A Novel Prodrug Strategy Based on Reversibly Degradable Guanidine Imides for High Oral Bioavailability and Prolonged Pharmacokinetics of Broad-Spectrum Anti-influenza Agents

We present orally administrable prodrugs (OSC-GCDIs) of guanidino oseltamivir carboxylate (GOC) based on guanidine cyclic diimide (GCDI) to treat influenza viruses. By concealing the guanidine group, which significantly limits the intestinal absorption, its prodrugs OSC-GCDIs demonstrate dramatic improvement of oral bioavailability. The most promising antiviral substance OSC-GCDI(P) readily forms covalent adducts with serum proteins via a degradable linker after the intestinal absorption. Subsequently, the active species, GOC, is released from the conjugate in a sustained manner, which greatly contributes to improving pharmacokinetic properties. Because of the remarkable improvements in both oral bioavailability and longevity of its active metabolite, OSC-GCDI(P) demonstrates outstanding therapeutic efficacy against both wild-type and oseltamivir-resistant (H275Y) influenza virus strains in a mouse infection model, even with a single oral administration.


Supplemental Experimental Procedures
Lipophilicity.
To assess of octanol-water partition coefficient (log P), the solution of each OSC-GCDI prodrug (5.0, 10, and 20 mM) was prepared in deionized water (DIW)-saturated n-octanol.1.0 mL of these solutions were then equilibrated at room temperature with an equivalent volume (1.0 mL) of octanol-saturated DIW by vigorous shaking for 1 min.After shaking, the octanol and aqueous phases were separated by centrifugation at 14,000 rpm for 10 min at 4°C.The aqueous phase was collected and filtered by PVDF filter (0.20 μm pore size).The total prodrug concentration in the aqueous phase ([prodrug]aq) was determined by HPLC analysis; To assess of octanol-buffer distribution coefficients at pH 7.4 (log D7.4), the solution of each prodrug (5.0, 10, and 20 mM) were prepared in sodium phosphate buffer (50 mM phosphate, pH 7.4)-saturated n-octanol.1.0 mL of these solutions were then equilibrated at room temperature with an equivalent volume (1.0 mL) of octanol-saturated sodium phosphate buffer by vigorous shaking for 1 min.After shaking, the octanol and buffer phases were separated by centrifugation at 14,000 rpm for 10 min at 4°C.The aqueous buffer phase was collected and filtered by a polyvinylidene difluoride (PVDF) filter (0.20 μm pore size).The total prodrug concentration in the aqueous buffer phase ([prodrug]pH 7.4) was determined by HPLC analysis.The total prodrug concentration in the octanol phase ([prodrug]oct) was obtained by mass balance.From these data, the octanol-water partition coefficient, P = [prodrug]oct/[prodrug]aq, and the octanol-buffer distribution coefficient at pH 7.4, D7.4 = [prodrug]oct/[prodrug]pH 7.4, were determined.

In vitro Parallel Artificial Membrane Permeability Assay (PAMPA)
Membrane permeability of OSC-GCDI prodrugs, GOC/HPMCP and OSC-GCDI(P)/HPMCP were evaluated using Parallel Artificial Membrane Permeability Assay Kit (PAMPA-096).Each prodrug and the supplied permeability controls were dissolved in DMSO at the concentration of 10 mM.Then 25 μL of each prodrug solution was mixed with 475 μL of 50 mM sodium phosphate buffer (pH 6.5) to be used as the test solution.GOC/HPMCP and OSC-GCDI/HPCMP particles are dissolved in 50 mM sodium phosphate buffer (5% DMSO, pH 6.5) to achieve a final concentration of 200 µM.In the acceptor plate, 300 μL of 50 mM sodium phosphate buffer (pH 7.4) was added.Subsequently, before adding 200 μL of each test solution to the donor plate, 5 μL of 4% lecithin in dodecane was directly applied to the well membrane of the donor plate.Donor plate was carefully placed into acceptor plate and incubate at r.t. for 16 -24 h.Donor plate was removed and each solution was collected and filtered by a PVDF filter (0.20 μm pore size).The total prodrug concentration in each plate was determined by HPLC analysis.
In the case of OSC-GCDI(P), partial degradation occurred due to the instability of OSC-GCDI(P) in the buffer.Therefore, the collected solution was further incubated in a 1 M KOH solution for complete hydrolysis.Then, the concentration of GOC was quantified instead.

