There is a strong need for safe drugs and vaccines to target emerging pathogens such as SARS-CoV-2. Although recent studies identified approved drugs that exhibit antiviral activities against SARS-CoV-2 [1
], current therapeutic treatment strategies for COVID-19 have limited effectiveness. There are currently no oral medications given emergency use authorization from the Food and Drug Administration to prevent SARS-CoV-2 infection or to treat COVID-19. There is an urgent need to identify safe, economical, orally deliverable approved drugs with activity against SARS-CoV-2 to prevent infection in at-risk populations, and to treat patients experiencing viral disease [3
]. Attempts to identify approved drugs with antiviral activity led to the discovery of more than 100 compounds that exhibit direct antiviral activity against SARS-CoV-2 isolates in vitro [2
]. Although the on- and off-target binding mechanisms that mediate anti-SARS-CoV-2 activity are not clear, two classes of molecules were previously found to effectively inhibit virus infectivity: protein biogenesis inhibitors (e.g., zotatifin, ternatin-4, PS3061) and ligands of the sigma-1 and sigma-2 receptors (e.g., haloperidol, clemastine, cloperastine) [7
Specific antihistamines with off-target antiviral activity may have repurposed utility for prevention and treatment of COVID-19 because of known safety profiles and widespread availability. Common antihistamines that exhibit off-target antiviral activity include hydroxyzine, azelastine and diphenhydramine [8
]. Mechanisms of action for drugs with direct anti-SARS-CoV-2 activity have important clinical implications in terms of dosing and drug interactions. Defining mechanisms that drive antiviral activity against SARS-CoV-2 will provide rationale for drug combinations targeting distinct antiviral pathways [9
]. Drug combinations that target separate antiviral pathways are expected to inhibit drug resistant variants resulting from emerging mutations.
Coronaviruses replicate in a modified compartment derived from the endoplasmic reticulum (ER). The sigma receptor-1 is an ER resident chaperone that normally functions to modulate the ER stress response [10
]. Coronavirus infection activates pathways to facilitate adaptation of ER stress to virus proliferation. These pathways are thought to hijack the host cell ER stress response to modulate protein translation, ER protein folding capacity and ER-associated degradation. Targeting the ER stress response could elucidate coronavirus protein-host interactions and provide rationale for new therapeutic approaches to prevention and treatment of COVID-19.
In infected cells, the sigma-1 receptor was shown to link the SARS-CoV-2 replicase/transcriptase complex to the ER membrane by binding directly to nonstructural protein 6 (NSP6) [7
]. Although sigma-1 receptor ligands exert antiviral activity against non-coronaviruses and coronaviruses [10
], it is not known if agonist or antagonist activities prevent SARS-CoV-2 infection. Understanding binding interactions of antiviral sigma-1 receptor ligands may provide the basis for drug development and optimization.
Although approved drugs that inhibit SARS-CoV-2 in vitro have been shown to bind sigma-2 receptors, the structure and functions of sigma-2 receptors are not well characterized or understood [11
]. Sigma-1 and sigma-2 receptors are unrelated in sequence and structure. The sigma-2 receptor is an ER resident membrane protein thought to be involved in hormone, calcium and neuronal signaling [12
]. The sigma-2 receptor regulates cholesterol transport and contributes to cholesterol homeostasis [13
]. In infected cells, the sigma-2 receptor was shown to bind directly to SARS-CoV-2 ORF9c [7
], suggesting that sigma-2 receptor ligands may block host protein:virus protein interactions. Recently, the sigma-1 and -2 ligand PB28 that had sub-nanomolar in vitro SARS-CoV-2 inhibitory effects was found ineffective in vivo (33). The hypothesized cause for poor efficacy was that this compound induced high levels of phospholipidosis in vitro that resulted in virus inhibition that could not be achieved in vivo. Sigma receptor ligands, as a class, should not be discounted because of the poor performance of a single compound. In this work, we demonstrate phospholipidosis induction by sigma ligands does not correlate with inhibition of SARS-CoV-2 viral replication.
