Simplifying sample preparation in microbial genomic sequencing

  • Dr Paul Blainey is an Assistant Professor of Biological Engineering at the Broad Institute at the Massachusetts Institute of Technology (MIT), where he develops and applies new molecular, microfluidic and optical tools in biological and biomedical research. He and Dr Soohong Kim, a post-doctoral scientist at the Broad Institute, have developed a ground-breaking microfluidic system to dramatically increase the efficiency of sample preparation for high-throughput nucleic acid sequencing.

    Sequencing bacterial genomes can help us determine why bacteria cause disease, how infections are spread between patients, how bacteria become resistant to antibiotics, and which drugs different bacteria are resistant to. Bacterial genomics can also identify chemicals synthesised by bacteria that may be useful as new drugs. The cost of bacterial genome sequencing has decreased by six orders of magnitude over the last ten years to about $5 using the most efficient technology. This incredible cost reduction enabled genome sequencing to emerge as a ubiquitous research tool in infectious disease and public health research.

    Despite the advancements, current DNA sequencers cannot directly read DNA in its natural state and require DNA to be extracted from cells and biochemically processed to produce a readable sequence library. Compared to the improvements in sequence data production that enabled the cost of the sequencing operation itself to fall to $5, methods for turning DNA samples into sequence libraries have lagged. Currently, sample preparation costs ranges from $100 to $300 including the total cost of labour, consumables, and equipment. Due in part to the sample preparation bottleneck, genome sequencing is not routinely used in hospitals to track the emergence and spread of antibiotic resistant bacteria, a pressing public health threat.

    While many steps need to be taken for the routine deployment of pathogen genome sequencing in patient care and tracking outbreaks, the high cost for sample preparation is a key issue. Recognising the opportunity to apply microfluidic technology to make a major advance in sample preparation efficiency, and ultimately cost, has led the Blainey Group at the Broad Institute, MIT – headed by Dr Paul Blainey – to take on this problem.

    Sample preparation
    Conventional genome sequencing methods require laboratories to lyse cells, purify their DNA, and extensively process it into a readable form. Crucially, the DNA must be fragmented and tagged with adaptors before the actual sequencing process can commence. Such laborious techniques have been somewhat relieved by the addition of liquid handling robots that partially automate processing. However, standard automation systems do not integrate the entire sample preparation process, but rather require extensive sample pre-processing, and require users to monitor the robots’ progress and perform some hands-on sample and reagent transfers. For example, one microfluidic system called NeoPrep (Illumina) only partially automated the sample preparation process, and is now scheduled to be discontinued.

    Sequencing microbial genomes can be as low as $5, but sample
    preparation can be twenty times this amount or more. This has led
    Dr Paul Blainey and colleagues to develop a new technology which can make sample preparation far more efficientQuote_brain

    The need for highly integrated sample preparation systems that dramatically reduce sample preparation cost led Dr Blainey and Dr Kim to develop a microfluidic sample preparation platform that combines all the key sample preparation steps on a single, micro-automation platform to generate sample libraries.

    Benefits of microfluidics
    Their microfluidic sample preparation device enables a batch size up to 96 samples on a single device (multiple devices can be run in parallel for larger batches) and integrates the key sample preparation steps: extraction, fragmentation, adaptation, size selection, and clean-up with minimal user intervention. At the same time, the technology significantly reduces reagent consumption, consumable plastics, hands-on time, need for large robotic liquid handlers, sample input quantity requirements, and sample contamination. Though microbial genomes are simpler to sequence given their smaller genome size (e.g. E. coli has millions of bases compared to billions in humans), sample preparation for whole-genome sequencing still involves up to ten steps. The new microfluidic device simplifies all of the steps required for DNA sequence sample processing and saves considerable staff and cost.

    How it works
    The microfluidic sample preparation device (see Figure 1 above) consists of a reactor (red), filter (yellow) and a reservoir (green) with separate input (black arrow) and sample input/output ports (red arrow). When constructing a sample library, the reactors mix different combinations of biochemical reagents for cell lysis, and DNA fragmentation, tagging, while the filters capture DNA-bound beads to purify the DNA fragments. The reactor and filter are reused multiple times during the sample processing steps as purified DNA from one step is returned to the reactor for the next biochemical treatment. By re-using each component of the device in a given processing run, the system efficiently utilises the microfluidic elements and enables a large number of samples to be processed at once in a compact device. The ability to flexibly reuse device elements also makes it possible to run different protocols on the same chip.

