Logo of narLink to Publisher's site
Nucleic Acids Res. 2007 Apr; 35(8): e60.
Published online 2007 Apr 2. doi:  10.1093/nar/gkm112
PMCID: PMC1885651

Carbodiimide-mediated cross-linking of RNA to nylon membranes improves the detection of siRNA, miRNA and piRNA by northern blot


The northern blot, or RNA gel blot, is a widely used method for the discovery, validation and expression analysis of small regulatory RNA such as small interfering RNA (siRNA), microRNA (miRNA) and piwi-interacting RNA (piRNA). Although it is straightforward and quantitative, the main disadvantage of a northern blot is that it detects such RNA less sensitively than most other approaches. We found that the standard dose of UV used in northern blots was not the most efficient at immobilizing small RNA of 20–40 nt on nylon membranes. However, increasing the dose of UV reduced the detection of miRNA by hybridization in northern blotting experiments. We discovered that using the soluble carbodiimide, EDC, to cross-link RNA to nylon membranes greatly improved the detection of small RNA by hybridization. Compared to standard UV cross-linking procedures, EDC cross-linking provided a 25–50-fold increase in the sensitivity of detection of siRNA from plants and miRNA or piRNA from mammalian cells. All types of hybridization probes tested benefited from the new cross-linking procedure. Cross-linking was dependent on a terminal phosphate and so, should be applicable to other related categories of small RNA.


Small regulatory RNAs, which direct the suppression of gene expression in eukaryotes, are now the subject of intensive investigation in many model organisms (1,2). A common feature of all such RNA is that their sizes always fall within a narrow, defined range of between 19 and 31 nt (1–3). For some, including the most prominent classes of siRNA and miRNA, that precise size is a direct result of biosynthesis catalysed by RNAse III-like enzymes known as DICERs (4–6). For others, such as piRNA [review by (3)] and repeat associated short interfering RNA (rasiRNA) repeat associated short interfering RNA (7), the mechanism of synthesis is not yet known. Nonetheless, the defined, discrete size of all of these small regulatory RNA species distinguishes them from catabolic fragments of longer cellular RNA and so, is a valuable identifying feature. Moreover, the precise size (i.e. single nucleotide resolution) of a small regulatory RNA has also been shown in certain cases to be indicative of a specific function (3,8).

As the interest in silencing-related small RNA has grown, there have been several improvements in methodology for their identification, validation and expression profiling (9). A widely used method is ‘northern blotting’ or ‘RNA-gel blotting’: denaturing polyacrylamide gel electrophoresis of RNA followed by electrophoretic or capillary transfer of the RNA onto a solid support and then detection by hybridization with a labelled, complementary nucleic acid probe. This procedure is less sensitive than RT-PCR-based approaches and has lower throughput than cDNA cloning, ‘deep’ sequencing or expression analysis with microarrays. However, northern blotting is quantitative, technically straightforward, relatively inexpensive and is often employed to validate small RNA identified by higher throughput methods. It is the most convincing way to display the size distribution of small RNA and can simultaneously display the levels of mature and precursor forms of a miRNA. Small RNA northern blots using candidate-sequence probes are also useful for the discovery of siRNA because these small RNA appear to lack precise precursors analogous to those of miRNA which can be identified a priori by computer-based approaches (10). Other RNA such as piRNA and rasiRNA which do not arise from transcription of obvious inverted repeats are similarly difficult to predict and/or validate by in silico methods (3,7). Small RNA northern blotting can also be used to determine the amount and size of siRNA produced from transgenic constructs created to knockdown gene expression and so help to monitor and troubleshoot these experiments.

The standard small RNA northern blot method for small RNA (11) is identical in principle to that used successfully for many years to study longer RNA species such as mRNA (12). An important step is the UV-induced cross-linking of the sample RNA to a nylon 6,6 membrane support that increases the retention of the RNA during subsequent immersion of the membrane in hybridization, washing and probe-stripping solutions. UV-cross-linking is believed to take place primarily through the thymine bases of DNA, and, by inference, the uridine residues of RNA. Uncharacterized reactive species produced are thought then to form covalent cross-links with free amine groups at the nylon membrane surface (13,14). Although the exact mechanism is not known, it would seem that cross-linking via the nucleotide bases could compromise subsequent hybridization by consuming the required functional groups, reducing the longest length of uninterrupted sequence or limiting the free movement of the target polynucleotide chain. We hypothesized that any of these factors might be particularly problematic for the detection of small RNA such as siRNA and miRNA where even one cross-linked base—presumably the minimum requirement for retention—could compromise hybridization with complementary nucleic acid probes.

