Journal of Antimicrobial Chemotherapy Advance Access published August 18, 2014

J Antimicrob Chemother
doi:10.1093/jac/dku309

Activity of the type I signal peptidase inhibitor MD3 against

multidrug-resistant Gram-negative bacteria alone
and in combination with colistin
Yoann Personne1, Michael A. Curtis2, David W. Wareham1 and Richard D. Waite1*

1
Centre for Immunology and Infectious Disease, Blizard Institute, Barts and The London School of Medicine and Dentistry, Queen Mary
University of London, 4 Newark Street, London E1 2AT, UK; 2Institute of Dentistry, Barts and The London School of Medicine and Dentistry,
Queen Mary University of London, 4 Newark Street, London E1 2AD, UK

*Corresponding author. Tel: +020-7882-2326; Fax: +020-7882-2181; E-mail: richard.d.waite@gmail.com

Downloaded from http://jac.oxfordjournals.org/ at Aston University on August 22, 2014

Received 22 May 2014; returned 23 June 2014; revised 26 June 2014; accepted 17 July 2014

Objectives: Effective treatment of Gram-negative bacterial infections is increasingly challenging due to the
spread of multidrug-resistant strains and a lack of new antimicrobials in development. Bacterial type I signal
peptidases (SPases) represent a highly conserved and essential target for inhibition by novel compounds.
SPases are required for the effective processing of membrane translocated proteins involved in core functions
related to metabolism, virulence and resistance. In this study we assessed the biochemical and functional
activity of a novel synthetic inhibitor (MD3) of SPases against a wide range of Gram-negative pathogens.
Methods: The activity and specificity of MD3 for recombinant Pseudomonas aeruginosa SPase (LepB) and a gen-
etically engineered LepB-regulatable strain were investigated. Antimicrobial activity of the compound alone and
in combination with outer membrane-permeabilizing agents (sodium hexametaphosphate, colistin) was also
determined against a collection of P. aeruginosa, Acinetobacter baumannii, Klebsiella pneumoniae and
Stenotrophomonas maltophilia isolates.
Results: MD3 was found to inactivate the P. aeruginosa LepB protein (IC50 10 mM), resulting in antimicrobial
effects potentiated in the presence of colistin. MD3 also demonstrated potent activity against wild-type and
multidrug-resistant strains of A. baumannii and S. maltophilia with MICs ranging from 0.5 to 14 mg/L in the
presence of subinhibitory concentrations of colistin.
Conclusions: MD3 is a novel inhibitor of bacterial SPase in a range of non-fermentative Gram-negative bacteria.
The antimicrobial activity is potentiated in combination with colistin and suggests that SPase inhibition warrants
further exploration as a basis for future mono or combination therapies.

Keywords: antibiotic synergy, novel therapies, MDR

Introduction processing of immature proteins in bacteria by cleaving the amino-

There is an urgent need for new agents for the treatment of
Gram-negative infections due to the global spread of multidrug-
resistant (MDR) strains and a dearth of novel synthetic compounds
beyond the early phases of development. The most pressing needs
are for drugs able to target the ESKAPE group of pathogens, which
include MDR strains of Pseudomonas aeruginosa, Acinetobacter
baumannii and carbapenem-resistant Enterobacteriaceae.1 The
development of compounds that target core metabolic processes
conserved amongst a wide range of MDR bacteria needs to be
accelerated and one target that has not been exploited in these
organisms is the type I signal peptidase (SPase). SPases are cyto-
plasmic membrane-bound enzymes, critical to the effective
terminal signal peptides of proteins translocated by the general
secretion (Sec) pathway.
Type I signal peptides have three conserved domains: a posi-
tively charged amino-terminus, a central hydrophobic region
and a hydrophilic carboxy-terminal domain containing the
SPase cleavage site.2 In P. aeruginosa there are predicted to be
more than 800 proteins (.14% of the proteome) that possess
type I signal peptides, including proteins involved in iron uptake,
flagellar biosynthesis, virulence (elastases LasA and LasB) anti-
microbial resistance (AmpC b-lactamase), global regulation and
many Sec-independent twin-arginine translocation (TAT) pathway
substrates.3 6 This suggests a direct role for SPases in cellular
housekeeping and pathogenesis, and this was confirmed in our

# The Author 2014. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved.
For Permissions, please e-mail: journals.permissions@oup.com

