Safe Polycationic Dendrimers as Potent Oral in Vivo Inhibitors of Mycobacterium tuberculosis: A New Therapy to Take down Tuberculosis

Article abstract of DOI:10.1021/acs.biomac.1c00355

The long-term treatment of tuberculosis (TB) sometimes leads to nonadherence to treatment, resulting in multidrug-resistant (MDR) and extensively drug-resistant (XDR) tuberculosis. Inadequate bioavailability of the drug is the main factor for therapeutic failure, which leads to the development of drug-resistant cases. Therefore, there is an urgent need to design and develop novel antimycobacterial agents minimizing the period of treatment and reducing the propagation of resistance at the same time. Here, we report the development of original and noncytotoxic polycationic phosphorus dendrimers essentially of generations 0 and 1, but also of generations 2-4, with pyrrolidinium, piperidinium, and related cyclic amino groups on the surface, as new antitubercular agents active per se, meaning with intrinsic activity. The strategy is based on the phenotypic screening of a newly designed phosphorus dendrimer library (generations 0-4) against three bacterial strains: attenuated Mycobacterium tuberculosis H37Ra, virulent M. tuberculosis H37Rv, and Mangora bovis BCG. The most potent polycationic phosphorus dendrimers 1G0,HCl and 2G0,HCl are active against all three strains with minimum inhibitory concentrations (MICs) between 3.12 and 25.0 μg/mL. Both are irregularly shaped nanoparticles with highly mobile branches presenting a radius of gyration of 7 ?, a diameter of maximal 25 ?, and a solvent-accessible surface area of dominantly positive potential energy with very localized negative patches arising from the central N3P3 core, which steadily interacts with water molecules. The most interesting is 2G0,HCl, showing relevant efficacy against single-drug-resistant (SDR) M. tuberculosis H37Rv, resistant to rifampicin, isoniaid, ethambutol, or streptomycin. Importantly, 2G0,HCl displayed significant in vivo efficacy based on bacterial counts in lungs of infected Balb/C mice at a dose of 50 mg/kg oral administration once a day for 2 weeks and superior efficacy in comparison to ethambutol and rifampicin. This series of polycationic phosphorus dendrimers represents first-in-class drugs to treat TB infection, could fulfill the clinical candidate pipe of this high burden of infectious disease, and play a part in addressing the continuous demand for new drugs.

Full text of DOI:10.1021/acs.biomac.1c00355

pubs.acs.org/Biomac
Article
Safe Polycationic Dendrimers as Potent Oral In Vivo Inhibitors of
Mycobacterium tuberculosis: A New Therapy to Take Down
Tuberculosis
Serge Mignani,* Vishwa Deepak Tripathi, Dheerj Soam, Rama Pati Tripathi,* Swetarka Das,
Shriya Singh, Ramakrishna Gandikota, Regis Laurent, Andrii Karpus, Anne-Marie Caminade,
Anke Steinmetz, Arunava Dasgupta,* Kishore Kumar Srivastava,* and Jean-Pierre Majoral*
Cite This: Biomacromolecules 2021, 22, 2659−2675
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: The long-term treatment of tuberculosis (TB) sometimes leads to
nonadherence to treatment, resulting in multidrug-resistant (MDR) and
extensively drug-resistant (XDR) tuberculosis. Inadequate bioavailability of the
drug is the main factor for therapeutic failure, which leads to the development of
drug-resistant cases. Therefore, there is an urgent need to design and develop
novel antimycobacterial agents minimizing the period of treatment and reducing
the propagation of resistance at the same time. Here, we report the development of
original and noncytotoxic polycationic phosphorus dendrimers essentially of
generations 0 and 1, but also of generations 2−4, with pyrrolidinium, piperidinium,
and related cyclic amino groups on the surface, as new antitubercular agents active per se, meaning with intrinsic activity. The
strategy is based on the phenotypic screening of a newly designed phosphorus dendrimer library (generations 0−4) against three
bacterial strains: attenuated Mycobacterium tuberculosis H37Ra, virulent M. tuberculosis H37Rv, and Mangora bovis BCG. The most
potent polycationic phosphorus dendrimers 1G0,HCl and 2G0,HCl are active against all three strains with minimum inhibitory
concentrations (MICs) between 3.12 and 25.0 μg/mL. Both are irregularly shaped nanoparticles with highly mobile branches
presenting a radius of gyration of 7 Å, a diameter of maximal 25 Å, and a solvent-accessible surface area of dominantly positive
potential energy with very localized negative patches arising from the central N3P3 core, which steadily interacts with water
molecules. The most interesting is 2G0,HCl, showing relevant efficacy against single-drug-resistant (SDR) M. tuberculosis H37Rv,
resistant to rifampicin, isoniaid, ethambutol, or streptomycin. Importantly, 2G0,HCl displayed significant in vivo efficacy based on
bacterial counts in lungs of infected Balb/C mice at a dose of 50 mg/kg oral administration once a day for 2 weeks and superior
efficacy in comparison to ethambutol and rifampicin. This series of polycationic phosphorus dendrimers represents first-in-class
drugs to treat TB infection, could fulfill the clinical candidate pipe of this high burden of infectious disease, and play a part in
addressing the continuous demand for new drugs.
global total: India (26%), Indonesia (8.5%), China (8.4%), the
Philippines (6.0%), Pakistan (5.7%), Nigeria (4.4%), Bangla-
desh (3.6%), and South Africa (3.6%).2 Despite the great
reduction of incidence and mortality rate over the years, the
current challenge is the increase in the prevalence of
multidrug-resistant TB (MDR-TB) and extensively drug-
resistant TB (XDR-TB), a form of TB that is resistant to at
least four of the core anti-TB drugs since the introduction of
several types of drugs. In 2016, an estimated 490 000 people
worldwide developed MDR-TB.2 Consequently, TB represents
INTRODUCTION
■
Throughout history, the disease of tuberculosis (TB) has been
known by different names: consumption, phthalmia, and white
́
plague. Cristobal Rojas (1857−1890) was suffering from
tuberculosis and painted the famous “La Miseria” in 1886.1 In
it, he depicts the social aspect of the disease and its relationship
with living conditions at the end of the 19th century.
Tuberculosis (TB) is a contagious, airborne infectious disease
caused most commonly by infection with various pathogen
strains of mycobacteria, such as Mycobacterium tuberculosis
(MTB), and is a major public health burden. MTB is a small
aerobic nonmotile bacillus. According to the World Health
Organization (WHO), an estimated 10 million people fell ill
with TB in 2019. This number has been declining very slowly
in recent years.2 There were close to 1.4 million and 208 000
TB deaths among HIV-negative and HIV-coinfected patients,
respectively.2 Eight countries accounted for two-thirds of the
Received: March 19, 2021
Revised: April 28, 2021
Published: May 10, 2021
https://doi.org/10.1021/acs.biomac.1c00355
Biomacromolecules 2021, 22, 2659−2675
© 2021 American Chemical Society
2659

