MluI (New England Biolabs; R0198S)

XhoI (New England Biolabs; R0146S)

BamHI (New England Biolabs; R0136S)

EcoRI (New England Biolabs; R0101S)

pENTR-1A® (Invitrogen; A10462)

cPPT forward sequencing primer - (GTGCAGGGGAAAGAATAGTAG)(Sigma-Aldrich)

Gateway™ LR Clonase™ II enzyme mix (Invitrogen; 11791020)

DH5α chemically competent cells (Invitrogen; 18258012)

One Shot™ Stbl3™ competent cells (Invitrogen; C737303)

ampicillin antibiotic (Sigma-Aldrich; A9393)

LB agar (Sigma-Aldrich; L3027)

LB broth powder (Sigma-Aldrich; L3147)

agarose gel powder (Sigma-Aldrich; A4718)

HEK293T cell line (VWR; MSPP-CRL3216)

Dulbecco’s Modified Eagle’s Media (DMEM) (Sigma-Aldrich; D6429)

Fetal calf serum (FCS)(Invitrogen; 10082147)

Penicillin/Streptomycin (Pen/Strep) (Sigma-Aldrich; P4333)

Phosphate buffered saline (PBS) (Sigma-Aldrich; D8662)

Trypsin EDTA dissociation media (Sigma-Aldrich; T4299)

Opti-MEM I reduced serum medium (Invitrogen; 31985070)

pMD2.G (VSV-G envelope) (Addgene; #12259)

pCMVd8.74 (gag-pol, tat, rev) (Addgene; #22036)

PsPAX2 (Addgene #12260)

RETRO-TEK HIV-1 p24 antigen ELISA (Zeptometrix; 0801200)

KAPA SYBR FAST qPCR kit (Roche; KK4600)

branched PEI (MW ~25,000) (lentivirus production) (Sigma Aldrich; #408727)

0.45 µM PVDF filter (Sigma-Aldrich; P1938)

pHGT1 Ad5 helper plasmid (MTA Harvard Medical School, USA)

Benzonase (Sigma-Aldrich; E8263)

AAV pro 293T cells (Takara; 632273)

Sodium chloride (Sigma-Aldrich; S9888)

Polyethyleneimine (PEI) (Polysciences; 24765)

0.45 µm filter membrane (Millipore; SCHVU05RE)

Cell lifter (Corning; 3008)

0.45 µm syringe filters (Sartorius; 17598)

0.22 µm centrifuge tube filter (Costar; 8160)

10 ml syringe (Terumo; SS+10ES1)

50 ml syringe (BD Plastipak; 300865)

Disposable needles (BD Microlance; 301155)

Benzonase nuclease (Sigma; E1014-25KU)

Amicon Ultra-15 centrifugal filter units (Millipore; UFC910024)

Slide-A-lyzer dialysis cassette 10,000 MWCO (Thermo Scientific; 66810)

5 ml FACS tubes (Falcon; 352053)

Sodium deoxycholate (Sigma; D6750-100G)

Glycine (Sigma; G8898-500G)

POROS™ CaptureSelect™ AAV Resin (Thermo Scientific; A36739)

ÄKTAprime plus (High performance liquid chromatography - HPLC system (GE Healthcare; 11001313)

DPBS (1X) (Gibco; 14190-094)

PBS tablets (Sigma; P4417-100TAB)

Luna Universal Probe qPCR MasterMix (NEB; M3004S).

96-well PCR plate 0.1 mL format.

MicroAmp® Optical Adhesive Film (Applied Biosystems).

HeLa cell line (Sigma; 93021013)

Huh7 cell line (JCRB Cell Bank; JCRB0403)

HepG2 cell line (Sigma; 85011430)

RPMI-1640 media (Sigma-Aldrich; R8758)

ultra-pure LPS-EB (InvivoGen; tlrl-3pelps)

Activin A (PeproTech; AF-120-14E)

BMP2a (PeproTech; AF-120-02)

Cobalt Chloride (Sigma-Aldrich; 60818)

Lithium Chloride (LiCl ₂) (Sigma-Aldrich; L9650)

Estradiol (Sigma-Aldrich; E2758)

Nutlin-3a (Sigma-Aldrich; SML0580)