Here we show that UV cross-linking is, to an extent, antagonistic to hybridization of small RNA and improved detection of small RNA cannot be achieved simply by increasing the dose of UV. Instead, the UV step can be omitted and the water-soluble carbodiimide, 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC) used to cross-link RNA to nylon membranes at least as efficiently. EDC-cross-linking likely proceeds via a 5′ terminal phosphate and results in an immobilized RNA that is much more amenable to detection by hybridization to complementary nucleic acid probes. We show that greatly improved detection of siRNA, miRNA and piRNA is possible with this modification. Given the similarities in end chemistry of most small regulatory RNA, this improvement to small RNA northern blotting should have widespread applications.


Plant growth conditions

Arabidopsis thaliana (Columbia) were grown under long-day glasshouse conditions at ∼22°C. Flowers were harvested from 5- to 6-week-old plants. Untransformed Nicotiana benthamiana and the transgenic line GFP16c/GFPi (15) were grown from seed for 4 weeks at 25°C under 16 h light/8 h dark in a Sanyo Environmental Test chamber.

Mammalian cell lines, cell culture and tissue

Mouse embryonic stem (ES) and HeLa (ECCAC-Sigma) cells were grown under standard conditions of 5% CO2 at 37°C. Undifferentiated ES cells were grown on gelatinized tissue culture plasticware in high glucose DMEM (GIBCO) supplemented with 20% fetal bovine serum (FBS) (Autogen Bioclear), 0.1 mM non-essential amino acids, 2 mM L-glutamine, 0.06 mM β-mercaptoethanol, 100 U ml−1 penicillin, 100 μg ml−1 streptomycin, and 103 U ml−1of LIF (Chemicon). ES cells were allowed to differentiate spontaneously for 4 days in media containing no LIF.

HeLa cells were grown in EMEM (Sigma) supplemented with 10% FBS (Autogen Bioclear), 0.1 mM non-essential amino acids and 2 mM L-glutamine. All tissue culture reagents were purchased from Invitrogen unless otherwise stated.

Animals were culled using a Schedule 1 method in accordance with the UK Home Office Regulations. Brain, heart, liver and testes tissue were harvested from 4-month-old male rats (Fischer) and 3-month-old male mice (C57BL/6), and mammary tissue was harvested from 3-month-old female mice (BALB/c) and frozen immediately in liquid nitrogen.

RNA extraction


Tissue was frozen in liquid nitrogen, ground to a fine powder with a mortar and pestle, homogenized in TRI-Reagent® (Sigma) and total RNA extracted according to the manufacturer's instructions with the following additional steps. Samples were extracted with an equal volume of Tris/HCl- buffered phenol/chloroform (Sigma) after chloroform extraction and prior to precipitation with isopropanol. After dissolving the nucleic acid pellets in DEPC-treated water, they were treated with TurboDNAse® (Ambion) in the manufacturer's buffer with ∼1 U per 100 μg of RNA at 37°C for 1 h. Samples were then extracted with an equal volume of Tris/HCl- buffered phenol–chloroform (Sigma), precipitated with an equal volume of isopropanol and one-tenth volume of 3 M sodium acetate (pH 6) and re-dissolved in DEPC-treated water.


Cultured mammalian cells were lysed and homogenized directly in TRI-Reagent® (Sigma) and RNA extracted as above with modified manufacturer's instructions. HeLa cell RNA was treated with TurboDNAse® as described above for RNA from plant tissue. Animal tissue frozen in liquid nitrogen was ground to a fine powder, homogenized in TRI-Reagent® (Sigma) and RNA extracted as for HeLa cells.

RNA was dissolved in DEPC-treated water and quantified by measuring the absorbance at 260 nm using a Nanodrop® UV spectrophotometer (Labtech). Deionized formamide (Sigma) was then added to a final concentration of 50% and the RNA stored at −20°C until use.

Dephosphorylation of total RNA involved treatment of the RNA with 5 U of bacterial alkaline phosphatase (Invitrogen) per μg of total RNA at 60°C for 1 h.

Enrichment for low molecular weight RNA was by differential polyethylene glycol (PEG) precipitation as described in ref. (11).

