A High Throughput Assay for Measuring Secreted Protein Based on a de novo Fluorescent Reporter Reveals Regulatory and Structural Insights in Salmonella Type Three Secretion System

Abstract

Intracellular protein production in bacteria is limited by the need for lysis and costly purification. A promising alternative is to engineer the host organism for protein secretion. While the Salmonella enterica serovar Typhimurium (S. Typhimurium) Type 3 Secretion System (T3SS) has been utilized for protein secretion, its study and eventual applicability for recombinant protein production is constrained by the lack of high-throughput assays to quantitatively measure secretion titer. Developing such assays is challenging, as proteins must remain unfolded for secretion, limiting the use of several common reporter proteins. In this work, we develop a high-throughput secretion assay using mini-Fluorescent Activating Protein (mFAP). mFAP forms a chromophore only upon addition of an exogenous substrate, allowing secretion and subsequent fluorescence detection. We demonstrate mFAP secretion via the T3SS with an N-terminal secretion tag and show that the fluorescent signal in the secreted fraction is rapid and linear over three orders of magnitude. Using this assay, we screen S. Typhimurium strains with secretion-enhancing mutations, identifying a constitutively active strain and reveal temporally controlled secretion dynamics. We also show that this assay may be applicable to other secretion systems, providing a universal tool for tracking heterologous protein secretion.

Keywords: fluorescence assay; protein secretion; recombinant protein production; type III secretion system (T3SS).

Figure 1:

SptP-mFAP2a produces a fluorescent signal after secretion via the T3SS. A) S. Typhimurium strains secreting mFAP are separated into whole culture and supernatant by centrifugation. Each sample is added to a 96-well plate, incubated with 100 μM DFHBI and fluorescence is read on a plate reader. B) A strain harboring the SptP-mFAP2a secretion plasmid was grown for 8 hours, a sample of the secreted fraction was collected after centrifugation, and mFAP fluorescence was induced with DFHBI. The fluorescence of the induced samples was measured alongside secretion samples from a secretion incompetent strain (ΔprgI), and a strain harboring the non-fluorescent SptP-DH secretion plasmid. Error bars reflect the standard error the mean of three biological replicates. C) Western blot of the secreted fraction from 1B. Figure 1A created in BioRender. Summers, J. (2025) https://BioRender.com/c83r875

Figure 2:

The mFAP assay produces a linear signal and develops rapidly. A) Purified SptP-mFAP (1–100 mg/L) was diluted into the secreted fraction of S. Typhimurium strains lacking secretion plasmids. Fluorescence was measured after DFHBI addition, with linear fits generated in rich (LB-L) and defined (NCE) media. R² values indicate fit quality, with error bars showing standard deviation from three technical replicates. B) LB-L samples were measured over 10–90 minutes. Each color represents a different SptP-mFAP concentration (10–100 mg/L), with error bars indicating standard deviation from three technical replicates.

Figure 3:

The mFAP assay robustly measures relative protein expression and secretion across different conditions. A) Strains expressing different secretion tags fused to mFAP were compared to SptP-mFAP using fluorescence (dark gray) and semi-quantitative western blot (light gray). Error bars represent standard error from three biological replicates. Significance is marked by * (p < 0.05) and ** (p < 0.01) using Welch’s t-test. B) A strain expressing SptP-mFAP was grown, with whole-culture samples collected over 2–8 hours. Fluorescence and relative protein expression (via western blot) were compared to the 8-hour time point. Error bars represent standard error from three biological replicates, with significance marked by * (p < 0.05).

Figure 4:

The mFAP assay identifies a constitutively active secretion strain. Strains harboring combinations of genomic mutations (prgI::S49R, ΔsipBCD, ΔinvE, and ΔhilE, left axis) were transformed with the SptP-mFAP secretion plasmid and either no additional plasmid or the activation plasmid (bottom axis). Strains were grown in either base LB-L or LB-ES, and expression and secretion samples were collected and measured with the mFAP assay. Relative expression and secretion scores are relative to the WT, no plasmid, base LB-L condition and reflect the average of three biological replicates. A) Protein expression data. B) Protein secretion data.

Figure 5:

Measuring secretion dynamics over time. A) Strains harboring the SptP-mFAP secretion plasmid and either no additional plasmid, the hilA expression plasmid, ΔhilE or were grown, and secretion samples were taken every hour between 4 and 10 hours. Fluorescence of the secretion samples were measured with the mFAP assay. Error bars represent the standard error of the mean for three biological replicates. B) The density of the cultures as measured by OD₆₀₀

Figure 6:

mFAP can be used to track secretion through other secretory pathways. E. coli strains BL21(DE3) harboring an expression plasmid for mFAP fused to either OsmY or YebF were grown and secretion samples were collected at the end of culture. Secretion samples were measured for background subtracted fluorescence using the mFAP assay. Error bars represent the standard error of the mean of three biological replicates.