| 1. |
Prepare a 1X TAE, 1.2% LMP (low melting point)
agarose gel. |
|
A. |
TAE buffer is the best for both RNA gels and for nucleic
acid recovery. It only has sufficient buffering capacity during
extended runs if there’s recirculation however. I used an Owl model
A5 Buffer Puffer™ horizontal gel electrophoresis system. Its
"passive" recirculation is sufficient. |
|
B. |
Ideally a LMP agarose gel less than 1% would provide the
best separation for RNAs 1.5 kb and up. In my experience, at 1% a
LMP agarose gel is too fragile to survive virtually any
manipulation. |
|
C. |
Do not use any kind of denaturant in the gel or the
running buffer. They all smear the RNA at least a little bit.
Especially do not use formaldehyde. It works fine for Northern
blots, but it does something irreversible to the RNA that prevents reverse
transcription. |
| 2. |
Equilibrate gel and running buffer at 4°C. |
|
A. |
The gel is run at 4°C to prevent overheating which can
smear bands and melt the gel. I have my electrophoresis apparatus
set up in a 4°C cold room. |
| 3. |
Heat denature RNA sample and RNA ladder in
formamide and 10mM EDTA at 65°C for 10 minutes. Place on ice, add
loading dye, and load onto gel immediately. |
|
A. |
For a ladder I used Ambion’s Millenium™ and Century™ RNA
markers. I added 0.5 ul of each ladder (1 ug/ul) and 0.2 ul of 0.5 M
EDTA to 8 ul molecular biology grade formamide. |
|
B. |
For my yeast sample I added 50 ul poly(A) RNA at 1ug/ul
and 3 ul of 0.5 M EDTA to 100 ul formamide. |
|
C. |
I added 1 ul loading dye (0.25% Bromophenol Blue, 0.25%
Xylene Cyanol, 20% Ficoll) to the ladder and 15 ul to the yeast
sample. |
|
D. |
I find that RNA can overload an agarose gel if the lane
isn’t wide enough. I recommend well width of at least 1 mm per
microgram of sample RNA. For the 50 ug of RNA I loaded, I taped six
teeth together to produce a 5 cm wide lane in the middle of the gel. |
| 4. |
Run at 9 V/cm until the desired separation
length is reached. |
|
A. |
I ran at 260 V for 2 hours. The bromophenol blue
migrated about 10 cm. |
| 5. |
Excise ladder lane. Stain with ethidium
bromide, and visualize with a UV transilluminator. |
|
A. |
The purpose of this step is to decide where to cut your
length fractions. I started at 300 bps and worked my way up. |
| 6. |
Excise your desired length fractions from the
gel. |
|
A. |
In order to do this, I taped a sheet of metric graph paper
to my bench, and secured the gel (still in the gel tray) over it.
Using the graph paper lines as a guide, I cut thirty 2 mm slices with a
scalpel. I had originally designed a cookie cutter-like device to
precisely cut several slices at once. At 2 mm, however, the device
tended to mash the gel preferentially to cutting it. So unless
you’re cutting larger slices or you have a real cool robot arm to work
with, I’d just cut the gel by hand. |
| 7. |
Melt each slice at 70°C for 10 minutes. |
| 8. |
Transfer to 42°C. Add 1 U of β-agarase per
200 ul of gel. Add exogenous doping control RNAs.
Vortex. Incubate at 42°C for 2 hours. |
|
A. |
I used agarACE from Promega. It probably doesn’t
matter where you get it from, but its activity goes down over time, so use
fresh enzyme. |
|
B. |
The doping control RNA is used to normalize the various
microarrays since each length fraction will have a different RNA recovery
yield and a different number of enriched genes. |
| 9. |
Concentrate each agarase reaction in a
Microcon-30 spin column. Wash twice with water and elute. |
|
A. |
I tried a number of methods to recover the RNA from the
agarase reaction. The Microcon-30 method had by far the best yield
for every length of RNA. |
|
B. |
The Microcon-30 columns do trap the β-agarase and some
undigested agarose. I find that neither of these have any effect on
reverse transcription, but they do effect UV absorbance in case you were
interested in measuring RNA concentration by OD. |
|
C. |
After eluting, I diluted each sample to 30 ul and used 10
ul per array. I recommend length separating enough RNA to do
duplicates or triplicates since not every microarray hybridization goes
perfectly. |