Skip to main content
NCBI home page
Nucleic Acids Research logoLink to Nucleic Acids Research
. 2003 Aug 1;31(15):e79. doi: 10.1093/nar/gng079

Fast and accurate determination of sites along the FUT2 in vitro transcript that are accessible to antisense oligonucleotides by application of secondary structure predictions and RNase H in combination with MALDI-TOF mass spectrometry

Angelika Gabler 1,*, Stefan Krebs 1, Doris Seichter 1, Martin Förster 1
PMCID: PMC169965  PMID: 12888531

Abstract

Alteration of gene expression by use of antisense oligonucleotides has considerable potential for therapeutic purposes and scientific studies. Although applied for almost 25 years, this technique is still associated with difficulties in finding antisense-effective regions along the target mRNA. This is mainly due to strong secondary structures preventing binding of antisense oligonucleotides and RNase H, playing a major role in antisense-mediated degradation of the mRNA. These difficulties make empirical testing of a large number of sequences complementary to various sites in the target mRNA a very lengthy and troublesome procedure. To overcome this problem, more recent strategies to find efficient antisense sites are based on secondary structure prediction and RNase H-dependent mechanisms. We were the first who directly combined these two strategies; antisense oligonucleotides complementary to predicted unpaired target mRNA regions were designed and hybridized to the corresponding RNAs. Incubation with RNase H led to cleavage of the RNA at the respective hybridization sites. Analysis of the RNA fragments by matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry, which has not been used in this context before, allowed exact determination of the cleavage site. Thus the technique described here is very promising when searching for effective antisense sites.

INTRODUCTION

Since its introduction by Zamecnik and Stephenson more than two decades ago (1), antisense technology has been widely used to specifically generate gene knockouts. The underlying strategy is conceptually elegant. Taking advantage of knowledge of the unique sequence of the mRNA transcribed from the gene of interest, small oligonucleotides of complementary sequences are synthesized, which enables them to interact by Watson–Crick base pairing with the corresponding mRNA. Although a number of RNA oligonucleotide derivatives and peptide nucleic acids are currently under investigation, the most common antisense compounds are chemically modified oligodeoxynucleotides (ODNs). One of the modes of antisense ODN action discussed is inhibition of protein synthesis via translational blockade; antisense ODNs that produce DNA–RNA hybrids can interfere in the initiation process and/or progression of the ribosomal apparatus and so induce translational arrest (2). The main reason for antisense ODN-mediated alteration of gene expression, however, is supposed to be degradation of the mRNA by endogenous RNase H (3,4). The ubiquitously expressed enzyme RNase H cleaves the RNA strand of DNA–RNA hybrids, and thus the target mRNA is no longer available as a template for protein synthesis. This mechanism of antisense ODN-mediated inhibition of gene expression is especially important in proliferative tissues as RNase H activity clearly correlates with the rate of DNA synthesis (5), and as few as 5–6 bp are often sufficient to be recognized as a DNA–RNA substrate by RNase H (6). Unlike other RNases, RNase H does not cleave RNA in a sequence-specific manner, and scission may occur at more than one point along the target RNA (7).

One of the most critical steps in antisense technology is the design of effective antisense ODNs, because intramolecular base pairing of the mRNA into stable secondary and tertiary structures makes large parts of the molecule inaccessible to interacting with complementary nucleic acids, whereas sequences free from intramolecular interactions may favor interaction with other nucleic acids (8,9). Furthermore, a linear correlation was found between the percentage of translation inhibition by RNase H and the percentage of unpaired target RNA (10); taken together, these observations clearly indicate that mRNA regions free from intramolecular base pairing are ideally suited as targets for antisense ODNs. Therefore, many researchers use computer programs which make predictions of the secondary structure of the target RNA. One of the most frequently used programs is ‘mfold’ [http://www.bioinfo.rpi.edu/applications/mfold/old/rna/ (11,12)]; this program calculates the global free energy minimum of an RNA molecule, thus leading to the prediction of single-stranded and double-stranded structures. However, several reports call the reliability of prediction of accessible and inaccessible sites on mRNAs by ‘mfold’ into question (13,14); thus computer prediction methods alone seem not to be sufficient for designing effective antisense ODNs. In order to circumvent these difficulties, we constructed antisense ODNs according to computational secondary structure predictions, and tested their ability to provoke RNase H-mediated degradation of the RNA in vitro. The resulting RNA fragments were subsequently analyzed not only by gel electrophoresis, but also by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF mass spectrometry), thus enabling us to determine the exact cleavage site.

