The engineering of a robust and sensitive stress hormone detection circuit is a crucial step in the development of the biomedical tattoo that monitors mental stress and offers an early warning to our users with a visible output. Through multiple design-test-learn cycles, we have managed to develop Tetstress, a novel synthetic transcriptional factor that effectively senses glucocorticoid, a well-established stress hormone whose subcutaneous level raises significantly upon chronic stress. In our Results and Proof of Concept pages, we have demonstrated the glucocorticoid-dose-dependent, visual signal produced by the Tetstress-based glucocorticoid-sensing circuit in engineered HEK-293T cells. In this Engineering Success page, we mainly discuss the design-test-learn cycles we went through during the development of the glucocorticoid-responsive transcriptional factor Tetstress (Figure 1).
Figure 1. The Design-Test-Learn cycles we went through during the development of an effective Tetstress.
Design: The most apparent approach to enable glucocorticoid sensing in mammalian cells is to rewire the endogenous glucocorticoid sensing pathways to ectopic gene expression. In humans, glucocorticoid receptor (GR) acts as a transcriptional activator in response to cortisol stimulation. Upon glucocorticoid stimulation, GRs translocate into the nucleus and bind specifically to the glucocorticoid response element (GRE), therefore initiating the expression of downstream genes. Utilizing GR as a glucocorticoid sensor, we designed chimeric promoters (PGRE3 and PGRE6) carrying either 3- or 6-copy GRE sequences upstream of a miniCMV promoter. Glucocorticoid-mediated activation of GR triggers the GR-PGRE binding and therefore activates downstream gene transcription (Figure 2).
Figure 2. Human glucocorticoid receptor (GR)-based genetic circuits. Schematic representation of the GR-based glucocorticoid sensing circuit. Overexpressed GR binds to glucocorticoid, moves into the nucleus, and binds to a synthetic GR responsive promoter (PGRE), activating downstream gene expression.
Build and Test: To evaluate the glucocorticoid-sensing performance of the GR-based glucocorticoid-sensing circuit, we co-transfected HEK-293T cells with the GR-expressing pNC046 plasmid and pNC047 (or pNC048) encoding PGRE3- or PGRE6-driven secreted alkaline phosphatase (SEAP)-expressing cassette. As shown in Figure 3, a mild upregulation of SEAP expression in cells treated with 100 nM glucocorticoid compared to the un-stimulated controls (1.92-fold for the PGRE3 group and 1.55-fold for the PGRE6 group) was observed. Cells treated with either 10 nM or 50 nM of glucocorticoids, however, showed no significant SEAP activation compared to the control cells.
Figure 3. Glucocorticoid responsive transcriptional activation mediated by the Human glucocorticoid receptor (GR)-based genetic circuits. PGRE-mediated SEAP production in glucocorticoid-stimulated HEK-293T cells co-transfected with pNC046 (PCMV-GR) and indicated reporter plasmids (pNC047 GRE3-Pmini-SEAP or pNC048 GRE6-Pmini-SEAP). SEAP production was measured 48 h post-DMEM or glucocorticoid stimulation; data shows mean±SD, n=3 independent experiments.
To understand the cause of the low sensitivity, we linked the Enhanced Green Fluorescent Protein (EGFP) to GR, which allowed us to observe whether GRs were correctly imported into the nucleus upon glucocorticoid stimulation as previously reported. Surprisingly, we observed a robust nucleus-alike localization of the GR-EGFP protein in pNC049 (PCMV-GR-EGFP) transfected cells that were not even treated with glucocorticoids (Figure 4, left panel). Confocal images of HEK-293T cells co-transfected with actin-labeling pXQ100 (PCMV-lifeact-mcherry) and pNC049 (PCMV-GR-EGFP) also showed that GR-EGFP proteins were majorly located in an actin-free, nucleus-alike location (Figure 4, right panel) in both untreated control cells and glucocorticoid-treated cells.
Figure 4. EGFP-labeled GR localization in HEK-293T cells w/wo glucocorticoid stimulation. For wide-field microscopy, cells were transfected with pNC049 (PCMV-GR-EGFP). For confocal images, cells were co-transfected with F-actin labeling pXQ100 (PCMV-lifeact-mcherry) and GR labeling pWY049 (PCMV-GR-EGFP). Photos were taken 48 h post-DMEM or glucocorticoid- administration, scale bar: 100 µm for wide-field microscopy and 10 µm for confocal microscopy. Data are representative images of 3 independent experiments.
The Ligand Binding Domain (LBD) of human GR participants in activities including ligand binding, subcellular localization, and transcriptional activation. Therefore, we wondered whether there were some unknown components in HEK-293T that affected LBD, causing the incorrect location full-length GR had. Hence, we constructed a fusion protein comprising the GRLBD domain and EGFP (Figure 5a) to test our hypothesis. To our delight, both wide-field and confocal fluorescent imaging showed a broadly spread EGFP signal in the nucleus and cytoplasm of untreated, pNC051 (PCMV-GRLBD-EGFP) transfected HEK-293 cells. In contrast, a nucleus localization of GRLBD-EGFP can be observed in cells stimulated by glucocorticoid (Figure 5b). Taken together, these results suggested that GRLBD might still functional as a glucocorticoid-sensing element.