It is clear that multiple sigma receptor ligands exhibit antiviral properties against SARS-CoV-2, but the relative roles of the sigma-1 receptor and sigma-2 receptor agonism and antagonism in modulating antiviral activities are not known. In this study, antiviral activities of highly selective ligands (Figure 1
) were measured to define mechanisms driving inhibition of SARS-CoV-2 infection in vitro: a sigma-1 receptor specific agonist (SA4503, cutamesine) [14
], sigma-1 receptor antagonist (CM304) [17
], sigma-2 receptor specific ligand (CM398) [18
], and a mixed affinity sigma-1/sigma-2 ligand (AZ66) [19
]. The benzothiazolone (CM304 and AZ66) and benzimidazolone (CM398) containing compounds were selected for their demonstrated selectivity for sigma receptors against other aminergic transporters or receptors [21
]. In addition, the specific compounds were chosen for their differential affinity at the two receptor subtypes, aiming to clarify the involvement of each receptor in the inhibition of SARS-CoV-2 infection in vitro.
3. Discussion and Conclusions
SARS-CoV-2, the causative virus of COVID-19 pandemic, belongs to a family of positive-sense single-stranded RNA (+ssRNA) coronaviruses (CoVs) that also cause illnesses ranging from common colds to severe diseases such as Middle East respiratory syndrome (MERS). There are 7 CoVs known to infect people: 229E, NL63, OC43, HKU1, MERS-CoV, and SARS-CoV that emerged in 2003 [3
]. CoV infection is known to activate pathways that facilitate adaptation of ER stress for viral replication [31
]. CoVs utilize host cell ER stress responses to modulate protein translation, ER protein folding capacity, ER-associated degradation (ERAD) including autophagy, and apoptotic cell death [32
]. It has been proposed that modulation of CoV induced ER stress responses may provide the rationale for new approaches to antiviral drug therapy.
Sigma receptors act as modulators of ER stress, functioning as ligand operated membrane bound chaperones at the ER-mitochondrial contact (mitochondrion-associated ER membrane) [34
]. Sigma-1 receptor ligands have been shown to exert antiviral activity against CoVs and non-CoVs, including Ebola, HCV, SARS-CoV, SARS-CoV-2, DENV, MERS-CoV, FLUAV (H5N1), HCV, HIV and HSV-1 [10
]. Sigma receptors were implicated as targets for antiviral drugs by mapping interactions between human proteins and 26 (of 29) SARS-CoV-2 proteins, and subsequent screening of approved drugs [7
]. Two sets of pharmacological agents effectively inhibited SARS-CoV-2 infectivity in Vero E6 cells: inhibitors of mRNA translation and predicted regulators of the sigma-1 and sigma-2 receptors. Non-selective sigma-1 receptor ligands, including the antihistamines clemastine and cloperastine, exhibited activity against SARS-CoV-2 in vitro. PB28 a sigma-1 and sigma-2 receptor ligand was highly efficacious in vitro [7
] but was toxic in vivo so the search for effective ligands was continued in this work [35
Mechanisms that drive anti-SARS-CoV-2 activity by sigma receptors are not well characterized. It is not understood if both sigma-1 or sigma-2 receptors are involved in antiviral activity, or if agonism, or antagonism of individual receptors mediate antiviral activity. A significant limitation to addressing the role of sigma receptors in SARS-CoV-2 inhibition (of entry, replication or infectious virus assembly/release) is the paucity of structural information available for the sigma-2 receptor, and absence of well characterized agonists and antagonists. Identification of ligands that exert antiviral activity by specific sigma receptor binding may provide the basis for use of existing drugs (repurposed) and for development of new drugs optimized for activity against CoVs.
Combining a sigma receptor ligand with antiviral drugs that bind distinct targets may provide additive or synergistic antiviral effects and decrease the likelihood of SARS-CoV-2 resistance to a single drug. Drug combinations are recommended for antiviral therapy of hepatitis C (e.g., combination of alpha interferon, simeprevir and ribavirin), and HIV [36
]. Combinations of drugs that bind host and/or viral proteins have the potential lessen the severity of COVID-19 by inhibiting virus replication and reducing symptoms. Administration of antiviral drug combinations to SARS-CoV-2 positive patients could determine hospitalization versus home-based care.
