Sample Homogenization Considerations for NextGen Sequencing
DNA sequencing has evolved significantly since the late 1970s, when shotgun cloned libraries were deciphered a couple hundred bases at a time using radioactive labels, hand poured gels, films, and a darkroom. As the demand for sequence data shifted from genes to genomes, or from kilobases to gigabases, gene cloning was excluded as the sophistication and throughput of DNA sequencing increased. In the sequencing lab, Escherichia coli lost its job.
The latest in Next Generation Sequencing (NGS) now allows for direct sequencing of individual DNA molecules or clonally amplified DNA. The primary systems used in NGS, manufactured by Pacific Biosciences and Illumina®, can generate gigabases to terabases of data per run (compared to megabases for the previous generation) by directly measuring bases incorporated during DNA polymerization. Gone are dideoxy terminators and pyrosequencing, replaced by fluorescent nucleotides that are detected during synthesis (i.e., SBS or sequencing by synthesis). Furthermore, the single-molecule real-time (SMRT™) technology of Pacific Biosciences measures the kinetics of base incorporation into a DNA chain, providing valuable insight into base modifications (i.e., DNA methylation) of native DNA templates (for review, see Reuter et al., 2015).
As always, DNA sequencing needs high purity and quality templates. Not only do the traditional parameters of DNA purity apply, but fragment size is also a critical characteristic. The chemistries of the major sequencing systems, those of Illumina® and Pacific Biosciences, require DNA fragment lengths at opposite ends of the spectrum. Shorter DNA fragments are needed for the Illumina® technology, while the PacBio™ RS II requires a higher molecular weight DNA of 45kb or greater. While it is relatively easy to harvest short DNA fragments, the isolation of large DNA molecules is more laborious and challenging.
Both Illumina® and Pacific Biosciences defer researchers to commercial kits for the initial isolation of DNA prior to their template preparation. DNA isolation methods, however, can significantly affect the size of the DNA fragments, with the disruption method having the greatest impact on fragment size. This is significant as commercial kits typically leave the disruption method up to the researcher. Common methods rely on hand-held homogenizers and mortar and pestles, chilled with liquid nitrogen, when a limited number of samples must be processed. Higher throughput methods almost exclusively rely on bead beating, whether in microplate formats with plate homogenizers or with oscillating bead beaters that hold disruption tubes. For NextGen sequencing, the number of samples that require processing doesn’t necessarily match up with the easiest method for disruption. This is especially true for samples destined for the PacBio™ RS II.
Bead beating is a popular method for homogenizing samples, because it is fast and can be adapted to high throughput laboratories. Compared to manual grinding, bead beaters have the capability to completely disrupt hundreds of samples in a matter of minutes, allowing for quicker DNA isolation protocols. Analysis of DNA isolated by bead beating performed by OPS Diagnostics, however, has shown that DNA is significantly fragmented as the result of processing (Fig. 1). Bioanalyzer results of DNA prepared by bead beating using the Synergy™ 2.0 Plant DNA Extraction Kit show the DNA to be between 2-7 Kb.
Figure 1. DNA isolated using Synergy™ Plant DNA Isolation Kit as measured by a Bioanalyzer.
While smaller fragments can be sequenced on the PacBio™ RS II, for instance 4 Kb fragments generated by bead beating were used for a full genome sequence of Neurospora crasa (Kim et al., 2014), fragments of this size lower the efficiency of the Pacific Biosciences sequencers. Alternatively, highly fragmented DNA generated by bead beating can be read with the Illumina® technology.
Illumina® offers a wide range of sequencers, all of which require DNA fragments between 150-300 bp. Although bead beated DNA is still too large for the Illumina® sequencers, the DNA can be further fragmented, by such means as sonication, before library preparation.