Figure 5. Glucocorticoid-stimulated translocation of GR Ligand Binding Domain (LBD) in HEK-293T cells. (a) Schematic representation of the GRLBD-EGFP construct. The LBD domain of human GR was fused to the N-Terminus of EGFP with a GGGSG linker. (b) EGFP-labeled GRLBD localization in HEK-293T cells w/wo glucocorticoid stimulation. For wide-field microscopy, cells were transfected with pNC051 (PCMV-GRLBD-EGFP). For confocal images, cells were co-transfected with F-actin labeling pXQ100 (PCMV-lifeact-mcherry) and GRLBD labeling pNC051 (PCMV-GRLBD-EGFP). Photos were taken 48 h post-DMEM or glucocorticoid administration, scale bar: 100 µm for wide-field microscopy and 10 µm for confocal microscopy. Data are representative images of 3 independent experiments.
Learn: The imaging results of GR-EGFP surprisingly revealed that GRs might not be correctly located in HEK-293T cells, which is consistent to the SEAP assay showing a low sensitivity and poor transcriptional activation function. Luckily, the truncation of GR into GRLBD successfully rescued the glucocorticoid-responsive translocation of GRLBD-EGFP fusion protein. We thereby followed up this lead to design a new GRLBD-based glucocorticoid-sensing circuit.
Design: Building on the knowledge we obtained from the first engineering cycle, we then designed Tetstress, a novel synthetic transcriptional factor based on GRLBD. To reconstitute the DNA binding function of the lost GR DBD, we supplemented a DNA-binding tetR domain on either the C-terminus or the N-terminus of GRLBD. With this well-established DNA binding protein, we can also rewire the GR signaling to the well-established PTetO7 promoter. Furthermore, since the lost GR NTD domain carries a transcriptional activation function, we designed several variants adding the VP64 transcriptional activator as a replacement (Figure 6).
Figure 6. Design of tetR-GRLBD-based glucocorticoid responsive circuit. tetR was fused to GRLBD on either its N-terminus or C-terminus. Optional VP64 domains were included to further improve transcriptional activation efficiency.
Build and Test: To assess the function of the different Tetstress variants, we co-transfected the tetR reporter plasmid pXQ164 with a Tetstress variant (pNC054 Pcmv-tetR-GRLBD, pNC057 Pcmv-tetR-VP64-GRLBD, pNC056 Pcmv-GRLBD-tetR or pNC058 Pcmv-GRLBD-tetR-VP64) into HEK-293T cells. A major improvement of glucocorticoid response in both tetR-GRLBD and GRLBD-tetR configurations (5.21-fold for tetR-GRLBD and 22.50-fold for GRLBD-tetR compared to their respective non-treated controls) can be seen from the result. Meanwhile, the corresponding VP64-supplemented variants did not show significant enhancement in glucocorticoid response as we wished, which might be caused by the high background activation (Figure 7a). Further characterization of pNC054 (Pcmv-tetR-GRLBD) transfected cells illustrated that the tetR-GRLBD configuration resulted in a strong SEAP activation and a high basal expression, which significantly narrowed the dynamic range (approximately between 0-50 nM) of the circuit (Figure 7b). For the GRLBD-tetR configuration, HEK-293T cells co-transfected with pXQ164 (PTetO7-SEAP) and pNC056 (Pcmv-GRLBD-tetR) showed a much lower basal leakage compared to the tetR-GRLBD configuration, resulting in a significantly wider dynamic range (wider than 0-100 nM) and a better dose dependence (Figure 7c).
Figure 7. Glucocorticoid responsive transcriptional activation mediated by the synthetic Tetstress transcriptional factor (a) Glucocorticoid-activated SEAP production in HEK-293T cells co-transfected with pXQ164 (PTetO7-SEAP) and a tetR-GRLBD variant (pNC054 Pcmv-tetR-GRLBD, pNC057 Pcmv-tetR-VP64-GRLBD, pNC056 Pcmv-GRLBD-tetR or pNC058 Pcmv-GRLBD-tetR-VP64). SEAP production was measured 48 h post-DMEM or glucocorticoid stimulation; data shows mean value, n=2 independent experiments. (b-c) Glucocorticoid-dose-dependent SEAP production in HEK-293T cells mediated by the optimal tetR-GRLBD in (a). Cells were co-transfected with pXQ164 (PTetO7-SEAP) and a tetR-GRLBD variant (pNC054 Pcmv-tetR-GRLBD for c, pNC056 Pcmv-GRLBD-tetR for (c). SEAP production was measured 48 h post-DMEM or glucocorticoid stimulation. Data shows mean±SD, n=3 independent experiments.
