Pharmaceutically controlled designer circuit for the treatment of the metabolic syndrome

Abstract

Synthetic biology has significantly advanced the design of genetic devices that can reprogram cellular activities and provide novel treatment strategies for future gene- and cell-based therapies. However, many metabolic disorders are functionally linked while developing distinct diseases that are difficult to treat using a classic one-drug-one-disease intervention scheme. For example, hypertension, hyperglycemia, obesity, and dyslipidemia are interdependent pathologies that are collectively known as the metabolic syndrome, the prime epidemic of the 21st century. We have designed a unique therapeutic strategy in which the clinically licensed antihypertensive drug guanabenz (Wytensin) activates a synthetic signal cascade that stimulates the secretion of metabolically active peptides GLP-1 and leptin. Therefore, the signal transduction of a chimeric trace-amine–associated receptor 1 (cTAAR1) was functionally rewired via cAMP and cAMP-dependent phosphokinase A (PKA)-mediated activation of the cAMP-response element binding protein (CREB1) to transcription of synthetic promoters containing CREB1-specific cAMP response elements. Based on this designer signaling cascade, it was possible to use guanabenz to dose-dependently control expression of GLP-1-Fc_mIgG-Leptin, a bifunctional therapeutic peptide hormone that combines the glucagon-like peptide 1 (GLP-1) and leptin via an IgG-Fc linker. In mice developing symptoms of the metabolic syndrome, this three-in-one treatment strategy was able to simultaneously attenuate hypertension and hyperglycemia as well as obesity and dyslipidemia. Using a clinically licensed drug to coordinate expression of therapeutic transgenes combines drug- and gene-based therapies for coordinated treatment of functionally related metabolic disorders.

The metabolic syndrome is a combination of disorders and risk factors including hypertension, hyperglycemia, obesity, and dyslipidemia that show an extremely complex and little-understood interdependent pathophysiology and collectively increase the risk for cardiovascular diseases that remain the prime cause of mortality worldwide (1–8). Currently available therapies independently target each of the individual disorders and risk factors, but a well-coordinated collective treatment strategy does not exist (2). We have therefore designed a synthetic signaling cascade in which a clinically licensed small-molecule drug fine-tunes expression of therapeutic transgenes. The combination of drug- and gene-based therapies enables simultaneous treatment of all key metabolic syndrome pathologies. Guanabenz (Wytensin) is a clinically licensed antihypertensive drug that is thought to activate alpha-2-selective adrenergic receptors in the central nervous system, thereby reducing the sympathetic outflow to heart, kidney, and peripheral vasculature and decreasing the systolic and diastolic blood pressure (9). Recently, guanabenz was also identified as an agonist of the trace amine-associated receptor 1 (TAAR1) (10). TAARs are G protein-coupled receptors located in the neural presynaptic membrane and some lymphocytes that respond to low-abundance endogenous amines and have recently come into the limelight as potential drug targets for neuropsychiatric diseases (11, 12).

In this study we take advantage of guanabenz’s specific off-target effect to link antihypertensive treatment to expression of peptide hormones that restore glucose and lipid homeostasis in mice developing symptoms of the metabolic syndrome. Synthetic biology-inspired treatment strategies that simultaneously target several functionally interconnected metabolic disorders may improve future gene- and cell-based therapies.

Results

Design and Characterization of a Guanabenz-Triggered Synthetic Signaling Cascade.