Stability of OSC-GCDIs in buffers
The stability of the GCDI prodrugs were examined by incubating prodrugs in 50 mM sodium phosphate buffer (pH 2.0, 6.5, 7.4, and 8.0) at 37 °C.Briefly, OSC-GCDI prodrug was dissolved in DMSO to be 10 mM concentration, then 10 μL of each prodrug solution was mixed with 990 μL of the phosphate buffer.

Stability of OSC-GCDIs in serum
Stability of the GCDI prodrugs were examined by incubating prodrugs in human and BALB/c serum at 37 °C and 200 μM final concentration.Briefly, each prodrug was dissolved in DMSO to be 20 mM concentration, then 10 μL of each prodrug solution was mixed with 990 μL of serum.At different time intervals, 100 μL of the samples were taken and mixed with 100 μL of MeCN.Samples were centrifuged at 14,000 rpm for 10 min at 4 °C and the supernatant was immediately frozen (−80 °C).After thawing and shaking, the samples were filtered using a polytetrafluoroethylene (PTFE) filter (0.20 μm pore size) and analyzed by HPLC.

Albumin binding of OSC-GCDIs
OSC-GCDI(P) (0.2 mg/mL or 10 mg/mL) was mixed with HSA (1 mg/mL) in 50 mM Tris-HCl buffer (pH 6.8).The sample was shaken for 20 min at room temperature.Unbound OSC-GCDI(P) was removed by washing three times with distilled water (DW) using Vivaspin 500 ultrafiltration units (10 kDa cutoff).Following centrifugation, the HSA samples were placed in the condition of 8 M urea, 50 mM Tris-HCl (pH 6.8), and 5 mM DTT at 37 °C for 60 min.Then, 50 mM Tris-HCl (pH 6.8) was added until the urea concentration was less than 1 M. Trypsin Gold was then added to achieve a final protease-to-protein ratio of 1:20 (w/w).The sample was incubated at 37 °C for 60 min.The mass-to-charge ratio of the digested sample was measured by MALDI-TOF (Matrix-Assisted Laser Desorption/Ionization Time-of-Flight) MS using α-cyano-4-hydroxycinnamic acid as the matrix.
Release profiles of GOC from the OSC-GCDI prodrugs conjugated to human serum albumin OSC-GCDIs were dissolved in Dulbecco's Phosphate-Buffered Saline (DPBS) containing HSA (20 mg/mL) to be concentration of 1 mM, then the mixture was shaken for 10 min at 37 °C.Unbound OSC-GCDIs were removed by washing twice with DPBS using Vivaspin 500 ultrafiltration units (10 kDa cutoff).Following centrifugation, the samples were redissolved in 1 mL DPBS.Before incubation at 37 °C, 50 µL sample was taken and mixed with 50 µL of 1 M KOH in methanol solution to completely hydrolyze the conjugates and determine the initial drug concentration bound to HSA.Subsequently, the samples were incubated at 37 °C and after 1, 2, 4, 24, and 72 h incubation, 50 µL of the samples were taken and mixed with 50 µL MeCN.Then, the concentration of the released GOC was quantified by HPLC.

In vitro microsomal stability
Sprague Dawley rat microsomes were obtained from Gibco TM .NADPH was purchased from Sigma-Aldrich (Munich, Germany).Liver microsomes at 0.5 mg/mL concentration (final volume of 200 μL) were incubated in phosphate buffer (pH 7.4), containing 3 mM MgCl2.OSC-GCDI prodrug was dissolved in DMSO to be 40 mM concentration, then each prodrug solution was mixed with prepared buffer to achieve a final concentration of 100 μM (0.25% DMSO).Test compounds were incubated at 37 °C and started by the addition of the 1mM NADPH as cofactor.The samples were taken at 0, 30, 60, and 90 min.The reaction was terminated by addition of ice-cold acetonitrile/methanol (1:1, v/v).After collection, samples were centrifuged (10 min, 16,000 rpm), then the centrifuged supernatant was directly analyzed by using HPLC analysis.Metabolic half-time (t1/2) was calculated by using the equation of first order kinetics.
Pharmacokinetic parameters were estimated using non-compartmental analysis implemented with the Pheonix WonNonlin software (Pharsight Corporation). 2Key parameters, including time point indicating maximum concentration (Tmax), maximum plasma concentration (Cmax), terminal half-life (T1/2), area under the concentration-time curve to the last measurable time point (AUClast), or extrapolated to infinity (AUC∞), clearance (CL), steady-state volume of distribution (Vss), mean residence time within the detection period (MRTlast) or extrapolated to infinity (MRT∞) were calculated.