Phorbol 12-myristate 13-acetate (PMA) (Sigma-Aldrich; P8139)

LY294002 (StemCell Technologies; #72152)

human recombinant LIF (StemCell Technologies; #78055)

Purmorphamine (Sigma-Aldrich; SML0868)

Hydrogen peroxide (H ₂O ₂) (Sigma-Aldrich; 216763)

2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) (Sigma-Aldrich; NIST1614)

Ionomycin (Sigma-Aldrich; I0634)

Luciferase assay lysis buffer (0.65% NP40 (Sigma-Aldrich; NP40), 10 mM Trizma® base (Sigma-Aldrich; T1699) (pH 8.0), 1 mM Ethylenediaminetetraacetic acid (EDTA) (Sigma-Aldrich; E9884)(pH 8.0), 150 mM NaCl(Sigma-Aldrich; S9888))

luciferase assay buffer (25 mM Trizma® base (Sigma-Aldrich; T1699) pH 7.8, 1 mM 1,4-Dithiothreitol (DTT) (Sigma-Aldrich; 10197777001), 1 mM EDTA (Sigma-Aldrich; E9884), 1% Triton™ X-100 (Sigma-Aldrich;), 8 mM MgCl ₂(Sigma-Aldrich; M8266), 3 ml glycerol (Sigma-Aldrich; G5516), 1.25 mM rATP (Promega; E6011), 0.5% BSA (Sigma-Aldrich; A2153))

Luminometer such as the GloMax (Promega; E9032)

Bio-rad protein assay (Bio-rad; #5000006)

CD1 mice (Charles River Laboratories)

4% isofluorane (Abbott Laboratories)

33-gauge Hamilton needle (Fisher Scientific, Loughborough)

bupivacaine hydrochloride (Advanz Pharma)

D-luciferin, potassium salt (Gold Biotechnologies; LUCK)

cooled charged-coupled device camera (IVIS machine) (Perkin Elmer)

Living Image Software (v4.1) (Perkin Elmer)

Lentivirus cloning. A simple two-step cloning process was developed to generate lentiviral TFAR vectors using Gateway® cloning. We first generated the parental lentiviral vector by cloning the Gateway® (GW) Destination recombination site into a blunted unique MluI restriction site of a 2 ^nd generation HIV-1 based lentiviral backbone upstream of a bicistronic 3xFLAG-FLuc-F2A-eGFP reporter gene cassette to produce pLNT-GW-Luc-eGFP. We then generated a shuttle vector containing the adenovirus E1A minimal promoter (MP) sequence cloned into a unique XhoI restriction site in the Gateway® shuttle vector, pENTR-1A. This shuttle plasmid (pENTR-MP) was used as a sub-cloning vector to introduce de novo synthesized regulatory sequences (Aldevron, North Dakota, USA) upstream of the MP (existing TFARs listed in Table 1) into BamHI/ EcoRI restriction sites. Vectors were confirmed by Sanger sequencing using a cPPT forward primer (GTGCAGGGGAAAGAATAGTAG) then recombined into pLNT-GW-Luc-eGFP using Gateway® cloning ( Figure 1A).

AAV cloning. The AAV8-Gateway®-Luc-T2A-eGFP was completely de novo synthesized (Aldevron). This AAV plasmid consisted of a Gateway® sequence, placed upstream of the Luc-2A-eGFP reporter cassette. AAV TFARs were generated by Gateway® cloning as with the lentiviral vectors ( Figure 1B).

Both lentivirus and AAV vectors were generated by transfection of human embryonic kidney, HEK293T, producer cells with accessory plasmids as previously detailed ² . Lentiviral titer was generated based on the p24 ELISA assay whereas AAV was titered by quantitation of capsid protein by viral genome copy numbers per ml ^{10, 11} .

All newly generated TFARs were comprehensively evaluated for predicted responses to known activators/inhibitors on multiple cell lines, including HEK293T, HeLa, Huh7, HepG2, and primary human fibroblasts, prior to being employed in vivo ² . The reporters, in vitro agonists, concentrations, and duration of activation are outlined in Table 1.

