Evaluation of NanoLuc substrates for bioluminescence imaging of transferred cells in mice Journal of Photochemistry Photobiology, B: Biology

NanoLuc luciferase recently gained popularity due to its small size and superior bioluminescence performance. For in vivo imaging applications, NanoLuc has been limited by its substrate furimazine, which has low solubility and bioavailability. Herein, we compared the performances of recently reported NanoLuc luciferase substrates for in vivo imaging in mice. Two substrates with improved aqueous solubility, hydrofurimazine and fluorofurimazine, were evaluated along with three stabilized O -acetylated furimazine analogues, the hikarazines. All 5 analogues, when tested in vitro , displayed greater signal intensity and reaction duration, in comparison to the standard NanoLuc substrate, furimazine. The two best-performing analogues from the in vitro study were selected for further in vivo testing. The NanoLuc/fluorofurimazine pair demonstrated the highest bioluminescence intensity, post intravenous administration. It was found to be around 9-fold brighter compared to the NanoLuc/ furimazine and 11-fold more intense than the NanoLuc/hikarazine-003 pair, with an average of 3-fold higher light emission when the substrate was injected intraperitoneally, in a subcutaneous model. Excitingly, despite the fact that NanoLuc/fluorofurimazine emits mostly blue light, we prove that cells trapped in mice lungs vasculature could be visualised via the NanoLuc/fluorofurimazine pair and compare the results to the AkaLuc/Aka- Lumine system. Therefore, among the tested analogues, fluorofurimazine enables higher substrate loading and improved optical imaging sensitivity in small animals, upgrading the use of NanoLuc derived bioluminescent systems for deep tissue imaging.


Introduction
Bioluminescence imaging (BLI) is based on a biochemical reaction that is dependent on the oxidation of a luminophore containing substrate (luciferin) by luciferase enzymes with light emission as a result. Luciferase enzymes and their substrates are extensively used as effective, non-invasive screening tools in diverse research fields and have become a prominent method for live-cell visualization. Improving the performance of luciferase enzymes, substrates and their mutants for BLI is an important part in improving the tool box available for life science research [1].
Several thousand bioluminescent species are represented by ~700 genera among which 90% originate from marine organisms [2,3]. Luciferases from marine luminous organisms primarily utilise as substrates, coelenterazine [2], varguline [3] or analogues of these two. Coelenterazine (CTZ) is best-known as the substrate for luciferases derived from the sea pansy Renilla reniformis (RLuc), the luminous shrimp Oplophorus gracilirostris and the copepod Gaussia princeps (GLuc), yielding light with a fast enzymatic turnover and an emission maxima in the blue region (450-485 nm) of the visible spectrum.
NanoLuc (NLuc) is one of the smallest luciferases in size (19 kDa) and an important addition to the marine BLI-toolset [4]. Its small size, luminescence superiority and stability easily replaced other luciferases where increased sensitivity, fast response dynamics and low background auto-luminescence are vital [5]. The NLuc system is coupled with an optimized CTZ analogue, the 2-furanylmethyl-deoxycoelenterazine, known as furimazine, which when paired with NLuc gives ~2.5 million-fold brighter luminescence in mammalian cells and ~150-fold greater specific activity than firefly (Photinus pyralis) or Renilla luciferases [4].