Biomacromolecules
pubs.acs.org/Biomac
Article
a tough challenge and a public health threat. Moreover,
according to the WHO, the recent Covid-19 pandemic
threatens to reverse the progress made in recent years in the
fight against tuberculosis. As a result, the countries most
affected by tuberculosis are focusing more on Covid-19
detection than on TB detection. According to the WHO
annual report, TB could cause 200 000−400 000 more deaths
this year than the 1.4 million in 2019, despite the existence of a
cure.2
Several of these nanoparticles are coated, for instance, with
polyacrylic acid and polysaccharide (chitosan). PLG wheat-
germ agglutinin-coated PLG, poly(lactic-co-glycolic acid)
(PLGA), chitosan, and gelatin are the main nanocarriers
used, whereas rifampicin, isoniazid, pyrazinamide, and
ethambutol are the main drugs for both oral and pulmonary
deliveries. In addition, several surface-functionalized nano-
particles, such as gelatin, cationic lipids, and PLGA coated with
mannose, lactose, mycolic acids, and PEG, with carried drugs
such as isoniazid, rifampicin, and moxifloxacin were developed
for antimycobacterial drug delivery.
Effective treatment is difficult due to the specific structure
and chemical composition of the mycobacteria cell wall
hindering the entry of drugs, which renders many drugs and
antibiotics ineffective. However, the antibiotics isoniazid,
rifampicin (RIF), pyrazinamide, ethambutol, and streptomycin
are the major first-line pulmonary anti-TB drugs that are all
typically applied as 6 month treatment.3 The majority of anti-
TB drugs are administered orally except for Streptomycin,
Amikacin, Kamamycin, and Capreomycin, which are adminis-
tered intravenously. Also, natural compounds with potential for
the treatment of TB were recently highlighted4 as well as
vaccines.5 Poor patient compliance to prescribed regimens and
the emergence of resistant strains are the major problems
during TB long-term treatment, which involves continuous,
frequent multiple drug dosing. At this point, two strategies
have been developed: (1) shortening the treatment duration
and reducing the dosing frequency with the capacity to tackle
drug-resistant TB; and (2) developing new drug formulations
to reduce toxic side effects. Indeed, the introduction of long-
duration drug formulations that release antimicrobial agents
incorporating both hydrophilic and hydrophobic substances
in a slow and sustained manner, allowing the reduction in
frequency and total number of doses, represents an important
therapeutic strategy against TB.6 In this direction, several
stable nanoparticle-based formulations for the controlled and
sustained release of first-line anti-TB drugs such as isoniazid,
rifampicin, and pyrazinamide have been described and show
good efficiency in in vivo models. For instance, poly(lactide-co-
glycolide) (PLG) nanoparticles, lectin-functionalized PLG
nanoparticles, and solid lipid nanoparticles were developed
by the oral, subcutaneous, or aerosol route of administration.7
Recently, several nanoparticles for the delivery of anti-TB
drugs directly to the lungs via the respiratory route have been
analyzed. Globally speaking, the main advantages of using
nanoparticles, e.g., lipids, polymers, and proteins, in inhaled
anti-TB drug delivery are as follows: (1) direct drug delivery in
the lung targeting alveolar macrophages; (2) decreased risk of
systemic toxicity; (3) delivery of multiple drugs simulta-
neously; (4) shortened treatment regimen; and (5) improved
patient compliance.8 The cost of treatment, stability, and large-
scale production of drug formulations are the main issues to be
analyzed to bridge the gap between theory and clinical reality.9
Taken together, there is an urgent need for new efficacious
and affordable chemotherapeutics with higher efficacy, which is
the main cause for failures in clinical drug development.10,11
Recently, an interesting review highlighted the development
of nanoparticles in TB treatment.12 Two strategies have been
developed: (1) antimycobacterial activities based on intrinsic
activities; and (2) nanoparticles as nanocarriers of known
antitubercular drugs for oral or pulmonary administrations.
Silver, gold, iron, iron oxide, and gallium nanoparticles
targeting macrophages were used in vitro and in vivo with
intrinsic antimycobacterial, in addition to antibacterial activity.
Achievement of biodegradable profiles is the main challenge.
To the best of our knowledge, no report emphasized the
development of dendrimers, which can be considered as soft
nanoparticles, as drug per se against TB. Earlier works have
described the use of dendrimers as nanocarriers and have been
highlighted in several reviews.13,14 Dineshkumar et al.
described the long-duration release of rifampicin-loaded
PEGylated poly(amidoamine) (PAMAM) dendrimers,15
whereas Bellini et al. showed the high stability of the
rifampicin−PAMAM complex under physiological pH and
the rapid release of rifampicin molecules under an acidic
medium, which is similar to the acidic domains of the
macrophage where the Mycobacterium resides.16 Interestingly,
Jain et al. described the intracellular macrophage uptake of
rifampicin-loaded mannosylated PPI dendrimers,17 whereas
Singh et al. used the same dendrimers for delivery of
isoniazid.18 A recent review has emphasized these new ways
to treat tuberculosis using dendrimers as nanocarriers.19
Dendrimers are three-dimensional macromolecules with a
high degree of molecular uniformity and a perfectly controlled
shape, surface chemistry, and size in the range of 1−15 nm. A
wide range of literature about dendrimers is available today.20
They display dendritic branches composed of hydrophobic and
hydrophilic moieties radiating out from a central core unit.
Interior-layer generations, Gn (G), where n ranges generally
from 0 to 6 and more exceptionally to 12 and even 13, are
made of regularly repeating branching units attached to the
core. The dendrimer diameter increases nonexponentially,
whereas the number of surface groups increases exponentially
for each generation. Because of their highly branched, three-
dimensional architecture and versatile surface functionaliza-
tion, dendrimers have been engineered for use as nanodevices,
either nanocarrier drug approaches21 or drugs per se,22 as well
for the construction of dendrimer hybrid materials;23 for
catalytic properties,24 PAMAM dendrimers are the most
extensively considered.25
For several years, Majoral, Caminade, and co-workers have
developed biocompatible phosphorus dendrimers, which
represent a versatile platform in drug delivery, e.g., drug and
gene transfection delivery, as well as drugs themselves.26 The
high level of control that is possible over their architectural
design, allowing for tunable control of sizes and shapes of cores
and interiors as well as tunable surface functionality, is the
main advantage of phosphorus dendrimers. Phosphorus
dendrimers have been employed as drugs per se in different
therapeutic fields as, for instance, anticancer,27 antiprions,28
anti-Alzheimer’s,29 anti-inflammatory agents,30 etc. In addition,
a large variety of administration routes can be used with
dendrimers in general and phosphorus dendrimers in
particular: intravenous, intraperitoneal, ocular, transdermal,
oral, intranasal, and pulmonary routes.31 The latter is very
important in the case of TB because TB is caused by M.
tuberculosis bacteria that most often affect the lungs.
2660
https://doi.org/10.1021/acs.biomac.1c00355
Biomacromolecules 2021, 22, 2659−2675

Biomacromolecules
pubs.acs.org/Biomac
Article
Scheme 1. General Syntheses of 1G0,HCl−5G0,HCl Polycationic Phosphorus Dendrimers
Scheme 2. General Syntheses of 6G1,HCl−12G1,HCl Polycationic Phosphorus Dendrimers
Herein, we report the development of original noncytotoxic
polycationic phosphorus dendrimers as new antitubercular
agents active per se, meaning intrinsic in vivo activity. The
strategy is based on the phenotypic screening of a newly
designed phosphorus dendrimer library versus, for instance,
target-based32 or fragment-based approaches.33 These intrinsi-
cally active nanoparticles represent first-in-class drugs to treat
TB infection, to fulfill the clinical candidate pipe of this high
burden of infectious disease, and to play a part in addressing
the continuous demand for new antitubercular drugs. We
decided to develop nanoparticles in general and phosphorus
dendrimers in particular due to three important points based
on medical needs: (1) limited number of drugs to treat MDR-
TB and XDR-TB, (2) substantial side effects including toxicity
issues of drugs available, and (3) less effective second-line
drugs. The antimycobacterial screening has been conducted
against three bacterial strains: attenuated strain M. tuberculosis
H37Ra, virulent M. tuberculosis H37Rv, and against Mangora
bovis BCG.
2661
https://doi.org/10.1021/acs.biomac.1c00355
Biomacromolecules 2021, 22, 2659−2675

Biomacromolecules
pubs.acs.org/Biomac
Article
Scheme 3. General Syntheses of 5G1 and 13G1−20G1, 5G1,HCl, and 13G,HCl−20G1,HCl Polycationic Phosphorus Dendrimers
Scheme 4. General Synthesis of Dendrimers 6G2,HCl, 7G2,HCl, 6G3,HCl, 7G3,HCl, 6G4,HCl, and 7G4,HCl
thionyl chloride as the chlorinating agent,34 which is a precursor of
EXPERIMENTAL SECTION
■
dendrimers 1G0−5G0, by substitution of the chloride by diverse
amines in the presence of potassium carbonate at room temperature
(5a−e) in good yields. Finally, water-soluble compounds are obtained
(1G0,HCl−5G0,HCl) by adding 1 equiv of HCl diethylether per
terminal function (6 equiv for generation 0 compounds).
Synthesis of Dendrimers. Scheme 1 shows the preparation of
dendrimers of generation 0 (compounds 1G0−5G0 and of their
corresponding salt). In the first step, 4-hydroxybenzyl alcohol (2) is
grafted to the cyclotriphosphazene derivative N3P3Cl6 in the presence
of cesium carbonate at room temperature to afford compound 3. The
six benzylalcohol groups are then converted to benzylchloride 4 using
2662
https://doi.org/10.1021/acs.biomac.1c00355
Biomacromolecules 2021, 22, 2659−2675

