Skip to main content
NCBI home page
ACS Pharmacology & Translational Science logoLink to ACS Pharmacology & Translational Science
. 2021 Aug 30;4(5):1499–1513. doi: 10.1021/acsptsci.1c00078

Physiologically Based Pharmacokinetic/Pharmacodynamic Model for the Treatment of Dengue Infections Applied to the Broad Spectrum Antiviral Soraphen A

Katharina Rox †,‡,§, Maxi Heyner ∥,⊥, Jana Krull †, Kirsten Harmrolfs #, Valtteri Rinne ∇, Juho Hokkanen ∇, Gemma Perez Vilaro ○, Juana Díez ○, Rolf Müller ‡,#, Andrea Kröger ∥,⊥, Yuichi Sugiyama §,*, Mark Brönstrup †,‡,*
PMCID: PMC8506605  PMID: 34661071

Abstract

graphic file with name pt1c00078_0007.jpg

While a drug treatment is unavailable, the global incidence of Dengue virus (DENV) infections and its associated severe manifestations continues to rise. We report the construction of the first physiologically based pharmacokinetic/pharmacodynamic (PBPK/PD) model that predicts viremia levels in relevant target organs based on preclinical data with the broad spectrum antiviral soraphen A (SorA), an inhibitor of the host cell target acetyl-CoA-carboxylase. SorA was highly effective against DENV in vitro (EC50 = 4.7 nM) and showed in vivo efficacy by inducing a significant reduction of viral load in the spleen and liver of IFNAR–/– mice infected with DENV-2. PBPK/PD predictions for SorA matched well with the experimental infection data. Transfer to a human PBPK/PD model for DENV to mimic a clinical scenario predicted a reduction in viremia by more than one log10 unit for an intravenous infusion regimen of SorA. The PBPK/PD model is applicable to any DENV drug lead and, thus, represents a valuable tool to accelerate and facilitate DENV drug discovery and development.

Keywords: Dengue virus, soraphen, simulation, PBPK/PD, natural products, antivirals

Infections with the Dengue virus (DENV) are mosquito-borne diseases transmitted by Aedes aegypti and Aedes albopictus as the most important vectors endemic in subtropical and tropical regions. More than 390 million people are infected by DENV each year, and about 100 million people show clinical symptoms.1,2 Four different serotypes of DENV can cause a variety of disease manifestations ranging from asymptomatic over febrile to life-threatening such as Dengue hemorrhagic fever (DHF) or Dengue shock syndrome (DSS). In addition, the socioeconomic burdens of the disease are massive.3 Recently, a vaccine has been launched, but investigations on its clinical benefit for various patient populations are still ongoing.4−6 So far, no specific antiviral drug is available; severe infections are only treated symptomatically, e.g., with paracetamol and measures to restore the hydration balance. However, mortality rates for DHF and DSS are still around 20% and can be reduced to 1% upon intensive care treatment.7−10 Various small molecules, mainly nucleoside or protease inhibitors or repurposed drugs such as ivermectin or chloroquine, have proceeded to early clinical trials, but they could not meet the primary end point.11 Because there is clear evidence that higher viremia results in more severe disease and vice versa,12−14 a treatment that decreases DENV serum levels and a subsequent overreaction of the immune system represents a currently unmet medical need.

Soraphen A (SorA), a macrolide natural product produced by the myxobacterium Sorangium cellulosum, has been recently shown to have anticancer, immune-modulatory, and antiviral activities.15−19 SorA is a potent inhibitor of acetyl-CoA carboxylase (ACC),20 thereby interfering with fatty acid metabolism. The validity of ACC inhibition as a host-directed antiviral target has been shown with SorA in cellular models of hepatitis C virus (HCV) and human immunodeficiency virus (HIV).21−24 Synthetic inhibitors of the catalytic site of ACC have been recently shown to inhibit Flaviviridae like DENV, West Nile virus, or Zika virus in cellular models with low micromolar potency.25,26 As SorA exerts a superior inhibitory potency against ACC,20 we probed its efficacy against DENV in this study. In order to assess the potential of SorA beyond a tool compound as a drug, we report the in vitro absorption, distribution, metabolism, and excretion (ADME) properties and the in vivo pharmacokinetic (PK) studies, followed by its efficacy in a DENV animal infection model. The experimental data were used to develop and validate a physiologically based pharmacokinetic/pharmacodynamic (PBPK/PD) model for DENV infections in mice and also in humans to mimic the clinical scenario. While existing DENV PD models have been constructed as one-compartment models,27−29 the model presented here takes the whole physiology of the body during infections into account. To the best of our knowledge, the PBPK/PD model is the first that predicts the efficacy of DENV treatments using different dosing schemes.

Results

Soraphen A Is Effective against DENV in In Vitro and In Vivo Infection Experiments

We first tested the effect of SorA (Figure 1a) against the replication of DENV in the human hepatocyte-derived cell line Huh7 infected with DV2-Luc virus, a Dengue virus carrying the reporter gene of Renilla Luciferase, and wild-type DENV-2. In the two assays, ascending concentrations of SorA led to EC50’s of 4.7 nM as quantified by bioluminescence (Figure 1b) and 2.2 nM as quantified by RT-PCR (Figure S1), respectively. Cellular cytotoxicity was observed at a CC50 value of 62 μM, which gives a selectivity index of more than 13.000. The data demonstrate that SorA exerts a very potent and selective activity against DENV.

Figure 1.

Figure 1

Open in a new tab

Molecular structure, antiviral efficacy, and pharmacokinetic profiles of soraphen A. (a) Molecular structure of SorA. (b) Antiviral activity of SorA against DV2-Luc virus. Closed circles represent viral infectivity as detected by chemiluminescence activity, and crosses depict viability as detected by cellular ATP content. The data displayed in the graph are expressed as mean values of triplicates (±SEM) related to the solvent (DMSO). (c) SorA plasma levels in mice after a single subcutaneous dose of 10 mg/kg. (d) SorA plasma levels in mice after multiple dosing.

To assess the suitability of SorA as a drug and as a tool for the validation of a PBPK/PD model, we first tested its ADME properties: SorA had a rather low aqueous solubility at pH 7 of around 200 ng/mL (Table 1). The calculated pKa values of around 10.6 and 11.4 suggest that SorA is mainly undissociated at pH 7 and, thus, transfers biomembranes. In vitro metabolic stability data from SorA using mouse and human liver microsomes gave an in vitro intrinsic clearance of 921 μL/min/mg protein. The plasma protein binding in mouse was 86%, but much higher in humans with around 99%. Details on the metabolite identification as well as on CYP enzyme inhibition can be found in the Supporting Information (Figure S2, Tables S1–S3). Moreover, a (reversible) tautomerization of SorA from its lactol form to a ring-opened δ-hydroxy-ketone form is observed in mouse plasma but not in human plasma (Figure S3).

Table 1. ADME Properties and Molecular Properties for SorA.

parameter value
molecular weight [g/mol] 520.63
pKa value 0 (in vitro calculation) 10.60
pKa value 1 (in vitro calculation) 11.43
lipophilicity (octanol/water partition coefficient) 1.45
plasma protein binding [%] 86 (mouse), 98.2 (human)
total hepatic clearance (Clint,inc) [μL/min/kg] 921
solubility at pH 7 [ng/mL] 200
Open in a new tab

Given the high in vitro potency of SorA and a formulation which partially overcomes solubility issues suggested by in vitro data,17,19 we conducted pharmacokinetic (PK) studies in CD-1 mice to find the optimal dose and treatment regimen to maximize the time over EC50. Following intravenous (IV) administration of 10 mg/kg SorA, an initial concentration C0 of around 1.1 μg/mL was determined with a volume of distribution of 8.8 L/kg (Table 2). This indicates distribution not only to the blood system, but also to tissue which is crucial for a DENV infection as the virus is also located within tissues and cells.30 Taking into account the high in vitro potency and despite the metabolic lability (Cl = 386.3 mL/kg/min) and the short half-life of 0.26 h, subcutaneous (SC) route of administration yielded in plasma concentrations of >10 ng/mL for up to 8 h (Figure 1c), demonstrating that the SC route is suited to ensure sustained plasma levels of SorA above the EC50.

Table 2. PK Parameters of Soraphen A after Administration of 10 mg/kg IV to CD-1 Mice.

parameter value
C0 [ng/mL] 1155.10 ± 144.1
Vobs [L/kg] 8.76 ± 1.2
t1/2 [h] 0.26 ± 0.0
AUC [ng/mL × h] 438.13 ± 63.6
Clobs [mL/kg/min] 386.32 ± 61.1
Open in a new tab

For an in vivo infection experiment, we hypothesized that SorA plasma levels had to be constantly kept above the EC50 for several days. We discarded the IV route, as it is technically challenging, especially for repeated administrations. Instead, three administrations per day were given in form of an initial dose of 20 mg/kg intraperitoneal (IP) to achieve high plasma levels, followed by two SC doses of 25 mg/kg to ensure sustained plasma levels. A multiple dose PK study proved that this dosing scheme indeed led to plasma levels above the EC50 over 24 h (Figure 1d).

Next, we reached out to test SorA in an in vivo murine DENV infection model. For this purpose, IFNAR–/– mice on a C57Bl6/J background were infected with 107 pfu/ml of DENV-2 new Guinea C strain and treated with either SorA or vehicle, applying the dose regimen established in the multiple dose PK study. At 72 h after infection, mice were sacrificed, and viral RNA loads were determined by quantitative RT-PCR in spleen and liver, as these two organs are most affected by a DENV infection.31 Whereas a significant reduction of viral load by approximately one log10 unit was observed in the liver compared to vehicle-treatment, the viral load in spleen decreased significantly by approximately two-log10 units (Figure 2). Thus, we demonstrate that SorA is effective against DENV in vitro and in an in vivo murine infection model.