Furimazine and its analogues are generally lipophilic and their solubility in aqueous solutions has limited the maximum injectable dose into small animals (e.g. 1.3 μmol in a polyethylene glycol (PEG) [6] based buffer and 0.016 μmol in Phosphate Buffered Saline (PBS) [7]. Therefore, furimazine analogues with improved solubility and bioavailability are needed to improve NanoLuc sensitivity for in vivo imaging [8]. Moreover, the NanoLuc/furimazine system mainly emits blue light, further limiting the sensitivity of detection in deep tissue.
Many synthetic analogues of CTZ have been reported in recent years, mostly focusing on the modification of the C-2, C-5, C-6 and C-8 substituents of the imidazopyrazinone core [4,9], with the most effort on modifying the C-2 and C-6 position [5][6][7]. The novel analogues were mostly optimized in order to improve in vivo imaging with NanoLuc, leading to better compound stability and more efficient light penetration through tissues [5][6][7]. A chemical library of 135 CTZ-analogues, the imidazo[1,2-a]pyrazin-3-one, have been recently synthetized as stabilized O-acetylated precursors (hikarazines) with an original synthesis route [10,11]. Some examples show robust bioluminescent properties in terms of signal intensity and duration of the reaction. Particular examples turned out to be biochemically more efficient than furimazine with up to 2.5 increased light intensity and signal stability lasting up to 2 h [11]. Novel furimazine analogues, with increased aqueous solubility, were recently developed and named hydrofurimazine (HFz) and fluorofurimazine (FFz) [12]. These analogues enable in vivo applications by allowing the delivery of higher substrate doses via single intraperitoneal injection in mice leading to extended light emission in vivo for NanoLuc [13].
In the current work we investigated the BLI properties of novel furimazines (HFz and FFz) [11] and 3 different hikarazines (hikarazine-001, hikarazine-003 and hikarazine-097) [14][15][16] by evaluating their in vivo performance as substrates for imaging with NanoLuc, as the reference marine luciferase since recently they have demonstrated to be better suited for in cellulo and in vivo NLuc based imaging [14][15][16].
Although the above mentioned NanoLuc substrates have been reported to perform well in vivo, no direct comparison has been performed yet. In the present study, we first evaluated 5 recently reported analogues in vitro, through controlled conditions, using cells expressing a known amount of NanoLuc reporter, following with an in vivo comparison in a subcutaneous model. Secondly, since BLI in deep tissue is mostly dependent on the amount of generated photons and the emission wavelength, we tested the best performing NLuc/substrate pairing (NLuc/fluorofurimazine) in deep tissue, by imaging transferred bioluminescent cells that get entrapped in mice lungs, and compared it to the near-infrared AkaLuc/AkaLumine system [17], recently described by Iwano et al.

