Skip to main content

MIT Researchers Developing Amplification-Free 'Massively Parallel' Microfluidic Testing Platform


NEW YORK – Massachusetts Institute of Technology researchers said that they have developed a technique that could significantly boost protein biomarker test performance and improve molecular testing workflows.

Reporting the results of a study this week in the Proceedings of the National Academy of the Sciences, the researchers said that a nanofluidic enrichment system they've developed increased the proportion of nucleic acid and protein disease biomarkers a billionfold and avoided pitfalls that can sometimes trigger false positives when molecular tests amplify nucleic acids for detection.

Importantly, the technique achieved sensitivity levels better than those achieved by enzyme-linked immunosorbent assays, or ELISAs, used to detect protein biomarkers, Jongyoon Han, one of the lead developers and a researcher in the department of electrical engineering and computer science at MIT, said in an interview. As a result, the technology, called Holmes, has potential to one day become a protein detection platform, he added.

The MIT technology detected protein concentrations of about five orders of magnitude lower than the detection limit for a routinely used ELISA — which has a detection limit of about 1 picometer — with significantly shorter time, up to 60 minutes compared to several hours, or more, the researchers said.

Another current method for detection, PCR, requires "stringent and complex isolation and purification of nucleic acids," and the amplification process is subject to many pitfalls, Han said. He added that the MIT technology directly prepares clinical samples for detection by selectively collecting them into a microfluidic space, "which greatly simplifies the workflow and avoids many issues of PCR."

Their hierarchical nanofluidic molecular enrichment system directly concentrates nucleic acids or proteins in clinical samples by using electric fields to shuttle, manipulate, and collect them in a microfluidic space, Han said. Holmes has a hierarchical architecture with vertically stacked, "massively parallel" microchannels — tens of thousands of parallel microchannels — in a first stage of sample processing and a single microchannel in the final stage. Between these stages, the numbers of microchannels are scaled down tenfold to one hundredfold per stage as the target-molecule concentration increases.

Combining electrical signals and this hierarchical structure, Holmes selectively increases the proportion of detectable target biomolecules within a given volume and simultaneously depletes nontarget molecules in complex crude samples, enhancing overall detection, Han said. The researchers said that they used fluorescence to detect nucleic acids in urine and serum within 35 minutes, and HIV p24 proteins in serum within 60 minutes, but the platform can integrate many types of diagnostic detection techniques.

In their study, they achieved molecular concentration levels of between 1 and 100 attomolars, which are similar to levels achieved by PCR-based diagnostics and the emerging area of CRISPR-based diagnostics, Wei Ouyang, the second developer who is Han's colleague at MIT and also an author of the PNAS paper, said in an interview.

However, the process they use to make molecules more detectable contrasts sharply with those used in PCR and CRISPR diagnostics, both of which use amplification to significantly increase the number of copies of target molecules in a clinical sample, he said.

"If this technology can be scaled up and applied to an unmet need that can’t be addressed with enzymatic amplification, then it could certainly make a significant impact," said Shana Kelley, who runs a laboratory at the University of Toronto focused on developing molecules and devices that enable biological activities to be measured and manipulated. Kelley is not associated with the work at MIT.

"At the same time, the bar is quite high to displace conventional approaches, given advances in the automation of isothermal and other amplification strategies," she said.

PCR is heavily used and well-integrated across the clinical testing landscape, but it has enough challenges that it's worth exploring amplification-free diagnostic methods, Kelley said. PCR is difficult to automate in a cost-effective way and usually can't be moved out of a "pristine lab setting" because of the potential of false positives from contamination, she said.

The PCR process can also necessitate extensive sample preparation, which requires automation and slows down testing, she noted.

A device that is amplification-free and provides detection from crude samples that consist of many non-target molecules or cells "can be very powerful," Kelley said. Getting a diagnostic result from a crude sample removes bias that sample preparation might introduce, eliminates concerns about losing samples, and lowers the number of steps involved in testing, she said.

Han and Ouyang said that they recognize their technique may take many years of additional development along a path to commercialization. Nonetheless, they aim to eventually integrate their technique as part of a clinical diagnostic test. MIT is well placed to take new technologies such as this to market, and given the technique's potential, an entrepreneur outside the institute may also be interested in doing so, they said.

Han did not want to project a potential timeline for commercializing the technology, but he noted that the researchers anticipate that the platform will complement rather than displace well-established nucleic-acid diagnostic tests. By quickly increasing the concentration of target molecules, Holmes could be best applied to enhance workflows for PCR-based molecular tests, and tests that use CRISPR, nano-based biosensors, and sequencing, they said.

The team has begun exploring ways that Holmes could be integrated with other types of testing technologies.

Further, based on the results of their preliminary study in PNAS, Holmes has potential as a standalone platform used to detect proteins, they said, noting that the device is fabricated using low-cost polydimethylsiloxane and operates by use of direct current voltages and gravitational flows, making it potentially suitable for point-of-care use.

Similar to challenges faced by any new technology entering the crowed diagnostics space, finding an application that existing devices cannot handle is going to be an important part of the MIT platform's ability to become commercial, Kelley said.

She pointed out that any new platform that does highly sensitive detection of proteins and nucleic acids would need to compete with Quanterix's Simoa platform, an immunoassay that allows detection of proteins and nucleic acids at attomolar levels.

Nonetheless, if the MIT team can keep a small system footprint in detecting target proteins indicative of diseases, "it can be really powerful," Kelley said. The most sensitive protein detection systems have testing instruments that are too large for transportation and field use, she said.