Stem Loop Mediated Transgene Modulation in Human T Cells

Abstract

Controlling gene expression is useful for many applications, but current methods often require external user inputs, such as the addition of a drug. We present an alternative approach using cell-autonomous triggers based on RNA stem loop structures in the 3' untranslated regions (UTRs) of mRNA. These stem loops are targeted by the RNA binding proteins Regnase-1 and Roquin-1, allowing us to program stimulation-induced transgene regulation in primary human T cells. By incorporating engineered stem loops into the 3' UTRs of transgenes, we achieved transgene repression through Regnase-1 and Roquin-1 activity, dynamic upregulation upon stimulation, and orthogonal tunability. To demonstrate the utility of this system, we employed it to modulate payloads in CAR-T cells. Our findings highlight the potential of leveraging endogenous regulatory machinery in T cells for transgene regulation and suggest RNA structure as a valuable layer for regulatory modulation.

Figure 1

Stem loop designs and their regulatory activity on transgene expression in primary human T cells. (A) Conceptual schematic demonstrating stem loop mediated transgene expression modulation triggered by T cell stimulation in endogenous and engineered T cell contexts. Top, in resting T cells, endogenous regulatory RBPs, specifically Regnase-1 and Roquin-1, regulate the expression of transcripts encoding inflammatory factors by acting on stem loops located at the 3′UTR. Upon stimulation and activation of T cells, Regnase-1 and Roquin-1 are degraded, increasing expression of inflammatory factors. Bottom, this system can be harnessed to regulate transgenes of interest by inserting regulated stem loops in the 3′UTR of transgenes, resulting in engineered regulation. (B) Top, single loop designs based on the endogenous TNFα stem loop. Bottom, double loop designs based on the TNFα stem loop joined to the endogenous ICOS stem loop. (C) Top, characterization of CAR-positive cells 24 h after electroporation of CAR mRNAs bearing different stem loops in CD8+ T cells. Each data point in each experimental group represents an independent donor (n = 4). Bottom, CAR expression kinetics over time after mRNA electroporation. Data shown is pooled from two independent donors (n = 2). Bars represent SD (D) Representative flow plots showing CAR expression 24 h after mRNA electroporation in CD8+ T cells from donor ND580.

Figure 2

The impact of the WPRE on stem loop regulation and dependence on Regnase-1 and Roquin-1 activity. (A) Conceptual schematic depicting the dual-promoter test system and rationale of use. Top, reporter expression is constant while transgene expression is modulated through degradation (dotted). Bottom, transgene expression can be normalized to reporter expression. P = promoter, SL = stem loop. (B) Specific constructs tested with variable stem loops to determine the effect of the WPRE. (C) Comparison of normalized CAR expression (NE) with and without the WPRE using different stem loops in rested CD8+ T cells. Representative flow plots showing CAR and GFP expression from ND451 are shown on the right. 2-way ANOVA followed by Tukey’s multiple comparisons test was used for statistical analysis with significant P values shown for comparisons between loop and dead loop counterparts. (D) Characterization of regulation ratios for different stem loops with and without the WPRE in CD8+ T cells. Bars represent SD. Multiple unpaired t tests were used for statistical analysis. For (C–D), individual data points in each experimental group represent independent donors (n = 2 or 3). (E) Comparison of normalized CAR/GFP between mock unedited and Reg1 and Roq1 DKO CD8+ T cells. Representative flow plots showing CAR and GFP expression from ND578 are shown on the right. Individual data points in each experimental group represent independent donors (n = 4).

Figure 3

Regulation by stem loops enables stimulation-induced upregulation of transgene expression. (A) Conceptual schematic depicting regulation of Regnase-1 and Roquin-1 activity and downstream stem loop regulated transgene expression through TCR stimulation and potential regulation from stimulation through engineered receptors, such as a CAR. (B) Conceptual schematic of the stimulation scheme used, and time points used to measure CAR expression. (C) CAR019 MFI fold change dynamics over time after CD3/CD28 (left) and antigen (right) stimulation in CD8+ T cells. Data shown is from one representative donor of the experiment performed in technical triplicate in 3 independent donors with bars representing SD (D) Representative flow plots of CAR019 expression following no (left), CD3/CD28 (middle), and antigen (right) stimulation over time in ND578. Dotted traces at the 24 h time point indicate expression levels without stimulation.

Figure 4

Functional tuning and application of stem loop modulation to other transgenes in human CAR-T cells. (A) Constructs tested to determine whether expression can be tuned by varying promoter strength without affecting stem loop function. (B) CAR019 expression levels under different promoters in CD8+ T cells. Representative flow plots of CAR019 and GFP expression from ND580 are shown on the right. Individual data points in each experimental group represent independent donors (n = 4). Bars represent SD (C) Tumor killing after 18 h after coculture with CAR-T cells with different promoters. Data shown is from ND602 and is representative of two independent experiments using two independent donors (n = 2). Bars represent SD (D) Constructs tested to evaluate stem loop modulation of IL18 expression with constitutive CAR019 expression. (E) Constructs tested to evaluate stem loop modulation of Bcl-xL expression with constitutive CAR019 expression using tLNGFR as a control transgene. (F) IL18 production from CAR-T cells without stimulation, with anti-CD3/CD28 stimulation, and with antigen stimulation. Data shown is from ND616 and is representative of three independent experiments in three independent donors (n = 3). Bars represent SD. Multiple unpaired t tests were used for statistical analysis. (G) Bcl-xL expression in CD8+ CAR-T cells without stimulation, with anti-CD3/CD28 stimulation, and with antigen stimulation. Representative flow plots of Bcl-xL expression under different stimulation conditions from ND578 are shown on the right. Individual data points in each experimental group represent independent donors (n = 3). Bars represent SD 2-way ANOVA followed by Tukey’s multiple comparisons test was used for statistical analysis.