DNA Viral Size Fraction Metagenomics for Human Stool Samples

Summary

We present a flexible protocol for generating high-quality viral size fraction metagenomes (viromes) from human stool samples. Originally developed for soil viromics, here this protocol was successfully applied to stool samples, enabling robust characterization of the human gut virosphere.

Abstract

Understanding the healthy human virosphere (the viral component of the microbiome) requires accurate measurements of viral community composition across a diverse range of viral types. Building on prior experience with soil viral community ecology methods, here we demonstrate a series of laboratory approaches for enriching and extracting DNA from extracellular DNA viruses in human stool samples. A working primary protocol is presented, along with options for deviations at different steps. The general approach involves adding a liquid buffer (default: protein-enhanced phosphate buffered saline, PPBS) to facilitate removal of free viral particles from the stool matrix, centrifugation to separate the liquid fraction containing viral particles, filtration (default: 0.2 µm pore size) to remove most cells, concentration of viral particles (default: ultracentrifugation), removal of free nucleic acids with nucleases prior to virion lysis, and then DNA extraction for sequencing. Alternative techniques, including different buffers, filter sizes, and concentration methods, are also noted. Overall, multiple options for generating high-quality viromic DNA for sequencing are offered. Rather than tailoring the approach to specific equipment and resources, the protocol's flexibility should make it broadly applicable across labs with varying standard molecular biology equipment.

Introduction

Viromics, or the untargeted metagenomic sequencing of the viral size fraction of mixed communities, can provide insights into the diversity, evolution, and ecology of uncultured viruses. Since the early efforts to catalog gut viral communities over 20 years ago, which notably used methods developed for environmental samples like seawater¹, viromics approaches have continued to enable discoveries in viral ecology within the human microbiome². However, there is no single universal viromics protocol, and variations in techniques can impact sequencing results^3,4, ultimately influencing the interpretation of viral community dynamics and hindering inter-study comparisons. Optimized protocols, combined with a thorough understanding of how methodological choices shape virome profiles, are needed to robustly characterize the human virosphere.

While there is no universal viromics protocol, the general basis of viromics methods involves physically separating and concentrating extracellular virus-like particles (virions) based on their size and/or density, removing free nucleic acids (derived in part from cellular debris or compromised viral capsids) prior to virion lysis, and extracting DNA and/or RNA for sequencing. Some viromics techniques select for or against specific viral types, whether as an intentional choice by researchers or an unintended methodological limitation. Syringe or vacuum filtration is frequently used to remove microbial cells and other debris, and the filter size inherently affects the quality and composition of the recoverable virome. For example, filtration can increase virome purity⁵, but it may exclude larger viral particles^6,7. Virion concentration methods, such as ultracentrifugation, polyethylene glycol (PEG) precipitation, density gradient ultracentrifugation, and ultrafiltration, also have trade-offs among purity, throughput, and equipment costs^3,8. Multiple buffer chemistries are known to be effective for viromics. They may not have outsized impacts on virome characterization⁹, but storage and resuspension buffers can still affect virus recovery and the virome profiles generated from bioinformatic analysis^10,11. Lastly, as viral nucleic acid concentrations can be low in natural samples, many studies opt to amplify viral DNA before library construction (e.g., by multiple displacement amplification (MDA) or sequence-independent single-primer amplification (SISPA)). Yet some of these techniques introduce significant bias^12,13. Overall, while many viromics methods can successfully generate viromes, optimized protocols that acknowledge biases and are somewhat standardized will enable accurate and reproducible virome research.

This article presents a flexible workflow for generating viromes from human stool samples. Adapted from proven soil viromics techniques^14,15,16, this protocol includes several steps that facilitate modification, enabling researchers to adapt it to their own equipment and research needs. The standard soil viromics method outlined below, successfully applied to both soil and frozen stool samples, produces high-quality viromes without a viral DNA amplification step prior to library construction. The suggestions for protocol modifications have been tested to varying degrees in other studies, but their specific, downstream effects on human stool virome sequences were not examined here.