Electrophoresis and blotting of RNA

All RNA samples were separated by electrophoresis using either 10 or 15% polyacrylamide (19:1) gels cast in 7 M urea and buffered with 20 mM MOPS/NaOH (pH 7) using a Protean II rig (Bio-Rad). Tris-based buffers were avoided as it was anticipated that these might interfere with the EDC-mediated cross-linking step. The electrophoresis buffer was 20 mM MOPS/NaOH (pH 7). RNA markers were 32γP-end-labelled Decade® RNA markers (Ambion) prepared according to the manufacturer's instructions. After gel electrophoresis, gels were stained with ethidium bromide (EtBr) or SybrGold® (Molecular Probes) and imaged on a FLA-5000 system (Fuji) with Aida Image Analyser software to visualize and record the amount and distribution of the RNA.

For blotting, gels were placed over a sheet of nylon hybridization membrane (Hybond-NX®, Amersham/Pharmacia) that had been pre-wetted in distilled water. This was then sandwiched between pieces of 3MM® Whatman filter paper (three layers on each side), also pre-wetted in distilled water and placed in a ‘semidry’ electroblotter (SciPlas). Excess liquid and air bubbles were squeezed from the sandwich by rolling the surface with a clean pipette. Electrophoretic transfer of RNA from the gel to the membrane was carried out at 20 V at 4°C for 30–60 min.

Cross-linking of RNA to Hybond-NX®


Immediately prior to use, a solution of 0.16 M l-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC) (Sigma) was prepared in 0.13 M 1-methylimidazole at pH 8 (pH adjusted with HCl). In each experiment, the prepared volume of this cross-linking solution was just sufficient to saturate a single sheet of 3MM® Whatman filter paper that had been cut slightly larger than the area of the hybridization membrane. For example, to cross-link RNA transferred onto a piece of Hybond-NX® nylon membrane 20 × 16 cm (i.e. 320 cm2), 122.5 μl of 12.5 M 1-methylimidazole was added to 10 ml of DEPC-treated H2O and adjusted to pH 8 with 1 M HCl. To this, 0.373 g of EDC was added and the volume made up to 12 ml. The membrane was placed on the EDC-saturated 3MM® paper with the side onto which the RNA had been transferred facing up. This was then wrapped in SARAN wrap and incubated at 60°C for between 15 minutes and 2 hours. The membrane was then washed in distilled water to remove residual EDC solution prior to prehybridization. Membranes can be dried and stored at −20°C after removal of residual EDC with no detrimental effects (data not shown).


Following transfer, membranes were immediately placed in the cabinet of a commercial UV irradiation device (Stratalinker®, Stratagene, CA, USA) and a prescribed amount of UV energy delivered. In most experiments, this was 0.24 J.

Probe synthesis

In vitro transcription of DNA templates was carried out with 20–40 U of T7 RNA polymerase per reaction and α-32P UTP (800 Ci·mmol−1; 10 mCi·ml−1). Probes for TS and GFP siRNA were generated by in vitro transcription reactions from plasmid templates [500 μg ml−1 of pTSa linearized with Sal1 or pKSGFP5 as described in (11)]. These reactions were supplemented with 100 μM unlabelled UTP and carried out at 37°C. Where oligonucleotide or ‘mirVana’ templates were used, the only source of UTP was from the radioactively labelled solution ([final] = ∼3 μM). Templates to make probes for mmu-mir-292-as, mmu-mir-294, hsa-mir-21, hsa-mir-16 were prepared using the mirVana™ miRNA Probe Construction Kit (Ambion, USA) with oligonucleotides aagtgccgccaggttttgagtgtcctgtctc (mmu-mir-292as), aaagtgcttcccttttgtgtgtcctgtctc (mmu-mir-294), tagcagcacgtaaatattggcgcctgtctc (hsa-mir-16) and tagcttatcagactgatgttgacctgtctc (hsa-mir-21) respectively according to the manufacturer's instructions. Resulting templates were used for RNA probe synthesis following manufacturer's instructions. The probes for ath-mir-159b and pi-R1 were prepared by annealing oligonucleotides aagagctcccttcaatccaaacctatagtgagtcgtatta (ath-mir-159) or tgacatgaacacaggtgctcagatagctttcctatagtgagtcgtatta (pi-R1) with the oligonucleotide taatacgactcactatagg. These were used at 250 nM as templates for in vitro transcription with α-32P UTP by T7 RNA polymerase at 22°C.

All in vitro transcriptions were treated with TurboDNAse® (Ambion) prior to addition to the hybridization solution to eliminate the DNA template.

