PROTOCOL

Improved northern blot method for enhanced

detection of small RNA
Gurman S Pall & Andrew J Hamilton
Division of Cancer Sciences and Molecular Pathology, Faculty of Medicine, University of Glasgow, Western Infirmary, Dumbarton Road, Glasgow G11 6NT, Scotland, UK.
Correspondence should be addressed to A.J.H. (a.hamilton@clinmed.gla.ac.uk).

Published online 5 June 2008; doi:10.1038/nprot.2008.67

This protocol describes an improved northern blot method that enhances detection of small RNA molecules (o40 nt) including
2008 Nature Publishing Group http://www.nature.com/natureprotocols

regulatory species such as microRNA (miRNA), short-interfering RNA (siRNA) and Piwi-interacting RNA. Northern blot analysis
involves the separation of RNA molecules by denaturing gel electrophoresis followed by transfer and cross-linking of the separated
molecules to nylon membrane. RNA of interest is then detected by hybridization with labeled complementary nucleic acid probes.
We have replaced conventional UV-cross-linking of RNA to nylon membranes with a novel, 1-ethyl-3-(3-dimethylaminopropyl)
carbodiimide (EDC)-mediated, chemical cross-linking step that enhances detection of small RNA by up to 50-fold. This requires no
specialized equipment, is relatively inexpensive and is technically straightforward. Northern blotting can be done in 2 d, but
detection of a specific RNA can vary from minutes to days. Although chemical cross-linking takes longer (15 min to 2 h) than UV
cross-linking, improved sensitivity means shorter periods of exposure are required to detect signal after hybridization.

INTRODUCTION
Increased awareness and interest in the regulatory roles of non-
coding small RNA have led to the modification of several existing
technologies, such as real-time PCR and nuclease protection assays,
to detect low levels of small RNA in a better way1,2. In addition,
high-throughput methods such as those employing small RNA
cDNA cloning, deep sequencing and microarray expression analysis
have greatly accelerated the discovery and characterization of novel
small RNA3. In comparison with these, conventional northern blot
(also known as RNA gel blot) analysis is less sensitive and lower
throughput. However, it is unmatched in its ability to display the
size of small RNA accurately and can simultaneously display the
sizes and amounts of multiple small RNAs that share significant
sequence identity. For example, both the mature (B20 nt) and
precursor (B70 nt) forms of microRNA (miRNA) can be simulta-
neously detected and such a pattern provides strong support for the
contention that ones sequence is indeed a miRNA4.
As size and size complexity are critical parameters in the
validation and functional characterization of novel small RNA,
northern blotting has remained a popular and valuable analytical
method.
Northern blot analysis of small RNA involves the separation of
RNA molecules according to size using denaturing PAGE (dPAGE)
followed by transfer of the separated molecules onto, typically, a
nylon membrane. The transferred molecules are usually cross-
linked to the membrane using UV irradiation to reduce loss of
the sample RNA during subsequent hybridization and wash steps,
in which specific labeled nucleic acid probes complementary to the
RNA sequence of interest are allowed to hybridize with the
immobilized sample RNA. Other cross-linking methods such as
baking at 80 1C or alkaline-assisted fixation are used for immobi-
lizing RNA that are typically of mRNA size range but have not been
widely used to cross-link small RNA.
Cross-linking with UV is rapid, inexpensive and works well for
RNA 4100 nt in length. It is thought that UV produces reactive-
functional groups within the bases of RNA (principally Uridine),
which then react with free amine groups on the nylon membrane
surface to form covalent bonds57. However, the evidence for this is
indirect and the exact mechanism has not been demonstrated yet. If
one presumes that the nucleotide bases are involved in forming the
cross-link, then it is reasonable to think that this may reduce the
subsequent availability of that base for hybridization with a
complementary probe. Furthermore, as the number of bases
involved in cross-linking cannot be controlled in the UV-triggered
reaction, over-cross-linking might occur, hindering the accessibility
of the probe for hybridization with the target sequence. We thought
that such events may have an increasingly negative effect as the
length of RNA to be detected decreased and so is particularly
problematic for analyses of siRNA (short-interfering RNA) and
miRNA. Our everyday laboratory experience of using northern
blots to detect small RNA, such as siRNA and miRNA, with UV
cross-linking had lead to numerous frustrations with lack of
sensitivity and reproducibility. We could not improve the UV
cross-linking procedure by varying UV dose and so we explored
alternative methods for cross-linking. We adopted the premise that
cross-linking would be optimal if all the nucleotides of a small RNA
remained accessible for hybridization to a complementary probe.
Therefore, we investigated cross-linking through the terminal ends
of a small RNA. The RNase III-type nuclease activity involved in the
production of small regulatory RNA, such as miRNA and siRNA,
dictate a 5-terminal monophosphate and a 2, 3 cis-diol at the
3-terminus8. Different 3-end modifications of small RNA have
been described in mammals, insects and plants9,10, so we thought
that cross-linking through the 5-terminal monophosphate would
provide the most flexible technique to immobilize many different
small RNA.
1-Ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC) has
been used for cross-linking synthetic 5-phosphorylated oligodeox-
yribonucleotides to primary amines on different solid substrates
and this resulted in enhanced hybridization compared with nitro-
cellulose-bound DNA1113. We tested whether EDC could be used
for efficient cross-linking of small RNA from biological samples to
commonly available nylon membrane (Hybond NX) following