Data suggests that specific drugs that bind SARS-CoV-2, or interacting host proteins, also have the potential to prevent COVID-19. For example, hydroxyzine is a first-generation antihistamine that exhibited off-target binding to the SARS-CoV-2 host receptor ACE2 [37
] and the sigma-1 receptor. Usage of hydroxyzine (and structurally related antihistamines diphenhydramine and azelastine) was associated with reduced incidence of SARS-CoV-2 positivity in a population of more than 219,000 individuals in California [8
]. Hydroxyzine, diphenhydramine and azelastine exhibited direct antiviral activity against SARS-CoV-2 infection of Vero E6 cells in vitro. Since antihistamines act as nasal decongestants and cough suppressants, the on- and off-target binding properties of drugs such as diphenhydramine may have broad utility in prevention and treatment of COVID-19.
In this study, we defined selective sigma receptor ligands (Figure 1
) that drive antiviral activity against SARS-CoV-2. The dual specificity sigma-1 and sigma-2 receptor ligand AZ66 exhibited antiviral activity against SARS-CoV-2 induced cytotoxicity of Vero E6 cells (Figure 2
A). Since the sigma-1 receptor antagonist CM304 did not inhibit viral cytotoxicity (Figure 1
B), and the sigma-1 receptor agonist SA4503 (cutamesine) exhibited inhibitory activity (Figure 2
D), these data suggest that sigma-1 receptor agonism drives antiviral activity against SARS-CoV-2. Ligation of the sigma-2 receptor may drive antiviral activity independently, since the highly selective sigma-2 receptor ligand CM398 exhibited direct inhibitory activity against SARS-CoV-2 (Figure 2
C). Ligation of both sigma-1 and sigma-2 receptors may elicit higher levels of antiviral activity compared to receptor specific ligands, since AZ66 exhibited the greatest gap between SARS-CoV-2 inhibitory and cellular toxic concentrations (CC50
ratio, Table 1
We verified the ability of sigma receptor ligands to exhibit antiviral activity by infecting Vero E6 cells at a low MOI and quantifying viral replication by qPCR (Figure 3
A). The dual sigma receptor ligand AZ66 exhibited the more significant antiviral effects compared to the selective receptor ligands. AZ66 exhibited antiviral activity against SARS-CoV-2 by plaque assay (Figure 3
B). Induction of phospholipidosis by these compounds was measured in human lung epithelial cells (Figure 3
C). All compounds induced phospholipidosis to ~50% of the positive control. However, we were unable to identify a correlation between phospholipidosis and inhibition of virus replication and is consistent with findings from other groups [38
]. CM304 was a strong inducer of phospholipidosis yet was ineffective at inhibition of SARS-CoV-2. These data indicate that the antiviral activities of sigma ligands AZ66 and CM398 are driven by specific antiviral inhibitory mechanisms outside of phospholipidosis. These data are consistent with microscopic observation of AZ66 (sigma-1 and sigma-2 receptor ligand) and CM398 (sigma-2 receptor ligand) reducing cell rounding and death caused by SARS-CoV-2 infection (Figure 4
We mapped potential interactions between ligands and sigma receptors to gain insight in intermolecular interactions that promote antiviral activity against SARS-CoV-2. Identification of specific residues in sigma receptors that bind antiviral drugs may provide the basis for drug development strategies to optimize ligand binding.
We used a crystal structure of the human sigma-1 receptor (PDB 5HK1) as the basis for molecular docking of sigma-1 receptor ligands. Comparison of the posed orientations of a sigma-1 receptor agonist (SA4503) with an antagonist (CM304) complexed to the sigma-1 receptor revealed that the agonist (with antiviral activity) formed more intermolecular contacts with the receptor compared to the antagonist (without antiviral activity) (Figure 5
). These data provide the basis for site directed mutagenesis studies to define key ligand binding residues. These data suggest that drugs optimized for sigma-1 receptor agonist and SARS-CoV-2 antiviral activity may be achieved with analogs that form interactions with specific residues in the sigma-1 receptor ligand-binding site: I124, F133, V152, V162, and W164.
Since crystal structures are not available for the human sigma-2 receptor, we generated an atomic homology model based on the most similar solved structure, 3-β-hydroxysteroid-Δ8,Δ7-isomerase, known as Emopamil-Binding Protein (EBP). EBP, a transmembrane protein comprised of α-helices and loop regions that form a ligand binding site, was solved complexed to an inhibitor (Figure 6
]. We used molecular docking to simulate ligand binding of sigma-2 receptor specific ligand CM398, and sigma-1/sigma-2 receptor ligand AZ66, predicted to form intermolecular interactions with ligand-binding site residues M28, D29, L47, Y50, Y147, Figure 5
B. In addition to the anti-viral effects of AZ66, binding of the sigma receptors reduces nociception [17
]. The analgesic effect of AZ66 could provide novel treatment of SARS-CoV-2 related pain while inhibiting viral replication. These data provide the basis for mutagenesis and structure-activity-relationship studies to optimize sigma-2 receptor binding and antiviral activity against SARS-CoV-2.