MATERIALS AND METHODS

Generation of in vitro transcripts

Total RNA was isolated from pig ileum with Trizol (Life Technologies) according to the manufacturer’s protocol. cDNA synthesis was performed with random primers and M-MLV reverse transcriptase (Promega). Using gene-specific primers, PCR products of the FUT2 coding sequence were generated (primer sequences: 5′-TAA TAC GAC TCA CTA TAG GGA GAT GCT CAG CAT GCA GGC ATC T-3′ and 5′-CCC TCT AGA TCA GTG CTT AAG GAG TGG GGA-3′) and cloned into pCR2.1TOPO (Invitrogen), serving as template for subsequent PCR where an SP6 promoter was inserted at the 5′ end for in vitro transcription (primer sequences: 5′-ATT TAG GTG ACA CTA TAG AAG ATG CTC AGC ATG CAG GCA TCT-3′ and 5′-AAT TAA CCC TCA CTA AAG GGA GAT CAG TGC TTA AGG AGT GGG GAC AG-3′; the generated PCR product has a length of 1067 bp). A 25 µl aliquot of this PCR product was used as template for a 50 µl in vitro transcription of FUT2, using the RiboMAX Large Scale RNA Production System-SP6 (Promega).

Design of antisense ODNs

The coding sequence of the porcine FUT2 gene (GenBank accession no. AF136895) was analyzed by the computer program ‘mfold’ (11,12). The obtained data about predicted single-stranded regions along the FUT2 mRNA were chosen for design of antisense ODNs. These were unmodified phosphodiester 18mers (MWG Biotech).

RNase H digestion and subsequent analysis of ODN–RNA duplexes

A 5 µg aliquot of RNA from the in vitro transcription reactions was incubated with 1 U of RNase H (Promega) and antisense ODNs in varying concentrations for 1 h at 37°C in 1× reaction buffer [40 mM HEPES-KOH pH 8.0, 1 mM dithiothreitol (DTT), 10 mM MgCl2, 100 mM KCl, according to Matveeva et al. (15)]. Samples were electrophoresed through a denaturing polyacrylamide gel [7 M urea, 4.5% total acrylamide (ratio acrylamide:bisacrylamide = 19:1)] and visualized under UV light after staining with ethidium bromide to check the efficacy of RNase H cleavage caused by the various ODNs.

MALDI-TOF mass spectrometry

RNase H reaction products and control samples were desalted on a membrane filtration plate (Multiscreen Seq, Millipore, Bedford, MA) by washing twice with 20 µl of water and twice with 20 µl of 10 mM diammonium citrate. The samples were resuspended in a MALDI-compatible buffer (16) and further digested with 10 U of RNase T1 (LaRoche) at 37°C for 1 h in order to cleave the RNA fragments after each guanine. Then they were spotted on dried matrix droplets (saturated 3-hydroxy picolinic acid in 50% acetonitrile, 10 mg/ml ammonium citrate) on a Scout 384 stainless steel target plate (Bruker Daltonics) and analyzed on a linear Bruker Biflex III delayed extraction MALDI mass spectrometer (record in negative ion mode, 19 kV acceleration voltage, 17.45 kV IS/2 potential, 400 ns extraction delay, no. of shots 120 on average).

RESULTS

Design of antisense ODNs

To obtain secondary structure predictions of the porcine FUT2 mRNA, we applied the computer program ‘mfold’, version 3.1 (11,12). We made an input of the FUT2 coding sequence (1023 bases) and obtained 38 different folding structures calculated for linear RNA at 37°C, based on free energy minimization. We looked for sequence segments of eight or more consecutive single-stranded bases found in parallel in at least 30 of the 38 calculated secondary structures in order to enhance the likelihood of finding true single-stranded regions. Of the eight sequence segments which matched these criteria, we selected four by chance, and designed six antisense ODNs whose sequences were complementary to the selected regions (Table 1).

Table 1. Structure and position of oligodeoxynucleotides.

Oligonucleotide Positiona Sequence (5′→3′)a
ODN1 113–130 GAAGAAAGATGCCTGCAT
ODN2 222–239 GCGTCACCATCTGTAACT
ODN3 224–241 CTGCGTCACCATCTGTAA
ODN4 460–477 TGGTAGTTCTGCCAGGGG
ODN5 477–494 CCATCCAGTCGTTCAGGT
ODN6 904–921 GCGAAGTCTTTGGCGGGG
Open in a new tab

aAntisense ODN sequences are complementary to porcine FUT2, GenBank accession no. AF136895.

