(343d) Charge-Based Sensing Platforms for Exosome Diagnostics

RNA and protein biomarkers for diseases are enriched in extracellular vesicles (EV), such as lipoproteins and exosomes. These vesicles accommodate many membrane protein markers inherited from their parent cells. This heterogeneity of colocalized proteins allows more precise disease, location and stage identification than other liquid biopsy techniques. Multiple copies of the same proteins also allow high-affinity multivalent capture (avidity) with at least 3 logs drop in the dissociation constant. Our recent works on PON1 on HDL (Kumar, S. et al. Nat. Commun. 14: 557(2023)) and EGFR on sEV (Maniya, N. H. et al. Commun. Biol. 7: 677(2024)) indicate that enzymatic markers only remain active on the lipid membrane, thus rendering active EV enzymes a better representation than the dispersed inactive enzymes that may not originate from their active counterparts.

However, precision EV diagnostics still face two major challenges: low abundance (fM to pM) that requires high sensitivity and interference from dispersed proteins that yields false positives. We propose a charge-based sensing approach to overcome the specificity issue due to interference. As most interfering proteins are weakly charged, they do not produce a charge signal. In addition, we design antibody-conjugated silica nanoparticle (SiNP) charge reporters of the optimum size, such that the undocked reporters can be easily removed with a controlled washing step. This enhanced specificity improves the signal-to-noise ratio so that EV diagnostics can be conducted with untreated plasma. We address the sensitivity issue for this charge-based sensing strategy with an ion-depletion module based on permselective membranes (Slouka, Z. et al. Annu. Rev. Anal. Chem. 7: 317-335(2014)). Due to unidirectional counterion flux through the membrane, ionic strength polarization occurs across the membrane such that the sensing region is at DI condition (Debye length of a few hundred nanometers), compared with the bulk buffer of untreated plasma with an ionic strength that is 6 logs higher (< 1 nm Debye length). This allows the 50 nm SiNP reporters on top of the ~100 nm EV to be detected.

Two charge-based EV diagnostic technologies will be presented. One is a proteomic EV array based on membrane sensors, shown in Figure(a), that uses the same depletion membrane module as a sensor. With ion depletion, the bound SiNP reporters can effectively gate the ion current to produce a large signal in the "over-limiting" current region, as their charge and size can significantly affect the electroconvective instability responsible for the over-limiting current (Sensale, S. et al. The Journal of Physical Chemistry B 125: 1906-1915(2021)). A limit of detection of about 10 fM can be achieved. We have successfully detected multiple colocalized sEV markers such as CD99/NGFR for Ewing Sarcoma and MUCI/ITGβ3 for Ovarian cancer, in a cohort of 20 patients using this multiplexed array.

Our second charge-based EV technology is an integration of the field-effect transistor (FET) sensor with the ion-depletion module. Instead of using graphene or other expensive 2D semiconductors, we show that, with the depletion module, the sensitivity of a simple and inexpensive commercial extended-gate FET, as shown in Figure(b), can be improved 3 logs to yield a sub-fM limit of detection. As micron-sized FET sensors can be fabricated in massive quantities, this integration of ionic and electronic conducting materials promises an EV diagnostic platform for a massively large number of biomarkers. The extension includes a miRNA FET array downstream of the proteomic EV array to profile exosomal miRNAs (Ramshani, Z. et al. Commun. Biol. 2: 189(2019)) for specific EVs.