Specific antihistamines exhibit off-target sigma receptor binding activity, and also exhibit antiviral activity against SARS-CoV-2, including clemastine, cloperastine, astemizole, hydroxyzine, azelastine and diphenhydramine. Since diphenhydramine is the most commonly used antihistamine exhibiting antiviral activity, we asked if antiviral activity could be improved by combining a sigma receptor ligand with lactoferrin, an antiviral agent that binds distinct targets [28
]. We found that co-administration of 400 μg/mL of lactoferrin with diphenhydramine reduced SARS-CoV-2 induced cytotoxicity and decreased the EC50 (Figure 8
C,D). The antiviral enhancement effects of lactoferrin were more apparent at lower, therapeutically relevant concentrations of diphenhydramine (Figure 8
E). Combining lactoferrin with diphenhydramine resulted in synergistic effects on antiviral activity against SARS-CoV-2 (Figure 8
F). Compounds we found effective in Vero E6 were validated in their ability to reduce infectious SARS-CoV-2 production following infection of human lung epithelial cells (Figure 9
C,D). These data suggest that sigma receptor ligands or formulated combinations of over-the-counter products have the potential to inhibit virus infection and/or decrease recovery time from COVID. Lastly, concentrations that inhibited SARS-CoV-2 production were decoupled from phospholipidosis in human lung epithelial cells, suggesting a specific mechanism at the sigma receptors/virus interface. The candidates investigated in this work target sigma receptors that result in selectivity indices higher than remdesivir, a top candidate in large-scale in silico screens that showed efficacy in vitro and in vivo [6
4. Materials and Methods
4.1. Sigma Ligands and Other Drugs Used in this Study
AZ66, CM304, and CM398 were synthesized and characterized in the McCurdy lab at the University of Florida as previously reported [17
] with purities > 95% each (Supplementary Materials
). SA 4503 (cutamesine) was obtained from MilliporeSigma (St. Louis, MO, USA) at >98% pure according to the manufacturer and diluted in PBS to 2 mg/mL and frozen at −80 °C in aliquots to eliminate freeze thaw cycles. Lactoferrin from human milk was obtained from MilliporeSigma at >85% purity according to the manufacturer and diphenhydramine HCl was purchased from Spectrum Pharmaceuticals at ≥98% purity according to the manufacturer.
4.2. Virus Culture Methods
The SARS-CoV-2 strain used in this study was UF-1. It has been described previously [1
] and was isolated from a COVID19 patient at UF Health Shands Hospital via nasal swab. Virus experiments were carried out under a University of Florida Institutional Biosafety Committee-approved protocol in a Biosafety Level 3 laboratory at the Emerging Pathogens Institute. The accession number of the previously sequenced strain can be found under the following GenBank accession number: MT295464.1. Vero E6 cells purchased from ATCC were used to propagate virus using standard methods. Vero E6 cells were grown in DMEM + 2%FBS+ PenStrep. SAEC, H23 and H23-hACE2 cells were grown in RPMI + 10% FBS + PenStrep with 4 μg/mL of blasticidin to maintain ACE2 expression if needed. Cell were grown at 37 °C and 5% CO2
in a humidified incubator. An EVOS XL Core microscope was used to visualize cells in the BSL3.
4.3. Quantitation of Virus Replication by qPCR
SARS-CoV-2 was used to infect Vero E6 monolayers at an MOI of 0.01 in the presence of each treatment in biological and technical triplicate. At 2 days post-infection (dpi), the monolayers were scraped and harvested into viral lysis buffer (buffer AVL) from the QIAamp Viral RNA Kit (QIAGEN). The AVL buffer is a CDC approved method of viral inactivation. Samples were frozen at −80 °C and removed from the BSL3. RNA was purified according to the manufacturer’s recommendations. Reverse transcription and cDNA synthesis was accomplished using the iTaq Universal SYBR Green One-Step Kit (BioRad) and primers targeting the nucleocapsid (N) gene of SARS-CoV-2 (NproteinF- GCCTCTTCTCGTTCCTCATCAC, NproteinR-AGCAGCATCACCGCCATTG). qPCR was carried out on a BioRad CFX96. N protein copy levels were calculated using CT values from a standard curve generated using a control plasmid containing the N protein gene and are presented as genome equivalents (GE) (Integrated DNA Technologies).