Figure 2.

Figure 2

Open in a new tab

Soraphen A shows efficacy against DENV in vivo. 6–8-week old female and male IFNAR–/– on a C57BL/6J background mice (n = 9 for SorA group and n = 8 for vehicle group) were infected with DENV (107 pfu/ml) intraperitoneally. RT-PCR was performed and the fold change of DENV-RNA was determined for liver (a) and spleen (b) compared to vehicle. Statistical analysis was done using a two-tailed Mann–Whitney test using GraphPad Prism 8.4.2 software. *p = 0.0274, ***p = 0.0002.

Development of a PBPK/PD Model to Mimic In Vivo Infection Experiments In Silico

Based on the encouraging results from the in vivo experiment, we were interested whether a further reduction of viral load in spleen and liver and also in other organs was possible through different dosing regimens. Moreover, we were curious to test if an in silico PBPK/PD model was suitable to predict efficacy which was seen in vivo. Moreover, in case predictions match well, this approach can be used to minimize animal experimentation with highly pathogenic viruses. First, a PBPK model was constructed for SorA with the programs PK-Sim and Mobi,32 using input parameters from the in vitro ADME assays such as plasma protein binding or metabolic stability, structural information of SorA such as molecular weight and number of halogens present in the molecule, the in vivo dosing regimen as well as physiologic parameters of the animals (Table S7). A 10 mg/kg IV dose was simulated for SorA using in vivo PK data points for parameter optimization (Figure 3a). To incorporate the subcutaneous route of administration to the PBPK model, a compartment called subcutaneous fat tissue (SC-fat) was added. The SC-fat compartment was further divided into subcompartments named interstitial, intracellular, blood cells, and plasma. For physiological characteristics such as blood flow rates, lipoprotein ratio (tissue/plasma), or albumin ratio (tissue/plasma), the same parameters as used for ordinary fat tissue were applied. In order to account for transport of compounds within SC-fat, connections were defined between plasma subcompartment and blood cell subcompartment, intracellular subcompartment and interstitial subcompartment, as well as interstitial subcompartment and plasma subcompartment (Figure S4). Moreover, connections were made between the plasma subcompartments of SC-fat, venous-blood, and arterial-blood plasma compartment and the respective blood cell subcompartments to ensure exchange within blood circulation of the whole body. Then, the process of tautomerization of SorA in tissue as well as in plasma was added to the model as a dynamic equilibrium. In addition, passive transport processes with first-order kinetics for transport of SorA from SC-fat plasma to venous-blood plasma and from SC-fat interstitial to SC-fat plasma were initially set and then further optimized by using the in vivo PK data for the SC route. In both cases, parameter optimization was performed using the Levenberg–Marquardt algorithm (Table S8). As SorA is formulated as an oil-in-water (O/W) emulsion with a half-life of around 60 min, we also added this information to the administration scheme. To mimic the multiple dose administration of the in vivo experiment, 38 mg/kg IV was dosed into the venous blood compartment (corresponding to 25 mg/kg IP taking into account the lower bioavailability for IP compared to IV and similar kinetics) and then 25 mg/kg SC was dosed twice into the SC-fat compartment. The initial faster decrease of SorA in the simulation compared to the observed data is explained by the two different routes of administration used for the simulation and for the actual in vivo multiple dose PK study: whereas the IV route was used for the simulation, the IP route with the same formulation accounting for a slight retention in the peritoneum was used during the in vivo multiple dose administration study. With this model, a very good match of the simulated multiple dose curve with the experimental in vivo data points for the same administration scheme could be obtained (Figure 3b).

Figure 3.

Figure 3

Open in a new tab

PBPK modeling of soraphen A matches with in vivo PK data. A PBPK model was built for SorA using the physiology of mice. (a) The simulated concentrations after a single IV dose at 10 mg/kg (blue line) match the measured in vivo PK data for three individual animals (red, green and purple line) after single IV 10 mg/kg administration. (b) A multiple dose administration of SorA was simulated using 25 mg/kg IV at t = 0 h, followed by 25 mg/kg SC at t = 0.75 h and t = 8 h (blue line). Measured in vivo PK data using the same administration scheme are plotted (red, green and purple line) and match the simulated results.

In the next step, a PBPK/PD model was constructed to link the dose to PD effects. For this purpose, the PD model was built using information about the virus and SorA. Constants for infection rate, viral clearance rate or viral production rate were taken from the literature (Table 3).27−29 The PD model was based on the physiology of the animal using several compartments involved in production of viral particles. As SorA-specific data, we chose the experimental EC50 value of 4.7 nM for the modeling, as it was slightly higher than the one with wild-type DENV. The information that SorA acts on the viral production, but neither on the release nor on the entry process of viral particles into the cells, based on mechanistic data for HCV and HIV,21,24 was later incorporated into a combined equation for viral release, infection and production. Although DENV enters the human body through a mosquito bite, the virus is directly injected into the system in our models. We assumed that infection rates will not be constant for every affected organ, but will highly depend upon the velocity of transport of the virus to the cells of the respective organ. Transport of the virus is kinetically determined by the blood flow rate (Bfr). In mice, muscle has the highest blood flow rate compared to all other organs. To account for the different infection rates of susceptible organs, such as brain, liver, lung, kidney, spleen, and heart, we formed a ratio for the blood flow rate of the respective organ with the highest possible blood flow rate from muscle. An illustration of the interplay of all the factors is displayed in Figure 4. DENV particles from the bloodstream infect the organs brain, liver, lung, heart, kidney, and spleen in which DENV proliferates (black arrows). Different SorA levels in the respective organs influence proliferation and release of DENV particles to the bloodstream (red arrows). Moreover, the immune response helps with viral clearance in the cells. Yellow arrows show the amount of DENV particles that are released from the organs into the bloodstream.

Table 3. Constants Used for PBPK/PD Modeling Adapted from Clapham et al. and Ben-Shachar et al.(27−29).

constant value
kcellclear 4.17 × 10–5 1/h
kcellinfect 3.85 × 10–8 1/h
kcellinfectaugment 4.17 × 10–4 1/h
kclear 6 × 10–3 1/h
kclearaugment 2 × 10–3 1/h
kinfectimmun (only for human PBPK/PD model) 4.16 × 10–8 1/h
kprodvir 4170 1/h
Vir (for mouse PBPK/PD model) 1 × 107 (count)
Vir (for human PBPK/PD model) 5000 (count)
Open in a new tab

Figure 4.

Figure 4

Open in a new tab

PBPK/PD model for DENV and SorA. A PBPK/PD model was built for SorA using the physiology of the animal knockout mice and the human physiology. The interplay and links of the different main compartments within the PBPK/PD model are shown. DENV particles (yellow circles) infect the bloodstream (red). DENV particles are distributed via the bloodstream and infect the organs brain, lung, kidney, spleen, liver, and heart with the cell infection rate constant kcellinfect and at a later time point with the cell infection rate constant kcellinfect_augment as described in eqs 1–7 in the main text for the mouse PBPK/PD model and in eqs 22–28 for the human PBPK/PD model. Black arrows point to the organs. DENV particles are produced with the production rate constant kprodvir. SorA inhibits the process of production/release in the organs brain, lung, kidney, spleen, liver, and heart, marked by red arrows. These processes are described in eqs 8–14 (mouse PBPK/PD model) and in eqs 29–35 (human PBPK/PD model). Yellow arrows point to the viral particles released from the organs back to the bloodstream. The immune system is displayed as a blue star and is represented via the clearance rate constant kcellclear which determines the amount of viral particles cleared in the cells by the immune system in the mouse PBPK/PD model. These processes are described by eqs 15–21 in the main text for the mouse PBPK/PD model. For the human PBPK/PD model, the clearance rate constants kclear, kclear_augment, and kinfectimmun are added (dotted blue arrows) which characterize the clearance of viral particles by the human immune system. These processes are described by eqs 36–42 in the main text.

In detail, eqs 1–7 determine the amount of virus infecting cells over time (a definition of variables in these and all subsequent equations is given in the Methods section):

graphic file with name pt1c00078_m001.jpg 1
graphic file with name pt1c00078_m002.jpg 2
graphic file with name pt1c00078_m003.jpg 3
graphic file with name pt1c00078_m004.jpg 4
graphic file with name pt1c00078_m005.jpg 5
graphic file with name pt1c00078_m006.jpg 6
graphic file with name pt1c00078_m007.jpg 7

After infection of the cells in the different organs, virus will be produced and then released again into plasma. These two processes are described with eqs 8–14 for the organs blood, brain, liver, lung, kidney, heart, and spleen. As SorA has an effect on the production of viral particles, its organ concentration within the aforementioned organs is incorporated in eqs 8–14. The compartment-specific localization of DENV in organs is well-known from human pathology studies.30,33−36 Thus, for the respective organ, SorA concentrations in that respective compartment are taken into account (Table 4). The cellular viral load in organs after reproduction and ready for release can be described by eqs 8–14:

graphic file with name pt1c00078_m008.jpg 8
graphic file with name pt1c00078_m009.jpg 9
graphic file with name pt1c00078_m010.jpg 10
graphic file with name pt1c00078_m011.jpg 11
graphic file with name pt1c00078_m012.jpg 12
graphic file with name pt1c00078_m013.jpg 13
graphic file with name pt1c00078_m014.jpg 14

