 |
 |
GEO help: Mouse over screen elements for information. |
|
| Status |
Public on Oct 14, 2024 |
| Title |
Spinach, M2 DMS MaP, 170 mM, SSII |
| Sample type |
SRA |
| |
|
| Source name |
in vitro transcribed RNA
|
| Organism |
synthetic construct |
| Characteristics |
tissue: in vitro transcribed RNA
genotype: Pool of randomly mutagenized RNA produced through error-prone PCR of T7 template
treatment: DMS
|
| Treatment protocol |
DMS modification of in vitro transcribed RNA
In vitro transcribed RNA was probed as follows. 200 ng – 1 µg of RNA was folded in a buffer designed to promote modification of A and C bases only (200 mM Bicine pH 7.75 at room temperature, 100 mM KCl) or a buffer which promotes the additional modification of m1G and m3U (200 mM Bicine pH 8.37 at room temperature, 100 mM KCl). RNA was refolded in buffer by heating to 95°C for 3 min then snap cooling on ice for 1 min. Then MgCl2 was added to a final concentration of 1-5 mM and the RNA was left at room temperature for 10 min. Neat DMS was diluted to 1.7M with anhydrous ethanol and added to the folded RNA. DMS reactions was performed for 10 min at room temperature or 6 min at 37°C. Spinach was probed at room temperature due to its thermal instability. pUG RNA was probed at 37°C. Reactions were quenched by the addition of 4 volumes of 20% beta-mercaptoethanol in water on ice. RNA was purified from the quenched DMS reaction with RNA Clean and Concentrator 5 kit and eluted in 10µL water. RNA was either directly processed or stored at -80°C.
In cell DMS modification and purification
DMS modification of SH-SY5Y cells was performed as follows. Cells were grown to 85% confluence in a 10 cm dish. Then, the media was exchanged for a DMS probing media which consisted of 1:1 DMEM to F12 medium with 10% FBS and 200 mM Bicine adjusted to pH 8.37 with NaOH at room temperature. Probing media was prewarmed to 37C and cell media was exchanged with 5.4 mL probing media for 3 min at 37°C. Then 600µL of prewarmed 1.7 M DMS or prewarmed 600µL ethanol was added to the cells and incubated for 6 min at 37°C. DMS reaction was stopped by the addition of 6 mL 20% 2-mercaptoethanol and cells were placed on ice. We noticed that DMS probing caused a large portion of adherent cells to become detached from the plate. We therefore decanted the quenched DMS probing media off the plate into a 15 mL falcon tube and pelleted the detached cells for 5 min at 4C and 1,000 x g. To isolate the total RNA, 2 mL Trizol reagent was added to the plate for the cells which had not become detached. Then, the 2 mL Trizol reagent was transferred to solubilize the pelleted cells which had become detached during DMS probing. Total RNA was then isolated according to manufacturer’s instructions.
M2 BASH MaP
Spinach and polyUG T7 templates were randomly mutagenized by performing 24 rounds of error-prone PCR. Mutagenized Spinach and polyUG were then in vitro transcribed, folded and DMS modified as described prior with minor modifications. MgCl2 was added to a final concentration of 5 mM and Spinach was probed in the presence of 5µM DFHBI-1T. polyUG RNA was probed in the presence of 2µM NMM.
|
| Growth protocol |
HeLa cells were grown in DMEM containing 10% FBS with 100 U ml−1penicillin and 100 µg ml−1streptomycin under standard culturing conditions (37°C with 5.0% CO2). SH-SY5Y cells were grown in a 1:1 ratio of DMEM to F12 media containing 10% FBS with 100 U ml−1penicillin and 100 µg ml−1streptomycin under standard culturing conditions (37°C with 5.0% CO2).
|
| Extracted molecule |
other |
| Extraction protocol |
In vitro transcribed RNA was purified from the quenched DMS reaction with RNA Clean and Concentrator 5 kit and eluted in 10µL water. RNA was either directly processed or stored at -80°C
We noticed that DMS probing caused a large portion of adherent cells to become detached from the plate. We therefore decanted the quenched DMS probing media off the plate into a 15 mL falcon tube and pelleted the detached cells for 5 min at 4C and 1,000 x g. To isolate the total RNA, 2 mL Trizol reagent was added to the plate for the cells which had not become detached. Then, the 2 mL Trizol reagent was transferred to solubilize the pelleted cells which had become detached during DMS probing. Total RNA was then isolated according to manufacturer’s instructions.
