PrimalSeq: Generation of tiled virus amplicons for MiSeq sequencing

Version 3.2 (2019.10.16)

Nathan Grubaugh, Yale School of Public Health (grubaughlab@gmail.com)

Generated in collaboration by the Loman, Andersen, and Grubaugh labs.

For general use of the protocol and primer design, please cite:

Quick, J. et al. Multiplex PCR method for MinION and Illumina sequencing of Zika and other virus genomes directly from clinical samples. Nature Protocols 12 (6), 1261-1276 (2017)

https://www.nature.com/nprot/journal/v12/n6/abs/nprot.2017.066.html

For measuring intrahost virus genetic diversity and calling variants using iVar, please cite:

Grubaugh, ND. et al. An amplicon-based sequencing framework for accurately measuring intrahost virus diversity using PrimalSeq and iVar. bioRxiv 383513 (2018)

https://www.biorxiv.org/content/early/2018/08/05/383513

The general approach to this protocol is to amplify the virus genome in small (~400 bp) overlapping fragments using two highly multiplexed PCR reactions (where the overlapping segments are in separate reactions). The amplicons are combined after PCR and are the correct size for library preparation and paired-end 250 nt sequencing using the Illumina MiSeq.

Version 3 notes: This protocol is being updated to include considerations for measuring intrahost virus variants, improved reagents, and a more efficient computational pipeline (iVar). To use the previous version protocol, you can find it by going to:

File > Version history > See version history > Version 2 (April 8, 2018)

All of the primers are now listed in a separate spreadsheet. We currently have 400 bp amplicon schemes for Zika virus, West Nile virus (North America lineage I genotype), Usutu virus, and chikungunya virus (ECSA genotype). More will be made available soon. You can build your own primer sets by using Primal Scheme.

Considerations for measuring intrahost genetic diversity using this amplicon-based protocol:

1. Requires at least 1000 virus RNA copies going into cDNA synthesis. More is better. Try to normalize virus RNA copies between samples to make comparisons easier.
2. Process each RNA sample twice through the protocol to sequence as technical replicates. By calling variants only present in both replicates, it reduces the number of false positives (mainly from sequencing errors) and increases the accuracy of variant frequency measurements.
3. Obtain at least 400x nt coverage of each nucleotide position. Because of different amplicon efficiencies, this typically means that ~1M 250 nt paired-end reads are needed. Amplification of high input virus concentrations (>10,000 virus RNA copies) are more even and require fewer total reads.
4. During our validation process, the lowest intrahost variant frequency that we could accurately and consistently measure was 3%. Measuring lower than this requires additional input copies, coverage depth, and validation.
5. Beware of intrahost virus variants that exist within primer binding sites as they can decrease the amplification efficiency of that particular virus haplotype. Because the primer sites are trimmed and are covered by an overlapping amplicon, the variants within the primer sites can be accurately measured. All variants within the amplicon with a primer mismatch, however, can be significantly altered. This is the major limitation with any PCR protocol for virus population diversity analysis.
6. Use our data pipeline, iVar (intrahost variant analysis from replicates) to process and analyze the data. It will align to the reference (or call a consensus), trim primers, call variants, compare variants between replicates, and flag variants within primer sites.

Overview of tiled virus amplicon sequencing protocol

Preparation of cDNA (est. time: 1 hour)

Reagents:

        SuperScript IV VILO Master Mix

Generation of tiled amplicons (est. time: 5 hours)

Reagents:

Q5 High-Fidelity 2X Master Mix

Custom primers

Qubit High Sensitivity dsDNA kit

Library preparation and quantification (est. time: 3-4 hours)

Reagents:

        KAPA HyperPrep kit (¼ recommended reagents)

Mag-Bind TotalPure NGS (or make your own!)

BIOO Scientific NEXTflex Dual-Indexed DNA Barcodes (these are $$$, but last a long time^[a])

Qubit High Sensitivity dsDNA kit

High Sensitivity DNA Analysis Kit

KAPA Library Quantification kit (optional)

MiSeq Reagent kit v2 (500 cycle output) or v3 (600 cycle output)

KingFisher Flex

        Protocols: https://github.com/grubaughlab/Kingfisher_protocols

Data analysis

        iVar: www.github.com/andersen-lab/ivar

Note: Add no-template water controls at each of the cDNA and PCR steps to monitor for contamination. 