4.4. Sigma Ligand Cytotoxicity Reduction Assays
Vero E6 cells were seeded into 96-well CellBind treated plates (Corning) and allowed to attach overnight. Drugs were pre-aliquoted in DMEM + 2%FBS. Cells and drug dilutions were transported into the BSL3 laboratory where titered SARS-CoV-2 aliquots were diluted to produce a target MOI of 0.2 PFU/cell in solution at the final indicated drug concentrations. Triplicate monolayers were infected by replacing growth media with 100 μL of the drug/virus suspensions. At 72 h post infection, supernatants were harvested, and lactate dehydrogenase (LDH) release was assayed using the Cytox 96™ Non-Radioactive Cytotoxicity Assay (Promega, Madison, WI, USA). Assays were performed as recommended by the manufacturer to generate a formazan dye. The optical density at 450 nm was measured using a MultiSkan FC plate reader (Thermo Fisher Scientific, Waltham, MA, USA). Controls included total LDH release as measured by lysis of all cells, spontaneous release from uninfected cells, and media alone. The toxicity of sigma ligands alone were also determined in parallel to discriminate the amount of SARS-CoV-2-induced cytotoxicity occurring in the presence of a given treatment. After spontaneous and background subtraction, OD450 values were transformed to a percent of SARS-CoV-2 infected cells (100%) in the absence of any drug treatment to obtain percent of SARS-CoV-2-induced cytotoxicity. These experiments were carried out twice.
4.5. Plaque Reduction Assay
Vero E6 cells were plated in 24-well plates with triplicate replicates on different plates. Virus master mix was used to dilute down to 20-200 PFU/mL and aliquoted in separate tubes with drugs at the final indicated drug concentrations. The drug virus mixtures were immediately used to infect Vero E6 monolayers for 1 h with rocking every 10 min. Monolayers were then overlaid with MEM in 1.5% low-melt agarose containing drugs at the final concentrations indicated. Plaques were counted at 72 hours post infection and used to calculate the apparent reduction in viral concentration compared to the starting volume. Data presented is representative of two independent experiments.
4.6. Generation of ACE-2 Lentivirus Particles
The lentivirus containing ACE2 were generated by co-transfecting psPAX2, pMD2.G, and an ACE expression vector that also contained a blasticidin selection gene EX-U1285-Lv197 (GeneCopoeia). The plasmids were transfected into HEK293T cells using X-tremeGENE 9 (Roche Cat# XTG9-RO) as per the manufacturer’s instructions. Media was replaced with DMEM containing 2% (w/v
) bovine serum albumin (BSA) 18 h post transfection and then lentiviruses were collected after 24 and 48 h [40
4.7. ACE2 Transduction of NCI-H23 Cells and Monoclonal Cell Selection
NCI-H23 (aka H23) cells were obtained from ATCC (CRL-5800) and ACE2 lentiviruses were filtered through a 0.45 μm filter and used to transduce H23 cells using reverse transduction. Briefly, filtered virus particles are added to the H23 cell suspension with RPMI 1640 (Gibco Cat# 1185093) media supplemented with 10% FBS and 8 μg/mL polybrene (Sigma Cat# TR-1003-G). 72 h post transduction, media was changed to RPMI 1640 supplemented with 10% FBS and 4 μg/mL blasticidin S hydrochloride- (Gibco Cat# R21001). Cells were expanded in increasingly larger cell culture plates and ACE2 expression was confirmed by infecting with SARS-CoV-2 (strain Canada/ON/VIDO-01/2020) and flow cytometry. Single clone isolation from the H23-ACE2 cell pool was carried out by the array dilution method in 96-well plates. Single clones were collected 2-3 weeks after seeding and expanded in increasingly larger cell culture plates. After successful isolation, cells were maintained with complete media containing 2 μg/mL blasticidin.