Chimeric TAAR1, a combination of human and mouse receptor fragments, was particularly sensitive to guanabenz and produced a stronger cAMP second messenger response compared with the native counterparts (10). When rewiring the intracellular cAMP surge via cAMP-dependent phosphokinase A (PKA)-mediated activation of the cAMP-response element binding protein (CREB1) (13) to CREB1-specific synthetic promoters containing cAMP-responsive elements (CRE), guanabenz’s antihypertensive activity could be coupled to expression of a desired transgene (Fig. 1). Initial experiments using different cTAAR1 expression platforms (pKZY38, pUC57::cTAAR1 and pKZY73, P_SV40-cTAAR1-pA_SV40) showed that the chimeric receptor was efficiently transcribed (Fig. S1A) and produced (Fig. S1B) and triggered a guanabenz-dependent cAMP surge (Fig. S2) in mammalian cells. Because pKZY38 provided significantly lower basal expression and superior guanabenz dose-dependent transgene expression tunability when cotransfected with the reporter construct pCK53 (P_CRE-SEAP-pA_SV40; SEAP, human placental secreted alkaline phosphatase), we used it as the preferred cTAAR1 expression vector in all follow-up experiments (Fig. S3). Guanabenz-inducible transgene expression was functional in different pKZY38-/pCRE-Luc- (P_CRE-Luc-pA_SV40) cotransfected mammalian cell lines, suggesting that the synthetic signaling cascade is broadly applicable (Fig. S4). However, possible differences in availability and compatibility of the endogenous signal-transduction components with cTAAR1-mediated input resulted in a wide range of induction factors (Fig. S4). The best guanabenz-triggered expression performance was achieved in Hana3A, a HEK-293-derived cell line engineered for optimal receptor expression (14). Although basal expression was insignificant, maximum SEAP production was superior to isogenic vectors serving as a constitutive expression reference (pSEAP2-Control) (Fig. S5). This will represent a unique asset for biopharmaceutical manufacturing of difficult-to-produce protein pharmaceuticals and when gene-based therapies require high therapeutic effector levels (Fig. S5). The synthetic signaling device was fully reversible; product gene expression levels could be reliably switched between induced and repressed levels when alternating the presence and absence of the antihypertensive drug in the culture medium (Fig. S6).

Fig. 1.

Guanabenz-inducible designer cascade. Guanabenz binds to the chimeric trace amine-associated receptor (cTAAR1) and activates, via a specific G protein (Gα_s), the membrane-bound adenylyl cyclase that converts ATP into cAMP. When rewiring the resulting intracellular cAMP surge to cAMP-mediated activation of the cAMP-dependent phosphokinase A (protein kinase A, PKA) its catalytic subunits translocate into the nucleus, where they phosphorylate and actuate the cAMP-response element-binding protein 1 (CREB1). Activated CREB1 binds to synthetic promoters (P_CRE) containing cAMP-response elements (CRE) and induces P_CRE-driven transgenes in a guanabenz-adjustable manner.

Guanabenz-Inducible Transgene Expression in Mammalian Cells.

To assess adjustability of guanabenz-triggered transgene expression pKZY38-/pCK53-engineered Hana3A cells were exposed to increasing drug doses and SEAP levels were profiled every 24 h for up to 3 d (Fig. 2A). SEAP levels precisely correlated with increasing guanabenz doses up to 20 µM and exhibited reliable induction profiles over the entire concentration range (Fig. 2A). When exposing the engineered cells for various time periods to pharmacologic doses of 20-µM guanabenz, the SEAP production kinetics could be accurately programmed (Fig. 2B). Similar guanabenz-triggered expression kinetics could be visualized by fluorescence microscopy when using pCK91 (P_CRE-EYFP-pA_SV40) instead of pCK53 (Fig. 2C).

Fig. 2.

Guanabenz-inducible transgene expression in mammalian cells. (A) SEAP expression kinetics of HEK-293–derived Hana3A cells cotransfected with pKZY38 (pUC57::cTAAR₁) and pCK53 (P_CRE-SEAP-pA_SV40) and cultivated for 24, 48, and 72 h in the presence or absence of different concentrations of the antihypertensive drug guanabenz (Wytensin). Data are means ± SD; n = 3 independent experiments. (B) SEAP expression profiles of pKZY38-/pCK53-cotransfected Hana3A cells cultivated for different periods of time in the presence of 20 µM guanabenz. Data are means ± SD; n = 3 independent experiments. (C) EYFP-specific fluorescence micrographs of pKZY38-/pCK91- (P_CRE-EYFP-pA_SV40) cotransfected Hana3A cells cultivated for 48 h in the presence or absence of 20 µM guanabenz.

Guanabenz-Regulated Transgene Expression in Wild-Type Mice.

To validate guanabenz-adjustable transgene expression in vivo, 5 × 10⁶ microencapsulated pKZY38-/pCK53-transgenic Hana3A cells were implanted into mice that received different doses of the antihypertensive drug by injection (Fig. 3 A and B) or oral administration (Fig. 3 C and D). The corresponding serum SEAP levels of treated mice were profiled after 48 h (Fig. 3 A and C) and 72 h (Fig. 3 B and D), which confirmed dose-dependent transgene expression in animals at clinically relevant guanabenz concentrations that decreased the blood pressure in treated animals (Fig. S7).