Low throughput methods, such as manual grinding, can become highly laborious when multiple samples must be processed. When using the PacBio™ RS II, however, it may be the only option because fragments must be about 10 Kb in length. DNA isolated from insect tissue by MOgene, LC (St. Lois, MO), using manual cryogenic grinding and OPS Diagnostics’ CTAB Extraction Buffer, yielded DNA fragments larger than 45 Kb (Fig. 2) in preparation for sequencing in the PacBio™ RS II.
Figure 2. Insect DNA isolated using cryogenic grinding using CTAB Extraction Buffer as analyzed on an Agilent Tapestation (data courtesy of MOgene LC, St. Louis, MO).
Pacific Biosciences specifies that the ideal fragment should be 10 Kb, thus a controlled reduction in fragment size is necessary before sequencing. This can be accomplished by using a shredder, such as the Covaris g-TUBE™, which is a device containing orifices that cause DNA to be fragmented between 6-20 Kb. Centrifugal force pushes DNA samples through slits, shearing the DNA. Library preparation for sequencing can be started once the DNA is of an acceptable length.
Manual processing of samples is not practical in most high throughput laboratories. Consequently, grinding in liquid nitrogen with a CryoGrinder™, or mortar and pestle, begs for a higher throughput equivalent using homogenizers such as the 1600 MiniG™ and HT Homogenizer. Cryogenic processing using high throughput homogenizers can be done using the proper configuration of grinding balls and vials (see Bead Beating: A Primer, Appendix B).
To evaluate high throughput cryogenic methods, plant samples were homogenized cryogenically using a 1600 MiniG™. This type of processing employs a CryoCooler™, Cryo-block, polycarbonate vials (4 ml) and large stainless steel balls. The CryoCooler™ is charged with liquid nitrogen and serves as a workstation to fill vials with samples and grinding balls. Vials are then placed in the Cryo-block, which has also been chilled in the CryoCooler™. All preparation is done at cryogenic temperatures. The Cryo-block not only holds the vials, but also dissipates heat generated during grinding by acting as a heat sink. Though cooling is passive, samples generally remain frozen during this process. Results from such a process were analyzed on the Bioanalyzer and showed that the DNA was much less fragmented than a solution based homogenization. Fragment size ranged from 4 Kb and beyond the 17 Kb marker (the upper resolution of the Bioanalyzer).
Figure 3. DNA isolated by cryogenic bead beating as assessed on a Bioanalyzer.
Larger DNA fragments are not effectively resolved by the Bioanalyzer, however Figure 3 illustrates that less fragmentation occurs when using a cryogenic approach compared to homogenizing with buffer (Fig. 1). Cryogenic grinding by hand yields larger fragments, most being larger than 17 Kb when measured on a Bioanalyzer (data not shown). Consequently, there is a significant trade-off between throughput and fragment size using the current tools available.
Clearly sample preparation can impact the DNA used for NextGen sequencing. The methods used to homogenize samples for DNA isolation should depend upon whether Illumina® or Pacific Biosciences technologies are used. For Illumina® systems, there are many options for sample disruption, including bead beating and sonication. Due to the large fragment requirements of the PacBio™ system, the isolation of DNA must be accomplished without the shearing forces associated with most common homogenization methods. Crushing samples cryogenically, or more preferably employing enzymatic or chemical lysis, will yield higher molecular weight DNA. A hybrid method of using cryogenic high throughput grinding, though better than bead beating at ambient temperatures, may be a suitable option for some samples destined for PacBio™ sequencing.
Kim, K.E., P. Peluso, P. Babayan, P. J. Yeadon, C. Yu, W.W. Fisher, C-S. Chin, N.A. Rapicavoli, D.R. Rank, J. Li, D.E.A. Catcheside, S.E. Celniker, A.M. Phillippy, C.M. Bergman, and J.M. Landolin. 2014. Long-read, whole-genome shotgun sequence data for five model organisms. Scientific Data 1:140045 doi: 10.1038/sdata.2014.45.
Reuter, J.A., D.V. Spacek, and M.P. Snyder. 2015. High-throughput sequencing technologies. Molecular Cell, 58:586-597.