NATURE PROTOCOLS | VOL.3 NO.6 | 2008 | 1077

 PROTOCOL
transfer from a dPAG and found that detection of small RNA
isolated from plant or mammalian sources can be enhanced by up
to 50-fold in comparison with UV cross-linking. Different regimes
of hybridization conditions, including a range of different types of
complementary probes, showed similar enhancement of detection
after EDC cross-linking14. Our data was consistent with EDC
facilitating the formation of a covalent phosphoramidate bond
between 5-terminal phosphates on small RNA and primary amine
groups on the surface of the nylon membrane. Non-5-phosphory-
lated RNA should also be detectable using this method if they are
first phosphorylated (e.g., with T4 polynucleotide kinase). The
enhanced detection we observed was significant for RNA molecules
o40 nt in length. No great enhancement was detected for larger
2008 Nature Publishing Group http://www.nature.com/natureprotocols
molecules, for example, miRNA precursors B70 nt (ref. 14). From
this, we infer that the enhanced detection afforded by EDC cross-
linking declines as the size of the RNA of interest increases. Thus,
owing to its greater simplicity, UV cross-linking would remain the
method of choice for RNA greater than B70 nt.
Here, we have described the small RNA northern blotting
procedure as we use in it in our laboratory (see Fig. 1 for an
overview of the protocol, or Supplementary Fig. 1 online, which
provides more detail and can be used as a quick-reference proto-
col). We describe only the preparation of northern blots here and
would suggest that subsequent hybridization of probes to these
blots to detect specific RNAs be carried out using conventional
probe preparation and hybridization conditions. There are many
methods of both probe preparation and hybridization described in
general molecular biology laboratory manuals, such as Sambrook
and Russell15, or in information provided by commercial manu-
facturer(s) of the reagents used in northern blotting. In principle,
any methods previously used for UV-cross-linked blots should be
equally applicable for EDC-cross-linked blots.
Experimental design
RNA sample material. A wide range of reagents and protocols to
extract RNA, from plant or animal tissue, are available to isolate
good-quality RNA. We typically use TRI-reagent (Sigma) but have
successfully used the original Chomczynski and Sacchi protocol16,
RNAbee (Biogenesis) and Trizol (Invitrogen) to isolate intact RNA,
following manufacturers recommended protocols. We have used
total and low-molecular weight preparations of RNA (prepared as
described by Hamilton and Baulcombe)17 and found that cross-
linking with EDC greatly enhances detection of small RNA from
both types of preparation14. Presumably, there will be a limit to the
short RNA-binding capacity of nylon membrane and this might be
approached when using highly enriched fractions. It has been
suggested that UV cross-linking exploits free amine groups on
nylon membrane5. Therefore, EDC cross-linking is likely to have a
similar upper binding limit as UV cross-linking.
Choice of buffer. Tris has primary amine groups and, to avoid any
potential reaction with EDC, we use MOPSNaOH (pH 7) to
buffer our dPAGs instead of conventional TrisBorateEDTA
(TBE). MOPSNaOH (pH 7)-buffered gels appear to be slightly
less denaturing than TBE gels and often we find the longer
pre-miRNA molecules (B70 nt) run faster than expected at B50 nt.
To get accurate sizing of the larger molecules, the MOPSNaOH
(pH 7)-buffered gels can be run at higher temperatures to aid
denaturation. We have observed that pre-miRNA run at their
1078 | VOL.3 NO.6 | 2008 | NATURE PROTOCOLS
M
dPAG
RNA separation
3MM Whatman
ve electrode
Electroblot
dPAG
Transfer RNA
to nylon
+ve electrode Nylon membrane
RNA on nylon
EDC Cross-link
immobilize RNA
Incubated at 60 C
3MM Whatman
for <2 h
saturated with EDC
Predicted
presentation
RNA on nylon
Rinse cross-linked blot in excess dH2O
Northern blot can be directly hybridized
or air-dried and stored
Figure 1 | Small RNA northern blotting protocol overview. Chemical 1-ethyl-
3-(3-dimethylaminopropyl) carbodiimide (EDC) cross-linking facilitates the
production of a covalent bond between the terminal phosphate of the RNA to
a free amino group on the nylon membrane resulting in the small RNA
becoming tethered by one end. This allows the full sequence of the
small RNA to be accessible to hybridizing probes in contrast to UV cross-
linking which may lead to consumption of uridine residues during the cross-
linking process and hence compromise the length of the sequence available
for hybridization (see also Supplementary Fig. 1 for additional protocol
details). dPAG, denaturing polyacrylamide gel. M, g[32P] ATP-labeled
Decade marker.
predicted size when MOPSNaOH (pH 7) gels are run hot
with buffer preheated (5055 1C) at 400 V (run time is reduced
to B2 h). We have not thoroughly compared the efficiency of EDC
cross-linking post-MOPSNaOH (pH 7) or TBE-buffered dPAGE
directly. Others have separated RNA using TBE-buffered dPAGE
and then satisfactorily used our EDC cross-linking protocol18, but
no comparison with using a MOPS system was described.
Use of radioactive size markers. Running g[32P] ATP-labeled
Decade marker RNA is of obvious importance for the correct
estimation of the size of sample RNAs. They also provide a very
useful yardstick for monitoring the progress of the protocol:
 Imaging the marker before transfer (see Step 15) allows assess-
ment of the quality of the gel electrophoresis.
 Hand-held Geiger monitoring of the membrane (see Step 21)
allows simple monitoring of the progress of the blot and helps to
optimize the transfer time.
 Imaging of the membrane after washing away EDC solution and
comparing this with the image of the gel allows one to estimate
the degree of RNA loss up to this point and whether it is
disproportionate for particular sizes of RNA.