Fig. 3.

Guanabenz-adjustable transgene expression in wild-type mice. Animals were intraperitoneally implanted with 5 × 10⁶ pKZY38-/pCK53-transgenic Hana3A cells and received daily injections (A and B) or oral doses (C and D) of guanabenz. SEAP serum levels were profiled 48 (A and C) and 72 h (B and D) after treatment of the animals. Data are means ± SEM; statistics by two-tailed t test; n = 8 mice. ***P < 0.001 vs. control.

Validation of Guanabenz-Induced GLP-1-Fc_mIgG-Leptin Expression in Wild-Type Mice.

To design a drug-based gene therapy scenario for the treatment of the metabolic syndrome, we linked antihypertensive guanabenz input to the bifunctional fusion protein GLP-1-Fc_mIgG-Leptin, which combines the anorexic and insulin secretion-stimulating effect of the glucagon-like peptide 1 (GLP-1) (15, 16) with the lipid level-, food intake-, and body weight-controlling capacity of leptin (17, 18), whereas the IgG-derived Fc linker peptide manages efficient expression and stability of the bifunctional peptide hormone. After control experiments confirming guanabenz-triggered expression of GLP-1-Fc_mIgG-Fc-Leptin in vitro (Fig. S8), we implanted 5 × 10⁶ microencapsulated pKZY38-/pHY69- (P_CRE-GLP-1-Fc_mIgG-Leptin-pA_SV40) transgenic Hana3A cells intraperitoneally into wild-type mice that received daily oral guanabenz doses for up to 3 d. Control animals either received guanabenz but no implants to exclude unexpected side effects of the antihypertensive drug or implants but no guanabenz to eliminate pleiotropic induction of the designer circuits in microencapsulated cell implants. After oral administration of guanabenz to animals containing transgenic cell implants their blood GLP-1 and leptin levels raised significantly (Fig. 4 A and B). Within 24 h increased GLP-1 levels increased insulin secretion (Fig. 4C) and attenuated glycemic excursions in response to i.p. glucose tolerance tests (Fig. 4D). Plasma cholesterol and free fatty acid concentrations decreased after 3 d (Fig. 4 E and F), most likely due to the anorexic effect of GLP-1 and leptin, as illustrated by reduced food intake and consequently lower body weight (Fig. 4 G and H).

Fig. 4.

Guanabenz-inducible GLP-1-Fc_mIgG-Leptin expression in wild-type mice. Animals were intraperitoneally implanted with 5 × 10⁶ pKZY38-/pHY69-transgenic Hana3A cells and received daily oral doses of 1 mg/kg guanabenz for 72 h. Control mice received either no guanabenz and/or no implants. (A) Active GLP-1, (B) leptin, and (C) blood insulin levels were profiled in the blood of treated animals after 72 h. (D) Glucose tolerance test performed 24 h after implantation. (E) Total blood cholesterol and (F) free fatty acid levels as well as (G) food intake and (H) body weight were assessed 72 h after implantation. Data are means ± SEM; statistics by two-tailed t test; n = 8 mice. *P < 0.05, **P < 0.01, ***P < 0.001 vs. control.

Simultaneous Attenuation of Hyperglycemia, Obesity, and Dyslipidemia in ob/ob Mice.

After confirming that oral administration of guanabenz reduces the blood pressure (Fig. S7) of the animals we implanted 1 × 10⁷ microencapsulated pKZY38-/pHY69-transgenic Hana3A cells into ob/ob mice, which suffer from the key metabolic syndrome symptoms. Treated mice received daily oral doses of guanabenz for up to 3 d, whereas control animals received no antihypertensive drug or no implants. Three days after the start of treatment the serum levels of GLP-1 and leptin rose significantly (Fig. 5 A and B). Whereas GLP-1 boosted blood insulin levels (Fig. 5C) and improved glucose homeostasis (Fig. 5D), leptin decreased food intake and body weight with concomitant improvement of dyslipidemia (lower cholesterol and free fatty acid) (Fig. 5 E–H).

Fig. 5.