1 of 8
Personne et al.
recent study that demonstrated LepB to be the primary SPase of
P. aeruginosa and essential for viability, whilst its paralogue
PA1303 was found to be dispensable yet still involved in virulence
factor secretion.7 Importantly, LepB is highly conserved amongst
P. aeruginosa clinical isolates, is not a target of existing therapies
and is thus an attractive target for a new inhibitor. The suitability
of SPases as drug targets more widely is supported by literature
showing them to be ubiquitous amongst bacteria and essential
in all bacterial species examined so far.7 10 SPase enzymes oper-
ate via a unique Ser/Lys catalytic dyad enabling specific inhibition
and the active site is located on the outer surface of the cytoplas-
mic membrane, making it accessible to any inhibitor able to
traverse the outer membrane (OM) of Gram-negative organ-
isms.11 Furthermore, monomeric bacterial SPases are structurally
different from the multimeric SPases of eukaryotes, which gives
confidence that selective toxicity using novel inhibitors can be
achieved.2
SPases are not sensitive to classical inhibitors of serine, cysteine,
aspartic and metalloproteases.12 However, b-lactam analogues, as
well as arylomycin and lipoglycopeptide natural products, have
been shown to inhibit SPase activity.13 17 Unfortunately, only
modest antibacterial activity has so far been demonstrated
for these inhibitors, mostly against Gram-positive bacteria organ-
isms, and no studies to date have reported antibacterial activity
against P. aeruginosa strains that have not been genetically
sensitized.15,17,18
A synthetic b-aminoketone inhibitor, MD3 [1-(2,5-dichlorophe-
nyl)-3-(dimethylamino)propan-1-one] was identified during a
screening of a library of compounds for their ability to inhibit recom-
binant Escherichia coli SPase (LepB).19 It has antimicrobial activity
against Staphylococcus aureus, Haemophilus influenzae, Moraxella
catarrhalis and Mycobacterium tuberculosis, and E. coli can be sen-
sitized in the presence of the OM-permeabilizing agentpolymyxin
B nonapeptide (PMBN).19,20 However, the activity of MD3 has not
been fully assessed against MDR Gram-negative pathogens.
Following our previous molecular characterization of SPases in
P. aeruginosa,7 the aim of this study was to use MD3 as a known
SPase inhibitor to determine whether targeting SPase function in
this organism and other Gram-negative pathogens (A. baumannii,
Klebsiella pneumoniae and Stenotrophomonas maltophilia) is a
valid antimicrobial strategy that could be exploited in future anti-
biotic development programmes. Given the previous requirement
of PMBN to render E. coli susceptible to MD3, the location of
SPases on the outer surface of the cytoplasmic membrane and
the known formidable barrier of the OMs of the Gram-negative
pathogens studied, we also assessed the antimicrobial activity of
MD3 in the presence of OM-permeabilizing agents in order to estab-
lish the potential use of SPase inhibitors as part of a combination
therapy.
Materials and methods
Bacterial isolates, culture media and chemicals
The isolates used in this study along with their characteristics and reasons
for inclusion are listed in Table 1. Strains were routinely grown in
Iso-Sensitest medium (Oxoid Ltd, Basingstoke, UK) at 378C. For the main-
tenance of genetically modified strains, media were supplemented with
gentamicin (200 mg/L) and/or tetracycline (100 mg/L for P. aeruginosa
and 15 mg/L for E. coli) where required. Resistance determinants and epi-
demiological markers amongst clinical isolates studied were determined
2 of 8
by PCR and sequencing methods as previously described (Table 1).21,22
The SPase inhibitor MD3 was synthesized as previously described and pro-
vided by Professor Tanya Parish (IDRI, Seattle, WA, USA).20
MD3 target validation assay
The specificity of MD3 as an inhibitor of type I signal peptidases was deter-
mined using recombinant P. aeruginosa LepB and fluorescence resonance
energy transfer (FRET) technology using a previously described assay.7
Two intramolecularly quenched fluorescent substrates were evaluated;
LasB FRET (Dabcyl-VSPAAFAADL-EDANS, purchased from Peptide Protein
Research Ltd, UK) contains a decapeptide based on the cleavage region
of the signal peptide sequence of the P. aeruginosa LasB protein, while
SceD FRET (Dabcyl-AGHDAHASET-EDANS, purchased from AnaSpec)
spans the signal peptide cleavage site of the Staphylococcus epidermidis
SceD protein. For IC50 value determination, MD3 was tested in triplicate
at concentrations ranging from 3 to 38 mM with LasB FRET and from 6 to
Downloaded from http://jac.oxfordjournals.org/ at Aston University on August 22, 2014
100 mM using SceD FRET. IC50 values were calculated using GraphPad
Prism software and the sigmoidal dose response function. DMSO
(Sigma Aldrich) at 0.5% (v/v) was used as a negative control.
Construction of a lepB-regulatable strain
(DlepB/pHERD26T-lepB)
In order to place lepB under the control of an arabinose-inducible pro-
moter (PBAD), this gene was amplified from P. aeruginosa PAO1 and cloned
into the shuttle vector pHERD26T to generate plasmid pHERD26T-lepB.23
The primers involved in the construction and sequencing of this plasmid
are listed in Table S1 (available as Supplementary data at JAC Online). A
strain containing a chromosomal lepB deletion and harbouring lepB in
trans on pHERD26T-lepB (DlepB/pHERD26T-lepB) was generated and con-
firmed as previously described.7
The effect of MD3 on the growth of the lepB regulatable strain (DlepB/
pHERD26T-lepB) was determined by measuring OD at 595 nm (BioTek
ELx800 plate reader) after overnight growth at 378C in Iso-Sensitest broth
containing several different concentrations of L-arabinose (0.025%, 0.05%,
0.075% and 0.1%).
In vitro activity of MD3 versus Gram-negative
pathogens
The activity of MD3 was assessed against susceptible type strains and
MDR clinical isolates of P. aeruginosa, K. pneumoniae, S. maltophilia and
A. baumannii (see Table 2). MICs were determined by broth microtitre dilu-
tion using Iso-Sensitest broth in 96-well plates and a bacterial inoculum of
104 cfu/mL. Following incubation at 378C in air for 24 h, the lowest concen-
tration of MD3 that inhibited visual growth of bacteria (no turbidity) was
recorded as the MIC. If an MIC could not be determined visually at the
highest concentration of MD3 used (28 mg/L), viable counts were per-
formed to assess the percentage survival of bacteria and establish
whether MD3 had any intrinsic activity alone.
The MIC of MD3 was also determined in the presence of OM-permeabilizing
agents (SigmaAldrich) added at the following concentrations: sodium hex-
ametaphosphate (NaHMP), 4 mg/mL; and PMBN, 10 mg/L (E. coli) and 1 or
5 mg/L (P. aeruginosa). Colistin sulphate (Sigma Aldrich) was also used at
the subinhibitory concentrations listed in Table 2 to assess synergy with a
membrane-disrupting agent with intrinsic antimicrobial activity.
Statistical analysis
Data were analysed using the Students t-test (two independent samples
assuming unequal variances). Differences were statistically significant at
P, 0.05. The degree of spread of the data is shown as standard deviation.
Signal peptidase inhibition in Gram-negative bacteria JAC
Table 1. Bacterial strains and plasmids