Figure 2 | Enhanced detection of short RNA with 1-ethyl-3-(3-
dimethylaminopropyl) carbodiimide (EDC) cross-linking versus UV cross-
linking. Identical 20 mg aliquots of total RNA from undifferentiated murine
embryonic stem cells were loaded in multiple lanes of the same denaturing
polyacrylamide gel (dPAG). The same volume of g[32P] ATP-labeled Decade
markers were loaded in each adjacent lane. After electrophoresis and transfer
to one sheet of nylon membrane, strips containing one individual sample RNA
lane and one adjacent marker RNA lane were cut and each was cross-linked
with 1,200 mJ UV (auto-cross-linking setting) or EDC at 60 1C for 15 min, 1 or
2 h. All membranes were then hybridized in the same hybridization solution
containing a probe complementary to mmu-miR-292-3p. The improvement of
detection of the microRNA after EDC cross-linking for 2 h was 30-fold. Blue
arrow, bromophenol blue dye front from 6 loading dye. Green arrow, xylene
cyanol dye in Decade marker loading dye. Black arrow, mmu-miR-292-3p.
2008 Nature Publishing Group http://www.nature.com/natureprotocols
M, g[32P]ATP-labeled Decade marker. S, RNA sample.
 Comparison of the marker image after hybridization and wash
with that obtained immediately after cross-linking allows one to
determine loss of RNA during the (usually) former steps.
 If one is reusing the blot for multiple probes (i.e., by sequential
probing, imaging and stripping), one can monitor the useful life
span of the blot by carefully comparing marker RNA images in
serial images (also accounting for radioactive decay). This is useful
as probe stripping usually employs high temperatures, which in
our experience accelerates loss of RNA from a membrane.
Choice of membrane. We routinely use the neutral nylon
membrane, Hybond NX (Amersham/Pharmacia). In principle,
other nylon 6,6 neutral membranes from different manufactures
should work as effectively, but we have not tested this. We also
tested the positively charged nylon membrane, Hybond N+ (Amer-
sham/Pharmacia), and found that EDC cross-linking afforded only
slightly enhanced short RNA detection compared with UV.
Optimization of EDC cross-linking parameters for your small
RNA of interest. The strategy described in Figure 2 legend can be
used to optimize EDC cross-linking conditions to achieve the best
detection levels for any small RNA of interest. In brief, identical
amounts of the same RNA sample are loaded on a 1015% dPAG.
The same volume of g[32P] ATP-labeled Decade markers is loaded
in each adjacent lane. After electrophoresis and transfer to one sheet
of nylon membrane, strips containing one individual sample RNA
lane and one adjacent marker RNA lane are cut and each cross-
MATERIALS
REAGENTS
. TRI-reagent (Sigma, cat. no. T9424) ! CAUTION Contains phenol and
guanidinium thiocyanate. Wear lab coat, gloves and safety spectacles when
handling.
. Diethylpyrocarbonate (DEPC; Sigma, cat. no. D5758)
. Deionized formamide (Sigma, cat. no. F9037)
. Acrylamide/bis (19:1) (Sigma, cat. no. A2917) ! CAUTION Neurotoxic.
Wear lab coat, gloves and safety spectacles when handling.
. Urea (Fluka, cat. no. 51461)
. Tetramethylethylenediamine (Sigma, cat. no. T9281)
. Ammonium persulfate (Sigma, cat. no. A3678)
. MOPS (Roche, cat. no. 11 124 684 001)
. Ethidium bromide (EtBr; Sigma, cat. no. E1510) (see REAGENT SETUP)
! CAUTION Intercalating agent. Carcinogenic and mutagenic. Wear lab coat,
gloves and safety spectacles when handling.
. Nylon membrane (Hybond NX; Amersham/Pharmacia, cat. no. RPN303T)
PROTOCOL
EDC
15 min 1h 2h UV MS MS MS MS
rRNA
and tRNA
60 nt
50 nt
40 nt
30 nt
20 nt
10 nt
Northern blot EtBr-stained 15% dPAG
linked with EDC at a range of times and temperatures to be tested,
for example, 60 1C for 15 min, 1 or 2 h (UV cross-linking can be
included to see the overall benefit of using EDC cross-linking). All
membranes are then hybridized in the same hybridization solution
containing a probe complementary to the small RNA of interest.
The improvement in detection, using EDC cross-linking, of the
small RNA of interest can be gauged by comparing the posthy-
bridization signal from each condition tested.
DNA oligo dot blot control. In small RNA northern blot
hybridizations, we routinely include a separate membrane onto
which we have pipetted dilutions of DNA oligonucleotides that are
complementary to our probe (DNA oligo dot blot: see TROU-
BLESHOOTING). These are simple to prepare but are very useful
positive controls that should give a strong positive signal if the
probe has been synthesized correctly and if hybridization condi-
tions were sufficiently tolerant. As DNA oligos are manufactured
least expensively without terminal phosphates, we use UV cross-
linking to fix these to nylon membranes. If the
DNA oligo dot blot control is positive and the northern blot
has weak or no signal, one can then infer that something has
gone wrong with the production of the small RNA northern blot
(see TROUBLESHOOTING).
. 3MM Whatman chromatography paper (Schleicher & Schuell,
cat. no. 3030 917)
. Decade RNA markers (Ambion, cat. no. 7778)
. g[32P] ATP (Amersham/Pharmacia, cat. no. AA0018) ! CAUTION Wear lab
coat, gloves and safety spectacles when handling.
. 1-Methylimidazole (Sigma-Aldrich, cat. no. M50834)
. 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC: Sigma, cat. no. 7750)
(see REAGENT SETUP) m CRITICAL All reagents listed here are the ones we
use, but these or equivalent reagents are available from other manufacturers.
EQUIPMENT
. Vertical gel system (protean II Bio-Rad system with gel plates 16  20 cm;
Bio-Rad) m CRITICAL We use equipment designed principally for protein
analysis. Similar equipment is available from a number of different
manufacturers.
. Semi-dry electroblotter (SciPlas) (see EQUIPMENT SETUP) m CRITICAL
We have previously used a completely wet electrotransfer unit satisfactorily,
NATURE PROTOCOLS | VOL.3 NO.6 | 2008 | 1079
 PROTOCOL
but this was slightly more difficult to set up properly and required several
liters of transfer buffer.
REAGENT SETUP
DEPC-treated water Add 1 ml DEPC to 1,000 ml distilled water. Shake
vigorously to mix. Incubate, in fume hood, at room temperature
(20 1C) overnight. Autoclave to inactivate residual DEPC and cool before use.
DEPC-treated water can be stored at room temperature. Providing the
DEPC-treated water remains contamination free, it can be stored and
used until finished.
503 MOPSNaOH (pH 7) 1 M MOPS prepared in distilled water (pH 7)
with NaOH. Store at 4 1C. This should be discarded if microbial and/or
ribonuclease contamination is suspected or once the buffer acquires a
noticeable yellow color.
dPAG Denaturing polyacrylamide (19:1) gel prepared with 1015% acrylamide, 7
M urea and buffered with 20 mM MOPSNaOH (pH 7) (see Table 1). For detailed
2008 Nature Publishing Group http://www.nature.com/natureprotocols
63 Loading dye 0.25% (wt/vol) Bromophenol blue dissolved in 30% (vol/vol)
glycerol. This can be stored at 20 1C for several months. Providing the
6 loading dye remains contamination free, it can be stored and used until
finished. Optional: we have found that including 1 MOPS buffer (the same as
the electrophoresis running buffer) in the samples can sometimes improve the
resolution of small RNA, such as miRNA.
EtBr working solution 0.5 lg ml1 Prepare fresh, dilute stock (10 mg ml1)
with 20 mM MOPSNaOH (pH 7) to use at 0.5 mg ml1.
EDC cross-linking solution 0.16 M EDC prepared in 0.13 M 1-methylimi-
dazole at pH 8. Add 245 ml of 12.5 M 1-methylimidazole to 9 ml DEPC-treated
water. Adjust pH to 8.0 with 1 M HCl (usually requires B300 ml). This can be
prepared 12 h before use and kept at room temperature. Immediately
before use, add 0.753 g EDC and make the volume up to 24 ml with DEPC-
treated water. This gives a working solution of 0.16 M EDC in 0.13 M
1-methylimidazole at pH 8 and is sufficient to saturate 320 cm2 (20  16 cm2)