Guanabenz-inducible GLP-1-Fc_mIgG-Leptin in (ob/ob) mice suffering from several key symptoms of the metabolic syndrome. Diseased animals were intraperitoneally implanted with 1 × 10⁷ pKZY38-/pHY69-transgenic Hana3A cells and received daily oral doses of 1 mg/kg guanabenz for 72 h. Control mice received either no guanabenz and/or no implants. (A) Active GLP-1, (B) leptin, and (C) insulin levels were profiled in the blood of treated animals after 72 h. (D) Glucose tolerance test performed 24 h after implantation. (E) Total blood cholesterol and (F) free fatty acid levels as well as (G) food intake and (H) body weight were assessed 72 h after implantation. Data are means ± SEM; statistics by two-tailed t test; n = 6 mice. *P < 0.05, **P < 0.01, ***P < 0.001 vs. control.

Discussion

Classic treatment concepts rely on exclusive drug- or gene-based interventions to address a single disease phenotype by targeting or complementing a particular pathologic compound. In recent years synthetic biology has greatly advanced the design of biomedical devices (19–22) that provide new diagnostic tools (23); offer novel concepts for the treatment of cancer (24, 25), immune diseases (26, 27), and diabetes (28); and pioneered prosthetic networks as a therapy for hyperuricemic disorders such as the tumor lysis syndrome and gout (29). Still, all synthetic biology-based treatment strategies target a single disorder at a time, taking advantage of a specific therapeutic nucleic acid or protein compound.

We have combined classic and synthetic biology-based treatment concepts by expanding the function of small-molecule drug compounds. While still operating systemically to control its original disease target the drug will also address its cognate receptor in engineered cell implants to trigger a dose-dependent therapeutic transgene expression response via a synthetic signaling cascade. The combination of drug- and gene-based therapies and the direct correlation of therapeutic drug dose and therapeutic transgene expression enables coordinated treatment schemes that tackle separate but functionally linked pathologies and so could eventually improve clinical success. Also, because the designer cascade converts the presence of a small-molecule compound into a sustained transcription response, the treatment scope could be further increased to multigene-based intervention strategies or, as shown here, by expression of multifunctional therapeutic fusion proteins. Taking advantage of the guanabenz-triggered expression of GLP-1-Fc_mIgG-Leptin the entire portfolio of disorders and risk factors of the metabolic syndrome—hypertension, hyperglycemia, obesity, and dyslipidemia—could be tackled simultaneously and successfully attenuated in relevant animal models of the corresponding human disease. Whereas GLP-1 and its analogs were approved for the treatment of type 2 diabetes by the Food and Drug Administration (FDA) in 2005 and are widely prescribed (30), therapeutic use of leptin in humans has only been successful in congenital leptin deficiency (31). Obese patients without this genetic disorder typically show increased leptin levels, and in these individuals leptin treatment does not result in substantial weight reduction owing to a phenomenon known as leptin resistance (32). Therefore, future clinical applications of this designer network may need to have leptin replaced by other anorectic peptides (33). Having the potential to address such a wide set of pathologies, the concept of drug-based gene therapies may improve treatment success and provide new therapies for multifactorial diseases.

Materials and Methods

Vector Design.

pcDNA3.1 (P_hCMV-MCS-pA_SV40; P_hCMV, human cytomegalovirus immediate early promoter; MCS, multiple cloning site; pA_SV40, simian virus 40-derived polyadenylation site; Invitrogen), pSBC-2 (P_SV40-MCS-pA_SV40; P_SV40, simian virus 40 early promoter; Addgene), pSEAP2-Control (P_SV40-SEAP-pA_SV40; SEAP, human placental secreted alkaline phosphatase; Clontech), pCRE-Luc [P_CRE-Luc-pA_SV40; Luc, firefly luciferase; P_CRE, promoter containing the cAMP-response element (CRE) activated by the CREB1; Clontech], pCK53 (P_CRE-SEAP-pA_SV40) (34), and pCK91 (P_CRE-EYFP-pA_SV40; EYFP, enhanced yellow fluorescent protein) (34) have been described previously. pKZY38 (pUC57::cTAAR1; cTAAR1, chimeric trace amine-associated receptor 1) (10) was synthesized (GenScript). pKZY73 (P_SV40-cTAAR1-pA_SV40) was cloned by excising cTAAR1 from pKZY38 (EcoR1/HindIII) and inserting it into the corresponding sites (EcoR1/HindIII) of pSBC-2. pHY63 [(pUC57::GLP-1-Fc_mIgG-Leptin) containing the fusion protein GLP-1-Fc_mIgG-Leptin (28); Fc_mIgG, mouse IgG-derived Fragment crystallizable region (28, 35); and leptin (GenBank accession number AAA64213)] was synthesized (GenScript). pHY69 (P_CRE-GLP-1-Fc_mIgG-Leptin-pA_SV40) was cloned by excising GLP-1-Fc_mIgG-Leptin (HindIII/XbaI) and inserting it into the corresponding sites (HindIII/XbaI) of pCRE-Luc.