Sequencing library preparation:
Reverse transcription
Superscript II: Reverse transcription with SSII was performed as follows. 200 ng of input RNA was mixed with 1µL 50µM random hexamer (HeLa total RNA) or 2µL of 1µM gene specific reverse transcription primer (in vitro transcribed RNAs) and annealed. Annealed RNA was added to reverse transcription master mix containing 6 mM final concentration of MnCl2. Reverse transcription reactions were incubated at 23°C for 15 min only when using random hexamers then 42°C for 3 hours.
Marathon RT: Reverse transcription with Marathon RT was performed with a modified protocol. First, 200 ng of input RNA was mixed with 2µL10 mM dNTP and 1µL 50µM random hexamer or 2µL of 1µM gene specific RT primer. The volume was brough up to 8µL and the RNA was annealed to the primer with the following thermocycler settings: 80°C for 1 min, 65°C for 5 min at which point the RNA was placed on ice for 2 min. Then, an RT master mix was created containing 4µL of 5x Marathon MaP buffer (250 mM Tris-HCl pH 7.5, 1 M NaCl), 4µLof 100% glycerol, 1µL of 100 mM DTT, 0.5µL of 20 mM MnCl2, 0.5µL of water. Next, 10µL of RT master mix was added to 8µL of annealed RNA followed by the addition of 2µL Marathon RT. Reverse transcription reactions were incubated at 23°C for 15 min only when using random hexamers then 42°C for 3 hours.
Reverse transcription reactions were inactivated by incubating at 75°C for 10 min. cDNA was purified with DNA clean and concentrator 5 columns by using 7 volumes of DNA binding buffer (140 µL) and eluted in 10µL water.
Sequencing libraries were generated via a 2 step PCR. Briefly, 1/5th of purified cDNA (2 µL) was amplified with step 1 forward and reverse primers. HeLa 18s rRNA samples included a 4-nucleotide barcode on the forward primer and a 7-nucleotide barcode on the reverse primer which enabled pooling of purified step 1 PCR products. AKT2 step 1 PCR primers were designed to amplify a ∼200 nucleotide segment of the AKT2 3’UTR. In vitro transcribed RNAs used non-barcoded step 1 forward and reverse primers which were specific to flanking sequences introduced into the RNA constructs. These flanking sequences are referred to as a Structure Cassette (SC). Step 1 PCR reactions were assembled on ice by adding to 2µL purified cDNA: 1.25µL of 10µM forward primer, 1.25µL of 10µM reverse primer, 9µL of water and 12.5µL of 2x Phusion HF master mix or 2x Phusion GC master mix for AKT2 step 1 PCR. Step 1 PCR reactions were run with the following thermocycler settings: 98°C for 2 min followed by 8-28 cycles of 98°C for 10 seconds, 65°C for 20 seconds, 72°C for 20 seconds, followed by a final extension of 2 min. PCR cycle number was optimized for each experiment separately. PCR reactions were purified with 1.8x AMPure XP beads according to manufactures instructions. Purified DNA was eluted in 25µL water.
Step 2 PCR was performed with NEB Next Multiplex Oligos for Illumina. Step 2 PCR reactions were composed of the following: 2-10 ng of purified step 1 DNA brough up to 12µL water, 4µL of i5 index primer or universal primer, 4µL i7 index primer, and 20µL of 2x Phusion HF or 2x Phusion GC for AKT2 library prep. Step 2 PCR reactions were run with the following thermocycler settings: 98°C for 2 min followed by 6-12 cycles of 98°C for 15 seconds, 65°C for 30 seconds, 72°C for 30 seconds, then 72°C for 5 min. PCR cycle number was optimized for each experiment separately. Step 2 PCR reactions were size selected with AMPure XP beads, eluted in 25µL of water, and quantified with Qubit 1x dsDNA HS kit. Libraries were pooled and sequenced on the NovaSeq 6000 PE 2x100, NovaSeq 6000 PE 2x150, MiSeq PE 2x150, Nextseq2000 P1 600 cycles, or MiSeq Micro 300 cycles.