instructions on how to prepare and pour dPAGs, refer to Sambrook and Russell15. of 3 MM Whatman paper. We use the EDC cross-linking solution immediately
c[32P] ATP-labeled Decade RNA markers Follow manufacturers and we have not tested how storage affects it.
(Ambion) instructions for labeling. ! CAUTION Decade markers radiolabeled EQUIPMENT SETUP
with g[32P] ATP work with appropriate shielding. Semi-dry electroblotter Place on ice to avoid overheating during transfer.

PROCEDURE
Extraction of RNA
1| Isolate intact RNA using TRI-reagent (or other available reagents; see Experimental design), following manufacturers
instructions, and resuspend RNA in DEPC-treated water. Typically, we attempt to resuspend in volumes of DEPC-treated water
that will give concentrations of B12 mg ml1 of RNA.

Assessment of quantity and quality of RNA

2| Measure RNA concentration by spectrophotometry at 260 nm (e.g., with Nanodrop spectrophotometer).
m CRITICAL STEP Typically, we require between 10 and 20 mg of total RNA to run per lane of the dPAG, but this can be adjusted
depending on the abundance of the RNA of interest.
PAUSE POINT RNA resuspended in DEPC-treated water can be stored at 80 1C. We have successfully used samples that have
been in storage for 41 year in our laboratory.
3| Add deionized formamide to a final concentration of 50% to help maintain RNA integrity19.
m CRITICAL STEP This dilutes your initial RNA sample concentration. We typically use 100% deionized formamide and thus a
volume equal to the DEPC-treated water used to resuspend the RNA (Step 1) is required. As a result, the concentration of the sample
(Step 2) is halved.
PAUSE POINT RNA resuspended in 50% deionized formamide can be stored at 20 1C. We have successfully used samples
that have been in storage for 42 years in our laboratory.
4| Check the quality of RNA by conventional agarose gel electrophoresis followed by EtBr staining. We typically run 0.51 mg
of sample RNA resuspended in 50% deionized formamide to assess quality. For detailed instructions of how to pour, run and
stain agarose gels refer to Sambrook and Russell15. The clear observation of distinct, 18S and 28S, rRNA and tRNA bands is a
good indicator that the RNA sample is of good quality and not degraded (Fig. 3a).