In vivo efficacy study
Female BALB/c mice (Orient Bio) aged six to seven weeks old were orally administered with OS-P, GOC•TFA, OSC-GCDI(P) or OSC-GCDI(P)/HPMCP at doses of 0.1, 1 and 10 mg/kg (n = 5).Four hours later, the mice were intranasally infected with maPR8 at a 50% mouse lethal dose (MLD50) of 10 for an additional 4 h.Subsequently, the mice were orally administered again with each antiviral agent at the same dose.These treatments (i.e., oral, b.i.d., doses of 0.1, 1 and 10 mg/kg) were repeated every day for the next four days.To assess in vivo therapeutic efficacy, both body weight changes and mortality were daily recorded for 14 days after virus challenge.Mock-infected, vehicle (0.5% CMC in PBS)treated mice and influenza virus-infected, vehicle-treated mice were included as controls.
For the comparison analysis of therapeutic efficacy with varying treatment frequencies, BALB/c mice were orally administered with OS-P, GOC•TFA or OSC-GCDI(P)/HPMCP at a dose of 10 mg/kg (n = 5).Four hours later, they were intranasally challenged with either maPR8 or with rgA/Korea/200953-60, harboring the H275Y mutation in NA, at an MLD50 of 5.The mice were then subjected to the same treatment regimen every day or twice a day for an additional 4 days.In separate groups, oral administration was additively conducted once on day 3 or 4 post-infection, or discontinued after the first single administration.Body weight changes and mortality were recorded daily for 14 days after virus challenge.The two groups, comprising mock-infected, vehicle (0.5% CMC in distilled water)-treated mice and influenza virus-infected, vehicle-treated mice, were included as controls.
In all animal experiments related to antiviral efficacy evaluation, mice exhibiting body weight decrease exceeding 30% were euthanized by CO2.These experiments were conducted in accordance with the guidelines of the Institutional Animal Care and Use Committee, with approval numbers 2023-6D-07-01, 2023-6D-09-03 and 2023-6D-12-01.

Fig. S3 .
Fig. S3.A proposed mechanism of OSC-GCDI hydrolysis into GOC.Predicted possible intermediates are indicated as dashed blue squares.

Fig. S4 .
Fig. S4.MALDI-TOF MS analyses of trypsin-digested human serum albumin (HSA) incubated with OSC-GCDI(P).(a)MALDI-TOF MS spectra of trypsin-digested human serum albumin (HSA) only (up), HSA incubated with a limited amount of OSC-GCDI(P) (middle; HSA : OSC-GCDI(P) = 5 : 1 (w/w)), and HSA incubated with an excess amount of OSC-GCDI(P) (bottom; HSA : OSC-GCDI(P) = 1 : 10 (w/w)).Non-conjugated sequences derived from HSA are indicated as blue circles and green squares.OSC-GCDI(P)-conjugated sequences are indicated as red triangles.The green square peaks of HSA only were shifted to the red triangle peaks with dashed green squares by incubation with OSC-GCDI(P).(b) Identification of some representative peaks in the spectra.Suggested structures of the covalent linkage between the sequence and OSC-GCDI(P) are shown.