End-labelling of oligonucleotides: (1 pmol per labelling) with or without LNA modifications were carried out with T4 polynucleotide kinase and γ32P ATP (6000 Ci·m mol−1; 10 mCi·ml−1). Pre-designed LNA oligonucleotides for hsa-mir-21 and hsa-mir-16 were obtained from Exiqon (Denmark). Unmodified oligonucleotide probe sequences were 5′ tcaacatcagtctgataagcta 3′ (hsa-mir-21) and 5′ cgccaatatttacgtgctgcta 3′ (hsa-mir-16) (Sigma).


Prehybridization and hybridization of both EDC and UV cross-linked membranes were carried out in 2× SSC, 1% SDS and 100 μg·ml−1 sheared, denatured salmon sperm DNA. The exact temperature used varied according to the experiment but was between 40 and 60°C. Post-hybridization washes were in 0.2 or 0.1× SSC, 0.2% SDS at temperatures ranging from 40 to 60°C. Membranes were stripped off the probe by placing them in a solution of 10 mM Tris/HCl (pH 8.5 at 25°C), 5 mM EDTA, 0.1% SDS at 100°C for 1 min. Removal of probe was assessed by re-exposure of the membrane for at least as long as any previous exposure. All radioisotopic images were recorded using a FLA-5000 system (Fuji) with Aida Image Analyser software.


Extended UV irradiation reveals a discrepancy between retention of small RNA and detection by hybridization

Total RNA extracted from wild-type A. thaliana (Columbia) flowers was separated by denaturing PAGE in six lanes with 20 μg RNA per lane. Each adjacent lane was loaded with γ32P-labelled marker RNA. RNA was transferred from the gel to a sheet of Hybond NX as described in methods and the membrane was cut into six strips, each with a single lane of A. thaliana RNA and a marker RNA lane. These strips were exposed to increasing doses of UV irradiation followed by hybridization with a probe complementary to the previously characterized miRNA, ath-mir-159b (16). The signal obtained from each marker lane indicated the degree to which the RNA was robustly linked to the membrane irrespective of conformation or chemical integrity of the RNA. Figure 1 shows that there was a more sensitive dependence on UV dose for RNA of >40 nt than for longer RNA. The optimum dose for RNA <40 nt in length was 0.48 J. In contrast, the optimum UV dose for subsequent detection by hybridization of the miRNA in the adjacent lanes was between 0.12 and 0.24 J. The dose delivered by the manufacturer's ‘autocross-link’ presetting is 0.12 J and is recommended for optimum detection of mRNA. Increasing the dose beyond 0.24 J progressively reduces the ability to detect the miRNA despite a continued increase in overall RNA retention.

Figure 1.
EDC cross-linking improves detection of plant siRNA. (a) RNA was extracted from leaves of transgenic N. benthamiana line GFP 16C/GFPi that transcribes a partial GFP inverted repeat (IR) or from non-transformed N. benthamiana (NT). From each sample, 5 ...

Improved detection of plant siRNA using EDC-mediated cross-linking

We tested whether small RNA from biological samples could be cross-linked to nylon hybridization membranes using the water-soluble carbodiimide EDC and also whether this improved their subsequent detection by hybridization compared to RNA cross-linked by UV. Untreated and dephosphorylated total RNA from a line of transgenic N. benthamiana that synthesizes GFP siRNA and from non-transgenic N. benthamiana was analysed. Half of each sample was loaded twice on the same gel. The gel was stained after electrophoresis to confirm equal loading (Supplementary Data, Figure 1) and phosphorimaging of both the gel and the membrane was carried out to demonstrate equal and efficient transfer of the γ32P-labelled marker RNA (Supplementary Data, Figure 2). The membrane was cut into two equal halves and, after UV or EDC cross-linking, the immobilized RNA was hybridized with an α-32P-labelled antisense GFP RNA probe. Figure 2a shows that slightly more (2–3-fold) marker RNA (20–40 nt incl.) was retained after cross-linking with EDC compared to UV. The RNA markers used (Decade®; Ambion) were labelled with T4 polynucleotide kinase and γ32P-ATP and so should react in the EDC-cross-linking step in the same way as siRNA and miRNA that have 5′ terminal phosphates. However, there was an ∼30-fold improvement in the detection of GFP siRNA by hybridization. EDC cross-linking of siRNA was dependent on terminal phosphorylation as the signal was almost completely eliminated by treatment of the RNA with alkaline phosphatase prior to electrophoresis.

Figure 2.
The effect of varying UV dose on the detection of small RNA. Here, 20 μg per lane of the same RNA preparation from A. thaliana flowers was separated by electrophoresis in six equivalent lanes (At). γ-32P ATP-labelled Decade® marker ...