Biomacromolecules
pubs.acs.org/Biomac
Article
Scheme 2 highlights the general preparation of dendrimers of
generation 1 (compounds 6G1−12G1 and of their respective salt).
The starting point is the generation 1 dendrimer (6), built from the
cyclotriphosphazene core.35,36 The nucleophilic substitution with
diverse primary amines (7a−g) bearing through a C2 or C3 linker a
cyclic amine as the substituent37 in N,N-diisopropylethylamine
(DIPEA) affords in good to excellent yields the dendrimers 6G1−
with water molecules, radii of gyration, and radial distribution
functions (rdf). Standard quality analysis showed stable simulations in
terms of potential and total energy, volume, temperature, and
pressure. The 1 μs simulation time sampled the conformational space
well as deduced from the similar torsional distributions of equivalent
rotatable bonds of the branches.
Bacterial Strains, Cell Lines, Growth Conditions, and
Animals. M. tuberculosis H37Ra, M. tuberculosis H37Rv, and M.
bovis BCG were collected from American type culture collection
(ATCC). Cultures were grown in a middlebrook (MB7H9) medium
supplemented with 0.2% glycerol, 0.05% tween 80, and 10% ADC.
Cultures were incubated at 37 °C. The J774A.1 murine macrophage
cell line and VERO cells were acquired from ATCC and cultured in
an RPMI-1640 medium containing 2 mM L-glutamine, 1 mM sodium
pyruvate, 4.5 g/L glucose, 1.5 g/L sodium bicarbonate, and 10 mM
HEPES, supplemented with 10% heat-inactivated fetal bovine serum
(FBS) at 37 °C and 5% CO2.
Antimycobacterial Activity. Microplate Alamar Blue Assay
(MABA). For the experiments, we used 96-well microplates with a
flat bottom and well-grown MTB H37Ra and M. bovis BCG cultures
of the mid-log phase (OD600 = 0.5−0.6) either in Sauton’s medium
supplemented with 3.5% glycerol and tween 80 or in MB7H9
medium. Initially, dendrimer dilutions were prepared according to
their solubility, and subsequent 2-fold dilution was performed in 0.15
mL of suspension cultures in each well of the microplate. Two control
wells were prepared, one containing only the suspension culture and
the other containing the suspension culture plus drugs (rifampicin
and streptomycin). Microplates were incubated at 37 °C for 2−3
days. After 48 h of incubation, 10% alamar blue dye solution
(Resazurin 0.03%) was added to all of the wells, microplates were
reincubated at 37 °C for 2−4 h, and fluorescence was measured using
a spectrophotometer in the range from 530 to 590 nm.
12G1. The corresponding water-soluble compounds (6G1,HCl
−
12G1,HCl) were obtained by adding 12 equiv of HCl in diethylether
to each dendrimer in quantitative yields.
Scheme 3 describes the general synthesis of dendrimers of
generation 1 (compounds 5G1, 13G1−20G1 and of their
corresponding salt). The first step is the simple condensation reaction
at room temperature between the first-generation dendrimer
functionalized with 12 aldehydes (compound 8),30,31 with several
amines 7a, b, e, f, h, i, j, k, and l already used in Scheme 2. The
second step is the reduction of the imine amines 9a, b, e, f, h, i, j, k,
and l with NaBH4 at room temperature; such a reaction occurs only
on the imines and not on the hydrazone linkages,38,39 affording
dendrimers 5G1 and 13G1−20G1 in moderate to good yields.
Protonation reaction was carried out using 12 equiv of HCl per
dendrimer, even if at least 24 N per dendrimer could be protonated,
affording 5G1,HCl and 13G,HCl−20G1,HCl in quantitative yield.
Scheme 4 depicts the general synthesis of dendrimers of generation
2 (compounds 6G2 and 7G2 and their respective salts), generation 3
(6G3 and 7G3 and their respective salts), and generation 4 (6G4 and
7G4 and their respective salts). The method of synthesis is the same
as the one shown in Scheme 2, but using 24 equiv of amine for
generation 2, 48 equiv for generation 3, and 96 equiv for generation 4.
The water-soluble phosphorus dendrimers are obtained by adding 1
equiv of HCl per terminal function, i.e., 24 for G2, 48 for G3, and 96
for G4.
Mycobacteria Growth Indicator Tube (MGIT) Assay. The
BACTEC MGIT 960 instrument is a fully automated system that
exploits the fluorescence of an oxygen sensor to detect the growth of
mycobacteria in culture. The MGIT tube contains 7 mL of the
modified middlebrook 7H9 broth base with OADC enrichment and
PANTA, a mixture of antibiotics. For this experiment, 500 μL of MTB
H37Rv cultures of the mid-log phase (OD600 = 0.5−0.6) was added to
each MGIT tube to which dendrimers at various concentrations were
added. The relative growth ratio between the drug-containing tube
and the drug-free growth control (GC) tube was determined by the
system’s software algorithm. The final interpretation and the
susceptibility results were reported by the M960 instrument
automatically. The drugs from the MGIT SIRE kit were used as a
positive control.
Activity against Drug-Resistant M. tuberculosis Strains. Activities
of the dendrimers were determined using the MABA assay against
drug-susceptible M. tuberculosis H37Rv ATCC 27294 (virulent
tubercle bacillus), INH-resistant M. tuberculosis ATCC 35822
(isoniazid-resistant), RIF-resistant M. tuberculosis ATCC 35838
(rifapicin-resistant), STR-resistant M. tuberculosis ATCC 35820
(streptomycin-resistant), and ETB-resistant M. tuberculosis ATCC
35837 (ethambutol-resistant).
Cytotoxicity Assays. For the cytotoxicity assay, Vero cells (African
green monkey kidney cell line) were added to a 96-well plate at a
concentration of 4 × 105 cells/well in a complete RPMI medium.
Vero cells were treated with dendrimers at different concentrations
(10- to 12-fold of MIC) and were incubated at 37 °C with 5% CO2
for 24 h to determine the response of dendrimer activity on cells. Two
control wells, one containing only the suspension culture as a negative
control and the other containing the suspension culture with ethanol
as the positive control, were used. After 24 h of incubation, the 10%
alamar blue dye solution (Resazurin 0.03%) was added to all of the
suspension wells, microplates were reincubated at 37 °C for 2−4 h,
and fluorescence was measured at 530/590 nm. The assays were done
in triplicate.
Molecular Modeling. Molecular modeling employed the Drug
̈
Discovery Suite release 2020.u2 commercialized by Schrodinger Inc
executed on a Linux workstation with the CentOS7 operating system.
Three-dimensional all-atom models were constructed manually in the
graphical interface Maestro, energy-optimized or -minimized to
convergence by programs Jaguar or Macromodel.40,41 Conformational
searches and molecular dynamics (MD) simulations were conducted
by programs Macromodel and Desmond, respectively.42 Molecular
mechanics (MM) all-atom molecular force field OPLS3 with an
implicit aqueous solvent and standard settings were applied unless
indicated otherwise.43 In more detail, special care was taken to
construct the correct geometry of the N3P3 core as OPLS3 did not
yield a planar conformation by MM energy minimization: cyclic
(NP(OH)2)3 was quantum-mechanically energy-optimized to con-
vergence in vacuo by the density functional theory (DFT) at the level
of the B3LYP-D3 theory applying the basis set 6-31G, ** polarization,
++ diffuse basis functions, and no symmetry. Fully protonated models
of 1G0 and 2G0 were derived from the optimized N3P3 core, the
geometry of which was subsequently frozen in MM energy
minimization, conformational searches, and MD simulations. Low-
frequency-mode conformational searches of all rotatable, noncyclic
bonds were executed for 1000 steps, minimizing every conformer for
2500 steps maximum, which only exceptionally did not yield a
converged geometry. The lowest energy conformers of fully
protonated 1G0 and 2G0 served to calculate solvent-accessible
surfaces and their properties as well as to set up MD simulations using
the explicit SPC water model at pH 7, 0.154 M NaCl, and neutralizing
the system by chlorine ions. Box margins of 20 Å in all three
dimensions with an optimized box orientation were applied. MD
simulations were prepared and submitted via the graphical interface
Maestro. Simulations on NVIDIA V100 graphical processing units
(GPUs) for 1 μs as NPT ensembles at 300 K and 1 atm constraining
the N3P3 core by weight 300 employed infinite boundary conditions,
the Nose−Hoover chain thermostat, and the Martyna−Tobias−Klein
barostat after standard pre-equilibration. Energy values and atomic
coordinates were recorded every 25 and 250 ps, respectively.
Trajectories were analyzed interactively via Maestro by calculating
Simulation Interaction Diagrams, hydrogen bonds of core or branches
Inhibition of Mycobacterial Growth and Survival in Infected
J774A.1 Cells. The J774A.1 murine macrophage cell line was
maintained as described previously. For infection, the J774A.1 cells
2663
https://doi.org/10.1021/acs.biomac.1c00355
Biomacromolecules 2021, 22, 2659−2675

Biomacromolecules
pubs.acs.org/Biomac
Article
Figure 1. General 2D structure of an example of a polycationic phosphorus dendrimer.
were seeded at the density of 5 × 104 cells/well in 24-well tissue
culture plates. The log-phase mycobacterial cultures were harvested
by centrifugation at 2000g for 10 min, washed two times with sterile
phosphate-buffered saline (PBS), and finally suspended in the RPMI-
1640 medium. The prepared bacterial culture was added in
macrophages at 1:10 MOI and allowed to undergo phagocytosis for
4 h at 37 °C and 5% CO2. After 4 h of phagocytosis, the medium was
removed and washed twice with an incomplete RPMI-1640 medium
and supplemented with 100 μg/mL amikacin for 2 h for killing of
extracellular mycobacteria. After 2 h of amikacin treatment, the
medium was removed and washed twice with an incomplete RPMI-
1640 medium. The mycobacteria H37Rv-infected J774A.1 cells were
treated with dendrimers at different concentrations, and cells were
incubated at 37 °C with 5% CO2 for 48 h to determine the response
of dendrimer activity on intracellular bacteria. After 48 h of dendrimer
treatment, the medium was removed, and cells were washed twice
with sterile PBS, and finally 0.05% sodium dodecyl sulfate (SDS) was
added to lyse the cells to release the bacteria. The intracellular
bacteria were calculated using BACTEC MGIT 960 systems for the
detection and recovery of mycobacteria by adding 0.1 mL of
harvested MTB H37Ra suspension in MGIT tubes and allow reading
tubes’ bar code through the BACTEC MGIT 960 system. The tubes
were monitored each day for fluorescence.
(approximately between 2 and 4 g) of muscle mass and their behavior
was found to be aggressive (data not shown).
Efficacy of 2G0,HCl in Balb/C Mice against Mtb H37Ra. To
determine the colony-forming units (CFUs) in lungs of infected (with
Mtb H37Ra) Balb/C mice, we used five groups (no drug, 2G0,HCl
with 33 mg/kg, 2G0,HCl 50 mg/kg, rifampicin 10 mg/kg, and
ethambutol 100 mg/kg) of Balb/C mice.
Female BALB/c mice, 5−6 weeks old, were taken for the
experiment. The mice were divided into five groups (n = 5) of 6
mice in each groups. All groups were injected with MTB H37Ra
intravenously. Animals were treated orally starting from seven days
post infection daily for 2 weeks. After 2 weeks, all of the mice were
sacrificed, and lungs and spleen were homogenized in 5 mL of normal
saline, serially diluted, and plated on Middlebrook 7H11 agar plates
supplemented with OADC for bacillary load detection. After
incubation for 4 weeks at 37 °C, colony-forming units (CFUs)
were enumerated and the data were averaged across experiments.
Statistical Analysis. Statistical analysis was performed using
GraphPad Prism 6.0 software (GraphPad Software, La Jolla, CA).
Comparison between three or more groups was performed using one-
way analysis of variance (ANOVA), with post hoc Tukey’s multiple
comparison tests. P-values of <0.05 were considered to be significant.
Preliminary Mechanism of Action Studies of 2G0,HCl. For
this, a sublethal concentration of 2G0,HCl (1.56 μg/mL) was used to
treat MTB H37Ra for 2 h. Equal amounts of whole MTB cell lysate
proteins were separated by 1-DE, followed by 2-DE on 12% sodium
dodecyl sulfate−polyacrylamide gel electrophoresis (SDS-PAGE), and
stained with Coomassie Brilliant Blue R-250. Stained spots were cut
out from the gels to process for matrix-assisted laser desorption/
ionization (MALDI) mass spectrometry (MS)−MS using the
standard gel digestion method with trypsin. Differently expressed
proteins were identified. Three biological and technical replicates
were used to derive the data.
Animal Experiments. Animal experiments were performed on 6-
week-old BALB/c mice procured from the National Laboratory
Animal Facility, CSIR-CDRI. The experimental protocols were
approved by the Institutional Animal Ethics Committee.
Determination of Maximum Tolerated Dose and Toxicity of
Dendrimers 2G0,HCl in Mice. Based on high antitubercular activities
of prepared dendrimers (Tables 2−6), the most efficient compound,
2G0,HCl, was selected for in vivo experiments. Balb/c mice (6−7
weeks) of 20 g weight were taken and were divided into 3 groups of 3
mice each based on the given dose of 2G0,HCl in mice. The oral dose
set (in water) for the experiment was 50 mg/kg for the first group, 33
mg/kg for the second group, and no drug (MD) in the third group.
Route for the administration of the drug was kept oral, and the mice
were examined daily. All of the animals in each group survived up to 3
months (100% survival after one-time daily administration during 2
weeks) and then were euthanized. In the everyday examination, we
found that all mice of group one and group two showed weight loss
RESULTS AND DISCUSSION
■
Synthesis of Dendrimers. Our strategy for synthesis was
to construct an original polycationic phosphorus dendrimer
library of 27 members to evaluate the influence of the cationic
moieties, the nature of the branches, and the size (generations:
2664
https://doi.org/10.1021/acs.biomac.1c00355
Biomacromolecules 2021, 22, 2659−2675

