Scalable Dual-Fluorescence Assay for Functional Interpretation of HNF-4α Missense Variants.

AIM

The study aimed to develop a scalable dual-fluorescence assay in cells to enable the functional interpretation of HNF-4α missense variants identified in exome sequencing, which can be used to guide clinical diagnosis.

METHODS

Using mOrange2 and GFP fluorescence proteins to track the expression of HNF-4α (HNF-4α-mOrange2) and reporter activity under the control of the HNF-1α promoter (pHNF1A-GFP), respectively, we designed a dual-fluorescence assay to evaluate the expression level, cellular localization, and transcriptional function of HNF-4α simultaneously in live cells. To assess the scalable characteristic of the assay, a small library containing five previously reported mutations and wild-type HNF-4α was constructed. Cells infected with this library were sorted into different populations through fluorescence-activated cell sorting (FACS) according to the transcription activity and expression abundance. Cloning and Sanger sequencing were used to detect the mutations of the different groups. High content screening (HCS) assay was used for the validation of individual mutants in the function and expression point of view.

RESULTS

HNF-4α-mOrange2 exhibited nuclear localization and transactivation capability on the HNF-1α promoter as physical HNF-4α does. The expression of HNF-4α-mOrange2 shows a 6-fold induction of GFP expression compared to the control without HNF-4α-mOrange2, which was significantly abolished by the known loss-of-function mutant M373R. The different performances of wild-type and mutant M373R made them distinguishable in the FACS system, empowering the scalable capability of this assay for classifying large numbers of variants combining functional stratification and sequencing. Further application of the assay in the small library showed that three cell populations were seen grouped as Normal (same transactivation as wild type), Reduced (reduced transactivation with normal or higher expression), and Reduced (reduced transactivation with lower expression). Subsequently, Sanger sequencing showed that wild-type HNF-4α was in the Normal group, two mutations (M373R and G79C) were enriched in the Reduced group, and three mutations (C115S, L272P, and F83C) belonged to the Reduced group. These results were validated by further imaging data using HCS assay for individual mutation.

CONCLUSIONS

Our study proposes a scalable and informative approach for the characterization of the variants in HNF-4α genes in a quantitative and high-throughput manner.