Strain or plasmid Relevant characteristics and/or genotype Reference/source

Strain
P. aeruginosa
ATCC 27853 antibiotic-susceptible type strain NCTC
PAO1 wild-type strainMPAO1 34
DlepB/pHERD26T-lepB PAO1 with chromosomal lepB deletion, expressing lepB in trans from this study
pHERD26T; TET-resistant, GEN-resistant
CF2004 MDR CF isolate, LES 35
CF2828 MDR CF isolate, LES 35
CF1092 MDR CF isolate, LES 36
PA11 MDR urinary isolate, carbapenem-resistant VIM-2 producer this study

E. coli

Downloaded from http://jac.oxfordjournals.org/ at Aston University on August 22, 2014

ATCC 25922 antibiotic-susceptible type strain ATCC
A. baumannii
ATCC 19606 antibiotic-susceptible type strain NCTC
AB14 MDR OXA-23 producer, UK epidemic clone 1 37
AB16 MDR OXA-23 UK epidemic clone 2 37
AB186 MDR OXA-23 producer UK burn clone 37
AB205 CST-resistant (MIC .256 mg/L), OXA-23 UK epidemic clone 1 strain 36
AB211 TGC-resistant, OXA-23 UK epidemic clone 1 38
AB219 CST-resistant (MIC32 mg/L), UK epidemic OXA-23 clone 1 36
AB292 MDR PFGE-defined OXA-23-like carbapenemase producer this study
S. maltophilia
NCTC 10258 antibiotic-susceptible type strain NCTC
Sm59 MDR environmental isolate 39
Sm60 MDR environmental isolate 39
Sms01 MDR respiratory isolate 39

K. pneumoniae
NCTC 9633 antibiotic-susceptible type strain NCTC
KP10 MDR ST147 NDM-1 producer this study
KP51 TEM-1, OXA-1, SHV-1 and CTX-M-15 producer 36
KPC-3 MDR TEM-1, SHV-11 and KPC-3 producer 40
KP7 MDR ST258 and KPC-2 producer this study
Plasmid
pEX100T-DlepB::Gmr pEX100T-lepB::Gmr with 700 bp of lepB removed; deletion construct used 7
for lepB chromosomal mutation analysis; AMP-resistant, GEN-resistant, sacB+
pHERD26T pUCP26 Plac replaced with 2.4 kb AdhI-EcoRI fragment of araC-PBAD cassette and oriT; TET-resistant 23
pHERD26T-lepB pHERD26 T carrying lepB; TET-resistant this study

NCTC, National Collection of Type Cultures; CF, cystic fibrosis; LES, Liverpool Epidemic strain; TET, tetracycline; GEN, gentamicin; CST, colistin;
TGC, tigecycline; AMP, ampicillin; sacB+, levansucrase-encoding gene.