Learn: Despite the significantly boosted glucocorticoid responsiveness of the Tetstress variants-based circuits, compared to the GR-based circuit, we still expect for higher sensitivity and signal intensity, especially for the GRLBD-tetR configuration. For this reason, we had a discussion with Mirta Viviani about our project, who is a Ph.D. student from Westlake University and experienced in engineering mammalian cell gene circuits (see integrated HP page for more information). She suggested several potential approaches like optimizing PTetO7, changing linkers between tetR and GRLBD, and further manipulating the subcellular localization of the Tetstress transcriptional factor. Meanwhile, our modeling team performed a sensitivity screening among all the parameters that can be changed by circuit design. Their results showed that the circuit’s sensitivity largely depends on the efficiency of nuclear import of Tetstress upon glucocorticoid stimulation (see Model page for more information).
Design: Building on the suggestions and findings we obtained in the previous engineering cycle, we then designed two new approaches that may improve Tetstress performance: 1) by optimizing the linkers between tetR and GRLBD, and 2) by integrating a nuclear export signal (NES) to manipulate the subcellular location of Tetstress. For the first approach, taking the tetR-GRLBD configuration as an example, we replaced the original 1xGS linker comprising a GGGSG sequence in the pNC054 (Pcmv-tetR-GRLBD) plasmid into a 3xGS ([GGGSG]3) or a 5xGS ([GGGSG]5) linker. Another configuration with no linker added between tetR and GRLBD was also generated (Figure 8a). For the second approach, a GSG-linker flanked NES was inserted between tetR and GRLBD in NES integration (Figure 8b).
Figure 8. Linker-optimized tetR-GRLBD variant and NES-integrating tetR-GRLBD variant (a) Linker-optimized tetR-GRLBD variant (pNC064 Pcmv-tetR-[no linker]-GRLBD, pNC054 Pcmv-tetR-GRLBD, pNC065 Pcmv-tetR-[3xGS]-GRLBD, or pNC066 Pcmv-tetR-[5xGS]-GRLBD). (b) NES-integrating tetR-GRLBD variant (pNC054 Pcmv-tetR-GRLBD, pNC067 Pcmv-tetR-NES-GRLBD, pNC056 Pcmv-GRLBD-tetR or pNC069 Pcmv GRLBD-NES-tetR).
Build and Test: While the removal of the linkers totally abolished the transcriptional activation activity of Tetstress, we were surprised to find that HEK-293T cells co-transfected with pXQ164 (PTetO7-SEAP), and one of these variants carrying GS linkers (pNC054 Pcmv-tetR-GRLBD, pNC065 Pcmv-tetR-[3xGS]-GRLBD, or pNC066 Pcmv-tetR-[5xGS]-GRLBD) showed a very similar glucocorticoid-stimulated SEAP expression level. We even observed an increasing basal SEAP production in configurations carrying longer GS linkers compared to the original 1xGS linker configuration (Figure 9a). Intriguingly, for the NES integrated Tetstress variants, while NES integration showed no significant improvement in the tetR-GRLBD configuration, a striking boost of glucocorticoid-stimulated SEAP production was observed in HEK-293T cells co-transfected with pXQ164 (PTetO7-SEAP) and pNC069 (PCMV-GRLBD-NES-tetR) compared to the cells transfected with the variant without NES (91.6x vs. 25x, compared to their untreated control). More importantly, the basal leakage SEAP expression remained similar between the two configurations (Figure 9b).
Figure 9. Nuclear export signal (NES) allowed further improvement of Tetstress performance. (a) Glucocorticoid-activated SEAP production in HEK-293T cells co-transfected with pXQ164 (PTetO7-SEAP) and a linker-optimized tetR-GRLBD variant (pNC064 Pcmv-tetR-[no linker]-GRLBD, pNC054 Pcmv-tetR-GRLBD, pNC065 Pcmv-tetR-[3xGS]-GRLBD, or pNC066 Pcmv-tetR-[5xGS]-GRLBD). (b) Glucocorticoid-activated SEAP production in HEK-293T cells co-transfected with pXQ164 (PTetO7-SEAP) and a NES-integrating tetR-GRLBD variant (pNC054 Pcmv-tetR-GRLBD, pNC067 Pcmv-tetR-NES-GRLBD, pNC056 Pcmv-GRLBD-tetR or pNC069 Pcmv GRLBD-NES-tetR). For panels a and b, SEAP production was measured 64 h post-DMEM or glucocorticoid stimulation. Data shows mean±SD, n=3 independent experiments.
Shown above is the engineering process through which we built, tested, and optimized a synthetic glucocorticoid-responsive transcriptional factor from the scratch. Though in-depth characterization of this transcriptional factor is still on-going, we have already demonstrated its effectiveness and sensitivity. The further application of this transcriptional factor on stress-induced tattoo can be found in our Results and Proof of Concept pages.