Preparation of cDNA^[b]

1.         Isolate viral RNA using Omega Viral DNA/RNA kit, Trizol, or equivalent.

2.         Many different cDNA synthesis kits can be used, but choose something that is relatively high-fidelity.       The current protocols uses SuperScript IV VILO Master Mix because the enzyme has low error rates and the protocol is fast and easy.

3.         Run the following cycles on a thermocycler:

4.         Store samples at 4°C (same day) or -20°C (up to a week) until ready for PCR.

PCR generation of tiled amplicons^[e]

1.         Validated primer schemes can be found here. Prepare two primer pools by mixing equal volumes of each 10 µM primer. Primers indicated by “*” should be pooled at a concentration of 50 µM and primers indicated by “**” should be pooled at a concentration of 100 µM to help normalize sequencing coverage.

2.        Prepare two PCR reactions for each sample (one for each primer pool):

3.         Run the following cycles on a thermocycler:

4.         Run 5 µL of each product on a 1% agarose gel. Each should produce a visible 400 bp band.

Post PCR cleanup (1.8:1 ratio of beads to sample) and quantification

1.        Allow Mag-Bind TotalPure NGS beads to equilibrate to room temperature, vortex until homogenous.

2.        Bring PCR product volume up to 25 µL with water (if not at volume already).

3.        Add 45 µL of beads to 25 µL of PCR product, mix well, and incubate at room temperature for 10 minutes.

4.        Place tubes on a magnetic stand and incubate until solution appears clear.

5.        Discard supernatant without disturbing the beads.

6.        While tubes are on the magnet, add 200 µL of 80% EtOH, incubate for 30 seconds, and discard the EtOH wash.

7.        Repeat previous 80% EtOH wash and remove as much EtOH as possible.

8.        Leave tubes on magnet and air dry for 5 minutes.

9.        Remove tubes from magnet and add 20 µL of nuclease-free water. Mix well by pipetting.

10.        Place tubes on magnet stand. When solution appears clear, remove supernatant without disturbing the beads and place into new tubes.

11.        Quantify the DNA concentration using the Qubit High Sensitivity DNA kit (or equivalent) from 1 µL of each product. Expected range = 10-100 ng/µL DNA. Sequencing from lower concentrations may still work.

12.        Note: If your lab has a KingFisher, you can download our automated protocols here: https://github.com/grubaughlab/Kingfisher_protocols (use ‘purification.bdz’ for this step)

Library preparation (using ¼ of vendor recommended reagents)

End-repair

1.        Combine 25-50 ng of PCR-amplified DNA from primer pool 1 and 2 together for a total of 50-100 ng in 12.5 µL (equal concentrations of each amplicon pool). QS to a total volume of 12.5 µL using nuclease-free water.

        Alternatively: proceed using 50-100 ng of primer pool product separately for library preparation. This allows for additional monitoring of cross-contamination. Data can be merged computationally post sequencing.

2.        Combine the following components from the Kapa Hyper prep kit for end repair:

3.         Run the following cycles on a thermocycler:

Adaptor ligation

1.        Dilute a working stock of NEXTflex Dual-Indexed DNA Barcodes 1:100 to obtain a concentration of 250 nM. Select unique barcodes for each sample. Try not to repeat barcodes from recent runs.

Note: Be careful to not cross-contaminate the adaptors by centrifuging all liquid from the caps and only opening one index at a time.

2.         Combine the following components:

3.        Incubate at 20°C for 15 minutes.

4.        Proceed immediately to cleanup.

Post ligation cleanup (0.8:1 ratio of beads to sample)

1.        Allow Mag-Bind TotalPure NGS beads to equilibrate to room temperature, vortex until homogenous.

2.        Add 22 µL of beads to 27.5 µL of ligation product, mix well, and incubate at room temperature for 10 minutes.

3.        Place tubes on a magnetic stand and incubate until solution appears clear.

4.        Discard supernatant without disturbing the beads.

5.        While tubes are on the magnet, add 200 µL of 80% EtOH, incubate for 30 seconds, and discard the EtOH wash.

6.        Repeat previous 80% EtOH wash and remove as much EtOH as possible.

7.        Leave tubes on magnet and air dry for 5 minutes.

8.        Remove tubes from magnet and add 20 µL of nuclease-free water. Mix well by pipetting.

9.        Place tubes on magnet stand. When solution appears clear, remove supernatant without disturbing the beads and place into new tubes - 15 µL will go into library amplification.