Results
Inhibition of the P. aeruginosa SPase LepB by MD3
The ability of MD3 to inhibit the enzymatic activity of recombinant
P. aeruginosa SPase LepB was determined using FRET technology.
In this assay, the internally quenched fluorescent peptide sub-
strate sequence was based on the signal peptide sequence cleav-
age region of P. aeruginosa preproenzyme elastase LasB, a
well-characterized secreted virulence factor. A characteristic of
type I signal peptides is that alanine residues are the most com-
mon amino acids at positions 21 and 23 relative to the cleavage
site and the LasB signal peptide cleavage site has been experimen-
tally identified as the peptide bond between the alanine residues at
positions 23 and 24 relative to the N-terminal methionine.5 The
peptide substrate Dabcyl-VSPAAFAADL-EDANS was designed to
span 10 amino acids of the LasB signal peptide cleavage region
(alanine residues 7 and 8 corresponding to positions 23 and 24
of the preproenzyme) and we have previously demonstrated dose-
dependent cleavage of this peptide by recombinant P. aeruginosa
LepB enzyme.7 When MD3 was incorporated into the assay, cleav-
age of this LasB FRET peptide by LepB (1 mM) was found to be inhib-
ited and an IC50 of 8 mM (R 2 0.99) was observed (Figure 1a).
Inhibition of LepB activity by MD3 was confirmed using a second

3 of 8
Personne et al.

(a) 150 LasB FRET (b) 150 SceD FRET

Activity (%)

Activity (%)
100 100

50 50

0 0
0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0
Log [MD3] mM Log [MD3] mM

Figure 1. Inhibition of P. aeruginosa recombinant LepB activity by MD3. IC50s generated using two different FRET peptides: (a) LasB, Dabcyl-VSPAAFAADL-
EDANS; and (b) SceD, Dabcyl-AGHDAHASET-EDANS. Assay performed using 1 mM LepB and 20 mM FRET peptide in a reaction buffer (50 mM Tris-HCl,
pH 7.5) containing 0.25% Triton X-100. The increase in fluorescence signal after 30 min of incubation at 378C was calculated for the untreated
control [0.5% (v/v) DMSO] and MD3 test reactions. Data normalized as percentage of fluorescence activity relative to the untreated control. Data
points shown are the means of three replicates. IC50s were calculated with GraphPad Prism software using the sigmoidal dose response function.

Downloaded from http://jac.oxfordjournals.org/ at Aston University on August 22, 2014

Table 2. Effect of MD3 on susceptible type strains and MDR clinical Gram-negative bacteria

MD3 MIC (mg/L) Colistin MD3 MIC (mg/L)

MD3 MIC in combination with Sensitization Colistin supplement in combination Sensitization
Bacterial strain (mg/L) NaHMP (4 mg/mL) factora MIC (mg/L) (mg/L) with colistin factora

P. aeruginosa
ATCC 27853 56 56 (28c) 1 1 0.75 28 2
PAOI 56 56 (28c) 1 1 0.75 56 1
CF2004 56 28 2 0.75 0.5 3.5 16
CF2828 56 28 2 0.75 0.5 56 (3.5 28c) 1
CF1092 56 14 4 1 0.75 14 4
PA11 56 56 1 0.75 0.5 7 8
A. baumannii
ATCC 19606 56 (88%b) 3.5 16 0.5 0.25 0.5 112
AB14 56 (81%b) 14 4 0.75 0.5 1.75 32
AB16 56 (76%b) 7 8 0.75 0.5 0.9 62
AB186 56 (80%b) 14 4 0.75 0.5 0.9 62
AB205 56 (97%b) 1.75 32 .2 2 7 8
AB211 56 (78%b) 7 8 0.5 0.25 3.5 16
AB219 56 (88%b) 3.5 16 .2 2 14 4
AB292 56 (70%b) 14 4 0.75 0.5 1.75 32
S. maltophilia
NCTC 10258 56 (93%b) 14 4 .2d 0.5 1.75 32
Sm59 56 (77%b) 14 4 .2d 0.5 1.75 32
Sm60 56 28 2 .2 0.5 7 8
Sms01 56 (61%b) 28 2 .2 0.5 3.5 16
K. pneumoniae
NCTC 9633 56 56 1 0.5 0.25 3.5 16
KP10 56 56 1 0.5 0.25 56 1
KP51 56 56 1 0.5 0.25 56 1
KPC-3 56 56 1 0.5 0.25 56 1
KP7 56 56 1 0.5 0.25 56 1