Biomacromolecules
pubs.acs.org/Biomac
Article
Table 1. Chemical Moieties on the Surface of Dendrimers of Generations 0 and 1 and from the Last Phosphorus Atom on the
Considered Branch
a
Table 2. Chemical Structure and Antitubercular Activity of Generation 0 Dendrimers
a
The 2-fold dilutions of each compound were screened. The experiments were repeated three times.
Gn, n = 0, 1, 2, 3, and 4), on their antimycobacterial activity.
Figure 1 shows the general two-dimensional (2D) structure of
an example of polycationic phosphorus dendrimer of
generation 1.
Pyrrolidino, N-methyl pyrrolidino, piperidino, N-methyl
piperidino, morpholino, imidazolino, 1-phenyl-piperazino,
pyrazine-2-carbonyl, and nicotinic groups were introduced on
the surface of phosphorus dendrimers to evaluate the influence
of the cationic moieties on their antimycobacterial activity.
Tables 1−6 represent all of the dendrimers prepared, classed
by generation, and tested. The construction of these
phosphorus dendrimers was also based on chemical feasibility
and chemical stability, as well as their aqueous solubility. As
shown in the Experimental Section and supplementary
2665
https://doi.org/10.1021/acs.biomac.1c00355
Biomacromolecules 2021, 22, 2659−2675

Biomacromolecules
pubs.acs.org/Biomac
Article
a
Table 3. Chemical Structure and Antitubercular Activity of Generation 1 Dendrimers
a
The 2-fold dilutions of each compound were screened. The experiments were repeated three times.
materials, the straightforward multistep synthesis was adopted
for the preparation of the phosphorus dendrimers of
generation 0 (Scheme 1, compounds 1G0,HCl−5G0,HCl),
generation 1 (Schemes 2 and 3, compounds 5G1,HCl
−
2666
https://doi.org/10.1021/acs.biomac.1c00355
Biomacromolecules 2021, 22, 2659−2675

Biomacromolecules
pubs.acs.org/Biomac
Article
a
Table 4. Chemical Structure and Antitubercular Activity of Generation 2 Dendrimers
a
The 2-fold dilutions of each compound were screened. The experiments were repeated three times.
a
Table 5. Chemical Structure and Antitubercular Activity of Generation 3 Dendrimers
a
The 2-fold dilutions of each compound were screened. The experiments were repeated three times.
a
Table 6. Chemical Structure and Antitubercular Activity of Generation 4 Dendrimers
a
The 2-fold dilutions of each compound were screened. The experiments were repeated three times.
20G1,HCl), generation 2 (Scheme 4, 6G2,HCl and 7G2,HCl),
generation 3 (Scheme 4, 6G3,HCl, 7G3,HCl), and generation 4
(Scheme 4, 6G4,HCl and 7G4,HCl). All compounds were first
synthesized and characterized in the neutral form, which are
named XGn, with X being the number of the compound and n
being the generation (number of layers) of the dendrimer. The
neutral dendrimers are generally less soluble in water; thus,
they are protonated (with HCl) to produce their correspond-
ing water-soluble dendrimers. These cationic dendrimers are
groups such as pyrrolidinium, N-methyl pyrrolidinium,
piperidinium, N-methyl piperidinium, morpholinium, imidazo-
linium, 2-methyl-imidazolinium, 1-phenyl-piperazinium, and
(2-methoxy)-1-phenyl-piperazinium. As shown in Table 1,
several linker types have been introduced between the
cyclotriphosphazene core ring and the diverse amino groups
on the surface.
We decided to develop a phenotypic screening strategy, in
contrast to the target-based strategy, although the latter has
been widely used in drug discovery since the molecular biology
revolution of the 1980s.44 Indeed, the main advantages of
phenotypic screens are as follows: (1) design and construction
of preselected compound collections with druglike properties
and optimizable chemical features to monitor efficacy and
druggability, in our case, construction of a biocompatible and
named XGn,HCl
.
As depicted in Schemes 1−3, first, we prepared a library of
dendrimers for generation 0 (6 amino groups on the surface)
and generation 1 (12 amino groups on the surface) composed
of 5 and 16 functionalized dendrimers, respectively (Table 1).
These dendrimers have on their surface diverse types of amino
2667
https://doi.org/10.1021/acs.biomac.1c00355
Biomacromolecules 2021, 22, 2659−2675

Biomacromolecules
pubs.acs.org/Biomac
Article
Figure 2. Three-dimensional models of fully protonated 1G0 (A) and 2G0 (B). The models are the lowest energy conformations issued from
conformational searches applying an implicit solvent model. Solvent-accessible surface area (SASA) arising from the entire dendrimer or N3P3 core
only is depicted as a mesh or solid, respectively; color ramp by potential energy from −0.3 in red over neutral in white to 0.3 in blue. The chemical
structure of the dendrimers is depicted as tubes with carbon, oxygen, nitrogen, phosphor, and hydrogen atoms indicated in green, red, blue, purple,
and white, respectively; nonpolar hydrogen atoms are not represented for clarity.
Figure 3. Snapshots of 1G0,HCl and 2G0,HCl from the molecular dynamics simulations in (A) and (B), respectively. Classical and aromatic
hydrogen bonds to water, ion pairs with chlorine, parallel or perpendicular π-stacking, and cation−π interactions are indicated by dashed lines in
yellow, light blue, magenta, cyan, and green, respectively. Dendrimer and water molecules are depicted in tube representation and chlorine ions as
spheres colored in plum; otherwise, the color coding is as in Figure 2. Nonpolar hydrogens are not depicted for clarity. Dotted lines in orange or
brown exemplarily highlight protonated tertiary amine distances giving rise to radial distribution function (rdf) maxima Max1 or Max2, respectively,
in the Supporting Information Figures F9C and F10C; dash-dotted lines in magenta exemplarily indicate distances of rdf in the Supporting
Information Figures F9A/B and F10A/B.
tunable phosphorus dendrimer collection within the dendrimer
space;45,46 (2) powerful assay development such as affinity,
cellular profiling, and functional genomics-based approaches to
screen the collections; (3) elucidation of the target(s) related
to the mechanism of action by several powerful well-known
technologies;38 (4) elucidation of off-targets based on
unwanted secondary phenotypes; (5) high attrition rate in
the anti-infective drug discovery;47 and (6) limited validated
targets in the TB field.32
Based on the very interesting anti-TB activities of the
protonated form of dendrimers 1G0, 2G0, 6G1, 7G1, 9G1,
11G1, and 20G1 (Tables 2 and 3), we decided to extend the
structure−activity relationships (SAR) to the construction of
libraries of generations 2 (G2, Table 4), 3 (G3, Table 5), and 4
(G4, Table 6) incorporating two phosphorus dendrimer types
each. These phosphorus dendrimers have protonated pyrroli-
dino and piperidino moieties, which showed good anti-TB
activities with dendrimers of generations 0 and 1 (Tables 2 and
3). Generations 2 and 3 are decorated on their surface with 24
or 48 protonated pyrrolidino or piperidino groups, while 96
pyrrolidinium and piperidinium groups are grafted on the
surface of a dendrimer of generation 4.
Furthermore, and importantly, 2G0,HCl is stable at room
temperature in an aerated aqueous solution at pH 5.5−6 for at
least 9 months without any chemical degradation. This point is
very important for in vivo experiments (vide infra) and
potential clinical development.48,49
Molecular Modeling Studies. Conformational space,
surface properties, and molecular dynamics of three-dimen-
sional all-atom models of 1G0,HCl and 2G0,HCl of the
developed phosphorus dendrimer library were explored by
classical force-field methods. Conformational searches identi-
fied the low potential energy conformations depicted in Figure
2. Both fully protonated 1G0 and 2G0 present a solvent-
accessible surface area (SASA) with very dominantly positive
or neutral potential energy arising to over 85% from the
branches. Very localized spots of less than 15% of the entire
surface have a negative potential energy arising from solvent-
2668
https://doi.org/10.1021/acs.biomac.1c00355
Biomacromolecules 2021, 22, 2659−2675