As we used IFNAR–/– mice in our model, the immune response of the animals is slightly hampered due to the fact that they lack the type I interferon response. To this extent only the CD4+-T cells, but not CD8+-T cells will respond to the infection. Moreover, the B cell response is hampered. The CD4+ cells mainly target virus located within cells in organs, such as liver or spleen.37 Thus, no equation is incorporated for viral clearance within the blood compartment. As the immune system will not respond immediately, but with a time delay, we reflected this in eqs 15–20. These equations display the amount of virus cleared in the cells during infection. As T helper cells have to get to the organs via the bloodstream, their velocity is dependent upon the blood flow rate of the respective organ.

graphic file with name pt1c00078_m015.jpg 15
graphic file with name pt1c00078_m016.jpg 16
graphic file with name pt1c00078_m017.jpg 17
graphic file with name pt1c00078_m018.jpg 18
graphic file with name pt1c00078_m019.jpg 19
graphic file with name pt1c00078_m020.jpg 20

Finally, eq 21 describes the viral load in plasma. In addition to the amounts of virus released from the different organs, the equation also incorporate a constant for the innate immune response clearing viral particles directly in plasma. This innate immune response mainly consists of macrophages and NK-cells which proliferate with time.

graphic file with name pt1c00078_m021.jpg 21

The PBPK/PD model was constructed based on these four sets of equations (in total 21 equations) that consider the whole physiology and different organs contributing to the production of viral particles. Next, the model served to predict the effects of SorA and vehicle-only treated animals in silico. First, the in silico PBPK/PD model predicted the time-dependent increase of viral load in all infected organs (Figure 5a). According to the predictions especially highly perfused organs such as the blood compartment per se, the kidney compartment and the lung compartment showed the highest viral loads. As we used the blood flow rate as a parameter for the infection rate, this explains why organs with high blood flow rates showed high viral loads in our simulation. Using the same administration scheme for SorA as applied in the experimental in vivo infection experiment, the PBPK/PD model resulted in a reduction of viral load in spleen by approximately two log10 units compared to vehicle-treated only animals (Figure 5b, c). Moreover, one log10 unit reduction of viral load in liver compared to vehicle-treated only animals was calculated. The log10 unit reductions for the lung, blood, heart, and kidney were in the same range as those for spleen, whereas viral load reductions in the brain were in the range as described for the liver compartment. These reductions in viral load are the result of exposure to SorA at the simulated concentrations in the corresponding compartments. The results with respect to viral load in the spleen and in liver were encouraging, as they were in very good agreement with the in vivo efficacy data. In summary, the constructed PBPK/PD model predicts viremia data in mice that correlate well with the in vivo data. Thus, this model can be used to investigate dosing schemes or select compounds in silico prior to conducting DENV challenge experiments in BSL-3 laboratories.

Table 4. Localization of DENV within Organs and Compartments Used for PBPK/PD Modeling.

organ location
spleen blood cells (macrophages)
heart blood cells and intracellularly
liver intracellularly
lung blood cells (macrophages) and intracellularly
kidney blood cells (macrophages)
brain intracellularly
blood blood cells (macrophages)
Open in a new tab

Figure 5.

Figure 5

Open in a new tab

PBPK/PD modeling of DENV infection upon treatment with soraphen A in a mouse model organism. (a) Viral loads following DENV infection in the blood, liver, spleen, lung, kidney, brain, and heart of untreated mice. (b) Simulated viral loads following DENV infection in the blood, liver, spleen, lung, kidney, brain, and heart of SorA-treated mice from 0 to 96 h. The pink curve displays the simulated drug concentrations over this time range. (c) Viral loads in liver and spleen of treated and untreated mice 48–96 h post infection. The pink curve displays the simulated drug concentrations over this time range. For liver, about one log difference upon treatment is calculated, whereas for spleen slightly more than 2log reduction in viral load is calculated. (a)–(c) Viral load in brain (black), blood (gray), liver (green), spleen (purple), lung (red), heart (blue), and kidney (yellow).

Transfer of the PBPK/PD Model to the Human System to Evaluate Compounds before Starting Clinical Trials

Next, the PBPK/PD model was adapted to the Homo sapiens species to mimic a clinical scenario. The animal physiology was exchanged to the human physiology using the program PK-Sim,32 whereas compound characteristics and reactions were the same as for the animal PBPK model. A significant difference to the mouse PBPK/PD model is that tautomerization of SorA does not take place in human plasma. Thus, the reaction accounting for tautomerization was not included for the human PBPK/PD model. Organism-specific compound characteristics, such as plasma protein binding or clearance, were replaced accordingly (Table 1). We assumed cell infection rates similarly to the mouse PBPK/PD model. To account for differences in cell infection rates of the susceptible organs, we formed a ratio for the blood flow rate of the respective organ with the highest possible blood flow rate from the portal vein in the human PBPK/PD model, in analogy to the mouse PBPK/PD model. Similarly to the mouse PBPK/PD model, DENV particles from the bloodstream infect the organs brain, liver, lung, heart, kidney in spleen in which DENV proliferates (Figure 4). Again, different SorA levels in the respective organs influence the process of proliferation and release of DENV particles to the bloodstream. However, in contrast to the mouse PBPK/PD model, the immune response helps in viral clearance in the cells, but also in the bloodstream. This process (dotted blue arrow in Figure 4) accounts for a substantial difference in the human immune system compared to the one in the animal knockout mice used for the in vivo infection experiment.

In detail, cell infection processes are characterized by eqs 22–28:

graphic file with name pt1c00078_m022.jpg 22
graphic file with name pt1c00078_m023.jpg 23
graphic file with name pt1c00078_m024.jpg 24
graphic file with name pt1c00078_m025.jpg 25
graphic file with name pt1c00078_m026.jpg 26
graphic file with name pt1c00078_m027.jpg 27
graphic file with name pt1c00078_m028.jpg 28

Furthermore, we used the same constants for the viral production and release capacity as for the mouse PBPK/PD model.27−29 This is reflected in eqs 29–35:

graphic file with name pt1c00078_m029.jpg 29
graphic file with name pt1c00078_m030.jpg 30
graphic file with name pt1c00078_m031.jpg 31
graphic file with name pt1c00078_m032.jpg 32
graphic file with name pt1c00078_m033.jpg 33
graphic file with name pt1c00078_m034.jpg 34
graphic file with name pt1c00078_m035.jpg 35

The human immune system differs substantially from the one in the animal knockout mice that were used for the PD efficacy trial. In the former, especially T helper cell responses as well as interferon responses, leading to an increased recruitment of macrophages and killer cells, are fully functional. It is described that the human immune response starts approximately 24 h after the first release of DENV from cells. As it is assumed that DENV initially needs around 24 h to be released from cells after infection, the human immune system will respond around 48 h after infection with DENV.38 Therefore, we added eqs 36–41 to account for the human immune response against DENV. Similar to the mouse PBPK/PD model, the equation was not applied to blood as an organ, as virus is not cleared via the T helper cell response in blood cells.

graphic file with name pt1c00078_m036.jpg 36
graphic file with name pt1c00078_m037.jpg 37
graphic file with name pt1c00078_m038.jpg 38
graphic file with name pt1c00078_m039.jpg 39
graphic file with name pt1c00078_m040.jpg 40
graphic file with name pt1c00078_m041.jpg 41

Finally, eq 42 describes the viral load in human plasma. In addition to the amounts of virus released from the different organs, the equation also incorporates a constant for the innate immune response clearing viral particles directly in plasma–similarly to the mouse PBPK/PD model. Moreover, we added a constant taking into account that the increased interferon levels will lead to an increased recruitment of macrophages over time. Viral clearance by the innate immune response was set to start approximately 5 days after first contact with DENV, as the innate immune system will react with a sufficient response when already high levels of DENV are sensed in plasma.14,39,40 Moreover, it is described that clearance by macrophages will augment due to an increased interferon response approximately 24 h after first contact between macrophages and DENV.38,41 Finally, we assumed that the viral load during the mosquito bite is around 5000 viral particles (=Vir).42−44

graphic file with name pt1c00078_m042.jpg 42

After construction of the human PBPK/PD model with eqs 22–42, the plasma levels of an untreated, DENV-infected individual were simulated. The model shows that plasma levels peak at around day 10 before being cleared by the immune response (Figure 6a). Remarkably, the simulation data show similar curve shapes and peak heights as described from clinical studies: patients are frequently enrolled in clinical studies when they already show symptoms such as fever. At this time point, viremia is measured daily. Based on the viremia level, clinical studies calculate back to the assumed time point of the mosquito bite and encountered that viremia levels always peak between day 9 and 10.45−47 The good match between predicted and clinically observed viremia gave us confidence to test different dosing schemes to model the potential effect of SorA in the human organism.

Figure 6.

Figure 6

Open in a new tab

PBPK/PD modeling of DENV infection with different treatment regimens for soraphen A in a human model organism to mimic the clinical scenario. DENV infection in humans was simulated using a PBPK/PD model. Calculated viral loads in the plasma are displayed (red curves, a–d) as well as SorA concentrations (blue curves, b–d) upon different treatment regimens: (a) Viral load simulation in untreated humans. (b) Viral load simulation in humans treated with SorA starting at day 0. SorA was administered at 2 mg/kg IV followed by two 2 mg/kg SC doses similar to the mouse PBPK/model. (c) Viral load simulation in humans treated with SorA starting at day 3. SorA was administered at 2 mg/kg IV followed by two 2 mg/kg SC doses similar to the mouse PBPK/model. (d) Viral load simulation in humans treated with an infusion therapy with SorA starting at day 3. (e) Impact of the three treatment regimens on viremia compared to untreated. All treatments lead to an earlier time point for peak viremia levels as well as lower viremia levels compared to untreated. (a)–(d) Viral load in plasma is depicted in red. SorA concentrations are shown in blue. (e) Viral loads: untreated, green (corresponding to (a)); early IV/SC scheme, red (corresponding to (b)); late IV/SC scheme (corresponding to (c)), blue; late infusion (corresponding to (d)), purple.