We also tested whether EDC improved the detection of natural siRNA corresponding to the TS SINE retrotransposon of the plant genus Nicotiana. It has been demonstrated previously that siRNA corresponding to this type of retrotransposon can be detected in leaves (8). Quantification of the phosphorimage shown in Figure 2b indicated that compared to UV, EDC enhanced the cross-linking of marker RNA larger than 40 nt by only 1.2-fold but by 2.7-fold for marker RNA of 40 nt or less. EDC cross-linking improved the detection of TS SINE siRNA by over 30-fold. As with GFP siRNA, the cross-linking by EDC (but not UV) was dependent on terminal phosphorylation. TS siRNA were detected using EDC cross-linking from as little as 0.5 µg total RNA (Figure 2b). This signal increased in direct proportion to the amount of RNA loaded.

Improved detection of mammalian miRNA and temperature dependence with EDC cross-linking

To test whether EDC also improved the detection of mammalian miRNA, we compared the detection, after cross-linking with EDC or UV, of two different miRNA, mmu-mir-292as (mmu-miR-292-3p) and mmu-mir-294, that were originally shown to accumulate in murine ES cells (17). Because miRNA generally accumulate as a single molecular species, this also allowed us to examine more accurately the effect of varying the time and temperature of EDC cross-linking efficiency. Low molecular weight RNA from differentiated murine ES cells was separated by electrophoresis, the gel stained with EtBr (Supplementary Data, Figure 3) and blotted onto Hybond NX® membrane. In this experiment, the UV treatment cross-linked the γ32P-labelled RNA markers as efficiently as the optimum EDC treatment (Figure 3). However, as for the plant siRNA, there was substantially enhanced detection of both mmu-mir-292as (50 fold) (Figure 3a) and mmu-mir-294 (Figure 3b) using EDC cross-linking compared to UV. Mmu-mir-294 was not convincingly detected using UV cross-linking, so the manifest improvement in detection was not quantified. The optimum conditions appeared to differ slightly for each miRNA but are between 50 and 60°C for 1–2 h. Surprisingly, the EDC cross-linking of the γ32P-labelled marker RNA varied less with time and temperature compared to the miRNA.

Figure 3.
EDC cross-linking improves detection of mammalian miRNA. Identical 5-μg aliquots of low molecular weight RNA from differentiating murine embryonic stem cells were loaded in multiple lanes of the same denaturing polyacrylamide gel. The same volume ...

Improved detection of piRNA using EDC-mediated cross-linking

A new class of small RNA related to microRNA and siRNA, which accumulate in rodent testes, was recently described by several groups [reviewed in ref. (3)]. These are known as piRNA and are in the range of 29–31 nt in length. Figure 4 shows that EDC cross-linking improves the detection of one of these, pi-R1 (18) compared to UV cross-linking. The increase in sensitivity was 20–25-fold, slightly less than that observed for siRNA and microRNA. We observed similar improved detection using EDC compared to UV of the two other piRNA that we tested (data not shown).

Figure 4.
Improved detection of piRNA using EDC-mediated cross-linking. Total RNA was extracted from mammary gland (m), heart (h), liver (l), brain (b) and testes (t) of adult mice and rats. Here, 10 µg of each sample was run in duplicate on a 10% ...

Comparison of LNA, DNA and RNA probes with EDC versus UV cross-linking

Three methods that are commonly used to prepare hybridization probes for small RNA detection were compared for their compatibility with EDC versus UV cross-linked RNA.

  1. Pre-designed, oligodeoxynucleotides incorporating ‘locked nucleic acid’ (LNA®) modifications. The sequences were entirely complementary to the target miRNA and were end-labelled with γ-32P ATP by T4 polynucleotide kinase. LNA-containing oligonucleotides have been shown to offer much greater sensitivity for the detection of miRNA by standard small RNA northern blotting using UV for cross-linking (19).
  2. Unmodified oligodeoxynucleotides complementary to the target miRNA and that were end-labelled with γ-32P ATP by T4 polynucleotide kinase.
  3. RNA oligonucleotides transcribed in vitro with α-32P UTP from templates created with the mirVana® kit (Ambion). The RNA probes generated are fully complementary to the target miRNA but additionally have the 5′ leader sequence GGGAGACAGG.

Here, 5 µg of the same sample of HeLa cell total RNA was run in six lanes of the same denaturing gel. 32γ-P-end-labelled Decade® marker RNA was run in each adjacent lane. The gel was stained with EtBr to confirm equal loading (Supplementary Data, Figure 5). After blotting, membranes were cross-linked with either EDC or UV. Probes were prepared by the above three methods for the detection of two human miRNA: hsa-mir-21 (Figure 5a) and hsa-mir-16 (Figure 5b).