Prepare dPAG TIMING 22.5 h
5| We use the Bio-Rad protean II vertical gel system with 20  16 cm2 plates and 1.5-mm spacers. Assemble the plates
following manufacturers instructions. Make sure the glass plates are clean before assembly. In our laboratory, it is usual practice
to clean glass plates first with 70% ethanol followed by a rinse in distilled water before use.
6| Use (19:1) acrylamide/bis to prepare 10 or 15% denatur-
ing gel mix (as described in REAGENT SETUP and Table 1) and TABLE 1 | dPAG recipes.
pour gel. Insert well comb that will accommodate RNA sample 10% 15%
to be loaded. This may vary depending on the concentration of
40% Acrylamide (19:1) (ml) 15 22.5
the RNA sample. For detailed instructions on how to prepare Urea (g) 25.2 25.2
and pour dPAGs, refer to Sambrook and Russell15. 50 MOPSNaOH (pH 7) (ml) 1.2 1.2
m CRITICAL STEP 60 ml of Mix is sufficient to prepare a gel Water (ml) Make volume Make volume
200  160  1.5 mm. up to 60 ml* up to 60 ml*
PAUSE POINT Allow at least 2 h for gels to set before use. 10% Ammonium persulfate (ml) 360 360
Alternatively, gels can be wrapped in Saran and left overnight Tetramethylethylenediamine (ml) 21 21
at room temperature to set. *60 ml is sufficient to pour a gel 200  160  1.5 mm.

1080 | VOL.3 NO.6 | 2008 | NATURE PROTOCOLS

 PROTOCOL


Running dPAG TIMING 216 h
a
1 2
b
1 2

7| Remove well comb and prerun gel in running buffer,

20 mM MOPSNaOH (pH 7) (prepared as described in REAGENT
High abundance
SETUP) at desired voltage for at least 10 min. rRNA and tRNA
m CRITICAL STEP We prerun and run gels at the same 28S 60 nt
voltage. Gels can be run at high voltage (300400 V), in which 18S 50 nt
case the run will take B24 h, depending on polyacrylamide 40 nt
concentration. Alternatively, gels can be run overnight 5S
(B16 h) at 6090 V. tRNA 30 nt

8| During gel prerun, prepare RNA samples (typically, we run

1020 mg of RNA already resuspended in 50% deionized 20 nt
formamide, see Steps 13) by adding 6 loading dye
2008 Nature Publishing Group http://www.nature.com/natureprotocols

(prepared as described in REAGENT SETUP) to samples and heat

denature at 95 1C for 15 min followed by snap cooling on ice.
m CRITICAL STEP As well as inhibiting renaturation 10 nt Bromophenol
of RNA structures, snap cooling samples also makes sample blue dye front
loading easier.
9| We also run g[32P] ATP-labeled Decade RNA markers
(as described in REAGENT SETUP) for sizing bands and as a Figure 3 | Examples of RNA preparations suitable for northern blot analysis.
control (see Experimental design and TROUBLESHOOTING). (a) Total RNA assessed on a 1% agarose gel: Total RNA isolated from mouse
embryonic stem (ES) cells resuspended in 50% deionized formamide ran on a
Follow manufacturers instructions to prepare g[32P]
1% agarose gel buffered with 0.5 TrisBorateEDTA (TBE). Each lane (1 and
ATP-labeled Decade markers. 2) loaded with 1 mg of total RNA. (b) Total RNA assessed on a 15% dPAG:
10| Just before loading RNA samples (i.e., while they are on Total RNA isolated from ES cells resuspended in 50% deionized formamide ran
on a 15% dPAG gel buffered with 20 mM MOPSNaOH pH 7. Each lane (1 and
ice after denaturingStep 8), flush wells thoroughly with 2) is loaded with 10 mg of total RNA.
running buffer [20 mM MOPSNaOH (pH 7)] injected using a
syringe and needle.
m CRITICAL STEP This step is necessary because as soon as the well comb is removed from the gel and running buffer is overlaid,
urea begins to leach from the gel and accumulates in a barely visible layer over the bottom of each well. In our experience, loading
sample on top of this negatively affects the resolution of RNA during electrophoresis.
11| Load RNA samples, typically 1020 mg of RNA (see Step 8), and markers, as recommended by the manufacturer
(see Step 9), onto the gel and run the gel at the desired voltage (see Step 7).
m CRITICAL STEP On a 15% dPAG, the blue bromophenol loading dye front runs at the equivalent rate to a B10-nt long RNA. The
dye front for each sample should run in a near perfect straight line indicating a uniform electric field across the gel. Disturbance of
the dye front is a possible indicator that either the gel or the run has not been uniform.
12| Disassemble the gel from the gel apparatus. Remove one glass plate from the gel, mark orientation of gel by cutting corner.
! CAUTION Running buffer is contaminated with g[32P] ATP-radiolabeled marker and ureadiscard appropriately following
safety guidelines.