Test of RNase H activity induction by specific ODNs

Each antisense ODN was incubated with the FUT2 in vitro transcript and RNase H as described in Materials and Methods; the concentration of the oligonucleotides was 25 µM. After gel electrophoresis of the samples, no more intact in vitro transcript was visible in any of the samples, except a small amount in the case of ODN1 (Fig. 1). In contrast to ODNs 1, 2, 3 and 5, where distinct cleavage patterns of the RNA were visible, ODNs 4 and 6 led to a more complete degradation of the RNA, leaving only a smear of small RNA fragments. We therefore selected these two antisense ODNs for further investigation; moreover, we designed sense control and mismatch control ODNs for both antisense ODNs (Table 2). Hence ODNs 4, 6 and control oligos were again incubated with FUT2 RNA and RNase H, and their concentrations were scaled down in order to reach the point where maximum stringency in hybridization was obtained as well as activation of RNase H-mediated cleavage of the transcript. For ODN4, this was the case at a concentration of 0.025 µM, and for ODN6 at 0.05 µM, respectively, where the in vitro transcripts were cut only once (Fig. 2). How can the generation of these RNA fragments be explained? Given the hybridization site of ODN4 between bases 460 and 477 of the FUT2 mRNA (i.e. bases 357–374 of the coding sequence), and taking into account the additional bases, linked by the PCR primers, at both ends of the in vitro transcript, one would expect two RNA fragments of ∼380 and 690 bases, assuming that RNase H cleaved the FUT2 RNA only once within the hybridization site of ODN4. In the case of ODN6, there should be two RNA fragments of ∼830 and 240 bases. These assumptions are confirmed by the results shown in Figure 2; however, the exact positions of RNase H-mediated cleavage of the RNA are still unclear.

Figure 1.

Figure 1

Open in a new tab

Electrophoretic separation of FUT2 RNA after incubation with RNase H and ODNs (25 µM each). Lane M: 0.15–1.77 kb RNA ladder (Life Technologies). Lanes 1–6: ODN1, 2, 3, 4, 5 and 6, respectively. Lane C: control reaction without RNase H. The weak signal at ∼400 bases that appears in each lane is due to residual PCR product which was the template for the in vitro transcription.

Table 2. Structure of control oligodeoxynucleotides.

Oligonucleotide Position Sequence (5′→3′)
ODN4 sense 460–477 ACCATCAAGACGGTCCCC
ODN4 mismatcha 460–477 TGGgAGTTaTGCCcGGtG
ODN6 sense 904–921 CGCTTCAGAAACCGCCCC
ODN6 mismatcha 904–921 GCGcAGTaTTgGGCGtGG
Open in a new tab

aMismatched bases indicated in lower case.

Figure 2.

Figure 2

Open in a new tab

RNase H-mediated digestion of the FUT2 in vitro transcript in the presence of ODN4 and ODN6, respectively (oligo concentrations: ODN4 series, 0.025 µM; ODN6 series, 0.05 µM). Two bands of RNA fragments of ∼380 and 690 bases result when ODN4 was applied; in the case of ODN6, two bands of ∼240 and 830 bases are visible. These bands are not detectable when sense control ODNs (marked 4S and 6S, respectively) and mismatch control ODNs (marked 4M and 6M, respectively) were present in the reaction mix. Lane 4M shows one additional band at ∼480 bases; this must be due to an ODN4 mismatch-induced RNase H digestion that occurred at any position but not at the ODN4 hybridization site. Lanes 4C and 6C show the in vitro transcript after incubation with ODN4 and ODN6, respectively, but without RNase H. Lane M: 0.15–1.77 kb RNA ladder (Life Technologies). The faint band that is visible at ∼400 bases in each sample represents some residual PCR product that served as template in the in vitro transcription reaction.

Determination of the exact cleavage site of RNase H by MALDI-TOF mass spectrometry

To determine the exact cleavage site of RNase H, the generated RNA fragments were further analyzed by MALDI-TOF mass spectrometry. With this technique, resolution is best for short RNA fragments [up to ∼100 bases (17)], and we therefore carried out further fragmentation of the RNA in a base-specific manner by the use of RNase T1. RNase T1 cleaves unpaired guanines, thus each specific RNA sequence will result in a unique fragment pattern. We therefore digested untreated FUT2 RNA with RNase T1 in order to generate a FUT2-specific fragment pattern. When measured by MALDI-TOF mass spectrometry, a characteristic peak pattern could be detected. A change of this peak pattern after additional pre-digestion of the FUT2 in vitro transcript by RNase H in the presence of ODN4 and ODN6, respectively, and subsequent digestion by RNase T1, could easily be detected by comparison of these mass spectra (Fig. 3).

Figure 3.

Figure 3

Open in a new tab

MALDI-TOF mass spectra of FUT2 RNA fragments generated after incubation with RNase H and ODN with subsequent RNase T1 digestion. Reactions were done using ODN4 and ODN6, respectively, as well as ODN4 sense and ODN6 sense; the control reaction was lacking in RNase H. Peak 2099.5 only appeared with ODN4, whereas peak 2421.2 could only be detected with ODN6.