Biomacromolecules
pubs.acs.org/Biomac
Article
Table 7. Antitubercular Activity of 2G0,HCl against Relevant Single-Drug-Resistant Mtb Strains and versus Selected
a
Commercial Antituberculosis Drugs Rifampicin, Isoniazid, Ethambutol, and Streptomycin
MTB strain resistant to
S. No.
compound
M. tuberculosis H37Rv (μg/mL)
INH (μg/mL)
RIF (μg/mL)
ETB (μg/mL)
STR (μg/mL)
1
2
3
4
5
6
2G0,HCl
3G0,HCl
rifampicin
isoniazid
ethambutol
streptomycin
3.12
6.25
0.04
0.04
3.12
1.56
6.25
25
0.04
>64
NT
NT
6.25
12.5
>64
0.04
NT
6.25
25
0.19
0.04
>64
12.5
6.25
>64
0.04
0.04
0.78
>64
NT
a
RIF, rifampicin; INH, isoniaid; ETB, ethambutol; STR, streptomycin; NT, not tested.
exposed regions of the N3P3 core. Such surface patches give
rise to hydrogen bonds formed by the core to water molecules
steadily observed during molecular dynamics simulations.
These molecular dynamics simulations of 1G0,HCl and
2G0,HCl in the explicit aqueous solution evince irregularly
shaped nanoparticles of changing forms with a radius of
gyration of roughly 7 Å (Supporting Information Figures F5
and F6) and minimal and maximal dimensions of 14−15 Å and
about 25 Å, respectively (Supporting Information Figures F7
and F8). The core typically maintains hydrogen bonds to 3 or
4 water molecules throughout the entire simulation (Support-
ing Information Figures F5 and F6). The protonated tertiary
amines of the branches distribute rather evenly around the
central N3P3 core (Figure 3), and their spatial distribution
relative to each other is very similar for 1G0,HCl and 2G0,HCl
(Supporting Information Figures F9 and F10). The nitrogen
atoms tend to be at a distance of 7.5 Å when analyzing
branches above or below the core and assuming conformations
that extend roughly perpendicularly to the N3P3 plane
(Supporting Information Figures F5B, F9A, and F10A,B).
These conformations give rise to distances of about 14 Å when
analyzing nitrogen atoms above the N3P3 plane with respect to
those below (Supporting Information Figures F9C and F10C).
Branch conformations that strongly tilt toward the N3P3 plane
and approach nitrogen atoms of branches above and below the
plane again give rise to nitrogen distances of about 7 Å
(Supporting Information Figures F9C and F10C). Branch
conformations are highly mobile, switching frequently from
perpendicular to tilted and back, thereby distributing the
positively charged nitrogen atoms evenly around the core and
optimizing the electrostatic repulsion between the charged
groups. Furthermore, 1G0,HCl and 2G0,HCl show numerous
intramolecular interactions or form hydrogen bonds and ion
pairs with solvent molecules (Figure 3). Transitory ion pairs
formed with chlorine ions include two branches simultaneously
binding to one chlorine ion. Classical and aromatic hydrogen
bonds to water, intramolecular parallel or perpendicular π-
stacking, and intramolecular cation−π interactions are also
observed (Figure 3). Altogether, 1G0,HCl and 2G0,HCl show
very similar molecular dynamics and globally expose the same
appearance to potentially interacting macromolecules of their
biological environment.
with localized negative spots arising from the central N3P3 ring.
The protonated amines form transient ion pairs with chlorine
in MD simulations, while the central core typically interacts
with three to four water molecules. Altogether, they present a
very similar appearance to macromolecules of their biological
environment.
In Vitro Activity of Phosphorus Dendrimers and Their
Cytotoxicity. The entire phosphorus dendrimer collection
was first tested for its antimycobacterial activity against three
different strains such as attenuated and virulent M. tuberculosis
H37Ra and H37Rv, as well as Mycobacterium bovis BCG. The
toxicity of the phosphorus dendrimers was evaluated against
Vero cell lines, which were isolated from kidney epithelial cells
extracted from African green monkey. All of the assays were
performed in triplicate wells.
Within the generation 0 series, the antimycobacterial
activities assessed as minimum inhibitory concentration
(MIC) against the three strains H37Ra, BCG, and H37Rv in
order are as follows: 1G0,HCl (MICs in μg/mL = 12.5 for Ra,
12.5 for BCG, and 3.12 for Rv), 2G0,HCl (MICs in μg/mL =
25.0 for Ra, 25.0 for BCG, and 3.12 for Rv), 3G0,HCl (MIC =
6.25 μg/mL against Rv), and 5G0,HCl (MICs in μg/mL = 50
for Ra and 100 for BCG) in extracellular killing of
mycobacteria (Table 2). The dendrimer 4G0,HCl displayed
no antimycobacterial activity. 1G0,HCl has pyrrolidinium
groups on its surface, whereas 2G0,HCl has piperidinium
groups with the same linker 4-(aminomethyl)phenoxyl moiety.
The replacement of the pyrrolidinium or the piperidinium
group on the surface by either a morpholinium group
(compound 3G0,HCl) or a 4-phenyl piperidinium group
(compound 4G0,HCl) decreases the antimycobacterial activities
versus 1G0,HCl and 2G0,HCl. The increase in chain length of
1G0,HCl, leading to dendrimer 5G0,HCl with the 4-(((2-
aminoethyl)amino)methyl)phenoxy chain, decreased also the
antimycobacterial activities against the three strains. Interest-
ingly, the safety index SI defined as CC50 against the Vero cell
line divided by MICs is 16.02 versus Ra and 3.1 versus Ra for
2G0,HCl and 1G0,HCl, respectively. Table 7 shows the
antitubercular activity of 2G0,HCl against relevant single-drug-
resistant Mtb strains and versus selected commercial
antituberculosis drugs such as rifampicin, isoniazid, ethambu-
tol, and streptomycin. Very interestingly, when the activities of
2G0,HCl and 3G0,HCl were tested against SDR M. tuberculosis,
2G0,HCl was found to be active against all of the SDR strains,
whereas 3G0,HCl demonstrated a moderate activity against
rifampicin-, isoniazid-, and ethambutol-resistant H37Rv but
was inactive against streptomycin-resistant H37Rv. Conse-
quently, 2G0,HCl represents a particularly interesting and safe
anti-TB agent based on cellular assays.
These two most potent dendrimers have a very simple
structure: generation 0 bearing 6 groups on their surface, either
pyrrolidinium groups (1G0,HCl) or piperidinium groups
(2G0,HCl). They are irregularly shaped dendrimers of changing
form due to highly mobile branches. Their radius of gyration is
roughly 7 Å, and their minimal and maximal diameters are 14−
15 and 25 Å, respectively. Their SASA is dominantly positive
due to the rather evenly distributed protonated tertiary amines
2669
https://doi.org/10.1021/acs.biomac.1c00355
Biomacromolecules 2021, 22, 2659−2675

Biomacromolecules
pubs.acs.org/Biomac
Article
Figure 4. Intracellular activity of dendrimers against infected macrophages. LJ slants showing the growth of bacteria in the presence of dendrimer
2G0,HCl (A) and dendrimer 3G0,HCl (B) versus rifampicin (RIF); ND, no drug. (C) MICs and CC50s of both dendrimers.
To establish additional structure−activity relationships
(SARs) based on modification of generations, surface amino
groups, and branches, as highlighted in Table 2, we then
decided to synthesize the next generation, G1 (12 amino
groups on the surface), by introducing five different types of
linkers (branches) between the dendrimer core and the
different amino groups on the surface: NH-amino-ethyl-
amine (6G1,HCl, 7G1,HCl, and 11G1,HCl), NH-amino-propyl-
amine (8G1,HCl−10G1,HCl and 12G1,HCl), 4-(((2-
aminoethyl)amino)methyl)phenoxy (5G1,HCl, 14G1,HCl,and
19G1,HCl−20G1,HCl), and 4-(((2-aminopropyl)amino)-
methyl)phenoxy (13G1,HCl and 15G1,HCl−18G1,HCl). As a
general remark, several observations are apparent. All of the
protonated forms of these 16G1 dendrimers displayed lower
potency against the three selected strains versus dendrimers of
generation 0. Within the G1 series, the most interesting G1
dendrimers are again pyrrolidinium or piperidinium groups
bearing 6G1,HCl and 7G1,HCl. However, they showed moderate
antimycobacterial activities only against MTB H37Ra and M.
bovis BCG with MICs of 25 and 12.5 μg/mL, respectively,
similar to those of 1G0,HCl, 2G0,HCl, and 5G0,HCl, and no
activity against MTB H37Rv. The phosphorus dendrimers
6G1,HCl, 7G1,HCl, and 5G0,HCl have the same 2-(pyrrolidin-1-
yl)ethanaminium group or 2-(piperidin-1-yl)ethanaminium
group on the surface. The lengthening of the chain of
6G1,HCl and 7G1,HCl with the same amino group on the surface
decreased antimycobacterial activities, 8G1,HCl versus 6G1,HCl
and 9G1,HCl versus 7G1,HCl showing MICs of ∼25−50 μg/mL.
The replacement of the piperidinium group of 9G1,HCl by a
morpholinium group (10G1,HCl) with the same linker did not
improve the potency, whereas the introduction of a 1-
methylpyrrolidinium group substituted in position 2
(11G1,HCl) maintains moderate antimycobacterial activities
only against MTB H37Ra and M. bovis BCG with an MIC of
25 μg/mL. No activity has been observed by the introduction
of the (2-methoxy)-1-phenyl-piperazinium group in place of
pyrrolidinium or piperidinium groups (12G1,HCl versus
8G1,HCl or 9G1,HCl). The replacement of the NH-amino-
ethyl-amine or the NH-amino-propyl-amine chains by the 4-
(((2-aminoethyl)amino)methyl)phenoxy chain
(14G1,HCl,19G1,HCl, and 20G1,HCl) or by the 4-(((3-
aminopropyl)amino)methyl) phenoxy chain (13G1,HCl, and
18G1,HCl) did not improve the antimycobacterial activities
irrespective of the nature of the amino groups on the surface
(pyrrolidinium (6G1,HCl), piperidinium (19G1,HCl), morpho-
linium (20G1,HCl and 15G1,HCl), imidazolinium (13G1,HCl and
17G1,HCl), 1-methylpyrrolidinium group substituted in posi-
tion 2 (14G1,HCl), 2-methylpiperidinium (16G1,HCl), and (2-
methoxy)-1-phenyl-piperazinium (18G1,HCl) groups). For
generation 1, in contrast to generation 0, no effect of the
nature of the amino groups on the surface has been clearly
demonstrated. All of the dendrimers showed moderate to no
antimycobacterial activities. To summarize, the most relevant
G1 dendrimers are 6G1,HCl and 7G1,HCl with the shorter chain
(NH-amino-ethyl-amine) and the pyrrolidinium and piper-
idinium groups on the surface.
Based on the very interesting results for 1G0, 2G0, 6G1, and
7G1, we decided to construct a series of six dendrimers of
generations 2, 3, and 4 bearing on the surface only the
pyrrolidinium and piperidinium groups (Tables 4−6) on the
surface. The dendrimers 6G2,HCl and 7G2,HCl, 6G3,HCl and
7G3,HCl, and 6G4,HCl and 7G4,HCl showed no antimycobacte-
rial activities against all three strains, except for 6G2,HCl
showing poor anti-TB activities against only MTB H37Ra
with an MIC of 50 μg/mL but an SI of 0.8.
Interestingly, 2G0,HCl inhibited the mycobacterial growth
and intracellular survival in infected J774A.1 cells. 2G0,HCl
showed extracellular killing of MTB H37Rv with an MIC of
6.12 μg/mL. To assure the effect of dendrimers on intracellular
replicating mycobacteria, we tested the dendrimers 2G0,HCl
and 3G0,HCl. The effect of dendrimer 1G0,HCl on intra-
cellularly replicating bacteria was not determined due to its low
CC50. For analysis of the effect of dendrimers on intracellular
bacteria, we infected J774A.1 cells with MTB H37Rv and
treated with different concentrations of dendrimers ranging
from 0.78 to 25 μg/mL. After 48 h post infection, the cells
2670
https://doi.org/10.1021/acs.biomac.1c00355
Biomacromolecules 2021, 22, 2659−2675