Gene-specific BASH MaP:
Total RNA was treated with DNAse I for 30 min prior to BASH MaP treatment according to manufacturer’s instructions. Three replicates of 1 µg of total RNA was mixed with 1µL 50µM random hexamers and the volume was adjusted to 11 µL. RNA was annealed by heating to 85°C for 1 min, 65°C for 5 min, then placing on ice for 3 min. Reverse transcription was performed with SuperScript II as described previously8. Reverse transcription reactions were incubated at 23°C for 15 min and then 42°C for 3 hours. Reactions were inactivated by heating to 75°C for 10 min. Reverse transcription replicates were pooled and the cDNA was purified with the Zymo DNA Clean and Concentrator 5 kit and eluted in 10µL water.
|
| |
|
| Library strategy |
OTHER |
| Library source |
other |
| Library selection |
other |
| Instrument model |
Illumina MiSeq |
| |
|
| Description |
Duration: 8 min;temperature: 25C;buffer: 200 mM Bicine pH 8.3, 150 mM KCl, 5 mM MgCl2;ligand: DFHBI-1T 5 uM
M2 Spinach DMS MaP
|
| Data processing |
Sequencing data was processed using ShapeMapperV2.2 (https://github.com/Weeks-UNC/shapemapper2)
For HeLa 18s rRNA experiments, sequencing data was aligned to 12 unique 18S amplicons defined by the unique combinations of 4-nucleotide forward and 7-nucleotide reverse barcode sequences inserted during step 1 PCR. Sequencing data alignment was performed with ShapeMapperV2.2 with the following settings: --amplicon --output-parsed --output-aligned-reads --nproc 15 --output-counted-mutations --min-mutation-separation 0.
For Spinach, pUG, and AKT2 experiments, sequencing data was aligned to a single fasta sequence with ShapeMapperV2.2 with the following settings: --amplicon --dms --output-parsed - -min-mutation-separation 2. Deletions are ambiguously aligned and therefore ignored by invoking the --dms option.
Misincorporation signatures were calculated by aligning the data with ShapeMapperV2.2 with the following settings: --amplicon --output-counted-mutations --min-mutation-separation 0.
Counts of misincorporation types for each position within an RNA are tabulated by the --output-counted-mutations field. A minimum misincorporation separation distance of 0 was chosen to capture the full range of misincorporation types generated in BASH MaP and DMS MaP experiments.
ShapeMapper alignment with --output-parsed produces a .mut file which encodes the misincorporations within a sequencing read as a bitvector where 0 represents no misincorporation and 1 represents a misincorporation. This .mut file was then input into the program RingMapper which identifies positions with statistically high rates of co-occurring misincorporations7. The output of RingMapper was processed with the custom python script ring_pair_to_heatmap.py which filters all negative correlations and creates a square matrix which was then visualized in Prism GraphPad.
For M2 Spinach and pUG experiments, sequencing data was aligned to a single fasta sequence with ShapeMapperV2.2 with the following settings: --amplicon --dms --output-parsed --min- mutation-separation 2. The output parsed mutation file was then processed with the custom script mut_to_simple.py which retains only sequencing reads that cover the entire length of the RNA. Then the output .simple file was used to calculate a unique mutational profile for each point mutant with the script simple_to_M2_map.py. This script produces a square matrix with length equal to the length of the RNA and displays how often two mutations co-occur on the same sequencing read. To better visualize changes in nucleotide reactivity given a mutation at a certain position, Z-score normalization was performed with the script M2_Map_to_Zscores.py.
The normalized matrix was then visualized in Prism GraphPad as a heatmap.
RingMapper was run with default parameters and the output file was filtered to include only positive correlations between two G nucleotides. A second filter removed all correlations with Z-scores less than 2.0. Networks of structurally related nucleotides were visualized in Cytoscape with unique G bases represented as nodes and positive correlations represented as edges.
Minimum free energy RNA secondary structure modeling:
The baseline secondary structure model of Spinach was generated in mFold with default parameters. Population average DMS guided RNA secondary structure modeling was performed as described previously in the DANCE pipeline. Briefly, the python script DanceMapper.py was run on parsed mutation output files with the following settings for deriving population average data: --fit --maxc 1. Then, the population average reactivities were converted to free energy restraints for the RNA folding program RNAstructure through the foldClusters.py script. When processing BASH MaP data, the additional command --nog was used with foldClusters.py to ignore the reactivity data for G nucleotides. The resulting RNA secondary structures were visualized in RNA2D drawer. Mapping of per nucleotide DMS reactivity onto RNA2D drawer secondary models was performed through the custom python script color_reactivities_RNA2D_drawer.py. Conformation specific RNA secondary structure modeling was performed by increasing the maximum allowed conformations from 1 to 5 with -- maxc 5.
