Diaminopimelic acid (DAP) analogs bearing isoxazoline moiety as selective inhibitors against meso-diaminopimelate dehydrogenase (m-Ddh) from Porphyromonas gingivalis

Article abstract of DOI:10.1016/j.bmcl.2017.06.056

Two diastereomeric analogs (1 and 2) of diaminopimelic acid (DAP) bearing an isoxazoline moiety were synthesized and evaluated for their inhibitory activities against meso-diaminopimelate dehydrogenase (m-Ddh) from the periodontal pathogen, Porphyromonas

Full text of DOI:10.1016/j.bmcl.2017.06.056

Bioorganic & Medicinal Chemistry Letters xxx (2017) xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Diaminopimelic acid (DAP) analogs bearing isoxazoline moiety
as selective inhibitors against meso-diaminopimelate dehydrogenase
(
m-Ddh) from Porphyromonas gingivalis
Hongguang Ma , Victoria N. Stone , Huiqun Wang , Glen E. Kellogg a,d, Ping Xu b,c, Yan Zhang a,
a
b,c
a
⇑
a
Department of Medicinal Chemistry, Virginia Commonwealth University, 800 East Leigh Street, Richmond, VA 23298, USA
Philips Institute for Oral Health Research, Virginia Commonwealth University, 521 North 11th Street, Richmond, VA 23298, USA
Department of Microbiology and Immunology, Virginia Commonwealth University, 1101 East Marshall Street, Richmond, VA 23298, USA
Institute for Structural Biology, Drug Discovery and Development, Virginia Commonwealth University, 800 East Leigh Street, Richmond, VA 23298, USA
b
c
d
a r t i c l e i n f o
a b s t r a c t
Article history:
Two diastereomeric analogs (1 and 2) of diaminopimelic acid (DAP) bearing an isoxazoline moiety were
synthesized and evaluated for their inhibitory activities against meso-diaminopimelate dehydrogenase
Received 20 April 2017
Revised 19 June 2017
Accepted 21 June 2017
Available online xxxx
(m-Ddh) from the periodontal pathogen, Porphyromonas gingivalis. Compound 2 showed promising inhi-
bitory activity against m-Ddh with an IC50 value of 14.9 mM at pH 7.8. The two compounds were further
tested for their antibacterial activities against a panel of periodontal pathogens, and compound 2 was
shown to be selectively potent to P. gingivalis strains W83 and ATCC 33277 with minimum inhibitory
concentration (MIC) values of 773 mM and 1.875 mM, respectively. Molecular modeling studies revealed
that the inversion of chirality at the C-5 position of these compounds was the primary reason for their
different biological profiles. Based on these preliminary results, we believe that compound 2 has proper-
ties consistent with it being a lead compound for developing novel pathogen selective antibiotics to treat
periodontal diseases.
Keywords:
Diaminopimelic acid
Antibacterial activity
meso-Diaminopimelate dehydrogenase
inhibitor
Molecular modeling
Published by Elsevier Ltd.
Bacterial infections remain a major threat to human health with
considerable treatment costs and an array of related clinical com-
plications. With the mounting incidence and mortality rates of
infectious diseases, the current pace for discovery and develop-
ment of novel antibiotics is far from satisfactory.1 A confounding
factor is that, because maintenance of the human microbiome is
vital for keeping in good physical health,2 the widespread use of
broad-spectrum antibiotics in recent decades has perturbed the
ecological balance of this microbiome. These drugs, with their
unspecific effects on both pathogenic and non-pathogenic species,
cause further unexpected clinical symptoms such as opportunistic
infections, bacterial resistance and disorders of intestinal micro-
biota.4 Therefore, there is an urgent and continuing need to
develop more efficient and less toxic antibiotics that could target
a specific range of pathogenic species.
meso-diaminopimelic acid (m-DAP) plays a crucial role in the for-
mation of peptidoglycan cell wall of Gram negative bacteria
7
(Fig. 1). Therefore, inhibitors that selectively suppress enzymes
in this pathway may be able to act as novel antibiotics with selec-
tive anti-pathogenic action, satisfactory effectiveness and low tox-
icity. meso-diaminopimelate dehydrogenase (m-Ddh), the only
dehydrogenase present in the DAP biosynthetic pathway,8 acts as
a critical enzyme, through which NADPH dependently catalyzes
,3
the conversion of
lizing an intermediate imine.