Specifically, for the HIF response element (HRE) construct that can be used to determine the induction of hypoxia, HEK293T and HeLa cells were transduced with LNT-HRE-Luc-eGFP at a multiplicity of infection (MOI) of 10. Three days post-transduction, transduced cells were replated in two sets of triplicates, one for HRE activation and one for basal levels of luciferase activity. The following day one triplicate set of cells were exposed to the agonist, CoCl ₂ (100 mM/µl), in RPMI-1640 for 12 hours while the non-activated cells were only placed in RPMI-1640. A time course of 12, 24, 48, and 72 hours post agonist activation was performed. At each time point, the cells were lysed in a standard luciferase assay lysis buffer and stored for future luciferase assay analysis.

All procedures were performed under United Kingdom Home Office Project License 70/8030, approved by the ethical review committee, and followed institutional guidelines at University College London. All methods were performed in accordance with the relevant guidelines and regulations. Male and female outbred wild type CD1 mice were used and were supplied by Charles Rivers Laboratories. Mice were used for adult interventions as close as possible to 6 weeks of age and 25g.

IVC cages were used to house the mice, with a maximum of 5 per cage. Male and female mice were housed separately post weaning. Food and water were routinely monitored by the animal technicians at the facility. The temperature of the room was set at 25°C with light/dark cycles alternating every 12 hrs. Mice were bred onsite and time-mated in order to plan experiments as accurately as possible and avoid wastage of unused pups. A total of 34 mice were used in the studies, described below.

Neonatal intracranial and intravenous injections. Mice either received intracranial or intravenous injections, never both. Skilled technicians can achieve more than 99% success rate for intravenous injections and no failed intracranial injections. No anaesthesia was used as the procedure is very quick and involves a single needle puncture per mouse. For intracranial injections, mice (postnatal day 1 (P1)) received unilateral injections of concentrated lentiviral or AAV vector (5 μl in total) into the cerebral lateral ventricles using a 33-gauge Hamilton needle. The site of injection was located two-fifths of the distance between the eye and the lambda intersection of the skull as described and illustrated by Kim et al. ¹² . During the procedure, mice were held in an upright position, with their heads firmly placed between the handler’s thumb and index finger. The 33-gauge Hamilton needle was held perpendicular to the mouse head and injected at a depth of approximately 3mm. The vector was slowly controlled and injected free handed into the lateral ventricles. At P1, a 33-gauge needle can easily penetrate the skull and there is no reflux of injectate as the ventricular system is sufficiently voluminous to accommodate the extra volume. Biodistribution of vector has been confirmed using this methodology ⁹ . For intravenous injections, P1 pups received intravenous injections of lentiviral or AAV vectors into the superficial temporal vein (20 μl) using a 33-gauge Hamilton needle. This procedure was performed free handed over approximately 5 seconds. Pups were returned as soon as possible to the dam and the nest undisturbed as much as possible. Animals were monitored daily for distress or discomfort such as the presence of a dull/ruffled coat, poor posture (hunched), pale or sunken eyes, a change in normal behaviour, reduced food or water intake, dehydration, any discomfort when handled, a reluctance to move, and any significant weight loss. In this cohort, we noted no significant signs of pain or discomfort following vector administration.

Validation and disease modelling induction of somatotransgenic bioimaging

Effects of carrier gas on hypoxia-inducible factor (HIF) induction

All mice received neonatal intracranial injection of lentivirus vector containing the HIF response element upstream of a luciferase-eGFP bicistronic cassette (LNT-HRE-Luc-eGFP) at P1. After weaning, age-matched mice were randomly allocated to two groups (N=11 for each group based on power calculations) for imaging at 0, 2, 4, 6 and 8 hours. Randomization was achieved by using Microsoft Excel to generate a random column of numbers against mouse numbers and then sorting the list of mice using the SORT function. The groups received isoflurane mixed with either 100% oxygen or air. The group receiving 100% oxygen exhibited consistently lower HIF activation than the group receiving air (21% oxygen); this was significantly different at 6 hours ( P=0.0465, repeated measures ANOVA where one factor was with/without 21% oxygen and the repeated measure was time). See data availability section.

This data suggests that the carrier gas, itself, may influence physiological and pathological processes and supports the concept that conscious bioimaging may be preferable, where possible ⁹ .