First, the effect of SorA using the dosing regimen applied for the in vivo animal model (1× IV followed by 2× SC) was probed with a human equivalent dose of 2 mg/kg. It was simulated that SorA reduced viral load in plasma by more than one log10 unit (Figure 6b). In clinical practice, therapy will not start directly after a mosquito bite, but patients will present at the hospital after first symptoms of a DENV infection occur, which is 3 days after the mosquito bite at the earliest. Using the in vivo administration scheme reported above (1× IV followed by 2× SC) starting at day 3 after the mosquito bite, viremia levels were only reduced by nearly one log10 unit (Figure 6c). We then examined the effects of SorA on viremia upon a 60 min IV infusion every 8 h, starting on day 3 after the mosquito bite. In this scenario, a reduction in viremia by just 0.5log10 units was predicted. Thus, infusion therapy at a dose of 2 mg/kg did not substantially reduce viremia at a later onset of therapy. Nevertheless, viremia did not increase as rapidly as before onset of therapy: In a scenario of direct treatment (1× IV followed by 2× SC), viremia levels augmented with a delay of about 2–3 days, whereas late onset of therapy still led to a delayed increase of about 1 day (Figure 6d,e).

Discussion

Viremia is an important predictor for severe clinical outcomes of DENV infections, as a high viral load in plasma is associated with a strong immune response triggered by interleukins like IL-10 and interferons, that causes tissue damage such as vascular leakage.41,48,49 Therefore, we selected viremia in plasma or in cellular supernatants as a readout of efficacy for experimental and modeled treatments. We found that SorA inhibited DENV proliferation in a cellular assay with high potency and an EC50 in the low nanomolar range. This observation can be rationalized with the targeting of ACC of the host cell by SorA, a mechanism that also enabled the highly potent inhibition of HCV and HIV proliferation.21,50 The dependence of DENV proliferation on ACC has been demonstrated before by synthetic compounds, although their inhibitory effect has been markedly (ca. 50-fold) weaker.26 It appears that the unique molecular mechanism of action of SorA, preventing oligomerization of ACC rather than inhibiting the catalytic site, might be responsible for the efficacy. While the specific binding site has been exploited to develop synthetic ACC inhibitors against other diseases like hepatosteatitis or diabetes,51,52 the natural product SorA itself has been frequently applied as a tool compound but rarely used in animals. In this study, we report detailed information on the ADME properties of SorA that disclose its potential as a systemic drug. Hepatic metabolism and a lactol-hydroxyketone tautomerism in vivo contribute to limited maximal concentrations (Cmax = 1.1 μg/mL after 10 mg/kg IV) and a moderate half-life of 0.26 h in mice. In addition, SorA is prone to metabolism and is likely to inhibit two important CYP enzymes (CYP3A4 and CYP2C19) at concentrations above 1 μM. On the other hand, the high cellular potency of SorA assures that drug levels above the EC50 can be maintained with an appropriate repeated dosing regimen. This encouraged us to optimize the formulation of SorA compared to initial studies17 and to undertake in vivo infection experiments. We provide evidence that the treatment of DENV-infected mice with SorA results in significant reduction of viral load in liver and spleen. In this context, it is noteworthy that the synthetic inhibitor PF-05175157 showed efficacy in mice infected with the West Nile Virus, a related flavivirus.25,26 Together, these findings support the concept of ACC inhibition as a host-directed, broad spectrum antiviral therapy.

Several drugs have failed to meet the primary end point in clinical trials of DENV treatment.11 In order to assess the potential of anti-DENV compounds like SorA already at an early lead stage, we constructed a physiologically based pharmacokinetic (PBPK) model of SorA using the mouse physiology and validated it with pharmacokinetic in vivo data. PBPK models can serve to predict compound levels in humans, although it has to be emphasized that also human PBPK models have to be validated with in vivo data. Several groups have built PD models of DENV to describe viral load in plasma in humans during DENV infection.27−29,53 None of these PD models have been linked to the human physiology, nor have they been used to assess the potential of a small molecule inhibitor against DENV. The PBPK/PD model presented here establishes the link between physiology and DENV infection, based on PD modeling data from the literature. First, the model was constructed for the mouse physiology and validated with in vivo data with DENV infection and SorA treatment. Our mouse PBPK/PD data predicted a reduction in viral load in liver and spleen upon SorA treatment, which was indeed observed in the same range in the in vivo infection experiment. This encouraged us to extend the PBPK/PD model to the human physiology, taking into account that the human immune response differs from that of mice. Data of viremia level from clinical trials from different regions worldwide show that viremia peaks at day 8–10 after the assumed mosquito bite and that viremia level peak at 108–1010 copies/mL before viral clearance by the immune system takes place.12,14,39,45,47,54 A comparison of the viral plasma levels predicted by our PBPK/PD model with data of viremia levels of these clinical trials demonstrated that the dynamics of viremia in terms of day of the viremia peak, peak height, and viral clearance were well predicted and reflected by the model. A limitation of the current model is that infection rates of different tissues are proportional to blood flow rates but do not reflect the preference of DENV entry for specific cell types in vivo like dermal dendritic cells or macrophages, hepatocytes, or splenic endothelial cells.30,55 However, given that DENV uses a variety of entry mechanisms, and that quantitative information on cellular entry rates is not available, such a mechanism-driven refinement of the model is not straightforward to implement, albeit highly desirable.

In a next step, the efficacy of a SorA treatment of a human DENV infection was modeled. The transfer of the mouse PBPK model to the human PBPK model enabled to explore different dosing scenarios as well as different scenarios for the onset of therapy. As clinical trials with SorA have not been conducted yet, the human PBPK prediction data cannot be validated at this point. The model predicts efficacy of SorA in terms of viral load reduction in plasma. But to achieve a substantial decrease of viral loads, high dose administrations of SorA that start at an early stage of infection are required. However, given the mild or asymptomatic time course of most DENV infections, patients get hospitalized at a late stage of infection in clinical practice. Under such conditions, frequent intermittent infusions of SorA would be required to exert such an effect. The guidelines of the World Health Organization (WHO) for Diagnosis, Treatment, Prevention and Control of Dengue require an effective antiviral against dengue to substantially lower the viremia levels even at a late presentation of the patients at hospital.8 Moreover, a clinical study with balapiravir failed, as neither viremia nor the median time to the viremia level of 1000 copies/mL was reduced.56 Our calculations predict that SorA can in principle exert an effect even at a later onset of therapy with respect to both of these end points, although higher doses than a human equivalent dose of 2 mg/kg of SorA would be needed to achieve a substantial decrease. Taking the pronounced metabolic lability, the inhibition of CYP3A4, the significant but moderate efficacy in the mouse model, and the predicted efficacy in the human PBPK/PD model together, our combined experimental and modeling study suggests that it is advisable to find derivatives of SorA with even better pharmacokinetic properties.

We argue that the PBPK/PD model for DENV infection facilitates the decision of whether lead compounds are ready for development in terms of efficacy or require further optimization. Such a tool is helpful for planning clinical development programs, as clinical trials for diseases such as dengue fever are associated with difficulties in the timing of patient enrollment and viremia measurements and large variations in the rate of viral clearance.11

In summary, we provide the first data set of PBPK/PD models that demonstrates the potential for drug development based on the compound SorA for which we provide in-depth ADME and PK characterization as well as an in vivo proof-of-concept. By combining in vitro and in vivo data, we constructed a solid PBPK/PD model for mouse and human DENV infections that allows the prediction of treatment efficacy and to mimic a clinical scenario. The tool has been successfully applied using SorA data but can be readily adapted to profile other anti-DENV drugs in the future.

Methods

Test Agents and Virus

SorA was isolated as described in Gerth et al.16 Dengue virus (DENV) serotype 2 laboratory adapted New Guinea C (NGC) wild-type (WT) strain (Genbank Accession Number KM204118.1) was kindly provided by Dr. V. Deubel (Institut Pasteur, Paris, France). Dengue virus carrying Renilla Luciferase reporter (DENV-R2A) was kindly provided by Dr. R. Bartenschlager (Heidelberg University, Heidelberg, Germany).

ADME Assays

Detailed information can be found in the Supporting Information.

EC50 Determination of Soraphen A against Dengue

Huh7/Scr cells were seeded at a density of 1.2 × 104 cells/well in 96-well plates. One day later, cells were preincubated for 1 h at 37 °C with SorA at final concentrations ranging from 100 to 0.001 μM. Then, medium was removed and replaced with medium containing DENV-R2A virus (a dengue virus derivative carrying the reporter gene of Renilla Luciferase) and the respective concentration of SorA. After 4 h, medium was removed and replaced with fresh medium containing the respective concentration of SorA. Cells were kept at 37 °C for 72 h. To assess effects on viral infectivity, Renilla Luciferase expression was quantified with the Renilla Luciferase Assay System (Promega Corporation, Madison, WI). To assess effects on cell viability, ATP levels were measured with the CellTiter-Glo Luminescent Cell Viability Assay (Promega Corporation, Madison, WI). Luminescence was measured using a plate luminometer (FLUOstar OPTIMA, BMG LABTECH, Ortenberg, Germany). Mean relative light units (RLU) were plotted as percentage relative to control infections (solvent without SorA) for both infectivity and cell viability. Infections were carried out in triplicates (mean ± SEM; n = 3). Half maximal effective concentration (EC50) and half maximal cytotoxic concentration 50 (CC50) were estimated by nonlinear regression of log inhibitor vs normalized response and used to calculate the Selectivity Index (SI) value.

