Empirical Validation of Overlapping Virus lncRNAs and Coding Transcripts by Northern Blot

Abstract

Viruses, like their metazoan hosts, have evolved to utilize intricate transcriptional mechanisms to generate a vast array of both coding and noncoding RNA transcripts. The resolution of specific noncoding RNA transcripts produced by viruses, particularly herpesviruses, presents a particularly difficult challenge due to their highly dense dsDNA genomes and their complex, overlapping, and context-dependent network of transcripts. While new long read sequencing platforms have facilitated the resolution of some noncoding transcripts from virus genomes, empirical molecular validation of transcripts from individual regions is essential. Herein, we demonstrate that the use of strand specific northern blots is essential for true validation of specific viral noncoding RNAs, and provide here a detailed molecular method for such an approach.

1. Introduction

Over the last 15 years, the rapid advance of new sequencing technologies has greatly expanded our understanding of the vast complexity of host and pathogen transcriptomes and has led to the identification of thousands of new noncoding RNA transcripts [1, 2]. Through this work it has become clear that a large number unannotated regions of metazoan genomes are actively transcribed [3, 4]. However, perhaps most emblematic of such complexity are the highly dense dsDNA genomes of herpesviruses. Despite their small genomic sizes, ranging from 100 to 250 kb, herpesviruses carry up to 235 protein coding genes which are expressed bidirectionally [5]. Moreover, these viruses have demonstrated the capacity to generate hundreds of additional coding and noncoding RNA molecules through antisense, intergenic, and readthrough transcription, as well as alternative splicing and alternate promoter usage [6–9].

The vast majority of the noncoding RNAs that have been identified in viruses are derived from herpesvirus family members [10]. Thus, it has been presumed that these viruses harbor numerous other noncoding transcripts that have yet to be discovered. However, the transcriptional intricacy of herpesviruses has provided an enormous obstacle for resolution of individual noncoding RNA isoforms. Recently though, the advent of single molecule long read sequencing has provided a catalyst for global resolution of RNAs from many of these genomes. For example, we have recently used the TRIMD pipeline [11] to integrate parallel data sets from Pacific Biosciences SMRT Iso-Seq long read sequencing, Illumina short read RNA-Seq, and 5′ cap analysis of gene expression (deepCAGE) platforms to globally resolve transcript structures from murine gammaherpesvirus 68 (MHV68) [12]. Not surprisingly, 25 putative long noncoding RNAs (lncRNAs) were among the transcripts identified using this approach.

As with metazoan lncRNAs, ascribing a biological role for such newly discovered viral lncRNAs provides a significant challenge [13–15]. In the field, the study of viral lncRNA function typically begins with examination of expression levels at different time points, in varying conditions, or in specific cell types [16–18]. Assessment of lncRNA expression level is most frequently accomplished using qRT-PCR assays using primers purposely designed for a short region thought to span the transcript of interest. However, this method requires the a priori knowledge of every transcript spanning the region of interest—comprehensive information that is not always available using even the most robust long read platforms for discovery. Thus, it is our experience that detailed molecular validation through northern blot should be undertaken for any individual region under study.

Here we provide a specific example of this approach using parallel qRT-PCR and northern blot analyses of viral noncoding RNA isoforms discovered through integrated long read sequencing of RNA from MHV68-infected cells [12]. These putative noncoding transcripts, which range in size from 1.9 to 5.4 kb (Fig. 1a), lie directly antisense to open reading frames (ORFs) 63 and 64, and utilize multiple 5′ transcription initiation sites as well as alternative splicing. qRT-PCR analysis clearly identified robust expression of at least one of these transcripts (Fig. 1b), generating a single amplicon (Fig. 1c). Although northern blot analysis using probes directed to the same exact region validated the presence of the four transcripts predicted to be most abundant (Fig. 1d), this approach also revealed that a major surprise: the single most abundant RNA in this region is an approximately 10 kb transcript not previously identified through modern transcriptomics methods. Such information is essential not only for qRT-PCR-based assessment of noncoding RNA expression, but is also directly applicable to the design and validation of viral mutants. Below we provide a detailed molecular protocol for the northern blot validation of viral transcripts.

Fig. 1.