Assessment of gel run TIMING 30 min
13| At this point, the gel is sitting on glass plate, so you can use this as support to handle gel or, with care, the gel can be lifted
from the plate by hand. Place the gel in freshly prepared EtBr solution (prepared as described in REAGENT SETUP) for 510 min.
! CAUTION Do not use the electrophoresis running buffer from the gel run (Step 11) to prepare the EtBr staining solution.
Although this might be economical and convenient, we have occasionally experienced problems with the subsequent electrophoretic
transfer of the RNA from the gel to the nylon membrane that we have attributed to using recycled running buffer at this stage.
m CRITICAL STEP Flimsiness of gel increases with decreasing acrylamide concentration.
m CRITICAL STEP Steps 1315 are optional. If one is confident with dPAGE and the integrity of the sample, then one can proceed
directly from Step 12 to 16.
14| Remove gel from EtBr solution and then scan stained gel to assess the run. We use a Fuji FLA5000 scanner, but a UV light
box can also be used. A number of discrete, strongly staining bands 460 nt should be visible. These include tRNA (70110 nt)
and 5S RNA (B120 nt). These stained RNA can be used to assess the relative loading of RNA samples retrospectively if a
quantitative digital image is obtained. The clarity and sharpness of the tRNA bands is a good indicator of RNA quality and
integrity during the electrophoresis (Fig. 3b).
m CRITICAL STEP If the integrity of the tRNA bands is diminished (may stain weakly with a heavily stained smear down the lane),
then one should suspect the quality of the RNA sample and recognize that data acquired following the protocol to completion may
not be reliable.

NATURE PROTOCOLS | VOL.3 NO.6 | 2008 | 1081

 PROTOCOL

15| Thereafter, wrap the gel in Saran and place, face up, with 100 nt a b c d
imaging screen. Expose to detect signal from the g[32P] 90 nt
ATP-labeled Decade marker; typically, an exposure of 35 min 80 nt
is sufficient. Separation of the marker into its respective 70 nt
fragments is a good indicator that gel electrophoresis has also 60 nt
successfully separated the RNA samples (Fig. 4). This is useful 50 nt

if EtBr staining of RNA samples is weak. 40 nt

Transfer of RNA from dPAG onto nylon membrane 30 nt

 TIMING 1 h
16| Prewet nylon membrane (200  160 mm) with
distilled water and place it on the gel. Layer three sheets 20 nt
2008 Nature Publishing Group http://www.nature.com/natureprotocols

(200  160 mm) of 3MM Whatman (prewet with distilled

water) on top of the nylon membrane. With each layer added,
roll out any bubbles between the gel, membrane and Whatman.
10 nt
17| Lift the gel, nylon membrane and Whatman, and place on
the positive electrode of the semi-dry electroblotter with gel
uppermost. Add another three sheets of 3MM Whatman (prewet
Figure 4 | Use of g[32P] ATP-labeled Decade markers as a control for small
with distilled water) on top of the gel; with each layer added,
RNA northern blotting procedure. Signal from g[32P] ATP-labeled Decade
roll out bubbles. markers ran on 15% dPAG gel buffered with 20 mM MOPSNaOH pH 7 and then
m CRITICAL STEP The exact composition of the blot (e.g., the transferred onto nylon membrane followed by 1-ethyl-3-(3-
number of layers of 3MM) may be varied according to users dimethylaminopropyl) carbodiimide (EDC) cross-linking: (a) gel image,
preference; however, the key issue is to ensure that the mem- (b) gel image after transfer, (c) nylon membrane image after transfer and
brane is tightly and uniformly apposed to the gel and is between (d) nylon membrane image after EDC cross-linking.
the gel and the positive electrode.
18| Place negative electrode lid on to transfer sandwich: 3MMgelmembrane3MM (see Fig. 1, and Supplementary Fig. 1,
online).
19| Place transfer unit on ice and transfer RNA by applying 20 V for 3060 min.
m CRITICAL STEP Little or no g[32P] ATP-labeled Decade marker RNA (i.e., radioactivity) should pass through the membrane
to the 3MM or electrode. It is possible to over-transfer by blotting for too long. This will be indicated by significant amounts
of radioactive marker RNA being deposited on the 3MM below the membrane and/or the positive electrode of the transfer
unit. In our experience, over-transfer also leads to poor signal from the sample RNA. Simply peeling back the gel partially from the
membrane and monitoring each allows one to check the progress of the gel nondisruptively. The gel is gently rolled back into
place after this check and transfer resumed.
20| After transfer is complete, remove layers of 3MM Whatman sitting on top of gel without disturbing the gel and membrane.
Mark the position of the wells on the nylon membrane using pencil (does not run or wash off in subsequent hybridization steps).
21| Remove the gel and keep it for post-transfer exposure (see Step 23). The majority of the bromophenol blue should be on
the membrane, an indication that transfer has occurred. Additionally, the gel and membrane can be monitored using a Geiger
counter; the majority of counts should be on the membrane.

22| Carefully lift the membrane and avoid all contact with the side onto which the RNA has been deposited.
23| Optional: Wrap both the gel and membrane in Saran and expose both to an imaging screen for the same time as in Step 15.
Upon successful transfer, the majority of signal from the g[32P] ATP-labeled Decade marker should now be on the membrane
(compare Fig. 4b with c).
m CRITICAL STEP In our experience, wrapping the membrane in Saran wrap does not disturb the position of the transferred RNA.


EDC cross-linking TIMING 15 min to 2 h
24| Place the damp membrane with RNA side face up onto 3MM (210  170 mm) saturated in freshly prepared cross-linking
EDC reagent (prepared as described in REAGENT SETUP).

