February 15, 2018 (Vol. 38, No. 4)
Microbial Identification and Characterization Are Fast and Accurate
“If an 1890s physicist time-travelled to a physics laboratory today, he would be totally flummoxed. But, if Louis Pasteur were transported to a microbiology laboratory just a few years ago, he very likely would have rolled up his sleeves and gone to work. That’s because, compared to physics labs, not much has changed regarding microbial systematics. Microbial systematics—the identification and characterization of microorganisms—simply had to be modernized,” says Rita Colwell, Ph.D., founder of CosmosID.
The need for modernization became patently evident after 9/11 and the ensuing anthrax attacks. Dr. Colwell was director of the National Science Foundation (NSF) from 1998 to 2004, and remembers the post-9/11 anthrax scare well. “It took approximately six years to determine the source of the anthrax,” Dr. Colwell recalls. That was six years too long. She envisioned a system to identify a pathogen and its source in minutes.
After completing her term as director of NSF, “It became clear that microbial forensics and the microbiome were fundamental to all of microbiology and its many applications. The best path forward was to start a company to develop this technology.”
In 2008, when she founded CosmosID, the requisite technologies had converged to make microbial analysis immediate, practical, and profitable. The microbiome analysis platform she and her team have developed identifies and characterizes all microorganisms, including pathogens, and assesses antimicrobial resistance and metabolic properties, all within minutes.
“We simultaneously identify bacteria, viruses, fungi, and parasites in a given sample,” Dr. Colwell says. “DNA is extracted, sequenced, and analyzed. We provide actionable results for our clients in the form of tabular and graphical presentations. It’s very broad-reaching, with applications in medicine, agriculture, public health, and bioforensics.”
As CEO Manoj Dadlani adds, “We’re currently characterizing microbial profiles to understand the mixture of microorganisms comprising the microbiome, how this mixture of strains produces a specific disease, and what constitutes a healthy microbiome.” From that baseline, the team is correlating deviations to disease states.
To some extent, the microbial makeup of the microbiome is linked to behavior, culture, and geography, he notes. This emerging capability can identify potential targets for therapeutics. For example, CosmosID is focusing on irritable bowel syndrome, colitis, and Crohn’s disease in gastroenterology, as well as other diseases.
“Ultimately, this may lead to the development of a companion diagnostic,” Dadlani predicts. “One of our objectives is to develop clinical diagnostics for clinical trials and infectious disease research that will identify pathogens quickly and accurately, so physicians can prescribe the best combinations for treatments.” Other applications of the platform include pharmaceutical discovery, bioprocessing, microbiome analysis, public health, and food and water safety inspections.
CosmosID plans to develop a presence in Europe and Asia this year, starting with building a client base and partnering with laboratories in the regions. “The regulatory hurdles are more surmountable in Europe than in the U.S.,” Dr. Colwell says. This may lead to faster approvals.
A History of Achievement
Dr. Colwell has devoted her long career to microbiology. “My Ph.D. thesis analyzed phenotypes of marine bacteria.” That was in the early 1960s, when mainframe computers used punch cards and had to be wired for each run.
The computing power has changed dramatically since those early days, but one challenge has remained constant: obtaining funding for research. “For women scientists, this was more difficult than for the men.”
Despite the obstacles, Dr. Colwell has a solid track record of federal grants for her academic work. “But the funding we needed to develop my latest inspiration was too expensive for a regular NIH or NSF grant.” Then an angel investor offered to help launch the company with $2 million in funding. Soon afterward, CosmosID won contracts with the Department of Homeland Security.
Applications for Biotech
For drug developers, the ability to analyze microbes rapidly and accurately solves many problems. “The algorithms and application can be used for serial sampling to determine very early on whether a production batch is contaminated or heading in the wrong direction,” Dadlani says.
“It’s also useful for biomarker discovery and clinical trials support, to determine how the microbiome affects drug efficacy,” says Nur Hasan, Ph.D., vice president of R&D. Clinical trials can benefit by utilizing microbiome-based inclusion/exclusion criteria and, thereby, ensure enrollment of qualified eligible patient populations and increase likelihood of success.”
“We can identify, characterize, and eventually describe potential functions for all microorganisms present in a sample,” Dr. Colwell adds. “This information is important because many, if not most, infections are polymicrobial. Our work indicates that diseases are often linked to a given mixture of microorganisms. It is critical to understand, therefore, in the aggregate, just what those microorganisms are doing.
“We have developed a highly curated database for antibiotic resistance, pathogenicity traits, and metabolic-coding genes,” Dr. Colwell continues. That database may be the company’s greatest advantage.
The primary competition for microbiome analysis tends to be open-sourced software that often is developed by academic researchers. “Because those research groups depend on grant funding, the algorithmic tools aren’t updated routinely, and don’t include highly accurate, curated databases,” Dr. Hasan says.
“In our case, our platform is evolving over time, with constantly improved analytics and functionality, and periodic updates of our databases,” he continues. “This lets us provide high-quality diagnosis, detection, and cross-disciplinary microbiome screening and discovery.”
CosmosID’s development approach is different, too, Dr. Hasan says. Academic groups typically begin with genomic sequencing and then build analytic tools. “We built an analysis platform and a large, curated database first. Then we offered sequencing to ensure good data generation—a prerequisite for reliable and accurate data analysis.” This approach enhances quality and speed.
When the industry thinks of microbiome analysis, 16S ribosomal RNA sequencing usually comes to mind. It was the first technique developed and has significant limitations. Most notably, it analyzes only one of 4,000 genes to predict the identity of an organism. Consequently, the results are not sufficiently specific.
Whole-genome shotgun metagenomics, in contrast, is target-free and aims to sequence entire genomes of every organism in a sample. Therefore, it provides a more precise and unbiased biological representation of the microbial community in a sample (Figures 1&2). It identifies all microorganisms and the genes they carry (i.e., antibiotic resistance or virulence associated genes), allowing a more complete understanding of the microbiology of the sample being studied.
Tips for Better Analysis
To achieve desired results, however, study designs must improve. “Experimental design is critical,” Dr. Colwell stresses.
“Genomic analyses often are conducted without designing the overall study properly,” Dr. Hasan says. He advises establishing the objective and, especially, processing all the samples identically. “You must be very systematic, because specific questions must be asked and answered for every specific type of study.”
“It’s also important to collect and analyze enough samples to achieve statistically significant results,” Dr. Colwell adds. “This often isn’t considered early enough for subsequent analyses to advance the research.”
Knowing the Microbes Is Powerful
“The important part of any collaboration is to understand how powerful it is to know the microbial community present throughout drug development and production phases,” Dr. Colwell says. Knowing those communities and their interactions can significantly improve targeting accuracy and drug effectiveness.
The specificity and speed possible today were unheard of only a few years ago. It’s safe to say that Louis Pasteur, were he to enter Dr. Colwell’s laboratory today, would be delighted.