Figure 5.
A combination of LNA probe and EDC cross-linking provides the most sensitive detection of miRNA. Identical 5-μg aliquots of total RNA from HeLa cells were loaded in multiple lanes of the same denaturing polyacrylamide gel. The same volume of ...

The blots were first hybridized with the three differently synthesized hsa-mir-21 probes. After exposure, the blots were stripped, exposed to ensure complete removal of probe (Supplementary Data, Figure 6) and re-hybridized with the three differently synthesized probes for hsa-mir-16. No signal was detected with the end-labelled unmodified oligonucleotides probably because of the relatively low amount of total RNA used (5 μg). LNA probes with EDC-cross-linked RNA produced the strongest signal for both miRNA tested although, for hsa-mir-16, this was only slightly greater than if using a RNA probe for EDC-cross-linked RNA.


Recent discoveries of thousands of novel, small RNA and increasing appreciation of their diverse biological functions (1–3,9,20) have led to a demand for improved experimental procedures to study these molecules. Northern blot analysis remains one of the most convincing ways to determine the presence, size and abundance of a small RNA but it is generally regarded as one of the least sensitive analytical methods (9). We hypothesized that the UV cross-linking step in this method, which was originally developed for the efficient detection of mRNA, might not be optimum for RNA of only 20–30 nt in length. Therefore, we tested whether varying the dose of UV delivered by a standard UV cross-linking cabinet could improve the detection of a typical small RNA (a 21-nt miRNA). Figure 1 shows that the optimum UV dose for the detection of that miRNA was, in fact, very similar to the recommended dose for mRNA (∼0.12 J) and that increasing the UV dose resulted in progressively less sensitive miRNA detection. Despite the reduction in detection of miRNA by hybridization beyond 0.12 J, direct detection of the 20 and 30 nt radioactive RNA markers increased, reaching an optimum at a 4-fold higher UV dose. Thus, the optimum UV dose for the detection of small RNA by hybridization was not the same as the dose that maximally cross-linked the RNA. Several possible events might account for this including UV-degradation of the functional groups required for base-pairing, fragmentation of the RNA and excessive cross-linking limiting access for complementary probes. We tested three other UV-cross-linking cabinets (all from Stratagene) located within other departments in our institute and found there was no difference in their performance compared to the unit used in this study with respect to the detection of a miRNA or in the cross-linking of marker RNA (Supplementary Data, Figure 7). We concluded that when using UV as a cross-linker, there was a trade-off between increasing retention of small RNA on a membrane and progressively reducing the ability of that immobilized RNA to form stable double-stranded hybrids. Therefore we investigated alternative methods for cross-linking RNA to hybridization membranes.

Ideally, for detection by hybridization, small RNA would be efficiently coupled to a support in such a way that the entire polynucleotide was freely available for base-pairing with the complementary probe. In principle, functional groups not located on the heterocyclic bases and that are specific to the 5′-end or 3′-end would be the most suitable reactants. In animals, small regulatory RNA such as siRNA and miRNA generally have 5′ terminal monophosphates and a 2′, 3′ cis-diol at their 3′-terminus (21). However, some 3′ modifications of small RNA have been described in insects and plants (7,22), therefore cross-linking through the 5′-terminal phosphate would seem to offer the greatest flexibility.

EDC has been used to cross-link 5′-phosphorylated oligodeoxyribonucleotides to primary amines on controlled pore glass supports (23) and this resulted in superior hybridization properties compared to nitrocellulose-bound DNA (24). EDC has also been used to couple phosphorylated oligodeoxyribonucleotides that were directly pumped onto the surface of the negatively charged nylon membrane, Biodyne C® (25). However, it was not obvious that a similar process could be used for cross-linking small RNA from biological samples because the coupling efficiency was not indicated: efficient coupling would be critical for a successful northern blot adaptation. We tested whether EDC could be used as an efficient cross-linker of several different small RNA from biological samples to standard, neutral nylon hybridization membranes (Hybond NX®) after deposition by electroblotting from denaturing polyacrylamide gels. Figure 2 shows that detection of plant siRNA by hybridization was enhanced by ∼40-fold after EDC cross-linking compared to UV. Small (20–40 nt) radioactive, marker RNA was cross-linked only 2–3-fold more efficiently by EDC than UV in the same experiment. The proportionally greater enhancement of detection by hybridization may have resulted from reduced degradation and/or improved presentation of EDC-cross-linked RNA. This result also indicated that contaminating molecules likely to be present in RNA preparations from biological sources did not significantly interfere with cross-linking. The observed dependence on terminal phosphorylation is consistent with the expected formation of a phosphoramidate bond between the RNA and the nylon. For the experiment shown in Figure 2, this was likely to be 5′ phosphorylation according to the known properties of siRNA. We have not tested whether the detection of RNA with 3′-terminal phosphates would similarly benefit from EDC cross-linking. Staining of gels with EtBr prior to transfer did not reduce the enhancement of hybridization signal obtained with EDC cross-linking (data not shown). siRNA detection increased in proportion to the amount of RNA loaded up to the maximum tested (50 μg of total RNA, Figure 2b). This suggests that cross-linking sites on the membrane were not becoming saturated and that the membrane might possibly accommodate larger amounts of RNA without a reduction in cross-linking efficiency. Thus, very rare small RNA might be detected if sufficient quantities of RNA enriched for low molecular weight species were available.

