By Aaron Krol
January 2, 2015 | I came to the Simoa Accelerator, a small lab located inside the Quanterix headquarters in Lexington, Mass., at a good time: the first experiment of the day had finished in the early afternoon, and the Simoa HD-1 instruments were free for a new run. Two Quanterix research associates, Dipika Gemani and Adam Shepro, were occupying the space, working on a grant from General Electric and the NFL. Together, they showed me how to load Simoa with their latest round of blood samples, which the machine would search for proteins that might be linked to head trauma or concussion.
The Simoa HD-1 is a huge instrument, like a double-wide fridge you might find in a restaurant kitchen. That’s because the machine is essentially a lab unto itself, with an onboard computer, a robotics system for handling samples, and an optical system for reading out test results. For all that, Simoa’s central innovation is centered on something very small: a disc etched with over 200,000 microwells, each of them twenty times smaller than a fine grain of sand. The machine works by trapping single proteins, combined with fluorescent molecules, inside these wells. If a well lights up, Simoa knows that the protein you’re looking for is present in your sample.
This process is almost identical to an ELISA test, an assay bench scientists have been using since the early 1970s. ELISA works by attaching a fluorescent enzyme to an antibody specially targeted to a particular protein. If your sample contains a sufficient amount of that protein, it will pick up the fluorescent signals, providing a quick, visual confirmation. The difference with Simoa is that it’s done on a much smaller scale, down to individual molecules. By separating out each protein into its own microwell, Simoa gives its optical system a much smaller field to search, and that means it can detect proteins at tiny concentrations — as little as one femtogram per milliliter. A femtogram is a hard unit of measurement to get your head around, but 1 fg/mL is about the same as a single gram of protein spread across 400,000 Olympic swimming pools.
“There are so many proteins in the world that science has never been able to measure in blood, because they’re at such a low concentration that no technology would pick it up,” says Kevin Hrusovsky, Executive Chairman of Quanterix, which designs the Simoa instruments. Even if those proteins are medically relevant, perhaps signaling the early stages of a disease, standard tests won’t find them until their concentrations spike dramatically. “What we’re trying to get to is protein levels so low that, in a normal person, before they have any issues, you can see [a disease-related protein appear.]”
Quanterix Director of Strategic Marketing & Collaborations David Hanlon shows off the Simoa Accelerator. Image credit: Quanterix
At the accelerator, Gemani and Shepro placed blood samples and plastic containers of reagents in their compartments inside Simoa, while I watched over their shoulders. They also inserted one of the disposable plastic discs containing the crucial microwells. To catch proteins in those wells, Quanterix uses tiny magnetic beads, just large enough to fit one to a well. The antibodies in a Simoa experiment are fixed to those beads, so if a protein is caught, it can be pulled into a well magnetically and sealed inside.
The research team shut the Simoa doors and turned to a touch screen on the outside of the instrument. On the screen, they named their samples and experiment, tapping on graphics of the different sample vials, told Simoa what tests they wanted to run, and let it double check that they’d used the right reagents. (The containers are barcoded so the machine will know what it’s working with, and can alert users if they’ve made a mistake.) Three minutes after opening the doors, the machine was humming away. Processing 180 samples, it would wrap up this second experiment of the day in around four hours.
“Bleed-to-Read”
Quanterix opened the Simoa Accelerator as a way to convince pharma companies, universities, contract labs, and other potential customers that the instrument is worth the price tag. “It started kind of organically, from a need to help customers evaluate the technology,” says Julien Bradley, the company’s Senior Director of Sales & Marketing. “But it’s just caught on fire.”
While the Accelerator only officially opened in December, customers have been circulating through the Quanterix lab for at least a year, bringing their own samples and designing their own assays to give the Simoa machines a test run. According to Bradley, three-quarters of clients brought into the lab for hands-on experiments have gone on to buy instruments of their own. With such a high conversion rate, Quanterix is now increasing its investment in the Accelerator concept, doubling the capacity to accommodate four teams at a time.
Quanterix is nothing if not ambitious. While the seven-year-old company is only just getting beyond startup mode, with around 65 employees and a single building for offices, R&D and the Simoa Accelerator, its business is built on every level as if in anticipation of sudden, explosive growth. Manufacturing has been farmed out to industrial giants: the Simoa machines are made by STRATEC, a German company that builds lab robotics systems on a large scale, and the discs come from a division of Sony, which engraves them in its old DVD-manufacturing facilities. “[We have] the scale of a big company when we’re still pretty small,” says Hrusovsky. “We feel really good about how rapidly we can expand the number of instruments we make.”
And while most new technologies can expect a long gestation period in basic research before they’re ready to be used in medicine, Quanterix is pushing for clinical use as soon as possible. Simoa was built with every expectation that it would be placed in hospital laboratories for in vitro diagnostics (IVDs), standard, mass-produced tests on live patient samples. That’s the philosophy that led to inside-the-box robotics for sample preparation, and to the step-by-step interface on the machine’s touch screen. While these time-saving measures are just conveniences in a research setting, in a busy clinical lab they can be essential.
