essentially the
phase plate is produced from wax/soot deposited by candle flame on a
coverslip, which is then placed as near as possible to the back focal
plane of the objective. Having found the rear focal plane by the zero
parallax method (you reminded me of this last time we met) I machine a
ledge on the inside of the objective (with great care!) to take a
brass holder to take a standard 6mm, 8mm or 10mm coverslip at the
required distance from the rear lens (of course modern objectives with
the rfp inside the elements are too difficult for the amateur, but a
Zeiss DD and Watson Para both have unscrewable front components to the
barrel and seem suitable). I make several of these holders at one time
so that I can experiment with different types with minimum effort. I
cement the coverslip before smoking. I then make the phase ring by
nipping the brass holder in a split chuck in my Unimat and using a
cocktail stick to wipe away the unwanted wax/soot.
Anyway, a phase-contrast webcam would be fantastic (and allow for
ignoring #4, below). For defractive lenses, you can print out a
spiral-like pattern on your printer and then make the non-printed
portions transparent, and then use this as a defractive lense. But
this isn't the same, though still worth investigating some more.
(4) some way of electrically detecting when DNA hits a well (built
into the bottom plate), such that your gel would block it until you
use your comb to dig into it, such that when the DNA drains into it,
you can have a timestamp (wire it up to a microcontroller).
Finding an alternative to staining would be really, really awesome.
I wonder though: how much of a difference can you make in the gel
process by changing the general shape and gel cast? For instance, does
anyone have some numbers on what level of accuracy you can get with
gel casting, and can this be used to our advantage in other ways than
just extra wells at the end?
I recall the original Raman spectroscopy paper had an easy
do-it-yourself version that could be used. There are also a number of
typical spectroscope designs on the web for using a cereal box and CD
Please excuse me. Some of those are for mass specs, not photospec. My
mistake. The first one still certainly applies.
Consider that more viscous substances, e.g. agarose solution, will
require a wider capillary tube than the sort one would normally use
for, say, blood. How about a thin, clear drinking straw? You could use
paperclips for electrodes.
This could also be a really neat trick for doing PAGE, since those
gels are usually run vertically, with a little well dug in the top. A
drinking straw would provide external support for the gel.
> Then you just break off the very end of the tube to
> extract the band of interest.
Or, with a tube that was flexible like a drinking straw, snip the band
of interest out with scissors and slit the plastic open with a razor
I love this mailing list so much.
Not exactly. In capillary electrophoresis you're relying on the molecule's
affinity for the walls of the capillary, so it has to be really small (say
50 microns) and this makes it no good for cloning DNA. Typically you'd
have 20 microliters of reaction solution from your PCR, so this would be a
2.5 meter long band! vs a 2mm band in a 3mm diameter tube. Also by using
standard agarose it should behave the same as in a regular gel, which
makes people feel warm and fuzzy inside.