(22c) A Microfluidic Platform to Enable Single Cell Transcriptomic Analysis of Cells with Defective HIV Genomes

Defective HIV provirus represents 90% of the treatment-resistant reservoir in HIV infected people. However, its impact is poorly understood because of the low frequency of infected cells (1 in 10,000). Defective proviruses can also produce viral RNA and protein which can elicit chronic inflammation that is thought to place patients at risk for a variety of co-morbidities. Furthermore, defective viruses can interfere with broadly neutralizing antibody (bNAb)-mediated control of viral replication, masking the efficacy against rarer, intact viruses. Elimination of these rare cells could have a significant impact for both cure as well as reducing the multiple co-morbidities in aging patients with HIV infection. A platform that can identify and exclude defective genomes from workflows can also assist in the studies of mutations that confer resistance to antiretroviral therapies (ART) in reservoir sites. Current single cell technologies are limited in their efficiency to capture rare cell populations. To target this shortcoming, we developed a microfluidic single cell nucleic acid sorting platform to enrich for rare cell populations for downstream single cell sequencing.

We use an innovative strategy to isolate rare, infected cells from peripheral blood for downstream genomic analysis. Our method to overcome the barrier of identifying rare molecular phenotypes is based on a microfluidic platform incorporating a single cell droplet generator, picoinjector, and magnetically actuated sorting device. First, our droplet generator encapsulates single CD4+ T cells into water-in-oil droplets along with magnetic microbeads and a barcoded bead. These droplets are then injected with RT-PCR reagents targeting provirus with the picoinjector which uses liquid PEDOT:PSS electrodes for longevity. Finally, fluorescent droplets with the target genome are sorted out on the basis of magnetic actuation to sequence defective genomes and the infected cell transcriptome.

The three microfluidic device platform successfully analyzed viral HIV genomes at the single cell level. Droplets containing single cells were generated at a rate of 10³ droplets per minute with about 30% of the droplets containing one cell. Parallel operation of droplet devices enabled high throughput processing of millions of cells. The subsequent picoinjector had a 100% injection rate of RT-PCR agents. Droplets underwent the RT-PCR reaction off-chip in a 1-step reaction with no detectable evaporation. Successful droplet RT-PCR was demonstrated by the detection of as low as 1 – 2 copies of the target HIV gene in a droplet. Finally, the magnetic beads co-encapsulated with the cells enabled the manipulation and enrichment of fluorescent droplets in the third and final device to select for HIV+ droplets.

Overall, we have created a microfluidic platform that combines three devices to streamline the high throughput identification and enrichment of defective HIV reservoirs. Our platform has throughput that is competitive with commercial single cell processing platforms and enables enrichment of target cells via nucleic acid sorting for downstream single cell sequencing. Not only do we hope to use this microfluidic platform to understand the impact of defective HIV reservoirs on viral pathogenesis, persistence, and treatment control, but we also envision utility for this platform to study other disease models convoluted by the presence of rare disease-ridden cells.