a
The highest concentration of MD3 used was 28 mg/L MD3. As an MIC was not obtained for all organisms when challenged with MD3 delivered alone, the
lowest theoretical MIC is 56 mg/L MD3. Therefore the sensitization factor was calculated by dividing 56 by the MIC value obtained when MD3 was in
combination with NaHMP or colistin.
b
Evidence of intrinsic activity for MD3. Percentage reduction in cfu in test reaction after exposure to 28 mg/L MD3 (overnight incubation at 378C, no
added permeabilizer, comparison made with the untreated control).
c
Antibacterial activity, but no MIC observed. At the indicated concentrations the viable count in the test was 10% of the control cfu count.
d
Intrinsic activity observed for colistin, but not optical clearing: 0.75 2 mg/L colistin.

4 of 8
Signal peptidase inhibition in Gram-negative bacteria
peptide (Dabcyl-AGHDAHASET-EDANS) based on the signal peptide
sequence of the pre-SceD protein from S. epidermidis (cleavage site,
AS bond), also containing alanine residues at positions 21 and
23 relative to the SPase cleavage site and validated as a means
for assessing the in vitro activity of SPases of S. epidermidis and
S. aureus.16,24 This alternative peptide sequence was also cleaved
by recombinant LepB protein and MD3 inhibited this reaction with
an IC50 value of 11.9 mM (R 2 0.97) (Figure 1b). These data using
two different SPase cleavage sequences provide convincing evi-
dence that MD3 is an effective inhibitor of the enzymatic activity
of the P. aeruginosa LepB SPase.
Susceptibility of P. aeruginosa to MD3 is dependent on
lepB expression
The antimicrobial activity of MD3 was confirmed by testing it
JAC
absence of arabinose, confirming lepB as an essential gene in
P. aeruginosa growth (Figure 2), or in the presence of 1%
L-arabinose (data not shown), suggesting that lepB expression is
tightly regulated and that overexpression may be toxic to the cell.
MD3 at 28 mg/L was able to significantly reduce the growth
of the lepB-regulatable strain across all concentrations of
L-arabinose supplementation (P, 0.001), but was most marked
in the 0.025% 0.075% range (2- to 8-fold reduction compared
with the paired control culture) (Figure 2). These data also high-
light that, in the presence of MD3, growth can be returned to
almost control levels at the highest level of lepB expression exam-
ined (0.1% L-arabinose, only a 1.06-fold reduction compared with
the paired control culture) (Figure 2).
MD3 has antibacterial activity against P. aeruginosa in

Downloaded from http://jac.oxfordjournals.org/ at Aston University on August 22, 2014

against E. coli. This organism is known to be sensitized to MD3 in
the presence of 10 mg/L PMBN,19 and in our analysis an MIC of
12.5 mM (3.5 mg/L) for E. coli ATCC 25922 was observed.
MD3 alone was found not to be active against P. aeruginosa
PAO1 and ATCC 27853 with no inhibition of growth observed
when challenged with MD3 at 28 mg/L (100 mM) (Table 2). As
under-expression of a drug target is a known strategy for sensitiz-
ing bacteria and confirming on-target activity of experimental
agents, we constructed a lepB-regulatable derivative in strain
PAO1 (DlepB/pHERD26T-lepB). In this strain the lepB gene was
removed (amino acids 18252 are deleted from the chromosome
and replaced by a gentamicin resistance cassette) and was com-
plemented with a functional copy of lepB under the control of an
arabinose-inducible promoter provided in trans (pHERD26T-lepB).
Growth of this strain was found to be directly correlated with levels
of LepB expression, with dose-dependent decreases in growth
observed upon reducing the concentration of L-arabinose to
0.05% and 0.025% (Figure 2). No growth was observed in the
1.2
*** *** *** ***
1 (CF2004, CF1092 and PA11; sensitization factors 4 to
0.8
DMSO
OD595
the presence of OM permeabilizers
Translocation of compounds across the OM is a significant barrier
to the potency of many antibiotics against P. aeruginosa.25 We
assessed synergy between MD3 and a number of agents known
to increase the permeability of the OM. PMBN inhibits the growth
of P. aeruginosa at much lower concentrations than are seen with
E. coli 26 and did not facilitate antibacterial activity for MD3 against
P. aeruginosa at the concentrations of PMBN used (1 and 5 mg/L,
MIC 10 mg/L). However NaHMP, a chelating agent and effective
OM permeabilizer of P. aeruginosa,27 at 4 mg/mL inhibited the
growth of ATCC 27853 and PAO1 in the presence of 28 mg/L
MD3. Viable counts after overnight exposure to MD3 showed a
10-fold reduction in cfu/mL when compared with untreated
controls (Table 2). The MIC of the MD3/NaHMP combination ran-
ged from 14 to 28 mg/L (sensitization factor 2 to 4) for three
MDR cystic fibrosis isolates (CF1092, CF2004 and CF2828)
(Table 2). Synergy was also observed when MD3 was combined
with colistin, a drug with both intrinsic antimicrobial and cell-
permeabilizing properties.28 31 Interestingly, after exposure to
sub-MIC concentrations of colistin (0.5 0.75 mg/L) paradoxical
effects on the susceptibility of MDR and susceptible P. aeruginosa
isolates were observed with MICs ranging from 3.5 to 14 mg/L
16-fold), whilst little activity was observed against ATCC 27853

