Centrifugation-Assisted Immiscible Fluid Filtration for Dual-Bioanalyte Extraction

Abstract

The extraction of bioanalytes is the first step in many diagnostic and analytical assays. However, most bioanalyte extraction methods require extensive dilution-based washing processes that are not only time-consuming and laborious but can also result in significant sample loss, limiting their applications in rare sample analyses. Here, we present a method that enables the efficient extraction of multiple different bioanalytes from rare samples (down to 10 cells) without washing-centrifugation-assisted immiscible fluid filtration (CIFF). CIFF utilizes centrifugal force to drive the movement of analyte-bound glass microbeads from an aqueous sample into an immiscible hydrophobic solution to perform an efficient, simple, and nondilutive extraction. The method can be performed using conventional polymerase chain reaction (PCR) tubes with no requirement of specialized devices, columns, or instruments, making it broadly accessible and cost-effective. The CIFF process can effectively remove approximately 99.5% of the aqueous sample in one extraction with only 0.5% residual carryover, whereas a traditional "spin-down and aspirate" operation results in a higher 3.6% carryover. Another unique aspect of CIFF is its ability to perform two different solid-phase bioanalytes extractions simultaneously within a single vessel without fractionating the sample or performing serial extractions. Here we demonstrate efficient mRNA and DNA extraction from low-input samples (down to 10 cells) with slightly higher to comparable recovery compared to a traditional column-based extraction technique and the simultaneous extraction of two different proteins in the same tube using CIFF.

Figure 1.

Principle of Centrifugation-assisted Immiscible Fluid Filtration (CIFF). (A) CIFF comprises of two immiscible liquid phases, an aqueous phase (d: 1 g/mL) and a hydrophobic phase that is denser than the aqueous phase (FC-3283 fluorinated oil, d: 1.82 g/mL) underlying the bottom of the aqueous sample. Functionalized glass microbeads (d: 2.48 g/mL) are mixed with the aqueous sample to bind analytes of interest, then loaded into a reservoir (here we used a 0.2 mL PCR tube for demonstration). Due to the hydrophilicity of the glass microbeads, they remain trapped inside the aqueous phase even after vigorous mixing (B). Under centrifugation however, the increased centrifugal force allows the glass microbeads to overcome the lipophobic resistance between the glass beads and oil phase and partition according to their density into the bottom oil phase, which effectively removes non-target molecules behind in the aqueous phase (C). (D) Parallel CIFF processes can be performed using conventional multi-well PCR plates.

Figure 2.

Characterization of CIFF. (A) Schematic of the CIFF process. (B) Relation of centrifugal force and amount of glass beads to the “jumping” of beads in CIFF. The smaller the amounts of added beads, the higher the centrifugal force required for the beads to overcome the resistance from jumping into the oil phase. (C) Percentage of residual glass microbeads (relative to input) left in the aqueous phase after performing CIFF for different amounts of input beads. Beads were all spun at 10,000 RCF for this measurement. (D) Relation of centrifugal force and Triton X-100 concentration in the aqueous phase to the “jumping” of 10 mg beads in the CIFF process. (E) Percentage of aqueous phase carryover of 10 mg beads after performing CIFF with or without 0.1% Triton X-100 in the aqueous phase. This was compared to a traditional “spin down beads, aspirate supernatant, and resuspend” approach (without CIFF). Error bars denote the standard deviation from 3 technical replicates. Statistical significance is represented by * p ≤ 0.05, ** p ≤ 0.01 and *** p ≤ 0.001.

Figure 3.

DNA and RNA extraction performance in CIFF. (A) qPCR performance of LINE1 DNA extracted using CIFF only, compared to CIFF plus two additional washes with buffer AW1 and AW2 (Qiagen). (B) RT-qPCR performance of RPLP0 mRNA extracted using CIFF only, compared to CIFF plus one single wash with RNA washing buffer. The RNA sample was not diluted prior to RT-qPCR, thus one extra wash was necessary to further remove residual lysis/binding buffer contamination which causes inhibition of RT-qPCR (see Figure S2). (C) mRNA extraction performance of CIFF with one single wash compared to a conventional manual 3-wash operation. ND: Not detected. Error bars denote the standard deviation from 2 technical replicates. Statistical significance is represented by * p ≤ 0.05, ** p ≤ 0.01 and *** p ≤ 0.001, NS = not significant.

Figure 4.

Simultaneous extraction of two different proteins (Alexa Fluor 647 anti-mouse IgG and Alexa Fluor 488 anti-rabbit IgG) from the same aqueous phase using dual-CIFF. The differential densities of mouse IgG-conjugated glass microbubbles and rabbit IgG-conjugated glass microbeads allow them to travel in opposite directions (floating vs. sinking) under centrifugation and simultaneously extract different target analytes to their respective top and bottom oil layers. Scale bar: 100 μm.