L-tetrahydrodipicolinate (THDP) to m-DAP by uti-
9
,10
One strategy to develop inhibitors
against m-Ddh is to design and synthesize small molecules that are
mimics of m-DAP. Consequently, a large number of m-DAP analogs
,5
9,11,12
have been previously reported;
however, only a few of them
showed notable activities. In particular, compound 1 (Fig. 1) that
contains an isoxazoline group had almost no activity against m-
Ddh, while compound 2, its diastereoisomer, showed pH-depen-
dent inhibition against m-Ddh from Bacillus sphaericus. Addition-
ally, it also showed moderate antibacterial activity against B.
It is known that the diaminopimelic acid (DAP) biosynthesis
pathway is the major source of L-lysine for most bacteria and some
other related species,6 but this pathway is not present in mam-
malian cells. As a biosynthetic precursor in the DAP pathway,
1
3
sphaericus in an in vitro study. The most highly studied DAP
enzymes are derived from B. sphaericus and Corynebacterium glu-
1
4,15
tamicum,
but, because Porphyromonas gingivalis is crucial in
⇑
Corresponding author.
E-mail address: yzhang2@vcu.edu (Y. Zhang).
the onset and development of periodontal diseases that result from
http://dx.doi.org/10.1016/j.bmcl.2017.06.056
960-894X/Published by Elsevier Ltd.
0
Please cite this article in press as: Ma H., et al. Bioorg. Med. Chem. Lett. (2017), http://dx.doi.org/10.1016/j.bmcl.2017.06.056

2
H. Ma et al. / Bioorganic & Medicinal Chemistry Letters xxx (2017) xxx–xxx
Fig. 1. The structures of m-DAP and its analogs.
an unbalanced microbiome ecology in oral cavities,16,17 we are tar-
geting meso-diaminopimelate dehydrogenase from that species in
this study. In order to define a lead compound from DAP analogs
to target m-Ddh, we synthesized compounds 1 and 2 in our labora-
tories, then evaluated their enzymatic inhibition against m-Ddh
from P. gingivalis. Next, we tested their antibacterial activities
against a panel of periodontal pathogens. Furthermore, molecular
modeling studies were carried out to help understand interactions
between these two compounds and the m-Ddh target protein.
The syntheses of target compounds were performed following
1
3
the reported procedure with modifications (Scheme 1). Protec-
tion of the amino and carboxyl groups of commercially available
L-allylglycine was carried out via its conversion to the N-Cbz
methyl ester (5). The key step in generating the isoxazoline skele-
Scheme 1. Synthesis route of m-DAP analogs 1 and 2 containing an isoxazoline ring
Fig. 2. Dose-dependent studies of compound 1 and 2 against P. gingivalis m-Ddh at
different pH values. (a) Compound 1 and (b) Compound 2 inhibition against P.
gingivalis m-Ddh at pH 7.8. (c) Compound 1 and (d) Compound 2 inhibition against
P. gingivalis m-Ddh at pH 10.5.
system. Reagents and conditions: (a) TMSCl, MeOH, 97%; (b) CbzCl, Et
CH Cl , 49%; (c) NaHCO , AcOEt, 37% for 7 and 42% for 8; (d) MeOH, 1 N NaOH, 90%
for 9 and 95% for 10; (e) TMSCl, NaI, MeCN, 35% for 1 and 43% for 2.