Fig. S6 .
Fig. S6.Pharmacokinetic properties of GOC•TFA in mice and rats.(a) Plasma concentrations and the pharmacokinetic parameters of GOC after administration of GOC•TFA in mice.ICR mice were administered with GOC•TFA intravenously (red circles) or orally (blue circles) at a dose of 5 mg/kg.Blood sample collection via tail vein began at 0.083 h when intravenously administered or at 0.25 h when orally administered.Collection continued at 0.5, 1, 2, 4, 8, and 24 h after treatment in both bases.Plasma concentration of GOC from three mice was determined by LC-MS/MS.The GOC concentration equivalent to an EC50 value (0.005 μM) against PR8 (A/H1N1) in MDCK cells is indicated by a red dashed line.Plasma concentrations of GOC from three mice were determined by LC-MS/MS.(b) Plasma concentrations and the pharmacokinetic parameters of GOC after administration of GOC•TFA into rats.SD rats were administered with GOC•TFA intravenously (red circles) or orally (blue circles) at doses of 5 and 10 mg/kg, respectively (n = 3).Blood sample collection via tail vein began at 0.083 h when intravenously administered or at 0.25 h when orally administered.Collection continued at 0.5, 1, 2, 4, 6, 8, 24, 32 and 48 h after treatment in both cases.The GOC concentration in plasma equivalent to the EC50 value (0.005 M) against PR8 (A/H1N1) in MDCK cells is indicated by red dashed lines.Plasma concentrations of GOC from three rats were determined by LC-MS/MS.The following pharmacokinetic parameters are defined; Tmax, time point indicating maximum concentration; Cmax, maximum plasma concentration; T1/2, terminal half-life; AUClast, area under the concentration-time curve to the last measurable time point; AUC∞, area under the concentration-time curve extrapolated to infinity; CL, clearance; Vss, steady-state volume of distribution; MRTlast, mean residence time within the detection period; MRT∞, mean residence time extrapolated to infinity; and Ft, oral bioavailability ((AUClast at oral administration/AUClast at intravenous administration)  100%).NA, not applicable; NC, not calculated.Values are presented as means ± SD from three different samples.

Fig. S7 .
Fig. S7.Pharmacokinetic properties of OSC-GCDIs in mice.(a) Plasma concentrations and the pharmacokinetic parameters of OSC-GCDI(C) (blue circles) and its metabolite GOC (red circles) in three different mice orally administered with OSC-GCDI(C).(b) Plasma concentrations and the pharmacokinetic parameters of OSC-GCDI(P) (blue circles; below detection limit) and its metabolite GOC (red circles) in three different mice orally administered with OSC-GCDI(P).The GOC concentration in plasma equivalent to an EC50 value (0.005 μM) against PR8 (A/H1N1) in MDCK cells is indicated by red dashed lines.The following pharmacokinetic parameters are defined: Tmax, time point indicating maximum concentration; Cmax, maximum plasma concentration; T1/2, terminal half-life; AUClast, area under the concentration-time curve to the last measurable time point; AUC∞, area under the concentration-time curve extrapolated to infinity; CL, clearance; Vss, steady-state volume of distribution; MRTlast, mean residence time within the detection period; and MRT∞, mean residence time extrapolated to infinity.NA, not applicable; NC, not calculated.Values are presented as means ± SD from three different mouse samples.

Fig. S8 .
Fig. S8.Therapeutic efficacy of OS-P, GOC•TFA and non-formulated OSC-GCDI(P) with varyingdoses in mice infected with A/H1N1 influenza virus.(a) Schematic representation illustrating the regimen for oral administration of antiviral agents into A/H1N1 influenza virus-infected mice.Six to seven weeks old female BALB/c mice were orally treated with increasing doses of OSC-GCDI(P) (0.1, 1 and 10 mg/kg) using OS-P and GOC•TFA as controls (n = 5 per group).Four hours later, mouseadapted PR8 (maPR8) was intranasally given to the mice.The same doses of each compound were orally administered again four hours post-infection.This twice-a-day (b.i.d.) administration continued for the next four days.Body weight changes and mortality were recorded daily for 14 days.Mock-infected ('No virus') and maPR8-infected groups ('A/H1N1 infection') served as controls.(b) Body weight changes (left) and survival rates (right) of maPR8-infected mice after OS-P administration.(c) Body weight changes (left) and survival rates (right) of maPR8-infected mice after GOC•TFA administration.(d) Body weight changes (left) and survival rates (right) of maPR8-infected mice after OSC-GCDI(P) administration.In panels (b) to (d), blue and red represent mock infection and maPR8 virus infection, respectively.White, grey and block circles represent 0.1, 1 and 10 mg/kg compound treatments, respectively, for five days into maPR8 virus-infected mice.Mice with a body weight reduction exceeding 30% were euthanized and counted as dead.In the left panels, values are presented as means ± SD from five mice.In the right panels, groups with overlapping survival rate curve are indicated with blue brackets.Statistical significance was determined by comparing time-course survival rates to the maPR8-infected group.*, P < 0.05; **, P < 0.01.