Pharmacokinetics (PK) Study

SorA was dissolved in 20% olive oil and in 80% of a mixture of lecithin and water. Mice were administered SorA at 10 mg/kg intravenously (IV) and subcutaneously (SC). In addition, SorA was administered at 25 mg/kg IV at t = 0 h and at 25 mg/kg SC at t = 0.25 and 8 h. About 20 μL of whole blood was collected serially from the lateral tail vein at time points t = 0.25, 0.5, 1, 2, 4, and 8 h post administration. For the multiple dose PK study, blood was collected serially at 15 min before and after and 4 h after the third administration and 4 h. After 24 h, mice were sacrificed and blood was collected from the heart. Whole blood was collected into Eppendorf tubes coated with 0.5 M EDTA and immediately spun down at 13 000 rpm for 10 min at 4 °C. The plasma was transferred into a new Eppendorf tube and then stored at −80 °C until analysis.

PK Sample Preparation and Analysis

Detailed information can be found in the Supporting Information. PK parameters were determined using a noncompartmental analysis with PKSolver.57

DENV Propagation for In Vivo Experiment

A total of 4.8 × 107 C6/36 cells (Aedes albopictus, ATCC CRL-1660) were seeded into a 300 cm2 cell culture flask and incubated for 48 h at 28 °C and 5% CO2. Then cells were infected with DENV at MOI of 0.1 in DMEM (Gibco) for 1 h at 33 °C and 5% CO2. The flask was tapped every 15 min to ensure equal infection. After 1 h, 48 mL of infection media was added. Five days post infection, the supernatant was harvested by first performing a debris-clearance at 1000g for 10 min followed by a sucrose cushion. For the sucrose cushion, 30 mL of supernatant containing the virus was underlaid with approximately 8 mL of 30% sucrose in HNE buffer and centrifuged in an ultracentrifuge at 130 000g and 4 °C for 3.5 h. During the sucrose cushion, the virus pelletized, so that the supernatant could be discarded and the purified virus could be resuspended in 500 μL of HNE buffer (5 mM HEPES, 150 mM NaCl, 100 μM EDTA, pH 7.4). The virus was stored at −80 °C, and the titer was determined via plaque assay.

Plaque Assay

A 12-well cell culture plate was seeded with 2.5 × 105 Vero-B4 cells (DSMZ no.: ACC 33) per well in 1.5 mL of DMEM, high glucose (Gibco) with GlutaMAX supplement and 10% FBS and incubated for 24 h at 37 °C and 10% CO2. A dilution series of the virus stock was prepared in a separate 96-well plate, and dilutions from 10–4 to 10–9 were used (to get final concentrations of 10–5–10–10) for the plaque forming unit (pfu) determination. After removing 500 μL from all wells of the cell culture plate, each dilution of 100 μL was transferred in replicates to the cells and incubated for 24 h at 37 °C and 10% CO2. The monolayer was washed with PBS before 1 mL of plaque media (DMEM, high glucose, GlutaMAX supplement (Gibco), 2% FBS (Gibco), 0.4% agarose) was added. Plaques were detected after 5 days at 37 °C and 10% CO2 by incubation with 100 μL MTT (5 mg/mL in PBS) for 3 h. The titer of the stock was determined as 3.5 × 109 pfu/ml.

Dengue Virus Infection Model

IFNAR–/– mice on a C57BL/6 background were bred under SPF conditions at the Helmholtz Centre for Infection Research in Braunschweig, Germany. The 6–8 week old mice (n = 10 per group) were anesthetized with ketamine (100 mg/kg body weight) supplemented with xylazine (10 mg/kg body weight) and infected intraperitoneally (IP) with 106 plaque forming units (PFU) of DENV-2 New Guinea C in 100 μL of 1× PBS. SorA was dissolved in 20% olive oil and in 80% of a mixture of lecithin and water. SorA was administered IP at 0.5 h, 24.5, and 48.5 h post infection at 25 mg/kg; and SC at 1.25, 8.5, 25.25, 32.5, 49.25, and 56.5 h post infection at 20 mg/kg. Animals were sacrificed 72 h postinfection (n = 9 for SorA group and n = 8 for vehicle-treated group). Spleen and liver samples were obtained and processed with Trizol reagent. Viral load in the tissue samples was quantified by qRT-PCR as indicated below. Statistical analysis was performed using the Mann–Whitney test and GraphPad Prism 8.4.2 software.

RNA Preparation from Tissue Samples

Tissue samples up to 100 mg were homogenized in 1 mL of Trizol [Life Technologies] by using a 1/4″ ceramic sphere with a FastPrep-24 homogenizer. RNA in the Trizol-aqueous phase was processed according to the manufacturer and further purified by additional phenol/chloroform/isoamylalcohol (25:24:1, pH 4) and chloroform/isoamylalcohol (24:1) extraction. Finally, isopropanol precipitated RNA was washed with 70% ethanol and resolved in RNase-free H2O. All purified RNA was quantified by measuring absorbance at 260 nm.

Quantitative RT-PCR

An amount of 2.5 μg of purified RNA was reversely transcribed with RevertAid Reverse Transcriptase (ThermoFisher) according to the manufacturer’s protocol with reverse primers specific for DENV-2 New Guinea C or GAPDH (Table S6). For quantitative PCR, 20 ng/μL of reversely transcribed RNA was used with the LightCycler480 Probes Master Mix (Roche), DENV-2 and GAPDH forward and reverse primers, as well as the hydrolysis probe according to the manufacturer’s protocol (Table S6). PCR was performed as follows: 95 °C for 5 min followed by 45 cycles at 95 °C for 10 s, 60 °C for 20 s and 72 °C for 1 s. Viral RNA extracted from animal tissues was quantified via the relative ΔΔCt method. ΔCt values were obtained using mouse GAPDH-mRNA as a reference. Data then was expressed relative to the average ΔCt of untreated samples (% relative to untreated control).

Animal Studies and Ethics Statement

For pharmacokinetic experiments, 4 week old outbred male CD-1 mice (Charles River, Netherlands) were used. For Dengue virus infection experiments, pathogen-free 6–8-week old female and male IFNAR–/– on a C57BL/6J background were used. IFNAR–/– mice were bred under SPF conditions at the Helmholtz Centre for Infection Research in Braunschweig, Germany. All animal experiments were performed in compliance with the German animal welfare law (TierSchG BGBl. S. 1105; 25.05.1998). The mice were housed and handled in accordance with good animal practice as defined by the Federation of Laboratory Animal Science Associations (FELASA). All animal experiments were approved by the Lower Saxony State Office of Consumer Protection and Food Safety. DENV infection experiments were performed in the BSL-3 facility of the Helmholtz Centre for Infection Research.

Equations for the Mouse PBPK/PD Model

Variables appearing in eqs 1–7:

Vircellmblood as the amount of virus in the blood (mice); Vircellmbrain as the amount of virus in the mouse brain; Vircellmliver as the amount of virus in the mouse liver; Vircellmlung as the amount of virus in the mouse lung; Vircellmheart as the amount of virus in the mouse heart; Vircellmkidney as the amount of virus in the mouse kidney; Vircellmspleen as the amount of virus in the mouse spleen; Virpls is the amount of virus in plasma; kcellinfect as cell infection constant; kcellinfectaugment is the cell infection constant taking into account cell infection rate augments due to more cells being infected over time; tcalc24 time starting after 24 h; Bfrmbrain as the blood flow rate within the brain in mice; ; Bfrmliver as the blood flow rate within the liver in mice, Bfrmlung as the blood flow rate within the lung in mice ; Bfrmheart as the blood flow rate within the heart in mice; ; Bfrmkidney as the blood flow rate within the kidney in mice; ; Bfrmspleen as the blood flow rate within the spleen in mice and Bfrmmuscle as the blood flow rate in muscle in mice.

Variables appearing in eqs 8–14:

Virmblood as the amount of virus in the blood compartment in mice after reproduction and ready for release; Virmbrain as the amount of virus in the brain in mice after reproduction and ready for release; Virmliver as the amount of virus in the liver in mice after reproduction and ready for release; Virmlung as the amount of virus in the lung in mice after reproduction and ready for release; Virmheart as the amount of virus in the heart in mice after reproduction and ready for release; Virmkidney as the amount of virus in the kidney in mice after reproduction and ready for release; Virmspleen as the amount of virus in the spleen in mice after reproduction and ready for release; kprodvir as viral reproduction constant; SorAmblood is the concentration of SorA in the blood compartment in mice according to the PBPK model calculations; SorAmbrain is the concentration of SorA in the brain compartment in mice according to the PBPK model calculations; SorAmliver is the concentration of SorA in the liver compartment in mice according to the PBPK model calculations; SorAmlung is the concentration of SorA in the lung compartment in mice according to the PBPK model calculations; SorAmheart is the concentration of SorA in the heart compartment in mice according to the PBPK model calculations; SorAmkidney is the concentration of SorA in the kidney compartment in mice according to the PBPK model calculations; SorAmspleen is the concentration of SorA in the spleen compartment in mice according to the PBPK model calculations; EC50SorA half-maximal effective concentration of SorA.