Cell Propagation for Life Cell Assays
Human embryonic kidney (HEK-293T) cells were grown in Dulbecco's Modified Eagle's Medium (DMEM) (Sigma-Aldrich), supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin. Cells were cultured in T-175 flasks (Thermo Fisher Scientific) until reaching confluence; growth media was aspirated and cells were washed with PBS (Sigma-Aldrich). The PBS was aspirated and the cells were removed from the flask by the addition of 3 mL of TrypLE™Express Trypsin (Life Technologies) and were incubated at 37 • C. Cells were centrifuged at 1500 RPM, washed, re-suspended in 5 mL of fresh growth media, and then counted using a BioRad TC20 cell counter. Cells were diluted to a desired final concentration and were plated in 96 well plates (Greiner Cell Star®) in total volume of 100 μL at 25000 cells/well. The culture was incubated at 37 • C with 5% CO 2 .

In Vitro Cell-Based Imaging
HEK-293T-NanoLuc cells were seeded in a 96-well black plate (Greiner Cell Star®) at a density of 25 000 cells per well. Post 24 h, cells were washed in PBS and imaged at the IVIS Imager (PerkinElmer) after addition of 6 different substrates profiled within this study; furimazine, hydrofurimazine (HFz), fluorofurimazine (FFz), all three provided by Promega and the hikarazine-001, hikarazine-003 and hikarazine-097 provided by Institute Pasteur, Paris, France. The hikarazines (hikarazine-001, hikarazine-003, hikarazine-097) were dissolved (1 mg) in a mixture of 0,2 mL DMSO (Sigma-Aldrich) and then diluted in a solution of 0,3 mL of acidic ethanol made from addition of 37% hydrochloric acid (Sigma-Aldrich), 100 μL on 12 mL of 100% ethanol (Sigma-Aldrich) to give a stock solution of 1 mg/mL which was then incubated at 50 • C for 2 h. Furimazine, hydrofurimazine (HFz) and flurofurimazine (FFz) were dissolved in 100% ethanol to give a 5 mM stock solution. All of the substrates were further diluted in DMEM medium to a final concentration of 10 μM (100 μL/well). Bioluminescence signal from wells was measured at the IVIS (Perkin Elmer) spectrum system every 5 min after substrate addition. All in vitro cell-based measurements were acquired after 1 min at 37 • C using a 30 s acquisition time with an open filter. Experiments were performed in triplicates and were repeated three times. Data was analyzed using Living Image 4.3 software (Perkin Elmer) by applying the appropriate region of interest (ROI).
As for the once-cell assay; HEK-293T-NanoLuc cells were plated in a black 96-well plate in seven dilutions starting at a density of approximately 100 cells per well as the starting cell-concentration. Further celldilutions were carried out with a ½ ratio; (50, 25, 12, 6, 3, 1, 0) cells per well. Luminescence images were acquired with the GloMax® Microplate Reader (Promega) immediately after substrates addition; furimazine, hydrofurimazine (HFz), fluorofurimazine (FFz), hikarazine-001, hikarazine-003 and hikarazine-097 with a final concentration of each substrate of 10 μM. For measurement of total luminescence, we used 1 s of integration time. Measurements were performed at room temperature 18-20 • C. Experiments were performed in triplicates and were repeated three times. All data were plotted using GraphPad Prism.

Substrate Formulation for Use in Whole-Animal Imaging
The hikarazines (hikarazine-001, hikarazine-003, hikarazine-097) (Institute Pasteur, Paris, France) were dissolved (1 mg AkaLumine (Sigma-Aldrich) was dissolved in PBS to a concentration of 33 mM (around 3 μmol/100 μL) for i.p. injection. These doses were chosen considering maximum solubility and tolerability in mice and maximum attainable signal based on previous findings.

In Vivo Bioluminescence Imaging of Subcutaneously Implanted NanoLuc Expressing Cells
Animal experiments were approved by the Bioethics Committee of Erasmus MC, Rotterdam, The Netherlands and performed in accordance with national guidelines and regulations established by the Dutch Experiments on Animal Act (WoD) and by the European Directive on the Protection of Animals used for scientific purpose (2010/63/EU). BALB/c nude (males) were obtained from Charles River Laboratory (The Netherlands). All mice aged 8 weeks were provided access to food and water ad libitum and were hosted in the animal facility at the Erasmus MC, Rotterdam, The Netherlands.
For the background assessment mice (n = 3 mice per tested analogue) were injected intraperitoneal (i.p.) with the following ana- For the subcutaneous skin model experiments, [8][9][10] week old BALB/ c nude (males) mice received a subcutaneous injection of 1 × 10 6 HEK-293T-NanoLuc which were prepared in PBS (Sigma-Aldrich) and matrigel (Corning) solution (50:50 ratio) in a final injectable volume of 50 μL. Mice received an injection of 100 μL of the different substrate intravenously (n = 6 mice per group) or intraperitoneally (480 μL) (n = 3 mice per group) right after implantation of cells. The size of the group was determined using power analysis. Mice were kept under isoflurane anaesthesia (1.5%) and a series of images were taken using an IVIS Spectrum with open filter binning = medium, field of view = 12.9 × 12.9 cm, f/stop = 1 and 1 s exposure time every 5 min for half an hour. Data analysis was performed by drawing ROIs in the images taken at the peak of bioluminescence emission.

Statistical Analysis
For experiments where more than two groups were compared, oneway ANOVA, followed by Tukey's t-test was used to determine significant differences among treated groups.