Comparison of northern blot and qRT-PCR analyses of noncoding RNA transcripts identified through modern transcriptomics. (a) Noncoding RNA transcripts identified using TRIMD pipeline integration of data from multiple transcriptomics platforms. A schematic of the MHV68 dsDNA genome is depicted to emphasize the location of ORF63 and ORF64 on the forward strand. TRIMD-identified RNA transcripts which lie antisense to ORF63 and ORF64 indicated, along with TRIMD score and transcript size. qRT-PCR amplicon and RNA probe site is shown in red. (b) qRT-PCR analysis of ORF63/64 antisense transcripts. qRT-PCR was used to amplify a 188 bp amplicon from the ORF63/64 transcript region. DNase-treated RNA was reverse transcribed using either random hexamers or gene- and strand-specific primer (GGCATTGTCCTCATCACCCAG), and real-time PCR was performed using forward (AAATTGGGCCTCTTATCTCCTGG) and reverse (GTTTGAATAATTGGCCGTCAGC) primers. Y-axis indicates raw C_t value. (c) Agarose gel visualization of qRT-PCR amplicons. Amplicons generated from (b) were visualized on a 3% agarose gel. (d) Northern blot analysis of ORF63/64 antisense transcripts. Northern blot was performed exactly as indicated in the accompanying protocol. Labels indicate putative noncoding RNA transcripts that match the size of northern blot bands

2. Materials

To avoid RNA degradation, prepare all buffers, reagents and materials under sterile conditions. Prepare DEPC-treated water by adding 1 mL Diethyl pyrocarbonate (MP Bio) to 1 L of ultrapure deionized water, incubating at room temperature overnight, and autoclaving. This protocol is designed to detect specific murine gammaherpesvirus 68 (MHV68) transcripts, but can be applied to any other host or virus transcripts.

2.1. RNA Extraction

1. TRIzol® Reagent (Thermo Fisher Scientific).

2. Chloroform.

3. Isopropanol.

4. 75% EtOH.

5. GlycoBlue (Invitrogen).

6. DEPC-treated water.

2.2. Formaldehyde Agarose Gel Electrophoresis

1. Ultrapure Agarose (Invitrogen).

2. 10× MOPS buffer: Dissolve 41.9 g MOPS, 8.2 g sodium acetate anhydrous (13.6 if it is trihydrate), and 3.7 g EDTA in 1 L water. Adjust pH to 7.0 and wrap in aluminum foil to protect from light. Store at room temperature.

3. Formaldehyde 37%.

4. 20× SSC: Dissolve 175.3 g NaCl and 88.2 g Sodium Citrate in 1 L water, autoclave and store at room temperature.

5. 2× RNA loading buffer: 95% formamide, 0.01% SDS, 0.01% Bromophenol Blue, 0.005% xylene cyanol, 0.5μM EDTA.

6. Millennium RNA Marker (Ambion). This ladder provides excellent resolution for transcripts from 0.5 kb to 9 kb. To avoid freeze-thawing, divide into 2μL aliquots and store at −80 °C.

2.3. Transfer, Blotting, Radiolabeling and Hybridization

1. Amersham Hybond N+ membrane (GE Healthcare).

2. Whatman Blotting paper GB003 (0.8 mm).

3. Whatman blotting paper 3MM.

4. ULTRAhyb hybridization buffer (Ambion).

5. Methylene blue solution: 0.02% methylene blue, 0.3 M sodium acetate.

6. DNA polymerase (NEB Q5 2× MM).

7. MAXIscript T7 Transcription Kit (Invitrogen).

8. CTP (α−³²P) 800 Ci/mmol 10 mCi/mL EasyTide 250μCi (PerkinElmer BLU508X250UC). Use fresh reagent if possible, as the half-life of ³²P is 14.3 days. If the reagent is nearing 14 days, double the quantity used for the labeling reaction.

9. Agarose gel running system (Thermofisher Owl B3), power supply.

10. UV Crosslinker.

11. Hybridization oven.

12. Autoradiography cassette, film, and developer.

13. Wash buffer 1: 2× SSC and 0.1% SDS.

14. Wash buffer 2: 1× SSC and 0.1% SDS.

15. Wash buffer 3: 0.1× SSC and 0.1% SDS.

3. Methods

To avoid RNA degradation, it is important to clean the pipettes and the work area with RNaseZap. If possible, use separate workspace and equipment for RNA work.

3.1. RNA Extraction

1. Infect one 10 cm dish (2 × 10⁶ cells per dish) or one well of a 6-well plate (2 × 10⁵ per well) of NIH 3T12 cells at multiplicity of infection (MOI) of 5 with MHV68.