Variables appearing in eqs 15–20:

Virmbrain2 as the amount of virus released (in mice) from the brain to plasma; with Virmliver2 as the amount of virus released (in mice) from the liver to plasma; with Virmlung2 as the amount of virus released (in mice) from the lung to plasma; with Virmheart2 as the amount of virus released (in mice) from the heart to plasma; with Virkidney2 as the amount of virus released (in mice) from the kidney to plasma; with Virspleen2 as the amount of virus released (in mice) from the spleen to plasma; kcellclear is the constant for viral clearance by the CD4+ T cell response; tcalc48 time delay before onset of CD4+ T cell response (t = 48 h) and Bfrmuscle as the blood flow rate in muscle.

Variables appearing in eq 21:

Virpls as the amount of viral load in the plasma; kclear clearance rate constant for viral particles from the plasma by the innate immune response; tcalc24 delayed time (t = 24 h) taking into account that viral clearance does not start immediately.

Equations for the Human PBPK/PD Model

Variables appearing in eqs 22–28:

Vircellhblood as the amount of virus in the human blood compartment; Vircellhbrain as the amount of virus in the human brain compartment; Vircellhliver as the amount of virus in the human liver compartment; Vircellhlung as the amount of virus in the human lung compartment; Vircellhheart as the amount of virus in the human heart compartment; Vircellhkidney as the amount of virus in the human kidney compartment; Vircellhspleen as the amount of virus in the human spleen compartment; Virpls is the amount of virus in plasma; kcellinfect as cell infection constant; kcellinfectaugment is the cell infection constant taking into account that more cells are infected over time; tcalc24 time starting after 24 h; Bfrhbrain as the blood flow rate in the human brain; Bfrhliver as the blood flow rate in the human liver; Bfrhlung as the blood flow rate in the human lung; Bfrhheart as the blood flow rate in the human heart; Bfrhkidney as the blood flow rate in the human kidney; Bfrhspleen as the blood flow rate in the human spleen and with BfrPV as the blood flow rate in the human portal vein.

Variables appearing in eqs 29–35:

Virhblood as the amount of virus in the human blood compartment after reproduction and ready for release; Virhbrain as the amount of virus in the human brain compartment after reproduction and ready for release; Virhliver as the amount of virus in the human liver compartment after reproduction and ready for release; Virhlung as the amount of virus in the human lung compartment after reproduction and ready for release; Virhheart as the amount of virus in the human heart compartment after reproduction and ready for release; Virhkidney as the amount of virus in the human kidney compartment after reproduction and ready for release; Virhspleen as the amount of virus in the human spleen compartment after reproduction and ready for release; kprodvir as viral reproduction constant; SorAhblood is the concentration of SorA in the human blood cell compartment; SorAhbrain is the concentration of SorA in the human brain compartment; SorAhliver is the concentration of SorA in human liver compartment; SorAhlung is the concentration of SorA in the human lung compartment; SorAhheart is the concentration of SorA in the human heart compartment; SorAhkidney is the concentration of SorA in the human kidney compartment; ; SorAhspleen is the concentration of SorA in the human spleen compartment; EC50SorA half-maximal effective concentration of SorA.

Variables appearing in eqs 36–41:

Virhbrain2 as the amount of virus released from the human brain compartment to plasma; Virhliver2 as the amount of virus released from the human liver compartment to plasma; Virhlung2 as the amount of virus released from the human lung compartment to plasma; Virhheart2 as the amount of virus released from the human heart compartment to plasma; Virhkidney2 as the amount of virus released from the human kidney compartment to plasma; Virhspleen2 as the amount of virus released from the human spleen compartment to plasma; kcellclear is the constant for viral clearance by the T helper cell response; kinfectimmun is the constant for viral clearance by cells responding to increased interferon levels; tcalc48 is the time delay (t = 48 h) before onset of the immune cell response.

Variables appearing in eq 42:

Virpls as the amount of viral load in plasma; Vir as the viral load in plasma at the time point of infection as a result of the mosquito bite; Virbrain2 is the virus released from the brain; Virheart2 is the virus released from the heart; Virkidney2 is the virus released from the kidney cells; Virliver2 is the virus released from liver cells; Virlung2 is the virus released from lung cells; Virspleen2 is the virus released from the spleen; kclear is the clearance rate constant for viral particles from plasma by the innate immune response; kclearaugment is the augmented clearance rate constant for viral particles from plasma by the innate immune response as a result of increasing interferon levels; tcalc120 delayed time (t = 120 h) taking into account that viral clearance does not start immediately; tcalc144 delayed time (t = 144 h) taking into account that viral clearance augmentation starts 24 h after viral clearance.

Software

PBPK and PBPK/PD models were built using open source software tool Open Systems Pharmacology version 9.1 the software (www.open-systems-pharmacology.org) using the software PK-Sim and MoBi under the GPLv2 License.58

Acknowledgments

The authors wish to thank Andrea Ahlers and Janine Schreiber for excellent technical assistance. K.R. thanks the German Center for Infection Research academy (DZIF academy) for a Laboratory Rotation Fellowship to enable her research at RIKEN. K.R. receives support from the German Center for Infection Research (DZIF) (TTU 09.710). The authors thank Dr. Hans Prochnow for providing GAPDH and DENV-2 primers. K.R. acknowledges Kota Toshimoto for scientific discussion on PBPK models.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsptsci.1c00078.

  • Data about the metabolite ID of SorA and the CYP enzymatic potential as well as descriptions of supplemental methods and supplemental figures and tables (PDF)

Author Contributions

K.R. conceived the study, performed animal PK studies, performed RNA extraction and RT-PCR, designed the PBPK/PD models, analyzed the data and wrote the manuscript. M.H. and A.K. performed animal infection experiments. A.K. analyzed animal infection data. J.K. prepared DENV for testing. K.H. and R.M. supervised in vitro ADME data. V.T. and J.H. conducted in vitro ADME studies. G.P.V. and J.D. conducted cellular antiviral assays with SorA. Y.S. conceived the part of the PBPK/PD model study and analyzed the data. M.B. conceived the study, analyzed the data, and wrote the manuscript.

The authors declare the following competing financial interest(s): Y.S. is a science advisory board member of SimCyp, Cetara. The other authors declare no conflict of interest.

Supplementary Material

pt1c00078_si_001.pdf (1.4MB, pdf)