Cell Culture and Transfections.

Human embryonic kidney cells (HEK-293; ATCC), human cervical adenocarcinoma cells (HeLa; ATCC), and HEK-293–derived Hana3A cells engineered for constitutive expression of RTP1, RTP2, REEP1, and G_αολϕ (14) were cultured in DMEM (Invitrogen) supplemented with 10% (vol/vol) FCS (PAN Biotech GmbH) and 1% (vol/vol) penicillin/streptomycin solution (Sigma-Aldrich). Chinese hamster ovary cells (CHO-K1; ATCC) were cultivated in ChoMaster HTS (Cell Culture Technologies) supplemented with 5% (vol/vol) FCS and 1% (vol/vol) penicillin/streptomycin solution. All cell types were cultivated at 37 °C in a humidified atmosphere containing 5% CO₂.

HEK-293 and Hana3A cells were (co)transfected using an optimized CaHPO₄-based protocol. In brief, 2 × 10⁵ HEK-293 or Hana3A cells seeded per well of a 12-well plate were (co)transfected with a total of 1.2 µg of DNA (for cotransfections, a 5:1 receptor/reporter plasmid ratio was used) diluted in 50 µL of 0.5 M CaCl₂ solution and subsequently mixed with 50 µL of 2× BES buffer (100 mM N,N-bis [2-hydroxyethyl]-2-aminoethanesulfonic acid, 280 mM NaCl, and 1.5 mM Na₂HPO₄, pH 6.95). The DNA-containing solution was added dropwise to the cells and the medium was replaced after 6 h. HeLa cells were transfected the same way except that the DNA-containing solution was incubated overnight. For transfection of CHO-K1 cells the DNA-containing solution was added to the well and centrifuged onto the cells (5 min at 1,200 × g) to increase transfection efficiency. After incubation for 4 h, the cells were treated with 0.5 mL of glycerol solution (ChoMaster HTS medium containing 15% glycerol) for 60 s. After washing once with PBS (Invitrogen) the cells were cultivated in 1 mL of ChoMaster HTS.

Reporter Gene Assays.

Production of the SEAP was quantified in cell culture supernatants (36) and mouse serum (37) as described previously. Luciferase was measured using the Tropix luciferase assay kit according to the manufacturer’s protocol (Applied Biosystems). EYFP expression was visualized using a LEICA DMI-600 microscope (Leica Microsystems) equipped with a DFC350FX R2 digital camera (Leica), a 10× objective, a 488 nm/509 nm (B/G/R) excitation and emission filter set, and Leica Application Suite software installed (version V2.1.0R1).

RNA Isolation and RT-PCR.

Total RNA was extracted from mammalian cells using the RNeasy kit (Qiagen) and RT-PCR was performed using the SuperScript II Reverse Transcriptase kit (Invitrogen) according to the manufacturer’s protocols. OHY128: 5′-CGGAACAGCGACTGGTCTAGGGAGG-3′ and OHY129: 5′-ATGGAAGATGGATGCGGAGCTCAGC-3′ were used to detect cTAAR1 transcripts in Hana3A cells. Actin mRNA (OMT12, 5′-CCAGTTCGCCATGGATGACG-3′ and OMT13, 5′-GCAGCTCAGTAACAGTCCGC-3′) was used as internal control.

Western Blot Analysis.