3
N, DMAP,
2
2
3
Please cite this article in press as: Ma H., et al. Bioorg. Med. Chem. Lett. (2017), http://dx.doi.org/10.1016/j.bmcl.2017.06.056

H. Ma et al. / Bioorganic & Medicinal Chemistry Letters xxx (2017) xxx–xxx
3
ton was the 1,3-dipolar cycloaddition of ethyl chlorooximidoac-
etate by treatment of 5 (with protected amino and carboxyl
groups) in the presence of a weak base that, in situ, generated
the expected diastereoisomers 7 and 8.18 These two compounds
prompted us to look for the binding interactions with m-Ddh. m-
DAP, compounds 1 and 2 were first sketched in Sybyl-X 2.0 and
assigned Gasteiger-Hückel charges. Then, each was immersed in
a square box of water molecules and these solvated systems were
2
were separated by flash column chromatography using Et O/hex-
ane system. Then, compound 7 was subjected to hydrolysis in
NaOH aqueous solution to afford the diacid 9. Subsequent removal
of Cbz group was carried out in the presence of TMSCl and NaI at
room temperature in acetonitrile to generate compound 1. Simi-
larly, hydrolysis of compound 8 gave the diacid 10, which was fur-
ther deprotected to give compound 2.
Both compounds were tested for their inhibition against m-Ddh
according to the reported procedures.17 Since earlier work by
Abbott et al. had shown the pH of the buffer solution was a crucial
9
,13
factor for inhibitory activity against B. sphaericus m-Ddh,
we
conducted the assay at pH 7.8 and pH 10.5, respectively (Fig. 2).
The results showed that compound 1 with a R configuration at C-
5
position had no activity (up to 1 mM) at physiological condition
(
pH = 7.8), while compound 2 exhibited significant inhibitory
activity against m-Ddh with an IC50 value of 14.9 mM. At the more
basic condition of pH = 10.5, compound 1 was still not active
against m-Ddh up to 3 mM while compound 2 had a steep reduc-
tion in potency with an IC50 value of 221 mM. These results were
in line with previous reports, suggesting that the configuration
inversion at the C-5 position in the isoxazoline ring diminished
the binding interaction between compound 1 and m-Ddh. Addi-
tionally, compound 2 could exert its enzymatic inhibition effec-
tively at physiological-like conditions, but was less effective at
more basic conditions. A conformational change or ionization of
the enzyme at high pH presumably accounts for the reduced enzy-
matic inhibition of compound 2.13
Fig. 3. The 3D structures of m-DAP, compounds 1 and 2 after energy minimization.
The parts in compounds 1 and 2 share similar conformation was defined as portion
A and the rest as portion B. All oxygen atoms are in red and nitrogen atoms in blue;
For carbon atoms, m-DAP is in yellow, compounds 1 in orange, compound 2 in
green.
Compound 2 was further evaluated for its potentially selective
whole-cell activity against specific pathogens with respect to a
panel of periodontal pathogens. In this assay, four strains were
chosen, including P. gingivalis strains W83 and ATCC 33277, which
possess the target m-Ddh, and Fusobacterium nucleatum strain
ATCC 25586 and Staphylococcus aureus strain N315, which do not
possess m-Ddh. Compound 2 was thus tested by a broth microdilu-
1
9
tion method. The assay results showed that compound 2 exhib-
ited inhibitory activities against P. gingivalis strains W83 and
ATCC 33277 with minimum inhibitory concentration (MIC) values
of 773 mM and 1.875 mM, respectively (Table 1). In contrast, there
was no visual inhibition of cell growth by the F. nucleatum strain
ATCC 25586 and the S. aureus strain N315 with concentrations of
up to 3 mM. These preliminary screens suggest that compound 2
shows selective growth inhibition against P. gingivalis. A likely
explanation for the significant enzymatic activity and relatively
lower cellular activity of compound 2 is that the highly hydrophilic
molecule does not easily penetrate through cell membranes. Thus,
increasing its hydrophobicity by modifying of the DAP skeleton
would appear to be a reasonable strategy for improving cellular
activity.
The different biological assay results of these two enantiomers
in their enzymatic inhibitions and antibacterial activities
Fig. 4. The alignment of backbone atoms of proteins 3BIO (blue) and 3DAP (orange).
The ligands in 3BIO and 3DAP are shown as sticks and colored in magentas and
green, respectively.
Table 1
The minimum inhibitory concentrations (MIC) of compound 2 against periodontal
pathogens.
Periodontal pathogens
MIC
m-Ddh
homologs
Table 2
P. gingivalis strain W83
P. gingivalis strain ATCC 33277
Fusobacterium nucleatum strain
ATCC 25586
773 mM
1.875 mM
No growth inhibition up
to 3 mM
Yes
Yes
No
The GOLD (CHEM-PLP) and HINT scores for the optimal binding poses of compounds 1
and 2.