0.6 and PAO1 (MICs 28 mg/L) (Table 2).

MD3
0.4
0.2
Antimicrobial activity of MD3 against other
Gram-negative bacteria
0
The activity of MD3 alone and in combination with NaHMP or
e

e
os

os

os

os

os

colistin was tested against selected strains of three other

in

in

in

in

in
ab

ab

ab

ab

ab

Gram-negative pathogensA. baumannii, K. pneumoniae and

ar

ar

ar

ar

ar
1%

5%

5%

No

S. maltophilia. A BLASTP search of A. baumannii, K. pneumoniae

75
0.

02
0

0.

and S. maltophilia genome sequences revealed the presence of

0.

0.

putative SPases in all three organisms. Both A. baumannii and

Figure 2. P. aeruginosa strain PAO1 is sensitized to MD3 upon reduction of
expression of lepB. The lepB-regulatable strain (DlepB/pHERD26T-lepB)
was used for this analysis. Expression of lepB is directed from a PBAD
promoter present on plasmid pHERD26T-lepB and thus controlled
through the indicated concentrations of L-arabinose. Control, 1% DMSO
(grey bars); MD3 used at a concentration of 28 mg/L (white bars).
Results are the means+SD of four biological replicates grown in 96-well
microtitre plates overnight at 378C. Paired cultures (+MD3) with
significantly different results are denoted with a triple asterisk
(***P,0.001, two-tailed t-test).
K. pneumoniae possess a single SPase homologue (sequence
identities with P. aeruginosa PAO1 LepB, 41% and 40%, respect-
ively), whilst S. maltophilia K279a harbours two SPase homolo-
gues (sequence identities to PAO1 LepB, 41% and 32%). All of
these proteins contain the characteristic catalytic Ser/Lys dyad,
and the four regions of significant sequence homology that
make up the conserved catalytic domain (referred to as boxes B
to E) associated with these enzymes, suggesting that MD3 may
also be capable of inhibiting these enzymes (Figure 3).

5 of 8
Personne et al.

When tested alone MD3 was found to have activity against
A. baumannii ATCC 19606 with an 88% reduction in the viable
count observed after exposure to 28 mg/L MD3. In combination
with OM permeabilizers, MICs as low as 3.5 mg/L (sensitization
factor of 16) and 0.5 mg/L (sensitization factor of 112) were
seen with NaHMP (4 mg/mL) and colistin (0.25 mg/L), respectively
(Table 2). We also observed evidence of antimicrobial activity
against seven clinical MDR strains (AB14, AB16, AB186, AB205,
AB211, AB219 and AB292) alone and MICs ranging from 1.75 to
14 mg/L (sensitization factors 4 to 32) in combination
with NaHMP (4 mg/mL) and from 0.9 to 14 mg/L (sensitization
factors of 4 to 62) with sub-MIC concentrations of colistin
(0.25 2 mg/L) (Table 2).
MD3 showed a similar profile of activity against S. maltophilia,
with intrinsic activity observed against NCTC 10258 and against
two of three MDR strains (Sm59 and Sms01). MICs in the presence
K. pneumoniae appeared to be the least susceptible to MD3. In
combination with 0.25 mg/L colistin an MIC of 3.5 mg/L was
observed against NCTC 9633. No activity was seen either alone
or in combination against the four clinical MDR strains tested
(KP10, KP51, KPC-3 and KP7) (Table 2).
Discussion
As the incidence of MDR Gram-negative bacteria increases there is
a critical need for novel drugs targeting different cellular pathways
to currently used antibiotics. In this study we show that SPase
inhibition in combination with membrane permeabilization may
be a useful approach to be developed for treating infections
caused by these organisms.
Using a lepB-regulatable derivative of P. aeruginosa we
observed that reducing LepB concentrations results in impaired