The main peak pattern characteristic for the FUT2 in vitro transcript digested by RNase T1 also appeared in reactions where the transcript additionally has been pre-incubated with specific ODNs and RNase H. However, in contrast to the control reaction and reactions with sense control ODNs, the samples that have been pre-treated with RNase H in combination with ODN4 and ODN6, respectively, showed new mass peaks, at ∼2099.5 Da in the case of ODN4, and at ∼2421.2 Da when ODN6 was used. Thus the RNA fragments underlying these two mass peaks must arise from RNase H-mediated digestion. From gel electrophoresis of the RNase H digestion products, we could already estimate that the RNase H cleavage sites are located nearly at, or even within, the hybridization site of the respective antisense ODN (Fig. 2). Furthermore, it is described that RNase H cuts the RNA portion of a DNA–RNA hybrid within, or immediately adjacent to, the heteroduplex region (18); therefore, RNase H cleavage in ODN4 and ODN6 reactions is supposed to occur within the respective hybridization site. In order to define RNase H cleavage sites exactly, the FUT2 RNA fragment pattern resulting from RNase T1 digestion was calculated by RnaseCut 1.01 software available at our homepage (http://www.vetmed.uni-muenchen.de/gen/forschung.html) (19). RNase T1-mediated digestion generates a 2′–3′ cyclic phosphate intermediate at the 3′ end of each digestion product, which is then transformed into a 2′ (or 3′) ‘linear’ phosphate; the conversion into the linear phosphate can be forced by using more enzyme or by prolonged digestion time (16 h), whereas almost exclusively cyclic phosphate products can be observed when using the reverse conditions (16). In order to save time, we wanted to generate cyclic phosphate products; therefore, with the conditions we applied, the observed RNA fragments were 18 Da less in mass because of the guanine with a cyclic phosphate having a mass of 327.2 Da, instead of 345.2 Da in the case of guanine with a linear phosphate. We therefore calculated the RNase T1 digestion products with respect to the detected cyclic phosphate intermediates. Figure 4 compares in detail the resulting fragments of the FUT2 RNA after RNase H/RNase T1 double digestion versus RNase T1 digestion only.

Figure 4.

Figure 4

Open in a new tab

Cleavage of the FUT2 transcript by RNase H and RNase T1 in comparison with cleavage by RNase T1 only. RNase H hydrolyzes the phosphodiester bonds of RNA hybridized to DNA to produce 5′-P- and 3′-OH-terminated products; RNase T1 cleaves RNA after each guanine, leaving 5′-OH- and 3′-P-terminated products (shown in detail only at relevant cleavage sites). Hybridization sites of respective ODNs are in italics. (A) Generation of a 2100.4 Da RNA fragment after ODN4-induced RNase H cleavage and subsequent RNase T1 cleavage. (B) Resulting RNA fragment of 2420.6 Da when ODN6 was used.

Figure 4A shows a calculated RNA fragment of 2507.6 Da that results from digestion of the FUT2 in vitro transcript by RNase T1 only. When subtracting the last base and phosphate group from this fragment, the corresponding mass is reduced to 2100.4 Da. The mass peak that we detected by MALDI-TOF mass spectrometry in the RNase H reaction after application of ODN4 was 2099.5 Da (Fig. 3), that is why we conclude that incubation with RNase H and ODN4 leads to cleavage of the FUT2 in vitro transcript between bases U464 and G465, leaving a 3′-OH terminated product. In Figure 4B, the origin of a calculated RNA fragment of 2482.6 Da after RNase T1 digestion of the FUT2 RNA is shown. If RNase H cleaves the FUT2 in vitro transcript in advance between bases G908 and C909 due to the presence of ODN6, then the 3′-terminal base G is hydroxylated (instead of being phosphorylated in the case of RNase T1 digestion), and therefore the mass of this RNA fragment after subsequent RNase T1 digestion is reduced to 2420.6 Da (measured: 2421.2 Da, Fig. 3). Thus the sites of RNase H cleavage induced by ODN4 and ODN6, respectively, could be analyzed. Supposing that the FUT2 in vitro transcript was completely digested by RNase H in the presence of ODN4 and ODN6, respectively, then the above-mentioned fragments of 2507.6 and 2482.6 Da generated in the control reactions should not be detectable in the respective ODN MALDI-TOF mass spectra. This is not the case; by chance, there exist other sequence segments along the FUT2 RNA that lead to these masses when digested by RNase T1. In Table 3, the RNA fragments generated by RNase H cleavage and RNase T1 cleavage are listed within the mass range of interest. Again the products resulting from RNase T1 digestion have been calculated by using RNaseCut 1.01 software with respect to the cyclic phosphate intermediates.