Compound 1
Compound 2
Staphylococcus aureus strain
N315
No growth inhibition up
to 3 mM
No
CHEM-PLP
HINT
41
245
51
477
Please cite this article in press as: Ma H., et al. Bioorg. Med. Chem. Lett. (2017), http://dx.doi.org/10.1016/j.bmcl.2017.06.056

4
H. Ma et al. / Bioorganic & Medicinal Chemistry Letters xxx (2017) xxx–xxx
Fig. 5. The binding modes of (a) compound 1 and (b) compound 2 in the binding pocket of m-Ddh from docking studies. Compounds 1 atoms: carbon (orange); compounds 2
atoms: carbon (green); amino acid residue atoms: carbon (yellow); oxygen (red); nitrogen (blue). The magenta dashed lines represent the hydrogen bond.
energy minimized (20,000 iterations) with the Tripos forcefield.
The resulting 3D structures of m-DAP, compounds 1 and 2 are dis-
played in Fig. 3. Overlaying these molecules showed that the con-
formations of portion A in compounds 1 and 2 were similar, while
the conformations of portion B were different due to the effects of
stereochemistry.
closer to the side chain of Asp90. This produces an unfavorable
interaction between these two negative centers, which is detri-
mental to the binding of compound 1. The two different configura-
tions of 1 and 2, particularly in portion B, produces variant binding
affinities and ultimately, different enzymatic activities of the two
compounds.
We are not aware of any report on the crystal structure of m-
Ddh (oxidoreductase, Gfo/Idh/MocA family member) from P. gingi-
valis in complex with compounds 1 or 2. However, the crystal
structures of m-Ddh from P. gingivalis in complex with another
compound, 4-(2-hydroxyethyl)-1-piperazine ethanesulfonic acid
In conclusion, two diastereomers of DAP analogs bearing an
isoxazoline moiety were synthesized and evaluated for their inhi-
bition against m-Ddh from the periodontal pathogen P. gingivalis,
and for their antibacterial activities. The results showed that only
compound 2, which has an S stereocenter at the C-5 position of
the isoxazoline ring, displays significant inhibitory activity against
m-Ddh under physiological conditions. Additionally, in our limited
survey, compound 2 also shows selective inhibition against P. gin-
givalis. Molecular modeling studies suggested that the inversion of
chirality at the C-5 position of isoxazoline ring between com-
pounds 1 and 2 leads to their different binding affinities and that
the hydrogen bond between compound 2 and m-Ddh plays a cru-
cial role in its enzymatic inhibition. These preliminary results sug-
gest that compound 2 may serve as a promising lead compound for
the discovery of novel antibiotics to treat periodontal diseases with
selective effectiveness and reduced toxicity. Further studies on
rational design and synthesis of m-DAP analogs are underway.
(
PDB ID: 3BIO) and compound 2 in complex with C. glutamicum
diaminopimelate dehydrogenase have been reported (PDB ID:
8
3
DAP). Comparing these two crystal structures, we found that
the binding site of compound 2 in C. glutamicum diaminopimelate
dehydrogenase was very similar to that of 4-(2-hydroxyethyl)-1-
piperazine ethanesulfonic acid in m-Ddh from P. gingivalis
(Fig. 4). Thus, we adopted 3BIO as our template protein and the
binding orientation of compound 2 in 3DAP for the target com-
pounds in m-Ddh. The protein was built by adding hydrogen
atoms, assigning Gasteiger-Hückel charges, and optimizing hydro-
gen coordinates by a 10,000 iteration minimization while holding
all heavy atoms as fixed (an aggregate) under the Tripos forcefield
2
0
(
TAFF) in Sybyl-X 2.0.