Downloaded from http://jac.oxfordjournals.org/ at Aston University on August 22, 2014

of 4 mg/mL NaHMP ranged from 14 to 28 mg/L (sensitization fac- or no growth, thus confirming that LepB is the primary SPase
tor range 2 to 4) and between 1.75 and 7 mg/L, (sensitization and that expression of PA1303 alone cannot support the growth
factor range 8 to 32) in combination with colistin (0.5 mg/L) of this organism. MD3 acts as an effective inhibitor of recombinant
(Table 2). P. aeruginosa LepB with an average IC50 value of 10 mM against

Figure 3. Alignment of SPase amino acid sequences. LepB (SPase) amino acid sequences (http://www.ncbi.nlm.nih.gov/) from the following strains used:
PAO1, P. aeruginosa PAO1; Ab, A. baumannii ATCC 19606; Kp, K. pneumoniae VA360; and Sm, S. maltophilia K279a. The sequences were aligned using
CLUSTAL W 1.83 (www.ch.embnet.org). Conserved boxes AE (shaded grey) were identified using sequence alignments and the following consensus
motifs, which are read with an upper-case underlined character denoting absolutely conserved, an upper-case character denoting conserved and x
denoting not conserved: box B, IPSGSMxPTLx; box C, RGDIVVFxxP; box D, YIKRxxGxPGDxV; box E, VPxGxYFxMGDNRDNSxDSR.12,41 Box A (transmembrane
region) is labelled as previously identified.41 The catalytic serine and lysine residues are boxed. Identical residues are also indicated (asterisks), as are
conserved residues (colons).