25| Wrap membrane and 3MM in Saran and incubate at 5060 1C for up to 2 h. We have found that the optimum cross-linking
time can vary for different short RNA and so can be optimized by the user (see Fig. 2). A good starting point is 1 h at 60 1C.
m CRITICAL STEP As a precaution, the RNA side should not be in direct contact with the saturated 3MM and excess cross-linking
reagent should not be washed over the RNA face of the membrane.
26| After cross-linking, rinse membrane in excess RNase-free distilled water to remove any residual cross-linking solution.

1082 | VOL.3 NO.6 | 2008 | NATURE PROTOCOLS

 PROTOCOL

27| Optional: To check for any loss of RNA during EDC cross-linking (Steps 2426), expose cross-linked northern blot from Step 27 to
imaging screen (as described in Step 15) and compare Decade marker signal to post-transfer signal (Step 23) (compare Fig. 4c with d).
PAUSE POINT After the membrane has been rinsed to remove the EDC solution, it can be air-dried and stored wrapped in
Saran at 20 1C until required for hybridization.
? TROUBLESHOOTING
 TIMING
Steps 14, RNA extraction, quantification, assessment of integrity: depends on the number of samples, can be stopped and resumed
Steps 5 and 6, preparation of dPAG: 20 min, allow to set for a minimal 2 h or can be poured on the day before use
Steps 715, prerun and running gel: 2 h or overnight 16 h
Steps 1623, transfer: 1 h
Steps 2427, EDC cross-linking: up to 2 h
m CRITICAL Steps 727 should be done sequentially
2008 Nature Publishing Group http://www.nature.com/natureprotocols

? TROUBLESHOOTING
Troubleshooting is generally well established for northern blotting15. Table 2 lists problems that might particularly arise from
the use of EDC rather than UV cross-linking after a successful separation of RNA samples using dPAGE. We also briefly mention
problems that can be mistaken for a failure in EDC cross-linking, but which may be due to steps that follow successful cross-
linking, for example, probe synthesis and hybridization. Comprehensive protocols and troubleshooting guides for these particular
steps are available elsewhere. As part of our controls, we run g[32P] ATP-labeled Decade RNA markers; these allow us to assess
each step of the protocol before and after cross-linking (Fig. 4, also see Experimental design and Table 2). To evaluate the
benefit of EDC cross-linking for novel small RNA, we run samples in duplicate and cross-link them with EDC or UV for comparison.

TABLE 2 | Troubleshooting table.

Problem
g[32P] ATP-labeled Decade RNA Markers not
deposited on membrane. In a successful
transfer, 490% of the marker RNA should
leave the gel and bind to the nylon
membrane. Note: monitor efficiency of trans-
fer by comparing marker signal in denaturing
polyacrylamide gel (dPAG) before and
after transfer with blot before cross-linking.
Failure in transfer will also result in
poor or absent signal on the blot (see below)
Marker RNA transfer successful (majority of
g[32P] signal from gel on nylon membrane
before cross-linking, see above), but lack of
signal on northern blot and poor (50%)
retention of g[32P] ATP-labeled Decade RNA
markers and control DNA oligo dot blot gives
signal. Note: monitor by comparing g[32P]
ATP-labeled Decade RNA markers signal
on nylon membrane before and after
cross-linking
Possible reason
Ribonuclease contamination of a solution or
equipment surface. Gel is not stained in fresh
solution of 1 MOPSNaOH buffer
Membrane is not positioned between
positive electrode and gel
Insufficient time allowed for transfer
(markers remain in gel) or too much time
allowed for transfer (markers detected on
3MM below nylon membrane)
EDC cross-linking failed due to RNase
contamination, reagent quality
or reagent preparation, although probe
synthesis and hybridization were successful
EDC is decomposed
1-Methylimidazole solution at wrong pH
EDC cross-linking buffer prepared too early
Nylon membrane decomposed
Not enough EDC solution used, membrane
drying out
Northern blot not EDC cross-linked for the
correct time or at the wrong temperature
Solution
Ensure solutions for transfer are freshly
prepared and nuclease free
Ensure setup is correct and equipment is
working
Carry out a test transfer monitoring labeled
markers with Geiger counter to optimize transfer
time for your laboratorys equipment
Make sure solutions and reagents for cross-
linking are nuclease-free. Store chemicals under
suppliers recommended conditions
Buy fresh reagent. We purchase 1 g aliquots
of EDC powder from Sigma and store these at
20 1C
Check pH meter is accurate
In our experience, EDC should be added
immediately before use, although the diluted
1-methylimidazole pH 8 solution can be
prepared a couple of hours earlier and kept at
room temperature
Nylon membrane should be stored protected
from light, in a sealed bag, at room temperature.
The quality of membrane deteriorates over time
Make sure 3MM is saturated, and ensure blot and
3MM wrapped well
Check equipment and optimize conditions for
EDC cross-linking

NATURE PROTOCOLS | VOL.3 NO.6 | 2008 | 1083

 PROTOCOL

TABLE 2 | Troubleshooting table (continued).