References

  1. Brady O. J.; Gething P. W.; Bhatt S.; Messina J. P.; Brownstein J. S.; Hoen A. G.; Moyes C. L.; Farlow A. W.; Scott T. W.; Hay S. I. Refining the Global Spatial Limits of Dengue Virus Transmission by Evidence-Based Consensus. PLoS Neglected Trop. Dis. 2012, 6, e1760. 10.1371/journal.pntd.0001760. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bhatt S.; Gething P. W.; Brady O. J.; Messina J. P.; Farlow A. W.; Moyes C. L.; Drake J. M.; Brownstein J. S.; Hoen A. G.; Sankoh O.; Myers M. F.; George D. B.; Jaenisch T.; Wint G. R. W.; Simmons C. P.; Scott T. W.; Farrar J. J.; Hay S. I. The global distribution and burden of dengue. Nature 2013, 496, 504–507. 10.1038/nature12060. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Castro M. C.; Wilson M. E.; Bloom D. E. Disease and economic burdens of dengue. Lancet Infect. Dis. 2017, 17, e70–e78. 10.1016/S1473-3099(16)30545-X. [DOI] [PubMed] [Google Scholar]
  4. Ferguson N. M.; Rodríguez-Barraquer I.; Dorigatti I.; Mier-y-Teran-Romero L.; Laydon D. J.; Cummings D. A. T. Benefits and risks of the Sanofi-Pasteur dengue vaccine: Modeling optimal deployment. Science 2016, 353, 1033–1036. 10.1126/science.aaf9590. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Fatima K.; Syed N. I. Dengvaxia controversy: impact on vaccine hesitancy. J. Glob. Health 2018, 8, 010312. 10.7189/jogh.08.020312. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Coudeville L.; Baurin N.; Shepard D. S. The potential impact of dengue vaccination with, and without, pre-vaccination screening. Vaccine 2020, 38, 1363. 10.1016/j.vaccine.2019.12.012. [DOI] [PubMed] [Google Scholar]
  7. Singhi S.; Kissoon N.; Bansal A. Dengue e dengue hemorrágico: aspectos do manejo na unidade de terapia intensiva. J. Pediatr. (Rio J.) 2007, 83, S22–S35. 10.1590/S0021-75572007000300004. [DOI] [PubMed] [Google Scholar]
  8. Organization W. H.Dengue: Guidelines for Diagnosis, Treatment, Prevention and Control; World Health Organization, 2009. [PubMed] [Google Scholar]
  9. Oliveira de Andrade S. M.; Maria Moreira Herkert C.; Venancio da Cunha R.; Rodrigues M. D.; Antonio Kato da Silva B. A New Approach to Reducing Mortality from Dengue. Open J. Clin. Diagn. 2014, 4, 12–16. 10.4236/ojcd.2014.41003. [DOI] [Google Scholar]
  10. Guzman M. G.; Gubler D. J.; Izquierdo A.; Martinez E.; Halstead S. B. Dengue infection. Nat. Rev. Dis. Primer 2016, 2, 1–25. 10.1038/nrdp.2016.55. [DOI] [PubMed] [Google Scholar]
  11. Low J. G. H.; Ooi E. E.; Vasudevan S. G. Current Status of Dengue Therapeutics Research and Development. J. Infect. Dis. 2017, 215, S96–S102. 10.1093/infdis/jiw423. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Duyen H. T. L.; Ngoc T. V.; Ha D. T.; Hang V. T. T.; Kieu N. T. T.; Young P. R.; Farrar J. J.; Simmons C. P.; Wolbers M.; Wills B. A. Kinetics of Plasma Viremia and Soluble Nonstructural Protein 1 Concentrations in Dengue: Differential Effects According to Serotype and Immune Status. J. Infect. Dis. 2011, 203, 1292–1300. 10.1093/infdis/jir014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Fox A.; Hoa L. N. M.; Simmons C. P.; Wolbers M.; Wertheim H. F. L.; Khuong P. T.; Ninh T. T. H.; Lien T. T. M.; Lien N. T.; Trung N. V.; Hien N. D.; Farrar J.; Horby P.; Taylor W. R.; Kinh N. V. Immunological and Viral Determinants of Dengue Severity in Hospitalized Adults in Ha Noi, Viet Nam. PLoS Neglected Trop. Dis. 2011, 5, e967. 10.1371/journal.pntd.0000967. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Tricou V.; Minh N. N.; Farrar J.; Tran H. T.; Simmons C. P. Kinetics of Viremia and NS1 Antigenemia Are Shaped by Immune Status and Virus Serotype in Adults with Dengue. PLoS Neglected Trop. Dis. 2011, 5, e1309. 10.1371/journal.pntd.0001309. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Bedorf N.; Schomburg D.; Gerth K.; Reichenbach H.; Höfle G. Antibiotics from Gliding Bacteria, LIV. Isolation and Structure Elucidation of Soraphen A1α, a Novel Antifungal Macrolide from Sorangium cellulosum. Liebigs Ann. Chem. 1993, 1993, 1017–1021. 10.1002/jlac.1993199301161. [DOI] [Google Scholar]
  16. Gerth K.; Bedorf N.; Irschik H.; Höfle G.; Reichenbach H. THE SORAPHENS: A FAMILY OF NOVEL ANTIFUNGAL COMPOUNDS FROM Sorangium cellulosum (MYXOBACTERIA). J. Antibiot. 1994, 47, 23–31. 10.7164/antibiotics.47.23. [DOI] [PubMed] [Google Scholar]
  17. Berod L.; Friedrich C.; Nandan A.; Freitag J.; Hagemann S.; Harmrolfs K.; Sandouk A.; Hesse C.; Castro C. N.; Bähre H.; Tschirner S. K.; Gorinski N.; Gohmert M.; Mayer C. T.; Huehn J.; Ponimaskin E.; Abraham W.-R.; Müller R.; Lochner M.; Sparwasser T. De novo fatty acid synthesis controls the fate between regulatory T and T helper 17 cells. Nat. Med. 2014, 20, 1327–1333. 10.1038/nm.3704. [DOI] [PubMed] [Google Scholar]
  18. Naini A.; Sasse F.; Brönstrup M. The intriguing chemistry and biology of soraphens. Nat. Prod. Rep. 2019, 36, 1394–1411. 10.1039/C9NP00008A. [DOI] [PubMed] [Google Scholar]
  19. Mamareli P.; Kruse F.; Lu C.; Guderian M.; Floess S.; Rox K.; Allan D. S. J.; Carlyle J. R.; Brönstrup M.; Müller R.; Berod L.; Sparwasser T.; Lochner M. Targeting cellular fatty acid synthesis limits T helper and innate lymphoid cell function during intestinal inflammation and infection. Mucosal Immunol. 2021, 14, 164–176. 10.1038/s41385-020-0285-7. [DOI] [PubMed] [Google Scholar]
  20. Vahlensieck H. F.; Pridzun L.; Reichenbach H.; Hinnen A. Identification of the yeast ACC1 gene product (acetyl-CoA carboxylase) as the target of the polyketide fungicide soraphen A. Curr. Genet. 1994, 25, 95–100. 10.1007/BF00309532. [DOI] [PubMed] [Google Scholar]
  21. Koutsoudakis G.; Romero-Brey I.; Berger C.; Pérez-Vilaró G.; Perin P. M.; Vondran F. W. R.; Kalesse M.; Harmrolfs K.; Müller R.; Martinez J. P.; Pietschmann T.; Bartenschlager R.; Brönstrup M.; Meyerhans A.; Díez J. Soraphen A: a broad-spectrum antiviral natural product with potent anti-hepatitis C virus activity. J. Hepatol. 2015, 63, 813–821. 10.1016/j.jhep.2015.06.002. [DOI] [PubMed] [Google Scholar]
  22. Singaravelu R.; Desrochers G. F.; Srinivasan P.; O’Hara S.; Lyn R. K.; Müller R.; Jones D. M.; Russell R. S.; Pezacki J. P. Soraphen A: A Probe for Investigating the Role of de Novo Lipogenesis during Viral Infection. ACS Infect. Dis. 2015, 1, 130–134. 10.1021/acsinfecdis.5b00019. [DOI] [PubMed] [Google Scholar]
  23. Villareal V. A.; Rodgers M. A.; Costello D. A.; Yang P. L. Targeting host lipid synthesis and metabolism to inhibit dengue and hepatitis C viruses. Antiviral Res. 2015, 124, 110–121. 10.1016/j.antiviral.2015.10.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Fleta-Soriano E.; Smutná K.; Martínez J. P.; Oró C. L.; Sadiq S. K.; Mirambeau G.; Lopez-Iglesias C.; Bosch M.; Pol A.; Brönstrup M.; Diez J.; Meyerhans A. The Myxobacterial Metabolite Soraphen A Inhibits HIV-1 by Reducing Virus Production and Altering Virion Composition. Antimicrob. Agents Chemother. 2017, 61, e00739–17. 10.1128/AAC.00739-17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Merino-Ramos T.; Vázquez-Calvo Á.; Casas J.; Sobrino F.; Saiz J.-C.; Martín-Acebes M. A. Modification of the Host Cell Lipid Metabolism Induced by Hypolipidemic Drugs Targeting the Acetyl Coenzyme A Carboxylase Impairs West Nile Virus Replication. Antimicrob. Agents Chemother. 2016, 60, 307–315. 10.1128/AAC.01578-15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Jiménez de Oya N.; de Esler W. P.; Huard K.; El-Kattan A. F.; Karamanlidis G.; Blázquez A.-B.; Ramos-Ibeas P.; Escribano-Romero E.; Louloudes-Lázaro A.; Casas J.; Sobrino F.; Hoehn K.; James D. E.; Gutiérrez-Adán A.; Saiz J.-C.; Martín-Acebes M. A. Targeting host metabolism by inhibition of acetyl-Coenzyme A carboxylase reduces flavivirus infection in mouse models. Emerging Microbes Infect. 2019, 8, 624–636. 10.1080/22221751.2019.1604084. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Clapham H. E.; Tricou V.; Van Vinh Chau N.; Simmons C. P.; Ferguson N. M. Within-host viral dynamics of dengue serotype 1 infection. J. R. Soc., Interface 2014, 11, 20140094. 10.1098/rsif.2014.0094. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Ben-Shachar R.; Koelle K. Minimal within-host dengue models highlight the specific roles of the immune response in primary and secondary dengue infections. J. R. Soc., Interface 2015, 12, 20140886. 10.1098/rsif.2014.0886. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Ben-Shachar R.; Schmidler S.; Koelle K. Drivers of Inter-individual Variation in Dengue Viral Load Dynamics. PLoS Comput. Biol. 2016, 12, e1005194. 10.1371/journal.pcbi.1005194. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Begum F.; Das S.; Mukherjee D.; Mal S.; Ray U. Insight into the Tropism of Dengue Virus in Humans. Viruses 2019, 11, 1136. 10.3390/v11121136. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Martina B. E. E.; Koraka P.; Osterhaus A. D. M. E. Dengue Virus Pathogenesis: an Integrated View. Clin. Microbiol. Rev. 2009, 22, 564–581. 10.1128/CMR.00035-09. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Eissing T.; Kuepfer L.; Becker C.; Block M.; Coboeken K.; Gaub T.; Goerlitz L.; Jaeger J.; Loosen R.; Ludewig B.; Meyer M.; Niederalt C.; Sevestre M.; Siegmund H.