10.        Note: If your lab has a KingFisher, you can download our automated protocols here: https://github.com/grubaughlab/Kingfisher_protocols (use ‘purification.bdz’ for this step)

Library amplification^[h][i]

1.         Combine the following components:

2.         Run the following cycles on a thermocycler:

3.        Proceed directly to cleanup or store at 4°C.

Post amplification cleanup (0.8:1 ratio of beads to sample)

1.        Allow Mag-Bind TotalPure NGS beads to equilibrate to room temperature, vortex until homogenous.

2.        Add 27.2 µL of beads to 34 µL of amplified product, mix well, and incubate at RT for 10 minutes.

3.        Place tubes on a magnetic stand and incubate until solution appears clear.

4.        Discard supernatant without disturbing the beads.

5.        While tubes are on the magnet, add 200 µL of 80% EtOH, incubate for 30 seconds, and discard the EtOH wash.

6.        Repeat previous 80% EtOH wash and remove as much EtOH as possible.

7.        Leave tubes on magnet and air dry for 5 minutes.

8.        Remove tubes from magnet and add 25 µL of Tris-EDTA or elution buffer. Mix well by pipetting.

9.        Place tubes on magnet stand. When solution appears clear, remove supernatant without disturbing the beads and place into new tubes.

10.        Note: If your lab has a KingFisher, you can download our automated protocols here: https://github.com/grubaughlab/Kingfisher_protocols (use ‘purification.bdz’ for this step)

Sequencing preparation

Library quantification and pooling

1.        Quantify the DNA concentration of each sample (1 µL) using the Qubit High Sensitivity DNA kit.

2.        Pool equal concentrations (e.g., 1-10 ng) of each library for sequencing.

3.        Check DNA fragment distributions of the pooled sample using the BioAnalyzer DNA 1000 kit. Peak fragment size from 400 bp tiled amplicons with proper ligated adaptors should be ~ 580 nt. If ~180 bp bands (adaptor dimers) still exist, perform post amplification cleanup again. 

4.        Quantify the DNA concentration of the pooled library (1 µL) using the Qubit High Sensitivity DNA kit.

5.        Note: At least 0.76 ng/µL is required to achieve 2 nM for library pooling. Libraries will need to be concentrated or re-amplified if less than this amount.

6.        Convert DNA libraries from weight to moles:

        Molecular weight [nM] = Library concentration [ng/µL] / ((ave. library size x 650)/1,000,000)

        Example: if ave. size of library is 580 bp and concentration is 2.5 ng/µL…

                (580 x 650) / 1,000,000 = 0.377

                2.5 / 0.377 = 6.6 nM

7.        Dilute the pooled library to 2 nM in 10 mM TE.

8.        (Optional) Ensure the library molar concentration using the Kapa Library Quantification kit.

9.        If sending your sample to a genomics core (i.e., not loading the MiSeq yourself), stop here.

Diluting the pooled library for sequencing

1.        Combine 10 µL of the 2nM pooled library to 10 µL of 0.1 N NaOH and mix. Incubate from 5 minutes at     room temperature to denature the dsDNA.

2.        Add 980 µL of HT1 (comes with the MiSeq kits). New concentration = 20 pM.

3.        Dilute to the desired concentration using the following volumes.

*PhiX control should also be denatured and diluted to 20 pM.

4.        Note: loading too high of a sample on a MiSeq leads to over-clustering and decreased quality, which may make the data unusable. Adding too low leads to under-clustering and may not generate enough data for sufficient sequencing coverage. In our hands, optimal cluster densities were reached using 10-12 pM with the MiSeq v2 kits and 14-16 pM with the MiSeq v3 kits. Loading concentrations should be empirically determined with each lab.

5.        Following loading instructions located in the MiSeq user guides.

Data processing and analysis

1.        Use iVar, follow the instructions on: www.github.com/andersen-lab/ivar