In Vitro Performance Comparison of Novel Analogues Paired with NanoLuc Luciferase
Bioluminescence properties of the chosen furimazine analogues (Fig. 1), profiled with NLuc were assessed in live HEK-293T cells that were stably expressing NLuc luciferase in order to determine the bestperforming NLuc/analogue pairs for in vivo testing. Cells were treated with substrates at a final concentration of 10 μM each and imaged in the IVIS imager at 5 min intervals after substrate administration for half an hour (Fig. 2a).
The imaging results in live cells showed brighter signals with the hikarazines (hikarazine-001, hikarazine-003 and hikarazine-097) than furimazine and its analogues hydrofurimazine (HFz) and fluorofurimazine (FFz) (Fig. 2b). We found that peak light emission was highest with hikarazine-003, followed by hikarazine-001, hikarazine-097 and fluorofurimazine. The signal of all tested analogues proved to be extended and stable over the 30 min imaging duration.
Since the photon emission was comparable among the 3 tested hikarazines, we selected the brightest analogue (hikarazine-003) for further in vivo evaluation among with fluorofurimazine (FFz).

In Vivo Bioluminescence Imaging of Subcutaneously Implanted Cells
Prior to testing the compounds in mice, we assessed background emission of all substrates, including the AkaLuc substrate, AkaLumine, in nude mice not expressing a luciferase reporter. We administered the substrates intraperitoneally, allowing us to estimate autoluminescence in the absence of reporter enzymes (NanoLuc/AkaLuc). All substrates were administered with the doses recommended by the literature as appropriate for each substrate. Negligible signals were detected with AkaLumine, whereas all tested NLuc analogues gave background signal of around 10 2 ph/s in the abdomen in nude mice ( Supplementary Fig. 1).
Given their intense in vitro brightness, we selected FFz and hikarazine-003 for further in vivo evaluation. Each substrate was administered via intravenous (i.v.) and intraperitoneal (i.p.) injection in mice. Around 1 × 10 6 HEK-293T cells were transferred in nude mice and injected subcutaneously in the lower flank, stably expressing NanoLuc luciferase.

In Vivo Deep Tissue BLI: NLuc/Fluorofurimazine vs. AkaLuc/ AkaLumine
To assess how well NLuc in combination with FFz performs in deep tissue, we evaluated the brightness from HEK-293T cells stably expressing NanoLuc and copGFP, and compared it to photon emission of cells expressing AkaLuc and copGFP. We used GFP to ensure equimolar expression of both bioluminescent enzymes. The luciferase expressing cells were localised in mice lungs after tail vein injection of approximately 5 × 10 5 HEK-293T expressing either NLuc-copGFP or AkaLuc-copGFP [20,21] (Fig. 6a). A dose of 4.3 μmol FFz gives more light, but was found to cause weight loss and was associated with organ damage in mice: later on, a dose was established which showed no toxicity in mice consisting of 1.3 μmol FFz in P-407 [12], which we implemented in this experiment. Mice were imaged 10 min after cell administration and a kinetic analysis was performed after i.p. injection of substrate (1.3 μmol/145 μL FFz; 3 μmol/100 μL AkaLumine) for 30 min. We were able to detect signals from both luciferase systems with a 1.5-fold higher photon emission arising from the AkaLuc/AkaLumine system (Fig. 6b). For the emission comparison, values at the time point of peak emission were used. The signals were corrected by a factor of 2.3 fold, given the higher-level of GFP expression of HEK-293T-AkaLuc expressing cells, determined prior the imaging session.