Table 3. Calculated FUT2 in vitro transcript fragment pattern after RNase H and RNase T1 digestion, sorted according to fragment masses.

FUT2 fragments (5′→3′)a Calculated mass (Da)
  RNase T1 only RNase H/ODN4 + RNase T1 RNase H/ODN6 + RNase T1
uucacccug 2813.8 2813.8 2813.8
ccuaccuug 2813.8 2813.8 2813.8
cuaccccug 2812.8 2812.8 2812.8
uaccuacg 2531.6 2531.6 2531.6
auuuccug 2509.6 2509.6 2509.6
ucuucacg 2508.6 2508.6 2508.6
auccccug 2507.6 – 2507.6
auccuccg 2507.6 2507.6 2507.6
ccacccug 2506.6 2506.6 2506.6
ccuuccug 2484.6 2484.6 2484.6
cucccccg 2482.6 2482.6 2482.6
cucccccg 2482.6 2482.6 –
cucccccgb – – 2420.6
aucaaug 2250.4 2250.4 2250.4
accacag 2248.4 2248.4 2248.4
accauug 2226.4 2226.4 2226.4
cccaacg 2224.4 2224.4 2224.4
ucuccag 2202.4 2202.4 2202.4
ccucacg 2201.4 2201.4 2201.4
cccccag 2200.4 2200.4 2200.4
ccccucg 2177.4 2177.4 2177.4
auccccub – 2100.4 –
aaccag 1943.2 1943.2 1943.2
ccaaag 1943.2 1943.2 1943.2
uuacag 1921.2 1921.2 1921.2
acuacg 1920.2 1920.2 1920.2
uaccug 1897.2 1897.2 1897.2
acuucg 1897.2 1897.2 1897.2
ccaucg 1896.2 1896.2 1896.2
cuaccg 1896.2 1896.2 1896.2
cucacg 1896.2 1896.2 1896.2
ucuuug 1875.2 1875.2 1875.2
cuccug 1873.2 1873.2 1873.2
cucccg 1872.2 1872.2 1872.2
caaug 1615 1615 1615
cacag 1614 1614 1614
caacg 1614 1614 1614
Open in a new tab

aFragments of >2813.8 Da and <1614 Da are not shown.

bDouble-hydroxylated fragments originating from RNase H digestion.

DISCUSSION

Data on the reliability of RNA secondary structure predictions by computer folding programs are contradictory; some researchers report problems in finding antisense ODN-accessible sites by the use of ‘mfold’ (13,14), whereas Zuker and Jacobson (20) could demonstrate high degrees of correspondence between structures predicted by this program and known structures under certain circumstances. They calculated by ‘mfold’ the secondary structures of different rRNAs with well-known secondary structures, and the identity between predicted and known structures was in the range of 27–70%. However, the degree of correspondence improved to 81% when the analysis was limited to structural domains that were present not only in the optimal predicted secondary structure, but also in all suboptimal structures. Thus one should avoid selecting structural domains with large differences in the calculated suboptimal structures when searching for reliable secondary structure predictions. This is why we only considered FUT2 sequence segments that were predicted to be single stranded in at least 30 of the 38 different foldings calculated by ‘mfold’. Single-stranded loops of hairpins on predicted RNA secondary structures were shown to be good target sites for antisense ODNs (15,21,22); that is why we selected only single-stranded regions for the design of complementary antisense ODNs. As could be demonstrated, all of the four predicted single-stranded sequence segments that we selected have been shown to be good target sites for antisense ODN-mediated RNase H cleavage (see Fig. 1). Thus when intensifying selection criteria, the use of ‘mfold’ can be very helpful in finding antisense-accessible sites on a given RNA sequence. The length of the antisense ODNs was fixed to 18 bases, as this size seems to be a compromise where specificity and cellular uptake are both satisfactory (23). As represented in Figure 1, application of the antisense ODNs at a concentration of 25 µM led to a more or less complete degradation of the RNA. Thus at this high concentration, the respective antisense ODNs hybridized to not only the complementary site, but also to many other sites of the FUT2 in vitro transcript, leading to multiple cleavage events by RNase H. When ODNs 4 and 6 were scaled down to 0.025 and 0.05 µM, respectively, an optimum was found where the FUT2 in vitro transcript was cut at the appropriate sites only, which was not the case when applying control ODNs (see Fig. 2). Hence, at these concentrations, maximum specificity could be reached.

