A Molecular Thermodynamic Model for Improving Point-of-Care Diagnostics Incorporating the Lateral-Flow Immunoassay and Aqueous Two-Phase Systems

Saturday, October 26, 2024

1:55 PM – 2:00 PM EST

Location: Room 342

Presenting Author(s)

Lillian Gong, N/A (she/her/hers)

Undergraduate Student
UCLA
Los Angeles, California, United States

Co-Author(s)

RC

Rohan Chawla

Undergraduate Student
University of California, Los Angeles, United States
EZ

Elaine Zheng

Undergraduate Student
UCLA, United States
ES

Emika G. Saito

Undergraduate Student
University of California, Los Angeles, United States
AL

Amy Lou

Undergraduate Student
UCLA, United States
MD

Mai Dang

Undergraduate Student
University of California, Los Angeles, United States
JL

Jiakun Lu

Graduate Student Researcher
UCLA, United States
MP

Matthew Prasetyo

Undergraduate Student
University of California, Los Angeles, United States

Primary Investigator(s)

Daniel T. Kamei, PhD

Professor of Bioengineering
UCLA, United States

Introduction: Although the lateral-flow immunoassay (LFA) has seen success as a point-of-care diagnostic, it suffers from low sensitivity when the target is present at low concentrations. To address this challenge, our group developed the Aqueous two-phase system (ATPS)-automated Concentration and Enhancement of the LFA (ACE-LFA), which utilizes ATPS separation on paper to both concentrate the sample and automatically and sequentially deliver signal enhancement reagents. This approach relies on the selective partitioning of 3,3’,5,5’-tetramethylbenzidine (TMB) and the target bacteria to opposite phases of the ATPS, which minimizes premature signal enhancement and ensures the delivery of the TMB enhancement reagent to the detection zone only after the bacteria has bound to the test line. Accordingly, the optimal combination of enhancement reagent and ATPS is one in which the enhancement reagent partitions significantly into the polymer-rich phase of a polymer-salt ATPS. Therefore, we developed a molecular thermodynamic model to predict this partitioning behavior. Using this model, we predicted the partitioning behavior of TMB and 3,3’-diaminobenzidine (DAB), another enhancement substrate, in a polymer-salt ATPSs and compared our predictions with experiments. We then confirmed our hypothesis that improved enhancement with darker test lines and minimal premature signal enhancement would be achieved when the enhancement reagent partitions extremely into the polymer-rich phase.

Materials and

Methods: The polymer-salt ATPS used in this paper was composed of 12 kDa molecular weight poly(ethylene glycol-ran-propylene glycol) (EOPO) and sodium citrate (SC) salt. After generating a phase diagram for this system using the cloud-point method, we introduced TMB and DAB into the ATPS and determined their partition coefficients using UV spectrophotometry to measure their concentrations in each phase at equilibrium. The partition coefficient was defined as the concentration of TMB or DAB in the top phase divided by its concentration in the bottom phase. To theoretically predict the partition coefficients, we assumed equilibrium and obtained expressions for the excess chemical potentials of TMB and DAB in each phase by applying the Universal quasichemical Functional-group Activity Coefficient (UNIFAC) model. An iterative MATLAB script was coded to solve the UNIFAC equations using the polymer concentrations of each phase determined from the measured phase diagrams. In addition, we designed an ACE-LFA diagnostic for a model target and varied the signal enhancement molecule in the ATPS to test the effect of the partition coefficient on test line intensity and premature signal enhancement. ACE-LFA tests were imaged in a controlled lighting environment, and line intensities were quantified with a custom Python script.

Results, Conclusions, and Discussions: Using parameters from the binodal curve, our model reasonably predicted the partition coefficients of TMB and DAB in the EOPO-SC ATPS (Fig. 1) for the screening of ATPS and enhancement reagent combinations. The ACE-LFA test strips exhibited a darker test line intensity for the positive test with TMB than the positive test with DAB (Figs. 2A and B) and demonstrated a positive correlation between partition coefficient and test line intensity for a positive sample (Figs. 3A, B, and C). These results are consistent with our hypothesis since TMB partitions more extremely into the polymer-rich phase than DAB, corresponding to a stronger signal enhancement since less reagent was lost to the leading phase. In contrast, DAB was present at a lower concentration in the lagging phase due to its lower partition coefficient, resulting in lower test line intensity after enhancement.

We have successfully established and demonstrated the use of our fully predictive thermodynamic model to calculate the partition coefficient of a small hydrophobic molecule in an ATPS with no fitted parameters. We have shown that more extreme partitioning corresponds to a greater signal enhancement of the test line and more sensitive ACE-LFA. The application of molecular thermodynamics to predict partitioning behavior has significant potential to identify optimal systems that can achieve improved detection limits at the point of care.

Acknowledgements (Optional): This work was supported by the National Science Foundation (NSF Grant 1707194 to DTK) and the Amgen Foundation (Amgen Scholars Program to MD).