Biomacromolecules
pubs.acs.org/Biomac
Article
Figure 5. Main structure−activity relationships between the generation of dendrimers and antimycobacterial activities.
were harvested and added to the intracellular bacteria on LJ
slants from treated and untreated cells. As shown in Figure 4,
after 4−5 weeks of incubation, the colony-forming units
(CFUs) were calculated, and it was observed that the MTB-
infected cells treated with 2G0,HCl showed no intracellular
bacteria at concentrations above 6.25 μg/mL. The dendrimer
3G0,HCl exhibited a marginal effect on intracellularly
replicating MTB H37Rv. In these assays, rifampicin was used
as a positive control.
Specificity of Active Dendrimers. To determine their
antimicrobial specificity, activities of 2G0,HCl and 3G0,HCl were
also determined against an ESKAPE panel consisting of
Escherichia coli ATCC 25922, Staphylococcus aureus ATCC
29213, Klebsiella pneumoniae BAA-1705, Acinetobacter bau-
mannii BAA-1605, and Pseudomonas aeruginosa ATCC 27853,
where both were found to be completely inactive, with MIC
evaluation indicating that their antimicrobial activity is
Mycobacterium-specific.
As a general statement, the size of dendrimers as well as the
nature of the terminal group plays key roles. Dendrimers of
generation 0 bearing pyrrolidinium or piperidinium groups
displayed good antimycobacterial activities against the three
selected strains. In this direction, moderate biological activities
were observed with generation 1 only with the shorter chain
and also with the pyrrolidinium and piperidinium groups on
the surface. Figure 5 represents the principal SARs among the
prepared dendrimers between the generation of dendrimers
and the antimycobacterial activities.
efficacy in Balb/C mice. Two weeks of oral treatment with 33
mg/kg 2G0,HCl (one-time-daily administration) reduced the
mean bacterial counts in lungs of infected mice significantly by
1.0 log10, while 50 mg/kg 2G0,HCl reduced bacterial counts by
∼1.5 log10 as compared to the untreated group. This outcome
is even better compared to rifampicin-treated groups, which is
not the case at the dose of RIF of 25 mg/kg versus 2G0.HCl at
33 mg/kg. Taken together, this important experiment
demonstrates the superior efficacy of 2G0,HCl in clearing
infection from various animal tissues in comparison to
ethambutol and rifampicin (Figure 6). 2G0HCl is considered
a very interesting lead to be optimized to improve its in vivo
efficiency.
Mechanism of Action Studies of 2G0,HCl: Preliminary
Investigations. Since 2G0,HCl demonstrated good in vitro, ex
vivo, and in vivo activities, it was imperative to elucidate the
mechanism of action by which killing of MTB H37Rv is
facilitated. MALDI MS−MS data showed differential regu-
lation of 15 specific proteins, which have been shown in Table
8 and Figure 7. In these preliminary investigational assays, we
did not quantify the amount of proteins. The aim is only to
evaluate the upregulation and the downregulation of proteins.
Within the 12 proteins observed, three were upregulated,
namely, SodA (Fe), unnamed Gi15607825 Rv0685, and
elongation factor TU Rv0685. Another 9 were downregulated,
namely, GAPDH Rv1436, alkyl hydroperoxidase reductase
subunit-c Rv2428, DnaK Rv0350, hypothetical protein
Rv0020c, Rv2145c, phosphopyruvatehydratage, DNA-directed
RNA polymerase α subunit, malate dehydrogenase, flavopro-
tein α subunit, and two-component response regulator protein
MprA. Detailed functional characterization of differentially
expressed proteins is required to identify the exact mechanism
of action of the dendrimers.
Taken together, these results fully confirm the interest of
2G0,HCl as a safe antitubercular agent active against resistant
strains. Consequently, 2G0,HCl was tested in an animal model
through oral administration.
Efficacy of 2G0,HCl in Balb/C Mice against Mtb H37Ra.
Since 2G0,HCl demonstrated safe and potent in vitro
antimycobacterial activity, we decided to evaluate its in vivo
2671
https://doi.org/10.1021/acs.biomac.1c00355
Biomacromolecules 2021, 22, 2659−2675

Biomacromolecules
pubs.acs.org/Biomac
Article
standard anti-TB drugs. Molecular modeling studies showed
that 1G0,HCl and 2G0,HCl have a radius of gyration of 7 Å and
minimal and maximal diameters of 14−15 and 25 Å,
respectively. Their SASA is dominantly positive due to the
rather evenly distributed protonated tertiary amines with
localized negative spots arising from the central N3P3 ring.
The protonated amines form transient ion pairs with chlorine
in MD simulations, while the central core typically interacts
with three to four water molecules. 1G0,HCl and 2G0,HCl are
active against three different strains, such as attenuated M.
tuberculosis H37Ra, virulent M. tuberculosis H37Rv, and
Mycobacterium bovis BCG. Importantly, the safety index values
are 16.02 and 3.1 versus Ra for 2G0,HCl and 1G0,HCl
,
respectively. Very interestingly, 2G0,HCl was tested against
single-drug-resistant M. tuberculosis and multiple strain
resistant to rifampicin, isoniazid, ethambutol, or streptomycin.
2G0,HCl showed activity against all single-drug-resistant strains,
which indicates that the dendrimers possess a potential new
mode of action. Moreover, our preliminary studies on their
mechanism of action suggested that 2G0,HCl acts on novel
targets and/or pathways. In addition, 2G0,HCl inhibited the
mycobacterial growth and intracellular survival in infected
J774A.1 cells. Consequently, 2G0,HCl represents a very
interesting and safe anti-TB agent and was selected for in
vivo experiments. A single daily oral administration for a
fortnight with 50 mg/kg 2G0,HCl very significantly reduces the
mean bacterial count in the lungs of infected mice by about
1.5 log10 compared to the untreated group. This bacterial
reduction is greater than that observed in the rifampicin- and
ethambutol-treated groups.
Taken together, all of the in vitro, in cellular, ex vivo, and in
vivo experiments demonstrated that the safe and chemically
stable G0 polycationic phosphorus dendrimer named 2G0,HCl
represents a new chemical entity to tackle tuberculosis.
Moreover, 2G0,HCl is chemically stable for up to 9 months in
aerated aqueous solution, which is one of the critical points for
potential clinical development. Interestingly, these original
phosphorus dendrimers should be developed in combination
Figure 6. In vivo efficacy of 2G0,HCl in Balb/C mice. Mean log10
CFU/mL in lungs of mice after 14 days post treatment with 2G0,HCl
(oral administration, once daily during 2 weeks), rifampicin, and
ethambutol. Six mice per group were infected intravenously with ∼106
CFU of M. tuberculosis H37Ra and treatment was started 7 days post
infection.
CONCLUSIONS AND PERSPECTIVES
■
TB is an old infectious disease mainly affecting the lungs
caused by M. tuberculosis. Compared with other diseases, TB is
one of the most important killers, and TB medication can be
toxic to the liver. In addition, most of the TB drugs are, in fact,
bioprecursor-type prodrugs.50 Development of MDR- and
XDR-TB warrants the introduction of new chemotherapeutic
tools in the arsenal of anti-TB drugs. Consequently, the search
for original anti-TB derivatives with a broad spectrum and with
a new scaffold is even more pertinent today. Very interestingly,
we developed original anti-TB compounds based on
biocompatible polycationic phosphorus dendrimers. These
nanoparticles open new avenues in the tuberculosis domain
and are chemically totally different from and not related to the
Table 8. List of MTB Proteins and Their Functions
expression level of
identified proteins
SodA (Fe)
proteins
role in MTB
upregulation
TB leads to increased oxidative burden (ROS, RNI) in the lung that aids in the activation of TB
prodrugs
SodA (Fe) destroys radi•c−als, which are normally produced within the cells and are toxic to
biological systems (O2 → H2O2)
unnamed Gi15607825 Rv0685
elongation factor TU Rv0685
GAPDH Rv1436
upregulation
upregulation
promotes GTP-dependent binding of aminoacyl-tRNA to the A-site of ribosomes during protein
biosynthesis
promotes GTP-dependent binding of aminoacyl-tRNA to the A-site of ribosomes during protein
biosynthesis
involved in the second phase of glycolysis
involved in oxidative stress response
downregulation
downregulation
alkyl hydroperoxidase reductase
subunit-c Rv2428
DnaK Rv0350
downregulation
acts as a chaperone; involved in induction by stress conditions, e.g., heat shock; possibly has an
ATPase activity
hypothetical protein Rv0020c
Rv2145c
phosphopyruvatehydratase
DNA-directed RNA polymerase α
subunit
downregulation
downregulation
downregulation
downregulation
signal transduction
cell division
anchorless surface-exposed proteins that bind to plasminogen
DNA-dependent RNA polymerase catalyzes the transcription of DNA into RNA
malate dehydrogenase
flavoprotein α subunit
two-component response regulator
protein MprA
downregulation
downregulation
downregulation
involved in the conversion of malate to oxaloacetate
the electron-transfer flavoprotein serves as a specific electron acceptor for other dehydrogenases
regulator, part of a two-component regulatory system
2672
https://doi.org/10.1021/acs.biomac.1c00355
Biomacromolecules 2021, 22, 2659−2675