2. At 6, 12 or 18 h post-infection, remove media and harvest cells in 1 mL TRIzol. Add TRIzol to the cells and transfer the solution into 1.5 mL Eppendorf tubes. If desired, at this stage TRIzol lysates can be stored at −20 °C for a few days or at −80 °C for up to a year.

3. Add 0.2 mL chloroform to 1 mL TRIzol lysate. Mix by shaking for 15 s, then place on ice for 5 min. Centrifuge at 16,000 × g for 15 min at 4 °C.

4. Transfer the top clear aqueous phase into a new 1.5 mL Eppendorf tube and add 0.5 mL isopropanol. Mix by inversion and then let stand at room temperature for 5–10 min. Centrifuge at 16,000 × g for 15 min at 4 °C.

5. Add GlycoBlue for visualization of the pelleted RNA.

6. Wash the RNA pellet with 1 mL of 75% ice cold ethanol and transfer it to 1.5 mL Eppendorf tube. Spin at 16,000 × g for 10 min at 4 °C.

7. Remove all of the EtOH by air drying, then resuspend the RNA in 50–100μL of DEPC-treated water. At this stage, purified RNA can be stored at −80 °C. For NIH 3T12 fibroblasts, this procedure yields approximately 1μg/μL of RNA in a volume of 50μL.

3.2. Formaldehyde Gel Electrophoresis

1. Prepare 1.0% agarose gel with formaldehyde: Microwave 73 mL DEPC-treated water and 1 g Ultrapure agarose. Allow the solution to cool to 55 °C, then add 10 mL of 10× MOPS buffer and 16.2 mL of 37% Formaldehyde. Cast the gel in a fume hood by using the thinner side of a 12 well comb (see Note ¹). Percent agarose can be adjusted from 0.8% to 1.5% depending upon the size of the target transcript; however, 1% agarose generally works well for transcripts that range from 0.5 to 9 kb.

2. Mix the RNA samples with 2× RNA loading buffer. 4–5μg total RNA in ~15–20μL final volume generates sharp, nicely visible bands seen by methylene blue staining. Optionally, EtBr can be added into the loading buffer for RNA visualization under UV. Heat the samples at 70 °C for 5 min then chill on ice.

3. Load each sample into one lane of the gel. If extra wells are available, avoid the wells near the edges and use the middle portion of gel to minimize the risk of aberrant migration of RNA.

4. Run the gel at 70–80 V for 3–4 h in 1× MOPS running buffer containing formaldehyde. The bromophenol blue runs around 0.3 kb (300 nts) and xylene cyanol runs around 4.0 kb. Use this indicator to determine how long to run the gel. For example, stopping the gel run when the bromophenol blue is 2–3 cm from the bottom of the gel should provide good resolution for 1.0 to 5.0 kb transcripts (see Note ²).

5. At this point, ribosomal bands on the gel can be visualized using a UV box or handheld UV light. To optimize fit to transfer membrane, excess gel can be removed using a razor blade.

3.3. Transfer to Membrane

1. Wash the gel four times with deionized water for 10 min each. Then wash with 1× SSC buffer for 15 min.

2. Prepare Hybond N+ membrane slightly larger than the size of the gel. Soak in 1× SSC for 2–3 min.

3. Set up the transfer system as shown in Fig. 2: First place 20 layers of dry GB003 Whatman paper, then 1 layer of dry 3MM Whatman paper. Soak one additional layer of 3MM Whatman paper in 1x SSC and then place on top. Next place the soaked Hybond N+ membrane, then place the gel onto the membrane. Remove any air bubbles by carefully rolling over the gel with a sterile pipette. Place two layers of soaked 3MM Whatman paper on top. Finally, place (as shown in the Fig. 2) the soaked final layer of 3MM Whatman paper with larger dimensions to serve as the wick for transfer buffer. 20× SSC will flow through this layer to the dry paper, enabling the transfer of the RNA to the membrane. Cover the transfer system with a lid to protect the transfer buffer from evaporation (do not place additional weight). Transfer without disturbance overnight (see Note ³).

4. The next morning, disassemble the transfer system. If the transfer worked, 20× SSC will be soaked to the lower layers of Whatman paper and the gel should be almost dry. Carefully remove the membrane, keeping track of which side was adjacent to the gel. Place in a glass container that has been cleaned with RNaseZap. Use a pencil to carefully mark the side of the membrane to indicate the location of the bromophenol blue (approximately 0.3 kb) and xylene cyanol (approximately 4 kb) bands. Gently soak the membrane in DEPC-treated water to remove running buffer salts.