In order to determine the exact cleavage sites produced by RNase H, we decided to use MALDI-TOF mass spectrometry. Knowledge of the cleavage site is of great importance when RNase H activity is to be maintained after chemical modification of the antisense ODN, which is a necessary step when in vivo application is desired. Normal phosphodiester oligonucleotides are readily degraded by nucleases when introduced into a biological system. The most commonly used modified oligonucleotides are phosphorothioate oligonucleotides; they are resistant to nucleolytic attack and support RNase H activity, but they have been shown to produce non-specific toxicity in both primates and rodents (24) and do not bind their target mRNA as tightly as unmodified oligonucleotides do (25), and thus one might choose other chemical modifications. Among these are new generation chemistries such as morpholino phosphorodiamidates (MFs) and 2′-O-methoxyethylribonucleotides (MOEs) that are resistant to nucleases and have a high affinity to hybridize to their target mRNA, but do not support RNase H-mediated cleavage of the DNA–RNA duplex (26,27). It is therefore advisable to develop gapped or chimeric oligonucleotides that contain a mixture of chemistries that allow for high affinity hybridization, resistance to nucleolytic attack and support of RNase H activity [as illustrated in Golden et al. (28)].

A common strategy to analyze the site of RNase H-mediated cleavage is to reverse-transcribe the resulting RNA fragments in a primer extension reaction with a labeled primer and resolve the newly synthesized fragments via electrophoresis (15,29,30). This is a troublesome and to some extent misleading method; strong RNA secondary structures can cause premature stops of reverse transcription reactions, thus generating artifactual fragments that do not result from RNase-mediated digestion. Hansen et al. (31) directly compared the applicability of primer extensions and MALDI-TOF mass spectrometry for the analysis of RNA. They searched for post-transcriptional modifications of 23S rRNA of diverse organisms. The detection of modifications by primer extension was of limited reliability because of premature termination of reverse transcription, probably due to very stable RNA secondary structures. In contrast, analysis of RNase-generated RNA fragments by MALDI-TOF mass spectrometry reliably detected almost all types of modifications, regardless of secondary structure. Thus they could demonstrate that MALDI-TOF mass spectrometry was much more suitable than primer extension for analyzing RNA. The unambiguous and easily practicable determination of RNase-generated RNA fragments by applying MALDI-TOF mass spectrometry has already also been validated by others (16,32–34). After RNase-mediated cleavage, the generated RNA fragments are directly analyzed; therefore, there is no fault-prone need to copy them. Moreover, analysis by MALDI-TOF mass spectrometry is more exact than by electrophoresis, as mass accuracy is high enough not only to determine the length of the sequence segment but also to draw conclusions about the sequence composition as the four bases differ in their masses (16). When applying MALDI-TOF mass spectrometry, most investigators use sequence-specific RNases for the degradation of the RNA (16,31,33,34). However, Hansen et al. (31) applied the non-sequence-specific RNase H in order to generate RNA subfragments of a smaller size suitable for subsequent digestion by sequence-specific RNases; they report the appearance of some unexpected peaks due to heterogeneity of the RNase H cleavage, indicating that their DNA oligonucleotides inducing RNase H-mediated cleavage have not been optimized in sequence and concentration. Optimization was not necessary because determination of RNase H-induced cleavage sites was not the purpose of their study. In contrast, Polo and Limbach (32) wanted to evaluate the efficiency and specificity of cleavage using RNase H. Therefore, they produced an oligoribonucleotide hairpin of 19 nt length and incubated it with RNase H as well as two different chimeric oligonucleotides [5′-r(NNN)d(NNNN) r(NNN)-3′]. The design of the chimeric oligonucleotides allows for a maximum number of four possible cleavage sites as RNase H recognizes only DNA–RNA hybrids; this restriction and the small size of the hairpin enable analysis of expected RNA fragments by MALDI-TOF mass spectrometry without the need of further treatment. However, in the context of searching for antisense ODN-accessible sites along a given RNA, the technique described by Polo and Limbach is not sufficient. In order to detect antisense sequences that can induce RNase H-mediated cleavage in vivo, the transcript must be synthesized in full length to form the same secondary structure as under physiological conditions. This is why we perform cleavage of the full-length transcript by RNase H under native conditions, and thereafter degrade the transcript in a sequence-specific manner by RNase T1, thus producing small RNA subfragments that can easily and unambiguously be analyzed by MALDI-TOF mass spectrometry. Therefore, with the technique presented by us, the analysis of RNA sites that are accessible to antisense ODN and RNase H in vivo by using MALDI-TOF mass spectrometry succeeded for the first time. Given the convincing results, this approach should become an attractive tool for other users of antisense technology.

Acknowledgments

ACKNOWLEDGEMENT

The authors thank Tierzuchtforschung e.V. München for supporting this project.