4.8. Analysis of Cell Surface ACE2 by Flow Cytometry
Healthy cells were detached from the monolayer using 0.5 mM EDTA in PBS and centrifuged at 1500 rpm for 3 min. The cell pellet was stained for 1 h at 4 °C with primary ACE2 antibody (R&D systems Cat #AF933, used at a concentration of 0.25 μg/106 cells). The cells were then washed twice with flow wash buffer (2% FBS in PBS) and stained with secondary Goat IgG APC conjugated antibody (R&D systems Cat# F0108, at recommended volume of 10 μL/106 cells), 1000x live-dead viability stain (Invitrogen Cat #L34958) and fixed with 2% PFA (diluted in flow wash buffer). The cells were analyzed using a Beckman CytoFLEX Flow Cytometer and the CytoExpert software.
4.9. TCID50 Assays in H23 Cells
H23 or H23-hACE2 cells were seeded at 1.5 × 105 cells in Corning CellBIND 24-well plates and allowed to attach overnight. The next day, SARS-CoV-2 was used to infect the cells at an MOI of 0.01 in the presence of mock treatment (PBS), AZ66 (50 μg/mL), CM398 (100 μg/mL), diphenhydramine (40 μg/mL), lactoferrin from human milk (400 μg/mL), or a combination of diphenhydramine (40 μg/mL) and lactoferrin (400 μg/mL). The TCID50s were performed by diluting 48 h supernatant from the H23 infections across 8 columns of Vero E6 cells in three independent experiments. Five days later the TCID plates were observed by microscopy for CPE. TCID50/mL in the original H23 infection culture supernatant were calculated by the method of Spearman-Kärber. The TCID experiments were carried out in technical triplicate as described above with individual TCID50/mL values and their average and standard deviation shown.
4.10. Inhibitory Concentration and Effective Concentration Calculations
Regardless of whether the assay was cytotoxicity or plaque reduction, CC50 values and EC50 values were calculated using the GraphPad Prism 9 software nonlinear regression module.
4.11. Molecular Docking of Sigma Receptor Ligands
Sigma receptor ligands were docked individually using AutoDock Vina 42
into the ligand binding site of the sigma-1 receptor (PDB 5HK1). The SMILES string of each compound was translated into 3 dimensional coordinates using the NCI/CADD translator (http://cactus.nci.nih.gov/translate/
, accessed on 1 October 2021). AutoDock Tools [41
] assigned hydrogen atoms and calculated atom charges for AutoDock Vina. Atomic coordinates for ligand PD144418 and solvent molecules were extracted from the sigma 1 receptor structure and each compound was docked to the ligand binding site using AutoDock Vina. The top 9 scoring orientations were evaluated by visual inspection with the highest scoring poses reported. PyMol (https://pymol.org/2/
, accessed on 1 October 2021) was used to measure interatomic distances and identify sigma-1 receptor residues implicated in ligand binding.
An atomic model of the human sigma-2 receptor was generated using SWISS-MODEL based on the most similar solved structure, 3-β-hydroxysteroid-Δ8,Δ7-isomerase, also known as Emopamil-Binding Protein (EBP), PDB 6OHT. EBP was solved complexed to an inhibitor, U18666A, and provided the basis for a putative sigma-2 receptor binding site. AutoDock Vina was used for molecular docking simulations of sigma-2 receptor ligands to the modeled human sigma-2 receptor. Figures generated with PyMol.
4.12. Phospholipidosis Assay
Phospholipidosis assay was performed with the acCELLerate GmbH (Hamburg, Germany) InstaCELL® Phospholipidosis assay kit per the manufacturer’s protocol using H23 cells. The cells were seeded in a 96 well plate at a seeding density of 1.5 × 104 cells per well and allowed to grow in an incubator at 37 °C, 5% CO2. After 24 h, cells were treated with 50 μg/mL of AZ66 and 100 for each of CM304, CM398 or SA4503, along with positive (sertraline, 5 µM) and vehicle (DMSO, 0.5%) controls, and incubated for 48 h. At the end of incubation, cells were washed with PBS buffer and stained with PLD staining solution (LysoID/Hoechst phospholipids staining solution) for 30 min. Thereafter the cells were washed with PBS buffer, and fluorescence was measured using a microplate reader (SpectraMax iD3, Molecular Devices LLC) at 540 nm excitation/680 nm emission and normalized against fluorescence at 340 nm excitation/ 480 nm emission.