Partial bile duct ligation in mice

All mice received neonatal intravenous injection of lentivirus vector containing the truncated GFAP promoter driving a luciferase-eGFP bicistronic cassette (LNT-GFAP-Luc-eGFP) at P1.

Mice were anaesthetized with 4% isoflurane at between 40–60 days after birth, shaved and disinfected around the surgical area, and a midline laparotomy (≈ 15 mm) performed just caudal to the sternum using aseptic surgery techniques. The median and left lobes of the liver were exteriorized and kept moist with sterile gauze. The bile duct was ligated with 6.0 silk suture to occlude outflow from the left and median lobes but not occlude bile outflow from the right and caudate lobes as described by Yang and colleagues ¹³ . The liver was returned to the abdomen and the wound closed in two stages using 6.0 silk suture; skin was closed using a subcuticular suture with buried knots (N=7). Sham operated mice received laparotomy only (N=5). Mice were recovered in a warmed chamber and monitored to make sure they were not under any unnecessary distress. Clinical record sheets were used to detail observations and post-operative treatments that were given. Pre- and immediate post-operative doses of topical bupivacaine hydrochloride (2 mg/kg) and subcutaneous 50 µl morphine (6.67 mg/ml) was administered for systemic post-operative analgesia ¹⁴ . Following surgery, mice were single housed and provided with moist food to aid consumption of food and prevent significant weight loss. Monitoring sheets recorded age, weight, disease and surgical status. Mice were monitored for approximately 1 hour post-surgery, and subsequently checked and weighed to monitor for overt signs of discomfort and significant weight loss. A loss of weight of more than 15% required humane euthanasia of the animal. In this cohort, the animals were not found to have negative post-operative outcomes, which is a significant improvement over performing full bile duct ligations which often result in pain, jaundice, and death to the animals. No animals within our cohort required euthanasia using partial bile duct ligation.

Whole-body bioluminescence imaging. Mice which underwent partial bile-duct ligations were anesthetized with isoflurane containing 100% oxygen, all other cohorts of mice described herein were imaged while conscious and free-moving. Mice received an intraperitoneal injection of 15 mg/mL of D-luciferin. Mice were imaged after 5 minutes using a cooled charged-coupled device camera (IVIS imaging machine) for between 1 second and 5 minutes. The regions of interest (ROI) were measured using Living Image Software and expressed as photons per second per centimeter squared per steradian (photons/second/cm ²/sr). The imaging chambers used to image conscious mice for the intracranial injected mice had the following dimensions; 5 cm × 5 cm × 6 cm. The Perspex box which was used to image the conscious mice which had received an intravenous injection is described in detail in Karda et al. ⁹ . Where possible, BLI was carried out in adult reporter rodents on three consecutive days to establish a robust median baseline; subsequent data points were expressed as fold-change over this internal standard for each individual animal.

Euthanasia. At experiment conclusion, mice were culled using rapid cervical dislocation to induce rapid loss of consciousness. To reduce the possibility of operator error, the main concern for animal welfare using this technique, the procedure was performed by adequately trained personnel with a high degree of technical proficiency only. Where possible, cervical dislocation was performed while mice were still under non-recovery isoflurane anaesthetic. Death was assured by decapitation.

In consultation with a statistician, power calculations are performed based on previous data using the free GPower (v3.0.10) online software. Using the probability of finding an effect set at 80%, the probability of incorrectly rejecting the null hypothesis when it is true set at 0.05 (alpha), and using a two-sided test, a power calculation was performed in order to determine the sample size required for each group under investigation. Using our previous data involving the activation of the Smad2/3 binding element using the potent agonist, activin A, a standard deviation of 1.75 was used. The mean of group 1 (control) was set at 5.13, and the mean of group 2 was set at 7.8 (experimental). Based on these values, the effect size is expected to be large (1.52 fold change). An effect size greater than 0.8 is indicative of a large effect size, indicating that 93% of the control group will be below the mean of the experimental group. The experiment was sufficiently powered with a sample size of eight experimental and eight control animals to detect a change in the total amount of photonic output between two groups with a power of 80%. For in vivo data, repeated measurements following agonist administration were analyzed with repeated measures analysis of variance using GraphPad Prism v8.0.0 with a significance level of P <0.01. Area under curve data for experiments with two groups were analyzed by Student’s one-tailed t-test. For areas under curve derived from two or more groups, one-way analysis of variance (ANOVA) with Newman-Keuls post-hoc multiple comparison test was performed.