Biomacromolecules
pubs.acs.org/Biomac
Article
Figure 7. Two-dimensional (2D) gel profile of MTB H37Ra treated with 2G0,HCl
.
therapy (multidrug regimens) to treat TB, for instance, with
isoniazid, rifampicin, or delamanid, which are currently used.
The objective is to increase treatment efficacy while avoiding
resistance (MDR-TB and XDR-TB).51 We are convinced that
the main goals concerning the development of new TB drugs
are to (1) reduce the duration of treatment, (2) effectively
treat MDR-TB, and (3) provide treatment for patients with
latent TB infection. Our original results may open new
perspectives in these challenging fields.
Rama Pati Tripathi − Medicinal and Process Chemistry
Division, CSIR-CDRI, 226031 Lucknow, India;
orcid.org/0000-0002-9256-3112; Email: rpt.cdri@
gmail.com
Arunava Dasgupta − Microbiology Division, CSIR-Central
Drug Research Institute, 226031 Lucknow, India;
Email: a.dasgupta@cdri.res.in
Kishore Kumar Srivastava − Microbiology Division, CSIR-
Central Drug Research Institute, 226031 Lucknow, India;
Email: kishore@cdri.res.in
Jean-Pierre Majoral − Laboratoire de Chimie de Coordination
du CNRS, 31077 Toulouse, France; LCC-CNRS, Université
de Toulouse, CNRS, 31400 Toulouse, France; orcid.org/
0000-0002-0971-817X; Email: jean-pierre.majoral@lcc-
toulouse.fr
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.biomac.1c00355.
■
sı
Preparation and characterization of compounds of
generation 0: 1G0, 2G0, 3G0, 5G0, 1G0,HCl, 2G0,HCl
,
3G0,HCl, and 5G0,HCl; generation 1: 5G1, 6G1, 7G1,
8G1, 9G1, 10G1, 11G1, 12G1, 13G1, 14G1, 15G1,
Authors
17G1, 18G1, 19G, 20G1, 5G1,HCl, 6G1,HCl, 7G1,HCl
8G1,HCl, 9G1,HCl, 10G1,HCl, 11G1,HCl, 12G1,HCl
13G1,HCl, 14G1,HCl, 15G1,HCl, 17G1,HCl, 18G1,HCl
,
,
,
Vishwa Deepak Tripathi − Laboratoire de Chimie de
Coordination du CNRS, 31077 Toulouse, France; LCC-
CNRS, Université de Toulouse, CNRS, 31400 Toulouse,
France
Dheerj Soam − Microbiology Division, CSIR-Central Drug
Research Institute, 226031 Lucknow, India
Swetarka Das − Microbiology Division, CSIR-Central Drug
Research Institute, 226031 Lucknow, India
Shriya Singh − Microbiology Division, CSIR-Central Drug
Research Institute, 226031 Lucknow, India
Ramakrishna Gandikota − Laboratoire de Chimie de
Coordination du CNRS, 31077 Toulouse, France; LCC-
CNRS, Université de Toulouse, CNRS, 31400 Toulouse,
France
Regis Laurent − Laboratoire de Chimie de Coordination du
CNRS, 31077 Toulouse, France; LCC-CNRS, Université de
Toulouse, CNRS, 31400 Toulouse, France
19G1,HCl, and 20G1,HCl; detailed characterization of
1G0,HCl and 2G0,HCl by NMR (1H, 31P),13C{1H},13C-
{31P}, 2D,13C,JMOD, 2D,13C HMBC, 2D, and13C
HMQC; detailed characterization of 1G0,HCl and
2G0,HCl by NMR (1H, 31P),13C{1H},13C{31P},
2D,13C,JMOD, 2D,13C HMBC, 2D, and 13C HMQC,
which also corroborate the full protonation of the
terminal pyrrolidine or piperidine end groups; figures
and statistics of noncovalent interaction counts, radii of
gyration, and radial distribution functions observed by
molecular dynamics simulations of 1G0,HCl and 2G0,HCl
(PDF)
AUTHOR INFORMATION
■
Corresponding Authors
Andrii Karpus − Laboratoire de Chimie de Coordination du
CNRS, 31077 Toulouse, France; LCC-CNRS, Université de
Toulouse, CNRS, 31400 Toulouse, France; orcid.org/
0000-0002-5760-3086
Anne-Marie Caminade − Laboratoire de Chimie de
Coordination du CNRS, 31077 Toulouse, France; LCC-
CNRS, Université de Toulouse, CNRS, 31400 Toulouse,
France
Serge Mignani − Laboratoire de Chimie et de Biochimie
Pharmacologiques et Toxicologique, PRES Sorbonne Paris
Cité, CNRS UMR 860, Université Paris Descartes, 75006
Paris, France; CQMCentro de Química da Madeira,
MMRG, Universidade da Madeira, 9020-105 Funchal,
Portugal; Email: serge.mignani@paris, serge.mignani@
staff.uma.pt
2673
https://doi.org/10.1021/acs.biomac.1c00355
Biomacromolecules 2021, 22, 2659−2675

Biomacromolecules
pubs.acs.org/Biomac
Article
(13) Pati, K.; Bagade, S.; Bonde, S.; Sharma, S.; Saraogi, G. Recent
therapeutic approaches for the management of tuberculosis:
Challenges and opportunities. Biomed. Pharmacother. 2018, 99,
735−745.
Anke Steinmetz − Sanofi R&D, Integrated Drug Discovery,
Centre de Recherche Vitry-Alfortville, 94403 Vitry-sur-Seine
Cedex, France
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.biomac.1c00355
(14) Skwarecki, A. S.; Milewskib, S.; Schielmann, M.; Milewska, M.
J. Antimicrobial molecular nanocarrier-drug conjugates. Nanomedicine
2016, 12, 2215−2240.
(15) Dineshkumar, P.; Panneerselvam, T.; Selvaraj, K. K.; Kumar, P.
V. Formulation of rifampicin loaded PEGylated 5.0G EDA-PAMAM
dendrimers as effective long-duration release drug carriers. Curr. Drug
Ther. 2017, 12, 115−126.
(16) Bellini, G. R.; Guimar, A. P.; Pachecoa, M. A. C.; Diasa, D. A.;
Furtadoc, V. R.; de Alencastro, R. B.; Hortac, B. A. C. Association of
the anti-tuberculosis drug rifampicin with PAMAM dendrimer. J. Mol.
Graphics Modell. 2015, 60, 34−42.
(17) Kumar, P. V.; Asthana, A.; Dutta, T.; Jain, N. K. Intracellular
macrophage uptake of rifampicin loaded mannosylated dendrimers. J.
Drug Targeting 2006, 14, 546−556.
(18) Singh, N.; Gautam, S. P.; Singh, H. L.; Dhiman, A.; Siddiqui,
G.; Verma, A. Isoniazid loaded dendrimer based nano carriers for the
delivery of anti-tuberculosis. Indian Res. J. Pharm. Sci. 2016, 3, 519−
529.
(19) Mignani, S.; Tripathi, R. P.; Chen, L.; Caminade, A.-M.; Shi, X.;
Majoral, J.-P. New ways to treat tuberculosis using dendrimers as
nanocarriers. Pharmaceutics 2018, 10, No. 105.
(20) Chis, A. A.; Dobrea, C.; Morgovan, C.; Arseniu, A. M.; Rus, L.
L.; Butuca, A.; Juncan, A. M.; Totan, M.; Vonica-Tincu, A. L.;
Cormos, G.; Muntean, A. C.; Muresan, M. L.; Gligor, F. G.; Frum, A.
Applications and Limitations of Dendrimers in Biomedicine. Molecules
2020, 25, No. 3982.
Author Contributions
All of the other coauthors contributed equally. The authors
declare the following competing financial interest(s): A.S. is a
Sanofi employee and may hold shares and/or stock options in
the company.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank the Director CSIR-CDRI for the support. S.M.
acknowledges the support of FCT-Fundacao para a Ciencia e a
Tecnologia (project PEst-OE/QUI/UI0674/2013, CQM,
Portuguese Government funds) and ARDITI through the
project M1420-01-0145-FEDER-000005Centro de Quimica
da MadeiraCQM+ (Madeira 14-20), and J.-P.M., A.-M.C.
R.L., and A.K. acknowledge the funds from Centre National de
la Recherche Scientifique (CNRS, France). The work was
funded through the Indo-French (CEFIPRA) project CEFI-
PRA-5303-2. This manuscript bears CSIR-CDRI communica-
tion number 10240.
(21) Madaan, K.; Kumar, S.; Poonia, N.; Lather, V.; Pandita, D.
Dendrimers in drug delivery and targeting: drug-dendrimer
interactions and toxicity issues. J. Pharm. BioAllied Sci. 2014, 6,
139−150.
REFERENCES
■
(1) Wikipedia. History of Tuberculosis, 2021. https://en.wikipedia.
org/wiki/History_of_tuberculosis#/media/File:Cristobal_Rojas_
37a.JPG.
(2) World Health Organization, G.S.. Global Tuberculosis Report,
2020. https://apps.who.int/iris/bitstream/handle/10665/336069/
9789240013131-eng.pdf.
(3) Kendall, E. A.; Azman, A. S.; Cobelens, F. G. MDR-TB
treatment as prevention: The projected population-level impact of
expanded treatment for multidrug-resistant tuberculosis. PLoS One
2017, 12, No. e0172748.
(4) Maiolini, M.; Gause, S.; Taylor, J.; Steakin, T.; Shipp, G.;
Lamichhane, P.; Deshmukh, B.; Shinde, V.; Bishayee, A.; Deshmukh,
R. R. The War against Tuberculosis: A Review of Natural Compounds
and Their Derivatives. Molecules 2020, 25, No. 3011.
(5) Mascola, J. R.; Fauci, A. S. Novel vaccine technologies for the
21st century. Nat. Rev. Immunol. 2020, 20, 87−88.
(6) Moodley, R.; Godec, T. R. STREAM., o. b. o. t., Trial Team.
Short-course treatment for multidrug-resistant tuberculosis: the
STREAM trials. Eur. Respir. Rev. 2016, 25, 29−35.
(7) Gelperina, S.; Kisich, K.; Iseman, M. D.; Heifets, L. The
Potential advantages of nanoparticle drug delivery systems in
chemotherapy of tuberculosis. Am. J. Respir. Crit. Care Med. 2005,
172, 1487−1490.
(8) Jawahar, N.; Reddy, G. Nanoparticles: A novel pulmonary drug
delivery system for tuberculosis. J. Pharm. Sci. Res. 2012, 4, 1901−
1906.
(9) Mignani, S.; Rodrigues, J.; Tomas, H.; Roy, R.; Shi, X.; Majoral,
J.-P. Bench-to-bedside translation of dendrimers: Reality or utopia? A
concise analysis. Adv. Drug Delivery Rev. 2018, 136−137, 73−81.
(10) Arrowsmith, J. Trial watch: Phase II failures: 2008−2010. Nat.
Rev. Drug Discovery 2011, 10, 328−329.
(11) Hay, M.; Thomas, D. W.; Craighead, J. L.; Economides, C.;
Rosenthal, J. Clinical development success rates for investigational
drugs. Nat. Biotechnol. 2014, 32, 40−51.
(12) Costa-Gouveia, J.; Aınsa, J. A.; Brodin, P.; Lucıa, A. How can
nanoparticles contribute to antituberculosis therapy? Drug Discovery
Today 2017, 22, 600−607.
(22) Gajbhiye, V.; Palanirajan, V. K.; Tekade, R. K.; Jain, N. K.
Dendrimers as therapeutic agents: a systematic review. J. Pharm.
Pharmacol. 2009, 61, 989−1003.
(23) Li, Z.; Hu, J.; Yang, L.; Zhang, X.; Liu, X.; Wang, Z.; Li, Y.
Integrated POSS-dendrimer nanohybrid materials: current status and
future perspective. Nanoscale 2020, 12, 11395−11415.
(24) Yamamoto, K.; Imaoka, T.; Tanabe, M.; Kambe, T. New
Horizon of Nanoparticle and Cluster Catalysis with Dendrimers.
Chem. Rev. 2020, 120, 1397−1437.
(25) Lyu, Z.; Ding, L.; Huang, A. Y. T.; Kao, L.; Peng, L.
Poly(amidoamine) dendrimers: covalent and supramolecular syn-
thesis. Mater. Today Chem. 2019, 13, 34−48.
(26) Caminade, A. M.; Turrin, C. O.; Majoral, J. P. Phosphorus
Dendrimers in Biology and Nanomedicine: Syntheses, Characterization,
and Properties; Jenny Stanford Publishing Pte. Ltd., 2018.
(27) Mignani, S.; El Brahmi, N.; El Kazzouli, S.; Laurent, R.; Sonia
Ladeira, S.; Caminade, A.-M.; Pedziwiatr-Werbicka, E.; Szewczyk, E.
M.; Bryszewska, M.; Bousmina, M. M.; Cresteil, T.; Majoral, J.-P.
Original multivalent gold(III) and dual gold(III)−copper(II)
conjugated phosphorus dendrimers as potent antitumoral and
antimicrobial agents. Mol. Pharmaceutics 2017, 14, 4087−4097.
(28) Solassol, J.; Crozet, C.; Perrier, V.; Leclaire, J.; Beranger, F.;
Caminade, A. M.; Meunier, B.; Dormont, D.; Majoral, J.-P.; Lehmann,
S. Cationic phosphorus-containing dendrimers reduce prion repli-
cation both in cell culture and in mice infected with scrapie. J. Gen.
Virol. 2004, 85, 1791−1799.
(29) Wasiak, T.; Marcinkowska, M.; Pieszynski, I.; Zablocka, M.;
Caminade, A. M.; Majoral, J. P.; Klajnert-Maculewicz, B. Cationic
phosphorus dendrimers and therapy for Alzheimer’s disease. New J.
Chem. 2015, 39, 4852−4859.
̀
(30) Blattes, E.; Vercellone, A.; Eutamene, H.; Turrin, C. O.;
Théodorou, V.; Majoral, J. P.; Caminade, A. M.; Prandi, J.; Nigou, J.;
Puzo, G. Mannodendrimers prevent acute lung inflammation by
inhibiting neutrophil recruitment. Proc. Natl. Acad. Sci. U.S.A. 2013,
110, 8795−8800.
2674
https://doi.org/10.1021/acs.biomac.1c00355
Biomacromolecules 2021, 22, 2659−2675