Protocol

All subjects recruited for stool sample collection provided informed consent. Sample collection protocols were approved by the University of California, Davis Institutional Review Board, and the studies conform to the Declaration of Helsinki. This protocol is adapted from Emerson et al. (2022)¹⁷.

1. Preparation of materials

1. Prepare protein-enhanced phosphate-buffered saline (PPBS, 10x PBS, 1% K-citrate, 150 mM MgSO₄, 2% bovine serum albumin (BSA))¹⁸.
   ​NOTE: Other commonly used buffers in viromics protocols are saline magnesium (SM)¹⁹ and amended potassium-citrate prime (AKC’) buffer¹¹.
   1. Add 10 mL 10x PBS, 10 g K-citrate, 18.05 g anhydrous MgSO₄ to a 1 L autoclavable, glass bottle and fill to 1 L with ultra-pure laboratory water. Adjust the pH to 6.5 (typically requires about 50 µL of 6 M HCl or about 100 µL of 2 M HCl). Leave a stir bar inside the bottle for use in Step 1.1.2.
   2. Autoclave the bottle with buffer and one empty 1 L autoclavable glass bottle. After autoclaving, once the buffer cools to room temperature, place it on a magnetic stir plate and slowly add 20 g BSA directly into the bottle, adding about 5 g of BSA at a time (to avoid foaming, minimize excessively agitating the solution, e.g., start with a slow rate of stirring and gradually increase).
   3. Filter the entire buffer volume through a sterile 0.22 µm vacuum filter into the empty autoclaved 1 L bottle from Step 1.1.2. Store the buffer at 4 °C until use.
      NOTE: PPBS buffer can be stored at 4 °C for about 3 months, but it is not stable long-term when stored above freezing.
2. Prepare the ultracentrifuge tubes.
   ​NOTE: This protocol uses ultracentrifugation to concentrate virions. Other common approaches include polyethylene glycol (PEG) precipitation, density gradient ultracentrifugation (e.g., cesium chloride gradients), and ultrafiltration.
   1. To clean the ultracentrifuge tubes, first, scrub the tubes using ultra-pure laboratory water. Then rinse the tubes with ultra-pure laboratory water.
   2. Fill the tubes with 1 M NaOH and let soak for at least 1 hour (can let soak overnight). Clean the plastic/rubber ultracentrifuge caps and their o-rings by immersing them in 1 M NaOH.
      NOTE: Do not immerse the metal screw top caps in NaOH.
   3. After soaking, rinse three times with ultra-pure laboratory water, followed by one rinse using nuclease-free, molecular biology grade water. Dry the tubes by inverting and gently tapping, or allowing them to dry completely overnight. Label each ultracentrifuge tube with a sample identifier.
3. Pre-cool the ultracentrifuge rotor at 4 °C by leaving the rotor in the fridge or cold room overnight.