Problem Possible reason Solution
Suspected artifactual bands on final image Over-cross-linking (i.e., for too long and/or Optimize conditions for EDC cross-linking
and/or high background signal within lanes at too great a temperature) can cause this reduce time and/or temperature of EDC
cross-linking

This problem can also be caused by Seek advice from troubleshooting guides for
inappropriate hybridization conditions for northern blot hybridization procedures
the probe used

g[32P] ATP-labeled Decade RNA marker signal EDC cross-linking successful but probe Ensure reagents are nuclease free
maintained but lack of signal on northern synthesis and hybridization is poor or failed.
2008 Nature Publishing Group http://www.nature.com/natureprotocols

blot and control DNA oligo dot blot gives
weak or no signal. Note: monitor by compar-
ing g[32P] ATP-labeled Decade RNA markers
signal on nylon membrane before and after
hybridization
RNase contamination of probe synthesis
reagents or hybridization solutions. The
northern blot may have been exposed to
nucleases rendering it unusable for
detection of RNA
Check integrity of northern blot by
hybridizing to detect known RNA that is
expressed in sample being analyzed

DNA oligo dot blot failed Check concentration and integrity of DNA oligos

Check whether UV cross-linker is working

Seek advice from troubleshooting guides for

northern blot hybridization procedures

ANTICIPATED RESULTS
The enhanced sensitivity achieved using EDC cross-linking will improve the detection of small RNA and thus correlation of
specific RNA with biological processes. Figures 24 demonstrate the expected results from each step of the small northern blot
protocol. Success of each step is important for the final enhanced detection of small RNA with EDC cross-linking. Figure 3
demonstrates resolution of total RNA after (a) agarose gel electrophoresis and (b) dPAGE. Figure 4 demonstrates the signal
from g[32P] ATP-labeled Decade marker in the dPAG after (a) electrophoresis, (b) gel and (c) nylon membrane after transfer and
(d) nylon membrane after cross-linking. Figure 2 demonstrates the enhanced detection of short RNA (mmu-miR-292-3p) with
EDC cross-linking versus UV cross-linking.

Note: Supplementary information is available via the HTML version of this article.
Published online at http://www.natureprotocols.com
Reprints and permissions information is available online at http://npg.nature.com/
reprintsandpermissions
1. Jiang, J., Lee, E.J., Gusev, Y. & Schmittgen, T.D. Real-time expression profiling of
microRNA precursors in human cancer cell lines. Nucleic Acids Res. 33, 53945403
(2005).
2. Lee, Y. et al. The nuclear RNase III Drosha initiates microRNA processing. Nature
425, 415419 (2003).
3. Berezikov, E., Cuppen, E. & Plasterk, R.H. Approaches to microRNA discovery. Nat.
Genet. 38 (suppl.): S2S7 (2006).
4. Lee, Y., Jeon, K., Lee, J.T., Kim, S. & Kim, V.N. MicroRNA maturation:
stepwise processing and subcellular localization. EMBO J. 21, 46634670
(2002).
5. Church, G.M. & Gilbert, W. Genomic sequencing. Proc. Natl. Acad. Sci. USA 81,
19911995 (1984).
6. Saito, I., Sugiyama, H., Furukawa, N. & Matsuura, T. Photoreaction of thymidine
with primary amines. Application to specific modification of DNA. Nucleic Acids
Symp. Ser. 6164 (1981).
7. Khandjian, E.W. UV crosslinking of RNA to nylon membrane enhances
hybridization signals. Mol. Biol. Rep. 11, 107115 (1986).
8. Elbashir, S.M., Lendeckel, W. & Tuschl, T. RNA interference is mediated by 21- and
22-nucleotide RNAs. Genes Dev. 15, 188200 (2001).
9. Vagin, V.V. et al. A distinct small RNA pathway silences selfish genetic elements in
the germline. Science 313, 320324 (2006).
10. Yang, Z., Ebright, Y.W., Yu, B. & Chen, X. HEN1 recognizes 21-24 nt small RNA
duplexes and deposits a methyl group onto the 2 OH of the 3-terminal
nucleotide. Nucleic Acids Res. 34, 667675 (2006).
11. Ghosh, S.S. & Musso, G.F. Covalent attachment of oligonucleotides to solid
supports. Nucleic Acids Res. 15, 53535372 (1987).
12. Gingeras, T.R., Kwoh, D.Y. & Davis, G.R. Hybridization properties of immobilized
nucleic acids. Nucleic Acids Res. 15, 53735390 (1987).
13. Godfroid, E., Min Hu, C., Humair, P.F., Bollen, A. & Gern, L. PCR-reverse line blot
typing method underscores the genomic heterogeneity of Borrelia valaisiana
species and suggests its potential involvement in Lyme disease. J. Clin. Microbiol.
41, 36903698 (2003).
14. Pall, G.S., Codony-Servat, C., Byrne, J., Ritchie, L. & Hamilton, A. Carbodiimide-
mediated cross-linking of RNA to nylon membranes improves the detection of
siRNA, miRNA and piRNA by northern blot. Nucleic Acids Res. 35, e60 (2007).
15. Sambrook, J. & Russell, D.W. Molecular Cloning: A Laboratory Manual 3rd edn.
Vols. 13 (Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York,
2001).
16. Chomczynski, P. & Sacchi, N. The single-step method of RNA isolation by acid
guanidinium thiocyanate-phenol-chloroform extraction: twenty-something years
on. Nat. Protoc. 1, 581585 (2006).
17. Hamilton, A.J. & Baulcombe, D.C. A species of small antisense RNA in
posttranscriptional gene silencing in plants. Science 286, 950952
(1999).
18. Gy, I. et al. Arabidopsis FIERY1, XRN2, and XRN3 are endogenous RNA silencing
suppressors. Plant Cell 19, 34513461 (2007).
19. Chomczynski, P. Solubilization in formamide protects RNA from degradation.
Nucleic Acids Res. 20, 37913792 (1992).

1084 | VOL.3 NO.6 | 2008 | NATURE PROTOCOLS