Single molecule probabilistic RNA secondary structure modeling:
A custom pipeline was developed to apply the DaVinci structural analysis method to BASH MaP56. This analysis method utilizes misincorporations to generate a folding constraint for each sequencing read. Folding constraints are then passed to ContraFold which folds each sequencing read as a unique RNA molecule. The complete analysis pipeline is as follows:
First, sequencing data is processed through ShapeMapper as described above with --amplicon --dms --output-parsed --min-mutation-separation 2 options enabled. The resulting .mut file is then filtered and converted to a .simple file with mut_to_simple.py. For DMS MaP experiments, the resulting .simple file is converted into a .bit file by running the simple_to_bit.py python script. For BASH MaP experiments, the resulting .simple file is instead converted into a .bit file with the script updated_simple_to_bit_rG42.py. This script requires an input fasta file and sets all upper-case G positions to be considered for base pairing. In addition, the script constrains all lowercase letters to be single stranded in the .bit file. If a sequencing read harbors a misincorporations at the position of one of the lowercase letters, then all lowercase letters in that sequencing read are instead considered for base pairing. The resulting .bit file is then passed as an argument to the fold-contrafold-uniq-bits-vectors2.py script. ContraFold was run either with normal settings or with --noncomplementary enabled. Typically, the fold-contrafold-uniq-bits-vectors2.py script is stopped after roughly 10,000 RNA molecules have been folded to save on computation time. The forgi vector encodings of the RNA secondary structures generated by ContraFold are output in the file forgi-vect-ser2.txt. Dimensional reduction is performed on the forgi vectors text file with the python script run-pca-on-forgi-vectors.py which reduces the vector space to 2 dimensions. Kmeans clustering is then performed on the 2-dimensional representation of RNA structures with the draw-kmeans-clusters.py script. Identification of a population average structure is done by choosing --num_clusters 1. Multiple conformational analysis is performed by varying the number of Kmeans clusters. The structure of the most representative of the Kmeans cluster is retrieved with the python script fetch_DaVinci_folded_copy.py and visualized with RNA 2D drawer.
Nucleotide reactivity normalization:
Mutation rates were normalized to reactivity values as previously described with the custom python script normalize_DAGGER_DANCE.py.
Tertiary folding constraint determination and implementation (DAGGER):
G’s likely to be engaged in tertiary interactions were identified by first selecting G’s in the bottom quartile for misincorporation rate. In situations when reactivity data was available for multiple conformations, G’s in separate conformations were treated as unique G’s and all unique G’s were pooled for reactivity normalization. From the bottom quartile of reactive G’s, G’s which displayed correlations to other lowly reactive G’s as identified by RingMapper were designated as engaged in a tertiary interaction.
The DaVinci analysis pipeline was modified to utilize tertiary constraints through setting tertiary G positions to lowercase G’s in the fasta input file for generation of .bit files by updated_simple_to_bit_rG42.py. The modified DaVinci analysis pipeline is reffered to as DAGGER.
All python scripts for analyzing data can be found at: https://github.com/SamieJaffreyLab/DAGGER_MaP
Assembly: custom fasta files for each RNA construct
Supplementary files format and content: Shapemapper output files include mutation rates for each nucelotide as output by ShapeMapper2. A complete description of the ShapeMapper output is available at (https://github.com/Weeks-UNC/shapemapper2). Mutate and MaP experiments processed files include a Z-score normalized mutate and map heatmap. See https://github.com/SamieJaffreyLab/DAGGER_MaP for detailed description of the heatmap file formats.
Library strategy: Amplicon sequencing
|
| |
|
| Submission date |
Jul 09, 2024 |
| Last update date |
Oct 14, 2024 |
| Contact name |
Samie Jaffrey |
| Organization name |
Weill Cornell Medicine
|
| Department |
Pharmacology
|
| Street address |
1300 York Ave
|
| City |
New York |
| State/province |
New York |
| ZIP/Postal code |
10065 |
| Country |
USA |
| |
|
| Platform ID |
GPL17769 |
| Series (1) |
| GSE271825 |
RNA tertiary structure and conformational dynamics revealed by BASH MaP |
|
| Relations |
| BioSample |
SAMN42385658 |
| SRA |
SRX25255758 |
| Supplementary file |
Size |
Download |
File type/resource |
| GSM8386318_Spinach_M2_DMS_MaP_Modified_map.Zscore.txt.gz |
22.0 Kb |
(ftp)(http) |
TXT |
SRA Run Selector |
| Raw data are available in SRA |
|
|
|
|
 |