    Low-cost, high-throughput bacterial genomics
    The Blainey Group recently tested the device, processing different types of cells including pathogens like tuberculosis cells and environmental micro-colony samples. Thanks to the device’s processing efficiency, these experiments were not only automated, but also produced high-quality data from reactions consuming only 1/200th the quantity of sample that is normally required, enabling the analysis of tiny samples that would not otherwise be possible. Requiring less quantity of a sample also enables earlier analysis of slow-growing pathogens like tuberculosis, which can take a month or longer to produce enough material for conventional sequencing approaches.

    Dr Blainey’s test analysis demonstrated the versatile nature of microfluidic platforms. Among the core benefits, he noted a reduction in input requirements of samples, which makes it more applicable to sequence hard-to-culture organisms and clinical samples obtained by minimally-invasive methodsQuote_brain

    When testing the device, the researchers also successfully constructed 400 Pseudomonas aeruginosa libraries from six patients. This bacterium can cause infections in patients with weak immune systems, and particularly those who have been hospitalised for more than a week. Antibiotic resistance is commonly found in Pseudomonas, which makes it an important subject of genomic research to determine new treatment strategies and clinical surveillance to track the occurrence and spread of resistance strains. Dr Blainey’s analysis demonstrated that acute Pseudomonas infections are nearly clonal, with all the cells from a given site of infection showing essentially identical genomic sequences. This differs from reports of chronic P. aeruginosa infections in cystic fibrosis patients where diverse P. aeruginosa strains are often found. The new findings show that high genotypic diversity is not a characteristic feature of all P. aeruginosa infections.

    An efficient, simplified workflow
    Overall, the test analysis demonstrated the versatility of the microfluidic platform and its high performance. Among the core benefits, the researchers noted a reduction in sample input quantity requirements, which makes the method more applicable to sequencing hard-to-culture organisms and convenient sampling approaches that yield low biomass quantities. In principle, this enables the platform to be used for most types of sequencing protocols, expanding its usage across a wide variety of applications – from clinical sample workflows for many disease areas to environmental surveillance. A particular advantage of the system is its ability to perform the same types of purification steps used in conventional protocols, allowing direct translation of conventional protocols to the new microfluidic platform. The improved automation, higher sample density and low-input capability sets the system up to be a serious contender for a wide variety of future uses.

    Though there is undoubtedly demand for such a device, and Dr Blainey and Dr Kim look to partner and commercialise the platform, the Blainey lab is currently focused on applying the system to process samples from a clinical trial testing a clinical protocol for decolonising patients with recurring infections by Methicillin-resistant Staphylococcus aureus (MRSA) – a bacterium that causes infections in humans which are difficult to treat due to its antibiotic resistance.

  • How did you first become involved in bio-engineering?
    I joined Stephen Quake’s group at Stanford University to learn microfluidic techniques and apply these to sequencing single bacterial and archaeal cells.

    Which other areas do you see the platform being applied to?
    Human genome sequencing, basic disease research, advanced clinical diagnostics, particularly of cancer and the immune system, and understanding how drug and other treatments are affecting patients.

    What was Dr Kim’s role in the development of the technology?
    Dr Kim designed, fabricated, and tested the device as well as the procedure for using it. He also led the biological studies featured in the Nature Communications article.

    Genomic sequencing is a powerful technology with the potential to prevent and treat disease on a large scaleQuote_brain

    Are you planning to commercialise the system as a platform in the future?
    We look to partner with industry professionals to make the technology widely available in an easy-to-use format.

    What are your opinions on the future of genomic sequencing?
    Genomic sequencing is a powerful technology with the potential to prevent and treat disease on a large scale as well as make healthcare more efficient. Continuing technical advances are needed to make sequencing easy enough and inexpensive enough for broad deployment so that the technology can realise its full benefit to patients and the way care is delivered.

  • Research Objectives
    This project aimed to make sample preparation for whole genome sequencing at least an order of magnitude more efficient using micro-scale automation of all the key sample processing steps.

    Department of Energy (DOE), National Institutes of Health (NIH), National Science Foundation (NSF), Burroughs Wellcome Fund

    Dr Slava Epstein

    Dr Paul Blainey trained in mathematics, chemistry, biophysics, microfluidics, and genomics before joining the Broad Institute and the Department of Biological Engineering at MIT as a faculty member in 2012. His laboratory integrates microfluidic, molecular, and imaging tools to address new challenges in single-cell analysis, genomic screening, and therapeutics development.

    Dr Soohong Kim is a postdoctoral scientist in Professor Paul Blainey’s lab. He studied Chemistry at the State University of New York and Biophysics/Physical Chemistry at the University of California. Since then, Dr Kim has trained as a methodologist and has been developing technologies to enhance the throughput and resolution of biological measurements.

    Paul Blainey, PhD
    Broad Institute,
    415 Main Street
    Cambridge, MA 02142

    T: +1 617 714 7320

  • Simplifying sample preparation in microbial genomic sequencing

Leave A Reply