“It’s a very different strategic approach,” says Hrusovsky. “We went after the IVD first, and built a box that is bleed-to-read. You can press a button, and you’ll get the answer.”
A major diagnostics company in France, bioMérieux, has signed on to file for IVD approval of Simoa with both the FDA and the European Medicines Agency. If all goes well, the Quanterix team estimates the first FDA-cleared tests for the Simoa could be available within two to four years. But they’re also hoping to leapfrog the approval process by pitching Simoa assays as laboratory developed tests (LDTs), a more loosely regulated market that has grown increasingly popular as a way to bring newer technologies to hospitals and doctors’ offices.
“The Root Cause”
In many ways, Simoa is an ideal instrument for labs that offer LDTs. These tests have some unique regulatory requirements, the most important of which is that they must be invented and performed entirely in a single lab. Because each new test on Simoa already has to be individually crafted, with new antibodies that will attach to your protein of interest, there’s plenty of room for in-house development. The instrument’s high throughput is also an advantage, moving tests that have so far been painstaking, case-by-case procedures onto a machine that can handle them in large volumes. It also has the benefit of working with blood or even saliva samples, in situations where an ELISA test, with its lower sensitivity, might require a tissue biopsy or fine-needle aspiration.
On the other hand, Simoa’s exquisite sensitivity could actually be a problem for labs planning to move their standard test menus onto the instrument. Even if a protein is known to signal disease at levels an ELISA assay would pick up, it might be perfectly benign at the minute levels Simoa can detect. Until new clinical studies can be run on these assays, customers just won’t know which proteins are medically relevant at ultra-low concentrations.
Still, LDT labs have often been the first ports of call for these kinds of studies, because providers who validate a new test internally have the chance to be first to market, bringing their assay to physicians before the FDA has cleared an equivalent test for mass production. DNA testing is a prime example: nearly every genetic test available to patients today is an LDT, validated not by the FDA but in the scientific literature or by individual labs’ research. Quanterix wants to duplicate this strategy for protein tests on Simoa.
“We think we’re going to do for proteins what Illumina did for DNA,” says Hrusovsky. Illumina is the market leader in DNA sequencing, whose products let scientists measure DNA faster, for less money, and at a far greater scale than would have been imaginable ten years ago. Yet, while medical centers are starting to embrace DNA sequencing as a diagnostic tool, Hrusovsky points out that a patient’s genetic code is only a poor indicator for most diseases. Your DNA doesn’t change much throughout your life, whether you’re sick or not, but the proteins your cells produce can change drastically.
“Once we get a really quantifiable technology for the protein, we’re going to get at the root cause of so much,” he says. “The protein is so much more influential to phenotypic behavior than the DNA.”
So far, DNA has the edge because of its convenience. Unlike an ELISA test, which is essentially a yes or no question about a single protein, a gene sequencer will read any DNA molecule it’s given, and high-throughput sequencers can run huge volumes of tests with relatively minor lab work. One way Quanterix hopes to make protein tests competitive is by using automation to boost the throughput. By using different fluorescent labels, Simoa can run up to 11 tests simultaneously, and on a big run it can accommodate well over a hundred samples at a time.
For early adopters, this has been one of the instrument’s biggest selling points. “We have customers running three shifts a day,” says Bradley, and that constant use means Quanterix is selling more than twice as many consumables per instrument as it expected.
It’s not yet clear which fields of medicine could most benefit from high-sensitivity protein assays, so Quanterix is trying to foster a wide range of customers and publications. Some uses for Simoa will be simple substitutions — taking current ELISA assays, for cancer-related proteins or risk factors for heart disease, and revalidating them on the Quanterix technology. Other projects are moon shots, hoping to find the first practical biomarkers for very difficult diseases.
The GE-NFL grant to explore a Simoa concussion test falls in this category, and Hrusovsky is broadly interested in neurological diseases — including some of the worst, like Alzheimer’s and Parkinson’s. “Any neurological disorder has certain markers,” he says. “We could be the player who helps quantitate, where today that whole field is so subjective.” An early sign of encouragement is that certain proteins from the cerebrospinal fluid, including the Alzheimer’s-linked tau and amyloid beta, seem to cross the blood-brain barrier in just high enough concentrations for Simoa to pick up in blood samples. While neither of these proteins is considered diagnostic, finding them in blood opens up new opportunities to study them as potential early warning signs.
Following up on such diverse areas of testing will require a large customer base, so Quanterix is hoping the Accelerator’s first year will be a busy one. With a bank of instruments ready to use in the lab, the company is open to new ideas and difficult problems.
“The power of the protein is untapped,” says Hrusovsky. “There are so many people suffering from diseases that we can get to much earlier.”
For more on Quanterix and Simoa, see “Quanterix Launches Single-Molecule Detection of Proteins, DNA.”
Update 1/5/14: The text of this article has been altered to credit the research associates who showed me around the Simoa Accelerator.