5. Crosslink the RNA to the membrane using UV light at 1200μJ/m². Repeat crosslink. Soak the membrane for 5 min in methylene blue solution. Wash the membrane twice with deionized water. At this stage, 18S rRNA (approximately 1.9 kb) and 28S rRNA (approximately 5.0 kb) and the RNA ladder will become visible. Carefully mark the side of the membrane with the positions of the marker and ribosomal bands with a pencil to ensure correct membrane orientation moving forward (see Note ⁴).

Fig. 2.

Setup of northern blot downward capillary transfer

3.4. Preparation of the Radiolabeled RNA Probe and Hybridization

1. Prepare the labeled probe by in vitro transcription by PCR amplifying the target region using primer containing T7 promoter sequence upstream of 5′ primer sequence (if desired, Sp6 promoter sequence can be used upstream of 3′ primer sequence to generate complementary strand probe for strand-specific control. The PCR product can be stored in −20 °C in aliquots for a few weeks. For the best results, the in vitro transcribed RNA probe should be used fresh each time (see Note ⁵).

   1. Prepare the PCR reaction by mixing 1μL of 10μM Forward primer with T7 promoter sequence at the 5′ (TAA TACGACTCACTATAGGG…), 1μL of 10μM Reverse primer, 1μL of the DNA template, 22μL of nuclease-free water, and 25μL of Q5 DNA polymerase 2× Mastermix.
   2. PCR amplify using the following cycling conditions: (1) 98 °C for 2 min (initial denaturation); (2) 98 °C for 15 s (denaturation); (3) 55 °C for 15 s (primer annealing temperature may be optimized for each primer set. However, the T7/Sp6 promoter sequences should not be included while calculating the T_m); (4) 72 °C for 60 s/kb (extension), Repeat steps 2–4 30 cycles; and 72 °C for 2 min (final extension).
   3. Run on agarose gel and purify amplicon using gel purification kit.
   4. Prepare in vitro transcription reaction by mixing Gel purified PCR amplicon (at least 300 ng), 1μL from each of ATP, GTP, UTP from the MAXIscript Transcription Kit, 1μL of T7 or SP6 Enzyme mix (depending on the promoter used), 2μL of 10× in vitro transcription buffer, Add up to 19μL with nuclease free water and 1μL of α−³²P CTP.
   5. Incubate at 37 °C for 2–4 h (1 h is generally enough but if the target region is small, increasing the incubation time and enzyme will yield higher activity probe).
   6. Add 1μL of DNase to the reaction, then incubate 20 min at 37 °C.
   7. Add 1μL of 0.5 M EDTA to stop DNase reaction and incubate at 95 °C for 5 min.

2. Prewarm the ULTRAhyb buffer at 60 °C for at least 30 min prior to use. Place the UV cross-linked membrane in the hybridization tube with the side that was adjacent to the gel facing inward. Add 5 mL of the prewarmed hybridization buffer to the tube. Prehybridize the membrane for at least 1 h; this can be accomplished while preparing the radiolabeled probe.

3. Add all of the probe to the hybridization bottle. When using the ULTRAhyb buffer, it is not necessary to change the buffer after prehybridization or remove unincorporated radionucleotides from the in vitro transcription reaction (see Note ⁶).

4. Rotate overnight at 60 °C in the hybridization oven. This temperature may be optimized for best results based on the length and GC content etc. of the probe being used (see Note ⁷).

5. Wash the membrane with wash buffer 1 for 20 min at 72 °C. Repeat with wash buffer 2, then repeat with wash buffer 3. All of the washes should be disposed in radioactive liquid waste containers (see Note ⁸).

6. Drip excess liquid from the membrane and then wrap with Saran Wrap and seal the sides. In a darkroom, place the membrane into an exposure cassette, followed by autoradiography film and intensifying screen. Expose at −80 °C. Depending on the abundance of the RNA, exposure time will need to be optimized from a few hours to several days. Overnight exposure is a good starting point for many RNAs, but with experience initial exposure times can be adjusted by assessing membrane radioactivity using a Geiger counter.

7. Develop the film in the dark room.

8. In our experience, stripping northern membranes is extremely difficult and generates a large amount of radioactive liquid waste. If a membrane must be reused for control probes, it is advised to begin with probes to expected low abundance targets, followed subsequent use of probes for expected higher abundance transcripts.