REFERENCES

  • 1.Zamecnik P.C. and Stephenson,M.L. (1978) Inhibition of Rous sarcoma virus replication and cell transformation by a specific oligodeoxynucleotide. Proc. Natl Acad. Sci. USA, 75, 280–284. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Haeuptle M.T., Frank,R. and Dobberstein,B. (1986) Translation arrest by oligodeoxynucleotides complementary to mRNA coding sequences yields polypeptides of predetermined length. Nucleic Acids Res., 14, 1427–1448. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Walder R.Y. and Walder,J.A. (1988) Role of RNase H in hybrid-arrested translation by antisense oligonucleotides. Proc. Natl Acad. Sci. USA, 85, 5011–5015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Minshull J. and Hunt,T. (1986) The use of single-stranded DNA and RNase H to promote quantitative ‘hybrid arrest of translation’ of mRNA–DNA hybrids in reticulocyte lysate cell-free translations. Nucleic Acids Res., 14, 6433–6451. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Sawai Y., Wada,K. and Tsukada,K. (1980) Change of nuclear ribonuclease H activity following deoxyribonucleic acid synthesis in liver of protein-deficient rat. Life Sci., 26, 1497–1503. [DOI] [PubMed] [Google Scholar]
  • 6.Shuttleworth J. and Colman,A. (1988) Antisense oligonucleotide-directed cleavage of mRNA in Xenopus oocytes and eggs. EMBO J., 7, 427–434. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Inoue H., Hayase,Y., Iwai,S. and Ohtsuka,E. (1987) Sequence-dependent hydrolysis of RNA using modified oligonucleotide splints and RNase H. Nucleic Acids Symp. Ser., 18, 221–224. [PubMed] [Google Scholar]
  • 8.Frauendorf A. and Engels,J.W. (1994) Interaction of linear and folded modified antisense oligonucleotides with sequences containing secondary structure elements. Bioorg. Med. Chem. Lett., 4, 1019–1024. [Google Scholar]
  • 9.Schwille P., Oehlenschlager,F. and Walter,N.G. (1996) Quantitative hybridization kinetics of DNA probes to RNA in solution followed by diffusional fluorescence correlation analysis. Biochemistry, 35, 10182–10193. [DOI] [PubMed] [Google Scholar]
  • 10.Jaroszewski J.W., Syi,J.L., Ghosh,M., GhoshmK. and Cohen,J.S. (1993) Targeting of antisense DNA: comparison of activity of anti-rabbit beta-globin oligodeoxyribonucleoside phosphorothioates with computer predictions of mRNA folding. Antisense Res Dev., 3, 339–348. [DOI] [PubMed] [Google Scholar]
  • 11.Zuker M., Mathews,D.H. and Turner,D.H. (1999) Algorithms and thermodynamics for RNA secondary structure prediction: a practical guide. In Barciszewski,J. and Clark,B.F.C. (eds), RNA Biochemistry and Biotechnology. NATO ASI Series. Kluwer Academic Publishers, Dordrecht, pp. 11–43. [Google Scholar]
  • 12.Mathews D.H., Sabina,J., Zuker,M. and Turner,D.H. (1999) Expanded sequence dependence of thermodynamic parameters improves prediction of RNA secondary structure. J. Mol. Biol., 288, 911–940. [DOI] [PubMed] [Google Scholar]
  • 13.Ho S.P., Bao,Y., Lesher,T., Malhotra,R., Ma,L.Y., Fluharty,S.J. and Sakai,RR. (1998) Mapping of RNA accessible sites for antisense experiments with oligonucleotide libraries. Nat. Biotechnol., 16, 59–63. [DOI] [PubMed] [Google Scholar]
  • 14.Sohail M., Akhtar,S. and Southern,E.M. (1999) The folding of large RNAs studied by hybridization to arrays of complementary oligonucleotides. RNA, 5, 646–655. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Matveeva O., Felden,B., Audlin,S., Gesteland,R.F. and Atkins,J.F. (1997) A rapid in vitro method for obtaining RNA accessibility patterns for complementary DNA probes: correlation with an intracellular pattern and known RNA structures. Nucleic Acids Res., 25, 5010–5016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Kirpekar F., Douthwaite,S. and Roepstorff,P. (2000) Mapping posttranscriptional modifications in 5S ribosomal RNA by MALDI mass spectrometry. RNA, 6, 296–306. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Krebs S., Seichter,D. and Forster,M. (2001) Genotyping of dinucleotide tandem repeats by MALDI mass spectrometry of ribozyme-cleaved RNA transcripts. Nat. Biotechnol., 19, 877–880. [DOI] [PubMed] [Google Scholar]