For immunohistochemical detection of HA-tagged cTAAR1 expression, 5 × 10⁶ Hana3A were collected 48 h after transfection of pKZY38 or pKZY73 and protein extracts were prepared as described before (34, 38). Sixty micrograms of protein were resolved on a 12% SDS polyacrylamide gel and electroblotted onto a polyvinylidene fluoride membrane (Millipore) on which HA-tagged cTAAR1 was visualized using a primary rabbit polyclonal anti-HA-tag antibody (Santa Cruz Biotechnology Inc.) and a secondary horseradish-peroxidase–coupled anti-rabbit IgG antibody (AbD Serotec). ECL-Plus Western blot detection reagents (Amersham) were used for the chemiluminescence-based signal detection performed with a Chemilux CCD camera (ImageQuant LAS 400 mini; GE Healthcare). Actin was used as a loading control (primary rabbit polyclonal anti-actin IgG; Sigma).

cAMP Assay.

Intracellular cAMP levels were determined using an AlphaScreen cAMP kit (PerkinElmer) according to the manufacturer’s protocol. In brief, per well of an OptiPlate-384 plate (PerkinElmer) 1 × 10⁴ cells were diluted in 5 µL of anti-cAMP acceptor beads and stimulated for 30 min by addition of different guanabenz concentrations. Then, 15 µL of biotinylated-cAMP/streptavidin donor beads were added and the sample was incubated for another hour at room temperature. Plates were then read with an Envision Plate Reader (2104 Multilabel Reader; PerkinElmer) using Wallac Envision Manager software (version: 1.12; PerkinElmer).

Animal Experiments.

Intraperitoneal implants were produced by encapsulating transgenic Hana3A cells into coherent alginate-poly-(L-lysine)-alginate beads (400 µm; 200 cells/capsule) using an Inotech Encapsulator Research Unit IE-50R (EncapBioSystems Inc.) set to the following parameters: 200-µm nozzle with a vibration frequency of 1,025 Hz, 25-mL syringe operated at a flow rate of 410 units, and 1.12-kV voltage for bead dispersion (28). Twelve-week-old female OF1 mice (oncins France souche 1; Charles River Laboratory) were intraperitoneally injected with 1 mL of DMEM containing 5 × 10⁶ encapsulated pKZY38-/pCK53-transgenic Hana3A cells. One hour after implantation, 200 µL of guanabenz (TOCRIS Bioscience) was administered orally or by injection. Control groups received 200 µL of PBS instead of guanabenz. The guanabenz dose varied from 0 to 30 mg/kg and was applied once per day for up to 3 d. Blood samples were collected 48 and 72 h after implantation and SEAP levels were quantified in the serum, which was isolated using microtainer SST tubes according to the manufacturer’s instructions (Beckton Dickinson). To assess the therapeutic potential of guanabenz-triggered GLP-1-Fc_mIgG-Leptin expression in animal models (B6.V-Lep^ob/J, ob/ob, and B6.129P2-Nos3^tm1Unc/J; Jackson Laboratory), 8-wk-old male mice were intraperitoneally implanted with 1 × 10⁷ microencapsulated pKZY38-/pHY69-transgenic Hana3A cells and received a daily oral guanabenz dose of 1 mg/kg. The systolic blood pressure of treated conscious mice was measured every day using tail-cuff plethysmography (NIBP LE-5002; Harvard Apparatus); mice were trained 2 wk before the experiment to accept the procedure. The glucose tolerance test of treated mice was initiated 4 h after implantation by fasting the animals for 16 h before they received an i.p. glucose injection (1 g/kg). Plasma glucose levels were monitored in tail-vein blood samples 0, 15, 30, 60, 90, and 120 min after glucose administration using a glucometer (Contour; Bayer HealthCare). After 3 d the treated animals were fasted for 4 h (unless otherwise indicated) and their body weight was determined before they were killed to profile the blood levels of free fatty acids (NEFA-HR[2] kits; Wako Chemical GmbH), cholesterol (LabAssay Cholesterol kit; Wako Chemical GmbH), insulin (mouse insulin ELISA kit; Mercodia), GLP-1 (active GLP-1 ELISA kit; Millipore), and leptin (leptin ELISA kit; Millipore). All experiments involving animals were performed according to the directive of the European Community Council (86/609/EEC), approved by the French Republic (69266310) and Institut Universitaire de Technologie (IUT) Département de Génie Biologique, and carried out by G.C.-E.H. at IUT.

Statistical Analyses.

Data represent means ± SD of three independent experiments. Comparisons among groups were made using analysis of Student t test and expressed as means ± SEM. Differences were considered statistically significant at P < 0.05.