Biomacromolecules
pubs.acs.org/Biomac
Article
(31) Mignani, S.; El Kazzouli, S.; Bousmina, M.; Majoral, J.-P.
Expand classical drug administration ways by emerging routes using
dendrimer drug delivery systems: A Concise overview. Adv. Drug
Delivery Rev. 2013, 35, 1316−1330.
based on in-vitro physicochemical parameters: key factor analysis
(Part 1). Drug Discovery Today 2019, 24, 1176−1183.
(50) Laborde, J.; Deraeve, C.; Bernardes-Génisson, V. Update of
Antitubercular Prodrugs from a Molecular Perspective: Mechanisms
of Action, Bioactivation Pathways, and Associated Resistance.
ChemMedChem 2017, 20, 1657−1676.
(32) Kana, B. D.; Karakousis, P. C.; Parish, T.; Dicke, T. Future
target-based drug discovery for tuberculosis? Tuberculosis 2014, 94,
551−556.
(51) Iacobino, A.; Fattorini, L.; Giannoni, F. Drug-Resistant
Tuberculosis 2020: Where We Stand. Appl. Sci. 2020, 10, No. 2153.
(33) Sabbah, M.; Mendes, V.; Vistal, R. G.; Dias, D. M. G.; Monika
̌
Záhorszká, M.; Mikusová, K.; Korduláková, J.; Coyne, A. G.; Blundell,
T. L.; Abell, C. Fragment-Based Design of Mycobacterium tuberculosis
InhA Inhibitors. J. Med. Chem. 2020, 63, 4749−4761.
(34) Cavero, E.; Zablocka, M.; Caminade, A. M.; Majoral, J. P.
Design of Bisphosphonate-terminated Dendrimers. Eur. J. Org. Chem.
2010, 14, 2759−2767.
(35) Slany, M.; Bardaji, M.; Casanove, M.-J.; Caminade, A.-M.;
Majoral, J.-P.; Chaudret, B. Dendrimer Surface Chemistry. Facile
Route to Polyphosphines and Their Gold Complexes. J. Am. Chem.
Soc. 1995, 117, 9764−9765.
(36) Launay, N.; Caminade, A. M.; Lahana, R.; Majoral, J. P. A
general synthetic strategy for neutral phosphorus-containing den-
drimers. Angew. Chem., Int. Ed. 1994, 33, 1589−1592.
(37) Padié, C.; Maszewska, M.; Majchrzak, K.; Nawrot, B.;
Caminade, A.-M.; Majoral, J.-P. Polycationic phosphorus dendrimers:
synthesis, characterization, study of cytotoxicity, complexation of
DNA, and transfection experiments. New J. Chem. 2009, 33, 318−326.
(38) Badetti, E.; Caminade, A.-M.; Majoral, J.-P.; Moreno-Manas,
̃
M.; Sebastián, R. M. Palladium(0) Nanoparticles Stabilized by
Phosphorus Dendrimers Containing Coordinating 15-Membered
Triolefinic Macrocycles in Periphery. Langmuir 2008, 24, 2090−2101.
(39) Garcia, L.; Roglans, A.; Laurent, R.; Majoral, J.-P.; Pla-
Quintana, A.; Caminade, A.-M. Dendritic phosphoramidite ligands for
Rh-catalyzed [2+2+2] cycloaddition reactions: unprecedented
enhancement of enantiodiscrimination. Chem. Commun. 2012, 48,
9248−9250.
̈
̈
(40) Schrodinger Release 2020-2; Schrodinger, LLC: New York, NY,
2020.
(41) Bochevarov, A. D.; Harder, E.; Hughes, T. F.; Greenwood, J.
R.; Braden, D. A.; Philipp, D. M.; Rinaldo, D.; Halls, M. D.; Zhang, J.;
Friesner, R. A. Jaguar: A high-performance quantum chemistry
software program with strengths in life and materials sciences. Int. J.
Quantum Chem. 2013, 113, 2110−2142.
(42) Bowers, K. J.; Chow, E.; Xu, H.; Dror, R. O.; Eastwood, M. P.;
Gregersen, B. A.; Klepeis, J. L.; Kolossvary, I.; Moraes, M. A.;
Sacerdoti, F. D.; Salmon, J. K.; Shan, B.; Shaw, D. E. In Scalable
Algorithms for Molecular Dynamics Simulations on Commodity Clusters,
Proceedings of the ACM/IEEE Conference on Supercomputing
(SC06), Tampa, Florida, USA, November 11−17, 2006.
(43) Harder, E.; Damm, W.; Maple, J.; Wu, C.; Reboul, M.; Xiang, J.
Y.; Wang, L.; Lupyan, D.; Dahlgren, M. K.; Knight, J. L.; Kaus, J. W.;
Cerutti, D. S.; Krilov, G.; Jorgensen, W. L.; Abel, R.; Friesner, R. A.
OPLS3: A Force Field Providing Broad Coverage of Drug-like Small
Molecules and Proteins. J. Chem. Theory Comput. 2016, 12, 281−296.
(44) Swinney, D. C.; Anthony, J. How were new medicines
discovered? Nat. Rev. Drug Discovery 2011, 10, 507−519.
(45) Mignani, S.; El Kazzouli, S.; Bousmina, M. M.; Majoral, J.-P.
Dendrimer space concept for innovative nanomedicine: a futuristic
vision for medicinal chemistry. Prog. Polym. Sci. 2013, 38, 993−1008.
(46) Schirle, M.; Jenkins, J. L. Identifying compound efficacy targets
in phenotypic drug discovery. Drug Discovery Today 2016, 21, 82−89.
(47) Payne, D. J.; Gwynn, M. N.; Holmes, D. J.; Pompliano, D. L.
Drugs for bad bugs: confronting the challenges of antibacterial
discovery. Nat. Rev. Drug Discovery 2007, 6, 29−40.
(48) Mignani, S.; Rodrigues, J.; Roy, R.; Shi, X.; Cena, V.; El
Kazzouli, S.; Majoral, J.-P. Exploration of biomedical dendrimer space
based on in-vivo physicochemical parameters: Key factor analysis
(Part 2). Drug Discovery Today 2019, 24, 1184−1192.
(49) Mignani, S.; Rodrigues, J.; Roy, R.; Shi, X.; Cen
Kazzouli, S.; Majoral, J.-P. Exploration of biomedical dendrimer space
̃
a, V.; El
2675
https://doi.org/10.1021/acs.biomac.1c00355
Biomacromolecules 2021, 22, 2659−2675