  • 18.Lima W.F. and Crooke,S.T. (1997) Cleavage of single strand RNA adjacent to RNA–DNA duplex regions by Escherichia coli RNase H1. J. Biol. Chem., 272, 27513–27516. [DOI] [PubMed] [Google Scholar]
  • 19.Krebs S., Medugorac,I., Seichter,D. and Förster,M. (2003) RnaseCut: a MALDI mass spectrometry-based method for SNP discovery. Nucleic Acids Res., 31, e37. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Zuker M. and Jacobson,A.B. (1995) ‘Well-determined’ regions in RNA secondary structure prediction: analysis of small subunit rRNA. Nucleic Acids Res., 23, 2791–2798. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Lima W.F., Monia,B.P., Ecker,D.J. and Freier,S.M. (1992) Implication of RNA structure on antisense oligonucleotide hybridization kinetics. Biochemistry, 31, 12055–12061. [DOI] [PubMed] [Google Scholar]
  • 22.Thierry A.R., Rahman,A. and Dritschilo,A. (1993) Overcoming multidrug resistance in human tumor cells using free and liposomally encapsulated antisense oligodeoxynucleotides. Biochem. Biophys. Res. Commun., 190, 952–960. [DOI] [PubMed] [Google Scholar]
  • 23.Probst J.C. (2000) Antisense oligodeoxynucleotide and ribozyme design. Methods, 22, 271–281. [DOI] [PubMed] [Google Scholar]
  • 24.Henry S.P., Templin,M.V., Gillett,N., Rojko,J. and Levin,A.A. (1999) Correlation of toxicity and pharmacokinetic properties of a phosphorothioate oligonucleotide designed to inhibit ICAM-1. Toxicol. Pathol., 1, 95–100. [DOI] [PubMed] [Google Scholar]
  • 25.Akhtar S., Kole,R. and Juliano,R.L. (1991) Stability of antisense DNA oligodeoxynucleotide analogs in cellular extracts and sera. Life Sci., 24, 1793–801. [DOI] [PubMed] [Google Scholar]
  • 26.Summerton J., Stein,D., Huang,S.B., Matthews,P., Weller,D. and Partridge,M. (1997) Morpholino and phosphorothioate antisense oligomers compared in cell-free and in-cell systems. Antisense Nucleic Acid Drug Dev., 2, 63–70. [DOI] [PubMed] [Google Scholar]
  • 27.Baker B.F., Lot,S.S., Condon,T.P., Cheng-Flournoy,S., Lesnik,E.A., Sasmor,H.M. and Bennett,C.F. (1997) 2′-O-(2-Methoxy)ethyl-modified anti-intercellular adhesion molecule 1 (ICAM-1) oligonucleotides selectively increase the ICAM-1 mRNA level and inhibit formation of the ICAM-1 translation initiation complex in human umbilical vein endothelial cells. J. Biol. Chem., 18, 11994–2000. [DOI] [PubMed] [Google Scholar]
  • 28.Golden T., Dean,N.M. and Honkanen,R.E. (2002) Use of antisense oligonucleotides: advantages, controls and cardiovascular tissue. Microcirculation, 1, 51–64. [DOI] [PubMed] [Google Scholar]
  • 29.Scherr M., LeBon,J., Castanotto,D., Cunliffe,H.E., Meltzer,P.S., Ganser.A., Riggs,A.D. and Rossi,J.J. (2001) Detection of antisense and ribozyme accessible sites on native mRNAs: application to NCOA3 mRNA. Mol. Ther., 4, 454–460. [DOI] [PubMed] [Google Scholar]
  • 30.Lloyd B.H., Giles,R.V., Spiller,D.G., Grzybowski,J., Tidd,D.M. and Sibson,D.R. (2001) Determination of optimal sites of antisense oligonucleotide cleavage within TNFα mRNA. Nucleic Acids Res., 29, 3664–3673. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Hansen M.A., Kirpekar,F., Ritterbusch,W. and Vester,B. (2002) Posttranscriptional modifications in the A-loop of 23S rRNAs from selected archaea and eubacteria. RNA, 2, 202–213. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Polo L.M. and Limbach,P.A. (1998) Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry for the analysis of RNase H cleavage products. J. Mass Spectrom., 12, 1226–1231. [DOI] [PubMed] [Google Scholar]
  • 33.Hahner S., Ludemann,H.C., Kirpekar.F., Nordhoff,E., Roepstorff,P., Galla,H.J. and Hillenkamp,F. (1997) Matrix-assisted laser desorption/ionization (MALDI) mass spectrometry of endonuclease digests of RNA. Nucleic Acids Res., 25, 1957–1964. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Hartmer R., Storm,N., Boecker,S., Rodi,C.P., Hillenkamp,F., Jurinke,C. and van den Boom,D. (2003) RNase T1 mediated base-specific cleavage and MALDI-TOF MS for high-throughput comparative sequence analysis. Nucleic Acids Res., 31, e47. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Nucleic Acids Research are provided here courtesy of Oxford University Press

RESOURCES