Discussion
In the present study we tested novel furimazine analogues paired with NanoLuc luciferase to determine the most optimal NLuc/luciferin pair for in vivo imaging. We report on photon yields of two novel, furimazine analogues, recently published by Su, Y. et al., named hydrofurimazine (HFz) and fluorofurimazine (FFz) and three different Oacetylated furimazine analogue derivates [9], known as 'hikarazines' (hikarazine-001, hikarazine-003 and hikarazine-097) and we confirm that many of these novel analogues improve the efficient brightness of NanoLuc in vitro and in vivo.
Our aim was to compare the best performing analogues in an in vivo setting based on different routes of administration. The comparison of    BLI systems using cells expressing a known amount of luciferase injected subcutaneously in animals has been largely used in the past [22][23][24][25]. As for coelenterazine based substrates, it is generally known that they produce higher light output when injected intravenously [26]. Yet for repeated measurements, intraperitoneal in vivo injections should be preferred, given the reproducibility of the technique being less prone to misinjections.
A recent study reported that furimazine possibly displays toxic side effects, when administered repeatedly in small animals with doses exceeding 40 μg of substrate per day [27]. Furimazine, inter alia, is poorly soluble in aqueous solution, whereas D-luciferin salt can be dissolved to high concentrations in simple buffered saline [12,28], and repeatedly administered via intraperitoneal administration (i.p.). The preferred furimazine administration route is via intravenous (i.v.) injection, since higher photon emission is achieved. However, signals appear faster after substrate injection and show a rapidly decaying kinetic, limiting its use in certain areas of molecular imaging where a broader imaging window (longer substrate half-life) is required [13].
Based on our in vivo results we can clearly highlight on the importance of substrate solubility and bioavailability as the number one limitation for sensitive in vivo bioluminescent imaging [12]. The pairs, NanoLuc/fluorofurimazine (4.3 μmol and 333 nmol) and NanoLuc/ hikarazine-003 (333 nmol), independent of the administration route, exhibited brighter (8 or 11-fold) in vivo BLI signals, than the standard NanoLuc/furimazine (333 nmol) pair. Moreover, it was clearly evident that the NanoLuc/fluorofurimazine (4.3 μmol and 333 nmol) pair was 11 or 3-fold brighter when administered intravenously or intraperitoneally, compared to the NLuc/hikarazine-003 pair (333 nmol). The photon generation from FFz is more prolonged and more intense, allowing in vivo tracking with the NLuc reporter over long time periods.
We also address in vivo coelenterazine analogues autoluminescence 17 by simply injecting the analogues in mice without the luciferase reporter being expressed. Such an approach provides accurate insight whether the compounds are prone to spontaneous light emission. Since dying cells expressing the NanoLuc reporter release the active luciferase in the bloodstream, which does not require any cofactors, and therefore can catalyze oxidation of its substrates and give background light. Our results demonstrate that NLuc substrate analogues (FFz, hikarazine-003) and furimazine itself also spontaneously oxidase, without the reporter being present in nude mice, and produce detectable light emission. This is important to take into account when performing imaging with furimazine and coelenterazine based analogues; performing a pre-scan and the proper positioning of animals is highly recommended in order to obtain accurate experimental results [12].
We next assessed the in vivo BLI performance of FFz with NLuc in deep tissue (lungs) and compared it to the near infra-red emission of  AkaLuc/AkaLumine pair, as a near-infrared BLI system reference with an emission peak at 650 nm.The luciferase expression was localised in mice lung vasculature by tail vein injection of NLuc and AkaLuc expressing cells [20,21]. AkaLuc/AkaLumine, yielded a 1.5-fold higher signal then FFz with NLuc. Nevertheless, here we demonstrate clearly that NLuc can be used for deep tissue imaging in small animals, when combined with the novel FFz.
Multiple BL reporters are derived from NLuc luciferase that uses furimazine, or its analogues as substrates, like Antares [6] Antares 2 and LumiScarlet [24,29]. These protein fusions exhibit emission wavelengths above 600 nm due to intramolecular bioluminescence resonance energy transfer (BRET), and therefore transmit light more easily through mammalian tissues. Moreover, using a different strategy, furimazine analogues can generate red-shifted emission when paired with NanoLuc [30,31], and therefore could enable more sensitive visualization in deep tissue. However, all of these bioluminescent systems do not show peak of emission in the near infrared region as the AkaLuc/AkaLumine system.
In conclusion, the combination of superior brightness and signal duration, given by FFz, enables bioluminescence imaging with NanoLuc luciferase, in superficial and deep tissue, making it applicable for studying diverse physiological events over a substantial period of time.

Funding
We acknowledge the funding for this work provided by the European Commission under the H2020-MSCA-RISE award grant number 777682 (CANCER) and under the H2020-MSCA-ITN award, grant number 675743 (ISPIC).

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.