-U.; Solodenko J.; Thelen K.; Telle U.; Weiss W.; Wendl T.; Willmann S.; Lippert J. A Computational Systems Biology Software Platform for Multiscale Modeling and Simulation: Integrating Whole-Body Physiology, Disease Biology, and Molecular Reaction Networks. Front. Physiol. 2011, 2, 2. 10.3389/fphys.2011.00004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Wang W.-K.; Sung T.-L.; Tsai Y.-C.; Kao C.-L.; Chang S.-M.; King C.-C. Detection of Dengue Virus Replication in Peripheral Blood Mononuclear Cells from Dengue Virus Type 2-Infected Patients by a Reverse Transcription-Real-Time PCR Assay. J. Clin. Microbiol. 2002, 40, 4472–4478. 10.1128/JCM.40.12.4472-4478.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Diamond M. S.; Edgil D.; Roberts T. G.; Lu B.; Harris E. Infection of Human Cells by Dengue Virus Is Modulated by Different Cell Types and Viral Strains. J. Virol. 2000, 74, 7814–7823. 10.1128/JVI.74.17.7814-7823.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Kularatne S. A. M.; Rajapakse M. M.; Ralapanawa U.; Waduge R.; Pathirage L. P. M. M. K.; Rajapakse R. P. V. J. Heart and liver are infected in fatal cases of dengue: three PCR based case studies. BMC Infect. Dis. 2018, 18, 681. 10.1186/s12879-018-3603-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Oliveira E. R. A.; Póvoa T. F.; Nuovo G. J.; Allonso D.; Salomão N. G.; Basílio-de-Oliveira C. A.; Geraldo L. H. M.; Fonseca C. G.; Lima F. R. S.; Mohana-Borges R.; Paes M. V. Dengue fatal cases present virus-specific HMGB1 response in peripheral organs. Sci. Rep. 2017, 7, 16011. 10.1038/s41598-017-16197-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Züst R.; Toh Y.-X.; Valdés I.; Cerny D.; Heinrich J.; Hermida L.; Marcos E.; Guillén G.; Kalinke U.; Shi P.-Y.; Fink K. Type I Interferon Signals in Macrophages and Dendritic Cells Control Dengue Virus Infection: Implications for a New Mouse Model To Test Dengue Vaccines. J. Virol. 2014, 88, 7276–7285. 10.1128/JVI.03827-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. St. John A. L.; Rathore A. P. S. Adaptive immune responses to primary and secondary dengue virus infections. Nat. Rev. Immunol. 2019, 19, 218–230. 10.1038/s41577-019-0123-x. [DOI] [PubMed] [Google Scholar]
  39. Vaughn D. W.; Green S.; Kalayanarooj S.; Innis B. L.; Nimmannitya S.; Suntayakorn S.; Rothman A. L.; Ennis F. A.; Nisalak A. Dengue in the Early Febrile Phase: Viremia and Antibody Responses. J. Infect. Dis. 1997, 176, 322–330. 10.1086/514048. [DOI] [PubMed] [Google Scholar]
  40. Yauch L. E.; Zellweger R. M.; Kotturi M. F.; Qutubuddin A.; Sidney J.; Peters B.; Prestwood T. R.; Sette A.; Shresta S. A Protective Role for Dengue Virus-Specific CD8+ T Cells. J. Immunol. 2009, 182, 4865–4873. 10.4049/jimmunol.0801974. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Sun P.; Kochel T. J. The Battle between Infection and Host Immune Responses of Dengue Virus and Its Implication in Dengue Disease Pathogenesis. Sci. World J. 2013, 2013, 843469. 10.1155/2013/843469. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Nguyen N. M.; Thi Hue Kien D.; Tuan T. V.; Quyen N. T. H.; Tran C. N. B.; Vo Thi L.; Thi D. L.; Nguyen H. L.; Farrar J. J.; Holmes E. C.; Rabaa M. A.; Bryant J. E.; Nguyen T. T.; Nguyen H. T. C.; Nguyen L. T. H.; Pham M. P.; Nguyen H. T.; Luong T. T. H.; Wills B.; Nguyen C. V. V.; Wolbers M.; Simmons C. P. Host and viral features of human dengue cases shape the population of infected and infectious Aedes aegypti mosquitoes. Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 9072–9077. 10.1073/pnas.1303395110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Raquin V.; Lambrechts L. Dengue virus replicates and accumulates in Aedes aegypti salivary glands. Virology 2017, 507, 75–81. 10.1016/j.virol.2017.04.009. [DOI] [PubMed] [Google Scholar]
  44. Novelo M.; Hall M. D.; Pak D.; Young P. R.; Holmes E. C.; McGraw E. A. Intra-host growth kinetics of dengue virus in the mosquito Aedes aegypti. PLoS Pathog. 2019, 15, e1008218. 10.1371/journal.ppat.1008218. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Vaughn D. W.; Green S.; Kalayanarooj S.; Innis B. L.; Nimmannitya S.; Suntayakorn S.; Endy T. P.; Raengsakulrach B.; Rothman A. L.; Ennis F. A.; Nisalak A. Dengue Viremia Titer, Antibody Response Pattern, and Virus Serotype Correlate with Disease Severity. J. Infect. Dis. 2000, 181, 2–9. 10.1086/315215. [DOI] [PubMed] [Google Scholar]
  46. Tsai J.-J.; Lin P.-C.; Tsai C.-Y.; Wang Y.-H.; Liu L.-T. Low frequency of asymptomatic dengue virus-infected donors in blood donor centers during the largest dengue outbreak in Taiwan. PLoS One 2018, 13, e0205248. 10.1371/journal.pone.0205248. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. de la Cruz-Hernández S. I.; Flores-Aguilar H.; González-Mateos S.; López-Martinez I.; Alpuche-Aranda C.; Ludert J. E.; del Angel R. M. Determination of Viremia and Concentration of Circulating Nonstructural Protein 1 in Patients Infected with Dengue Virus in Mexico. Am. J. Trop. Med. Hyg. 2013, 88, 446–454. 10.4269/ajtmh.12-0023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Diamond M. S.; Pierson T. C. Molecular Insight into Dengue Virus Pathogenesis and Its Implications for Disease Control. Cell 2015, 162, 488–492. 10.1016/j.cell.2015.07.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Tremblay N.; Freppel W.; Sow A. A.; Chatel-Chaix L. The Interplay between Dengue Virus and the Human Innate Immune System: A Game of Hide and Seek. Vaccines 2019, 7, 145. 10.3390/vaccines7040145. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Jump D. B.; Torres-Gonzalez M.; Olson L. K. Soraphen A, an inhibitor of acetyl CoA carboxylase activity, interferes with fatty acid elongation. Biochem. Pharmacol. 2011, 81, 649–660. 10.1016/j.bcp.2010.12.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Schreurs M.; van Dijk T. H.; Gerding A.; Havinga R.; Reijngoud D.-J.; Kuipers F. Soraphen, an inhibitor of the acetyl-CoA carboxylase system, improves peripheral insulin sensitivity in mice fed a high-fat diet. Diabetes, Obes. Metab. 2009, 11, 987–991. 10.1111/j.1463-1326.2009.01078.x. [DOI] [PubMed] [Google Scholar]
  52. Harriman G.; Greenwood J.; Bhat S.; Huang X.; Wang R.; Paul D.; Tong L.; Saha A. K.; Westlin W. F.; Kapeller R.; Harwood H. J. Acetyl-CoA carboxylase inhibition by ND-630 reduces hepatic steatosis, improves insulin sensitivity, and modulates dyslipidemia in rats. Proc. Natl. Acad. Sci. U. S. A. 2016, 113, E1796–E1805. 10.1073/pnas.1520686113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Zitzmann C.; Kaderali L. Mathematical Analysis of Viral Replication Dynamics and Antiviral Treatment Strategies: From Basic Models to Age-Based Multi-Scale Modeling. Front. Microbiol. 2018, 9, 9. 10.3389/fmicb.2018.01546. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Busch M. P.; Sabino E. C.; Brambilla D.; Lopes M. E.; Capuani L.; Chowdhury D.; McClure C.; Linnen J. M.; Prince H.; Simmons G.; Lee T.-H.; Kleinman S.; Custer B.; de Almeida Neto C.; Mendrone-Junior A.; Carneiro-Proietti A. B.; Sampaio D.; de A.; Loureiro P.; Lobo C.; Esther Lopes M.; Cerdeira Sabino E.; Capuani L.; Eduardo Ferreira J.; Oikawa M.; Losco Takecian P.; Di Lorenzo Oliveira C.; Custer B.; Busch M. P.; Kelly S.; Gonçalez T. T.; Brambilla D.; McClure C.; Glynn S. A. Duration of Dengue Viremia in Blood Donors and Relationships Between Donor Viremia, Infection Incidence and Clinical Case Reports During a Large Epidemic. J. Infect. Dis. 2016, 214, 49–54. 10.1093/infdis/jiw122. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Cruz-Oliveira C.; Freire J. M.; Conceição T. M.; Higa L. M.; Castanho M. A. R. B.; Da Poian A. T. Receptors and routes of dengue virus entry into the host cells. FEMS Microbiol. Rev. 2015, 39, 155–170. 10.1093/femsre/fuu004. [DOI] [PubMed] [Google Scholar]
  56. Nguyen N. M.; Tran C. N. B.; Phung L. K.; Duong K. T. H.; Huynh H.; le A.; Farrar J.; Nguyen Q. T. H.; Tran H. T.; Nguyen C. V. V.; Merson L.; Hoang L. T.; Hibberd M. L.; Aw P. P. K.; Wilm A.; Nagarajan N.; Nguyen D. T.; Pham M. P.; Nguyen T. T.; Javanbakht H.; Klumpp K.; Hammond J.; Petric R.; Wolbers M.; Nguyen C. T.; Simmons C. P. A Randomized, Double-Blind Placebo Controlled Trial of Balapiravir, a Polymerase Inhibitor, in Adult Dengue Patients. J. Infect. Dis. 2013, 207, 1442–1450. 10.1093/infdis/jis470. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Zhang Y.; Huo M.; Zhou J.; Xie S. PKSolver: An add-in program for pharmacokinetic and pharmacodynamic data analysis in Microsoft Excel. Comput. Methods Programs Biomed. 2010, 99, 306–314. 10.1016/j.cmpb.2010.01.007. [DOI] [PubMed] [Google Scholar]
  58. Lippert J.; Burghaus R.; Edginton A.; Frechen S.; Karlsson M.; Kovar A.; Lehr T.; Milligan P.; Nock V.; Ramusovic S.; Riggs M.; Schaller S.; Schlender J.; Schmidt S.; Sevestre M.; Sjögren E.; Solodenko J.; Staab A.; Teutonico D. Open Systems Pharmacology Community-An Open Access, Open Source, Open Science Approach to Modeling and Simulation in Pharmaceutical Sciences. CPT: Pharmacometrics Syst. Pharmacol. 2019, 8, 878–882. 10.1002/psp4.12473. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

pt1c00078_si_001.pdf (1.4MB, pdf)

Articles from ACS Pharmacology & Translational Science are provided here courtesy of American Chemical Society

RESOURCES