Fig. S10 .
Fig. S10.Effect of oral administration frequency on therapeutic efficacy of OSC-GCDI(P)/HPMCP in mice infected with wild-type influenza virus (a) Body weight changes of maPR8-infected mice after OS-P administration.(b) Body weight changes of maPR8-infected mice after GOC•TFA administration.(c) Body weight changes of maPR8-infected mice after OSC-GCDI(P) administration.In panels (a) to (c), blue and red circles represent mock infection and maPR8 virus infection, respectively.Sky blue, dark green, bright green, yellow and orange circles represent Groups 1 to 5, respectively.Mice with a body weight reduction exceeding 30% were euthanized and counted as dead.In the left panels, values are presented as means ± SD from five mice.In the right panels, groups with overlapping survival rate curve are indicated with blue brackets.Statistical significance was determined by comparing time-course survival rates to the maPR8-infected group.**, P < 0.01.

Fig. S11 .
Fig. S11.Antiviral efficacy of orally administered OSC-GCDI(P)/HPMCP in mice.(a) Schematic representation illustrating the regimen for oral administration of antiviral agents into A/H1N1 influenza virus-infected mice.Six to seven weeks old female BALB/c mice were orally treated with increasing doses of OSC-GCDI(P) at doses of 0.1, 1 and 10 mg/kg (n = 5 per group).(b) In vivo therapeutic efficacy of OSC-GCDI(P) formulated with HPMCP (OSC-GCDI(P)/HPMCP) in A/H1N1 virus-infected mice.Female BALB/c mice (six to seven weeks old) were orally treated with escalating doses of OSC-GCDI(P)/HPMCP, including 0.1 (white circles), 1 (grey circles) and 10 mg/kg (black circles) (n = 5 per group).Four hours later, A/H1N1 virus (maPR8) was intranasally challenged to the mice.The same doses of OSC-GCDI(P)/HPMCP were orally administered again four hours post-infection.The twice-aday (b.i.d.) administration continued for the next four days.Daily monitoring of body weight changes (left) and survival rates (right) were maintained for 14 days.Mock-infected (blue circles) and maPR8infected groups (red circles) served as controls.In the right panel, groups with overlapping survival rate curve are indicated with a blue bracket.

Fig. S12 .
Fig. S12.Effect of oral administration frequency on therapeutic efficacy of OSC-GCDI(P)/HPMCP in mice infected with an OS-resistant influenza virus harboring the H275Y mutation in NA.Female BALB/c mice were orally treated with OSC-GCDI(P)/HPMCP at a dose 10 mg/kg, with OS-P and GOC•TFA serving as controls (n = 5 per group).Four hours later, the OS-resistant A/H1N1 virus (rgA/Korea/09/2009∆53-60(H275Y), referred to as A/H1N1(H275Y)) was intranasally challenged to the mice.Each compound was additionally administered at different intervals: days 1 to 4 post-infection in Group 1, days 2 and 4 in Group 2, day 3 in Group 3 and, day 4 in Group 4. In Group 5, there was no additional treatment after the single administration.Daily recordings of body weight changes and mortality were maintained for 14 days.Mock-infected and A/H1N1(H275Y)-infected groups were included as controls.(a) Body weight changes of A/H1N1(H275Y)-infected mice after administration with OS-P.(b) Body weight changes of A/H1N1(H275Y)-infected mice after administration with GOC•TFA.(c) Body weight changes of A/H1N1(H275Y)-infected mice after administration with OSC-GCDI(P).In panels (a) to (c), blue and red circles represent mock infection and A/H1N1(H275Y) virus infection, respectively.Sky blue, dark green, bright green, yellow and orange circles represent Groups 1 to 5, respectively.Mice with a body weight reduction exceeding 30% were euthanized and counted as dead.In the left panels, values are presented as means ± SD from five mice.In the right panels, groups with overlapping survival rate curve are indicated with blue brackets.Statistical significance was determined by comparing time-course survival rates to the A/H1N1(H275Y)-infected group.*, P < 0.05; **, P < 0.01.

Table S1 .
In Determined by HPLC analysis of the remaining GOC or OSC-GCDI prodrugs at various time points.The calculation of the halflives was based on the assumption of pseudo-first order kinetics of the degradation.b Examined in Sprague Dawley rat microsomes at a prodrug concentration of 100 μM at 37 °C.b Examined in phosphate buffered saline (50 mM phosphate, 154 mM ionic strength, 37 °C).

Table S2 .
Effective permeability through the artificial membrane of GOC and OSC-GCDI(P) with and without HPMCP encapsulation.