Compounds 1 and 2 were next docked into the m-Ddh binding
Acknowledgements
2
1
site with GOLD 5.4. The docking models were analyzed based on
two criteria: 1) the binding orientation of compounds 1 and 2 in m-
Ddh being similar to that of compound 2 in the C. glutamicum pro-
tein and 2) high CHEM-PLP scores. We then ranked these docking
models using the HINT (Hydropathic INTeractions) free energy
The authors thank NIH for financial support. This work was par-
tially supported by R01DE023078 and Virginia Commonwealth
University CCTR Endowment Fund Multi-School Grant.
2
2
scoring tool. The docking models with the highest HINT score
of compounds 1 and 2 were chosen as the optimal docking models,
respectively (Table 2 and Fig. 5). Fig. 5 shows that the hydrophobic
interactions with Trp119, Phe144 and the electrostatic interactions
with the side chains of Ser149 and Thr169 in the m-Ddh/com-
pound 1 and m-Ddh/compound 2 complexes were almost the
same, i.e., mainly with portion A of 1 and 2. However, for portion
B of compounds 1 and 2, the interactions are very different: in
the m-Ddh/compound 2 complex (Fig. 5b), there is a hydrogen
bond to Asp120, but in the m-Ddh/compound 1 complex (Fig. 5a),
the inversion of chirality leads to two consequences, a) that hydro-
gen bond to Asp120 is lost and 2) the carboxyl group of 1 moved
References
1. Boucher HW, Talbot GH, Bradley JS, et al. Clin Infect Dis. 2009;48:1.
2
3
4
. Avila M, Ojcius DM, Yilmaz O. DNA Cell Biol. 2009;28:405.
. Johnson CL, Versalovic J. Pediatrics. 2012;129:950.
. Jernberg C, Lofmark S, Edlund C, et al. Microbiol SGM. 2010;156:3216.
5. Rashid MU, Weintraub A, Nord CE. Anaerobe. 2012;18:249.
6
7
8
. Scapin G, Blanchard JS. Adv Enzymol Ramb. 1998;72:279.
. Cox RJ. Nat Prod Rep. 1996;13:29.
. Scapin G, Cirilli M, Reddy SG, et al. Biochemistry. 1998;37:3278.
9. Misono H, Soda K. J Biol Chem. 1980;255:599.
10. Maniscalco SJ, Saha SK, Fisher HF. Biochemistry. 1998;37:14585.
11. Gelb MH, Lin YK, Pickard MA, et al. J Am Chem Soc. 1990;112:4932.
12. Lam LKP, Arnold LD, Kalantar TH, et al. J Biol Chem. 1988;263:11814.
13. Abbott SD, Lanebell P, Sidhu KPS, et al. J Am Chem Soc. 1994;116:6513.
Please cite this article in press as: Ma H., et al. Bioorg. Med. Chem. Lett. (2017), http://dx.doi.org/10.1016/j.bmcl.2017.06.056

H. Ma et al. / Bioorganic & Medicinal Chemistry Letters xxx (2017) xxx–xxx
5
1
1
1
1
1
1
4. Misono H, Soda K. Agr Biol Chem. 1980;44:227.
20. Ballesteros JA, Weinstein H. Integrated methods for the construction of three
dimensional models and computational probing of structure function relations
in G protein-coupled receptors. In: Sealfon SC, Conn PM, eds. Methods in
Neurosciences. San Diego, CA: Academic Press; 1995:366–428.
21. Gohlke H, Kiel C, Case DA. J Mol Biol.. 2003;330:891.
22. Kellogg GE, Abraham DJ. Hydrophobicity: is LogP o/w more than the sum of its
parts? Eur J Med Chem. 2000;651:661.
5. Misono H, Ogasawara M, Nagasaki S. Agr Biol Chem. 1986;50:2729.
6. Kassebaum NJ, Bernabe E, Dahiya M, et al. J Dent Res. 2014;93:1045.
7. Stone VN, Parikh HI, El-rami F, et al. PLoS One. 2015;10.
8. Kozikowski AP, Adamczyk M. J Org Chem. 1983;48:366.
9. Jorgensen JH, Turnidge JD. Manual of Clinical Microbiology. 9th ed. Washington,
DC: American Society for Microbiology; 2007.
Please cite this article in press as: Ma H., et al. Bioorg. Med. Chem. Lett. (2017), http://dx.doi.org/10.1016/j.bmcl.2017.06.056