TFAR design, synthesis, and sub-cloning

Cloning synthetic promoter into the lentiviral cassette.

Cloning synthetic promoters into the AAV cassette.

Generation of high-titer TFAR lentivirus.

Generation of high-titer TFAR AAV.

In vitro validation of HRE reporter constructs.

Lentiviral delivery of vectors containing transcription factor binding elements upstream of a bicistronic reporter cassette offer a unique method with which to interrogate biological pathways. HRE is used as a biological sensor of hypoxia, important in cancer and ischemia models of disease. Here we present data validating our biosensing constructs in vitro ( Figure 1 ) prior to in vivo ( Figure 2 ¹⁵ ) validation and experimentation.

In vitro validation of the pLNT-HRE-Luc-eGFP construct. Prior to using the LNT-HRE-Luc-eGFP vector to interrogate signaling profiles in disease models i n vivo, it was important to first validate this construct and establish its ability to be induced by agonists in vitro. CoCl ₂ is a chemical inducer of the hypoxia-inducible factor that mimics hypoxia. In the presence of CoCl ₂, a highly significant increase in luciferase-mediated bioluminescence fold-change (Δ bioluminescence) over non-activated cells was observed in vitro ( Figure 2 ¹⁵ ) prior to in vivo ( Figure 3 ¹⁵ ) validation and experimentation.

Response to oxygen content of carrier gasses during anesthesia. Following validation of the hypoxia response element in vitro, we chose to use it to ask whether in vivo HIF activity is modulated by the anesthetic carrier gas. 100% oxygen is frequently used as a carrier gas to ensure sufficient tissue oxygenation during anesthesia however, theoretically, it may downregulate HIF activation as this response element responds inversely to hypoxia, or the lack of oxygen. Mice underwent anesthesia with isoflurane using either air or 100% oxygen as the carrier gas and HRE signaling monitored at a series of time points using bioluminescence imaging. A change in bioluminescence (Δ bioluminescence) compared to time point 0 was noted over time with oxygen expectedly decreasing the expression of the HIF response element ( Figure 3A-B ¹⁵ ). This indicates the importance of selecting the correct carrier gas for anaesthetic induction when interrogating HIF-responsive signaling using luciferase expression and imaging.

Somatotransgenic bioimaging to assess GFAP activation in a model of cholestatic liver injury. The partial BDL model is an established model of cholestatic liver disease ¹⁷ . Currently, analysis of liver disease consists of terminal endpoint analysis of groups of animals at serial time points, and/or serial bleedings for analysis of serological biomarkers of liver function, particularly liver enzymes and total serum bile acids. We have previously studied signal transduction pathways of inflammation and WNT-signaling in this model ¹⁴ . Here, we asked whether expression of GFAP, a marker of activated hepatic stellate cells that are a major cell type involved in liver fibrosis, was altered following induction of liver injury. P1 mice received LNT-GFAP-Luc-eGFP, a lentiviral vector containing the luciferase-eGFP transgene driven by the bioresponsive glial fibrillary acidic protein (GFAP) response element. At 45 days of age mice were randomly allocated to two groups, one of which received partial bile duct ligation; controls received sham surgery without ligation of the bile duct ( Figure 4A–B ¹⁵ ). There was a consistent increase in bioluminescence 3 days after induction of the model versus sham controls. One particular mouse showed a remarkable, sustained increase in luciferase output following induction. This data illustrates the strength of continual non-invasive measurements, in that it is possible to identify consistent outliers, which would not be possible with terminal end-point analysis, due to a lack of data on expression kinetics.

Open Science Framework: Non-invasive somatotransgenic bioimaging in living animals. https://doi.org/10.17605/OSF.IO/XF7EZ ¹⁵

This project contains the following underlying data:

Data are available under the terms of the Creative Commons Zero "No rights reserved" data waiver (CC0 1.0 Public domain dedication).