Figure 3 shows that EDC cross-linking also greatly enhances the detection of miRNA from mammalian cells. This experiment also tested the effect of varying time and temperature on the EDC-cross-linking step. Similar optimum conditions (2 h at 50 or 60°C) were identified for hsa-mir-292as and hsa-mir-294. The same optimum conditions were found for two other miRNA tested (data not shown). In this experiment, EDC cross-linking did not significantly improve the retention of marker RNA compared to UV indicating again that increased miRNA detection sensitivity is largely the result of improved RNA quality and/or positioning on the membrane. The detection of piRNA was also improved using EDC cross-linking compared to UV although by only 20–25-fold (Figure 4) using these conditions.

Probes that incorporate LNA modifications have been shown to substantially enhance the detection of miRNA in standard northern blot protocols (i.e. with UV cross-linking) (19). Figure 5 shows that the most sensitive detection of two different mammalian miRNA was achieved with a combination of LNA probe and EDC cross-linking. However, for one of those microRNA, the use of EDC cross-linking with an inexpensive, in vitro transcribed RNA probe was nearly as effective.

In principle, neutral nylon membranes from other manufacturers might be as suitable as Hybond NX® but we have not extensively tested this. One positively charged membrane tested (Hybond N+®) gave substantially poorer results than the neutral membrane (data not shown). Although we did not establish an upper limit on the amount of small RNA that can be cross-linked using EDC to existing neutral nylon membranes (per mm2), saturation will inevitably occur at some point. Therefore, future improvements to the described process could possibly be made by fabrication of membranes with increased surface amine content. However, care would have to be taken not to reduce the initial, very efficient non-covalent adsorption of RNA at the point of blotting/transfer that neutral nylon membranes provide (data not shown).

We have briefly tested other standard hybridization conditions such as those with buffers containing high concentrations of SDS or formamide and found them to be equally compatible with EDC-cross-linked membranes (data not shown).

In summary, we have described a simple, inexpensive alternative to UV cross-linking that greatly improves the detection of small, terminally phosphorylated RNA such as siRNA, miRNA and piRNA by northern blotting. It is likely that the detection of other related small RNA will also benefit from this procedure.


Supplementary data is available at NAR online.

[Supplementary Material]


This work was supported by the BBSRC and Breakthrough Breast Cancer. We are grateful to Professor Duncan Graham (University of Strathclyde, Glasgow, UK) for helpful discussions and to Dr Adam West, (Glasgow University, UK) for critical reading of the manuscript. We thank Professor David Baulcombe (Sainsbury Laboratory, Norwich, UK) for the transgenic line GFP 16C/GFPi. Funding to pay the Open Access publication charge was provided by BBSRC and Breakthrough Breast Cancer.

Conflict of interest statement. None declared.


