Microfluidic devices for active and passive separation of submicron particles

Abstract

Cancer is the leading cause of death in the developed world and the second leading cause in the developing world. Early detection improves the survival rates of cancer patients dramatically, but current methods of diagnosis are invasive and lack both sensitivity and specificity. Microfluidics has the potential to efficiently isolate and analyze cancer markers from blood and other biological fluids directly and thus enable a minimally invasive liquid biopsy. In this work we have developed microfluidic devices for separation of submicron particles with an intended application in isolation of tumor-derived exosomes from biological fluids. Exosomes are small extracellular vesicles (30–150 nm) present in body fluids including blood and urine, and represent a rich source of biological information. Protein profiling and RNA analysis of exosomes show great promise for cancer diagnosis and treatment monitoring. Furthermore, exosomes are well-suited targets for microfluidic isolation as they are relatively abundant in biological fluids and require only small volumes for analysis (≤100 μL). The microfluidic devices developed herein are based on label-free isolation — physical properties such as size and density instead of specific binding events. This limits selection bias and removes the requirement for time-consuming incubation steps. Both active and passive separation principles were explored in this work. Passive filtration of submicron particles was achieved using PDMS crossflow filtration devices with precisely defined cutoff sizes using electron-beam lithography (EBL). Active separation with submicron cutoff size was achieved using acoustophoresis in a standing surface acoustic wave (SSAW) device. A three-dimensional hydrodynamic focusing design developed for on-chip alginate fiber synthesis was adopted for particle alignment in the SSAW device to enhance separation efficiency. The resulting hybrid SU-8/PDMS device is capable of separating submicron particles at higher flow rates and lower power than previously reported while increasing fabrication yields. Novel micro- and nanofabrication techniques for PDMS-based microfluidics have been developed in the process. This includes a fast-writing process for electron-beam lithography of SU-8 molds with nanoscale resolution compatible with soft lithography, thereby challenging the assumption of EBL not being suitable for rapid prototyping. Furthermore, a novel technique for making planar processing available to PDMS films was developed for the fabrication of PDMS membranes. The technique circumvents issues with PDMS swelling in commonly used solvents and thermal expansion during metallization by transferring highresolution metal patterns to PDMS and using them as masks for dry etching of through-hole membranes. The technique is complementary to soft lithography, which is limited to quasithree-dimensional structures. The final PDMS membranes can be embedded in PDMS devices directly using a standard oxygen plasma and the process has potential for large-scale production. Altogether, this work broadens the scope of PDMS fabrication to bring advanced lab-on-achip devices for early cancer diagnosis closer to reality. The focus on submicron separation is driven by the need for efficient isolation of exosomes as they are of great interest to cancer diagnostics and play to the strengths of microfluidics. The platforms developed here are currently being explored for isolation of exosomes from biological fluids.

I vesten er kreft den leiande dødsårsaka i dag. I den tredje verda er det den andre leiande årsaka. Pasientar som får påvist kreft på eit tidleg stadium, har langt større sjansar for å bli friske enn dei som oppdagar det seint. Dessutan sparer det pasienten for store plager og samfunnet for store kostnader. Dessverre er dagens verktøy for kreftdiagnostikk ofte lite spesifikke og krev store pasientinngrep. Dei siste tiåra har gjeve oss kunnskap som tilseier at vi kan oppdage kreft med å analysere ein enkel dråpe blod. Problemet er berre at informasjonen gøymer seg i celler, partiklar og arvestoff som er vanskelege å få tak i med vanlege metodar som til dømes sentrifugering. Ny teknologi basert på mikrofluidikk kan snu dette i tida som kjem. I doktorgradsarbeidet mitt har eg utvikla mikrobrikkar for å separere partiklar med storleik mindre enn ein mikrometer. Desse mikrobrikkane er mynta på isolasjon av naturlege nanopartiklar (30-150nm) som vert kalla for eksosom. Analysar av protein og arvestoff frå eksosom har allereie vist stort potensiale til bruk i kreftdiagnostikk og til overvaking av korleis pasientar responderer på behandling. Ved bruk av nanoteknologi har vi laga plastbrikkar som kan brukast til kontinuerlig isolering av partiklar på størrelse med eksosom frå væske. Støypeformene for desse plastbrikkane kan lagast i løpet av nokre minutt ved å skanne ein elektronstråle over eit elektronsensitivt material som gjer det enkelt å gå frå idé til prototype. Vi har òg utvikla mikrobrikkar som kan skilje submikrometer-partiklar frå kvarandre ved bruk av ultralyd. Såleis kan vi kontrollere korleis partiklane beveger seg inne i ein mikrokanal ved å endre intensiteten på lydbølgjene. No er vi i ferd med å teste desse plattformene for isolering av eksosom frå biologiske væsker.