2. Virion purification and concentration

1. Weigh out the stool into a sterile, DNase-, RNase-, and DNA-free 50 mL conical tube.
   NOTE: The amount of stool used may depend on the amount of sample material available. Here, 2 g of stool was used. If unused sample material needs to be stored frozen, consider aliquoting the entire sample before freezing to minimize future freeze-thaw cycles.
2. Add 8 mL of PPBS buffer directly into the conical tubes containing the stool (conical #1). Thoroughly resuspend by vortexing.
3. Shake the tubes at 300 RPM at 4 °C for 10 min.
   NOTE: If necessary, the shaking can be done at room temperature, but 4 °C is recommended.
4. Centrifuge samples in conical #1 at 4,000 x g at 4 °C for 10 min.
   NOTE: Use a swinging bucket rotor, if possible, to facilitate vertical density separation to make the next steps easier.
5. Carefully decant the supernatant into a new sterile 50 mL conical (conical #2) and store at 4 °C (the supernatant here contains the virions). Do not discard the conical tube that contains the pellet (conical #1), as it is used again in Step 2.6.
6. Repeat steps 2.2-2.5 two more times (so that the sample in conical #1 has gone through a total of 3x resuspension with buffer, 3x shaking, and 3x centrifugation) and continue to pool the supernatant into conical #2.
   NOTE: Three rounds should accumulate ~24 mL supernatant total.
7. Centrifuge the 50 mL conical tube containing accumulated supernatant (conical #2) for 8 min at 10,000 x g at 4 °C.
8. Perform a second round of centrifugation. First, decant the supernatant from conical #2 into a new 50 mL conical tube (conical #3). Centrifuge conical #3 for 8 min at 10,000 x g at 4 °C.
9. Using syringe filtration, sequentially filter the supernatant from conical #3, avoiding the pellet at/near the bottom of the tube. Pre-filter using a sterile 5.0 µm polyethersulfone (PES) syringe filter, followed by a sterile 0.20 µm PES syringe filter, directly into an ultracentrifuge tube.
   NOTE: The purpose of pre-filtering with 5.0 µm filters is to reduce clogging of the 0.20 µm filters. It may still be necessary to discard and replace the syringe filters if they become clogged. Filtration can be performed in a single step by stacking the 5 µm filter on top of the 0.20 µm filter, attached to a single syringe, and directly into an ultracentrifuge tube. PES membranes are preferred, as other membrane materials can retain virions (e.g., glass fiber) or clog faster (e.g., polyvinylidene fluoride, PVDF). 0.45 µm is another common filter size.
10. Check the mass of all ultracentrifuge tubes with their sample contents, and pair tubes in the ultracentrifuge based on mass. If necessary, add PPBS to the lighter tube of each pair to equalize weight (within a 0.05 g margin of error for paired tubes).
11. Ultracentrifuge at 112,000 x g at 4 °C under vacuum for 2 h 25 min.
    NOTE: Depending on the rotor used, the spin duration will differ (according to the sediment coefficient (s) = 58). Refer to the Table of Materials for the rotor used here. Before removing the tubes from the ultracentrifuge, note the orientation of each tube in the rotor to aid in locating the viral pellet. Sometimes, after ultracentrifugation, the viral pellet is not visible, and extra care should be taken when removing the supernatant in the following step.

3. Viral pellet resuspension and DNase treatment

1. Remove the supernatant either by decanting or using a pipette, taking care not to disturb the pellet, which can be fragile (the pellet contains the concentrated virions).
   NOTE: As noted in Step 2.11, the pellet may or may not be visible, but a 'hole punch-sized' brown or yellow pellet can often be seen.
2. Treat the samples with DNase.
   NOTE: If opting not to DNase treat samples, proceed to Step 4 and follow “If DNase treatment was not performed”. Refer to the Discussion for considerations on performing DNase treatment.
   1. Resuspend the pellet in 100 µL of nuclease-free, molecular biology grade water into a labeled 1.5 mL microcentrifuge tube.
   2. Make the DNase treatment master mix by adding 10 µL of DNase reaction buffer and 10 µL of DNase for each sample to be treated. Pipette to mix thoroughly.
      NOTE: Prepare enough DNase treatment master mix for extra reactions to account for potential volume loss during pipetting in Step 3.2.3.
   3. Add 20 µL of the DNase treatment master mix to the resuspended virions. Mix well by gentle pipetting. Incubate at 37 °C for 30 min.
   4. Add 10 µL of DNase stop solution to inactivate the nucleases. Briefly vortex.

4. Extraction of viral nucleic acids

1. Perform the DNA extraction using the DNA extraction kit listed in the Table of Materials.
   ​NOTE: Make two modifications to the manufacturer’s instructions. 1) After Step 1 in the kit procedure, heat the bead tube containing the virions in buffer CD1 at 65 °C for 10 min, then proceed with the extraction according to the manufacturer’s instructions. 2) During the DNA elution step (step 16 in the kit procedure), incubate the columns at room temperature for 5 min after adding C6.
   1. If DNase treatment was performed: After the nucleases have been inactivated (in Step 3.2.4), use the entire volume as input for the DNA extraction (i.e., in step 1 of the DNA extraction kit procedure, add this volume to the bead tube containing 800 µL of CD1). Make sure not to leave any solution behind (this solution contains the virions).
   2. If DNase treatment was not performed: Add 100 µL of nuclease-free, molecular-grade water to the ultracentrifuge tube containing the pellet from Step 3.1. Gently scrape the pellet and resuspend. Use the resuspended pellet as input for DNA extraction.