1. Zamore PD, Haley B. Ribo-gnome: the big world of small RNAs. Science. 2005;309:1519–1524. [PubMed]
2. Kim VN. Small RNAs: classification, biogenesis, and function. Mol. Cells. 2005;19:1–15. [PubMed]
3. Carthew RW. Molecular biology: a new RNA dimension to genome control. Science. 2006;313:305–306. [PubMed]
4. Bernstein E, Caudy AA, Hammond SM, Hannon GJ. Role for a bidentate ribonuclease in the initiation step of RNA interference. Nature. 2001;409:363–366. [PubMed]
5. Knight SW, Bass BL. A role for the RNase III enzyme DCR-1 in RNA interference and germ line development in Caenorhabditis elegans. Science. 2001;293:2269–2271. [PMC free article] [PubMed]
6. Ketting RF, Fischer SEJ, Bernstein E, Sijen T, Hannon GJ, Plasterk RHA. Dicer functions in RNA interference and in synthesis of small RNA involved in developmental timing in C-elegans. Genes Dev. 2001;15:2654–2659. [PMC free article] [PubMed]
7. Vagin VV, Sigova A, Li CJ, Seitz H, Gvozdev V, Zamore PD. A distinct small RNA pathway silences selfish genetic elements in the germline. Science. 2006;313:320–324. [PubMed]
8. Hamilton A, Voinnet O, Chappell L, Baulcombe D. Two classes of short interfering RNA in RNA silencing. EMBO J. 2002;21:4671–4679. [PMC free article] [PubMed]
9. Berezikov E, Cuppen E, Plasterk RHA. Approaches to microRNA discovery. Nat. Genet. 2006;38:S2–S7. [PubMed]
10. Sethupathy P, Megraw M, Hatzigeorgiou AG. A guide through present computational approaches for the identification of mammalian microRNA targets. Nat. Methods. 2006;3:881–886. [PubMed]
11. Hamilton AJ, Baulcombe DC. A species of small antisense RNA in posttranscriptional gene silencing in plants. Science. 1999;286:950–952. [PubMed]
12. Anonymous. Northern blotting: transfer of denatured RNA to membranes. Nat. Methods. 2005;2:997–998.
13. Church GM, Gilbert W. Genomic Sequencing. Proc. Natl. Acad. Sci. USA. 1984;81:1991–1995. [PMC free article] [PubMed]
14. Saito I, Sugiyama H, Furukawa N, Matsuura T. Photochemical ring-opening of thymidine and thymine in the presence of primary amines. Tetrahedron Lett. 1981;22:3265–3268.
15. Schwach F, Vaistij FE, Jones L, Baulcombe DC. An RNA-dependent RNA polymerase prevents meristem invasion by potato virus X and is required for the activity but not the production of a systemic silencing signal. Plant Physiol. 2005;138:1842–1852. [PMC free article] [PubMed]
16. Mette MF, van der Winden J, Matzke M, Matzke AJM. Short RNAs can identify new candidate transposable element families in Arabidopsis. Plant Physiol. 2002;130:6–9. [PMC free article] [PubMed]
17. Houbaviy HB, Murray MF, Sharp PA. Embryonic stem cell-specific microRNAs. Dev. Cell. 2003;5:351–358. [PubMed]
18. Girard A, Sachidanandam R, Hannon GJ, Carmell MA. A germline-specific class of small RNAs binds mammalian Piwi proteins. Nature. 2006;442:199–202. [PubMed]
19. Valoczi A, Hornyik C, Varga N, Burgyan J, Kauppinen S, Havelda Z. Sensitive and specific detection of microRNAs by northern blot analysis using LNA-modified oligonucleotide probes. Nucl. Acids Res. 2004;32:e175. [PMC free article] [PubMed]
20. Mattick JS, Makunin IV. Small regulatory RNAs in mammals. Hum. Mol. Genet. 2005;14:R121–R132. [PubMed]
21. Elbashir SM, Lendeckel W, Tuschl T. RNA interference is mediated by 21- and 22-nucleotide RNAs. Genes Dev. 2001;15:188–200. [PMC free article] [PubMed]
22. Yang Z, Ebright YW, 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. Nucl. Acids Res. 2006;34:667–675. [PMC free article] [PubMed]
23. Ghosh SS, Musso GF. Covalent attachment of oligonucleotides to solid supports. Nucleic Acids Res. 1987;15:5353–5372. [PMC free article] [PubMed]
24. Gingeras TR, Kwoh DY, Davis GR. Hybridization properties of immobilized nucleic-acids. Nucl. Acids Res. 1987;15:5373–5390. [PMC free article] [PubMed]
25. 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. 2003;41:3690–3698. [PMC free article] [PubMed]

Articles from Nucleic Acids Research are provided here courtesy of Oxford University Press
PubReader format: click here to try


Save items

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • Compound
    PubChem chemical compound records that cite the current articles. These references are taken from those provided on submitted PubChem chemical substance records. Multiple substance records may contribute to the PubChem compound record.
  • MedGen
    Related information in MedGen
  • PubMed
    PubMed citations for these articles
  • Substance
    PubChem chemical substance records that cite the current articles. These references are taken from those provided on submitted PubChem chemical substance records.

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...