5. DNA quantification

1. Quantify the extracted DNA with the fluorescence-based method of choice.
   NOTE: Virome DNA yields are rarely high enough for absorbance-based quantification; however, this can still be helpful for identifying chemical contaminants.
2. For short-term storage, store DNA at -20 °C after quantification. For longer-term storage, store at -80 °C. The DNA is now ready for library preparation and sequencing. Refer to the Table of Materials for the library preparation kit and sequencing platform used here.

Results

An exploratory effort applying this protocol to a single frozen stool sample (processed in technical triplicate for viromes) alongside two soil samples demonstrated its efficacy in producing high-quality viromes from different sample types. Using quality-filtered sequencing data and de novo assembled contigs generated by this protocol, putative viral sequences were identified using an open-source virus identification tool²⁰. Reads mapped to the identified viral contigs were then ...

Discussion

This protocol presents a relatively straightforward and adaptable method for generating viromes. Proven in soils^14,15,16,21, it is successfully applied here to human stool samples. This protocol will best recover extracellular, intact DNA viruses and is not designed to intentionally recover intracellular viruses or prophage (though some may be recovered from contaminating cells), very large DNA...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Bovine Serum Albumin (BSA) DNase- and Protease-free Powder	Fisher BioReagents	BP9706100	For PPBS recipe (Step 1.1)
DNeasy PowerSoil Pro Kit	QIAGEN	47016	DNA extraction kit (Step 4.1)
KAPA DNA HyperPrep library kit	Roche	n/a	Libraries generated by this kit were sequenced using the Illumina NovaSeq 6000 platform (2 x 150 bp).
Magnesium sulfate (MgSO4), anhydrous, ≥99.5%	Sigma-Aldrich	M7506	For PPBS recipe (Step 1.1)
Phosphate-Buffered Saline (PBS), 10x, pH 7.4, RNase-free	Invitrogen	AM9624	For PPBS recipe (Step 1.1)
Polyethersulfone (PES) membrane syringe filters, sterile, 28 mm, 0.2 μm	Corning	431229	For syringe filtration (Step 2.9)
Polypropylene conical centrifuge tubes, sterile, 50 mL	Olympus Plastics	28-108
Potassium Citrate Monohydrate (Crystalline/Certified)	Fisher Chemical	P222	For PPBS recipe (Step 1.1)
Qubit  Assay Tubes	Invitrogen	Q32856	For DNA quantification (Step 5.1)
Qubit 4 Fluorometer	Invitrogen	Q33238	For DNA quantification (Step 5.1)
Qubit dsDNA Quantification Assay Kits	Invitrogen	Q32850	For DNA quantification (Step 5.1)
RQ1 RNase-Free DNase	Promega	PR-M6101	Includes the reaction buffer, DNase, and stop solution
Supor hydrophilic polyethersulfone (PES) membrane syringe filters, sterile, 32 mm, 5.00 µm	Cytiva (Formerly Pall Lab)	28150-956	For syringe filtration (Step 2.9)
Ultracentrifuge rotor: Type 50.2 Ti Fixed-Angle Rotor, 50.000 rpm, 302.000 x g, 12 x 39 m	Beckman Coulter	337901	We also use the Beckman Coulter Type 70 Ti Fixed-Angle Titanium Rotor (Cat. 337922). Centrifuge for 3 hours if using the 70 Ti.
Ultracentrifuge tubes: 26.3 mL Polycarbonate Bottle with Cap Assembly, 25 x 89mm	Beckman Coulter	355618	Compatible with Type 50.2 Ti (Cat. 337901) and Type 70 Ti (Cat. 337922) Beckman Coulter rotors
Ultracentrifuge: Beckman Instruments Optima LE-80K Floor Ultracentrifuge	Beckman Coulter	365668	We also use the Beckman Coulter L8-70M Refrigerated Ultracentrifuge (Cat. 344196)