6 of 8
Signal peptidase inhibition in Gram-negative bacteria
two different peptide substrates, and this activity is comparable to
that previously observed using recombinant E. coli LepB (IC50 of
5 mM) in a different biochemical assay.19 We also demonstrated
on-target activity for MD3 using an engineered strain of P. aerugi-
nosa in which lepB was under-expressed, and improved activity in
the presence of two membrane-permeabilizing agents (NaHMP
and colistin) with the greatest synergy observed when MD3 was
combined with colistin against MDR strains of P. aeruginosa.
Importantly, this is the first study to find antibacterial activity
against the wild-type LepB enzyme of P. aeruginosa. At present
the number of SPase inhibitors is small, but it includes a number
of b-lactam analogues, arylomycins and lipoglycopeptide com-
pounds.13,15,17,18 Arylomycins are natural products that show
antibacterial activity; however, P. aeruginosa, like many other bac-
teria, are not susceptible to arylomycin due to a specific proline
residue within the target SPase that disrupts interactions with
the lipopeptide tail of the antibiotic. However, P. aeruginosa can
be rendered susceptible through mutation of this proline residue
in LepB.15,32
In this study we observed that, in general, MD3 together with
colistin was a more potent combination than MD3 with NaHMP. In
addition, the most dramatic antimicrobial activity was observed
against both type and colistin-susceptible MDR strains of
A. baumannii when they were challenged with MD3 in combin-
ation with colistin (MD3 MICs 0.5 3.5 mg/L, sensitization factor
range 16 to 112), although MD3 appeared to have some intrin-
sic activity alone. Similarly strains of S. maltophilia were also
sensitized to MD3 in the presence of colistin. These data suggest
that SPase inhibition in combination with colistin may be a valid
strategy for targeting these pathogens. Against colistin-resistant
strains of A. baumannii the activity of MD3 was enhanced with
addition of NaHMP (MICs of 1.75 and 3.5 mg/L for strains AB205
and AB219), suggesting that SPase inhibition in combination
with an alternative membrane-permeabilizing agent could be
pursued as an alternative strategy.
Whether the potent synergy observed using MD3 and colistin is
due to the OM-damaging activity of colistin, enabling MD3 to bind
more readily to the inner membrane-bound SPase target or
whether the relationship between the two drugs is co-operative
will only be established once MD3 is tested in combination with
OM permeabilizers that are more effective than NaHMP.
Interestingly it has been found in E. coli that gross defects in cell
architecture and membrane integrity occur upon depletion of cel-
lular LepB protein concentrations;33 therefore the development of
more potent SPase inhibitors could also have the dual benefit of
potentiating the activity of a partner drug during combination
therapy of Gram-negative bacteria.
Although MD3 has been reported to be biochemically unstable,
a synthesized breakdown product (MD4) was shown to retain both
biochemical and antibacterial activity, and thus optimization of
this chemical class should focus around this reactive analogue.19
Nevertheless, the potent activity of MD3 in combination with
colistin against MDR strains of A. baumannii and S. maltophilia
and the evidence of broader activity against P. aeruginosa and
K. pneumoniae emphasizes the need for new and more potent
SPase inhibitors. Either structurally adjusted analogues of existing
natural products or rationally designed synthetic inhibitors could
form the basis for future mono or combination therapies for
Gram-negative bacteria. Furthermore, as the antimicrobial thera-
peutic options are becoming more limited for healthcare
JAC
professionals, this study highlights that the identification of
effective novel drug combinations deserves further exploration
and suggests that experimental compounds should not necessar-
ily be discarded as inhibitors of Gram-negative bacteria if they are
ineffective as a monotherapy.
Acknowledgements
We would like to thank Professor Tanya Parish for the gift of MD3.
Funding
This work was supported by the BSAC through grant number GA2012_15R.
Downloaded from http://jac.oxfordjournals.org/ at Aston University on August 22, 2014
Transparency declarations
D. W. W. has received research grants from Wyeth/Pfizer, AstraZeneca
and Basilea, as well as funds to attend educational meetings and
conferences. D. W. W. has acted as an advisor to Merck Sharp & Dohme
and Forest. All other authors: none to declare.
Supplementary data
Table S1 is available as Supplementary data at JAC Online (http://jac.
oxfordjournals.org/).
References
1 Boucher HW, Talbot GH, Bradley JS et al. Bad bugs, no drugs: no ESKAPE!
An update from the Infectious Diseases Society of America. Clin Infect Dis
2009; 48: 1 12.
2 Paetzel M, Karla A, Strynadka NC et al. Signal peptidases. Chem Rev
2002; 102: 4549 80.
3 Lewenza S, Gardy JL, Brinkman FS et al. Genome-wide identification of
Pseudomonas aeruginosa exported proteins using a consensus computa-
tional strategy combined with a laboratory-based PhoA fusion screen.
Genome Res 2005; 15: 3219.
4 Gustin JK, Kessler E, Ohman DE. A substitution at His-120 in the LasA
protease of Pseudomonas aeruginosa blocks enzymatic activity without
affecting propeptide processing or extracellular secretion. J Bacteriol
1996; 178: 6608 17.
5 Kessler E, Safrin M, Peretz M et al. Identification of cleavage sites
involved in proteolytic processing of Pseudomonas aeruginosa preproelas-
tase. FEBS Lett 1992; 299: 2913.
6 Chaddock AM, Mant A, Karnauchov I et al. A new type of signal peptide:
central role of a twin-arginine motif in transfer signals for the delta
pH-dependent thylakoidal protein translocase. EMBO J 1995; 14: 2715 22.
7 Waite RD, Rose RS, Rangarajan M et al. Pseudomonas aeruginosa pos-
sesses two putative type I signal peptidases, LepB and PA1303, each
with distinct roles in physiology and virulence. J Bacteriol 2012; 194:
4521 36.
8 Cregg KM, Wilding I, Black MT. Molecular cloning and expression of
the spsB gene encoding an essential type I signal peptidase from
Staphylococcus aureus. J Bacteriol 1996; 178: 5712 8.
9 Date T. Demonstration by a novel genetic technique that leader peptid-
ase is an essential enzyme of Escherichia coli. J Bacteriol 1983; 154: 76 83.
10 Zhang YB, Greenberg B, Lacks SA. Analysis of a Streptococcus pneumo-
niae gene encoding signal peptidase I and overproduction of the enzyme.
Gene 1997; 194: 249 55.
7 of 8
Personne et al.
11 Auclair SM, Bhanu MK, Kendall DA. Signal peptidase I: cleaving the way
to mature proteins. Protein Sci 2012; 21: 1325.
12 Dalbey RE, Lively MO, Bron S et al. The chemistry and enzymology of the
type I signal peptidases. Protein Sci 1997; 6: 1129 38.
13 Black MT, Bruton G. Inhibitors of bacterial signal peptidases. Curr Pharm
Des 1998; 4: 13354.
14 Luo C, Roussel P, Dreier J et al. Crystallographic analysis of bacterial sig-
nal peptidase in ternary complex with arylomycin A2 and a b-sultam
inhibitor. Biochemistry 2009; 48: 8976 84.
15 Roberts TC, Schallenberger MA, Liu J et al. Initial efforts toward the
optimization of arylomycins for antibiotic activity. J Med Chem 2011; 54:
4954 63.
16 Bockstael K, Geukens N, Rao CV et al. An easy and fast method for the
evaluation of Staphylococcus epidermidis type I signal peptidase inhibitors.
J Microbiol Meth 2009; 78: 231 7.
17 Kulanthaivel P, Kreuzman AJ, Strege MA et al. Novel lipoglycopeptides
as inhibitors of bacterial signal peptidase I. J Biol Chem 2004; 279:
36250 8.
18 Roberts TC, Smith PA, Cirz RT et al. Structural and initial biological ana-
lysis of synthetic arylomycin A2. J Am Chem Soc 2007; 129: 158308.
19 Barbosa MD, Lin S, Markwalder JA et al. Regulated expression of the
Escherichia coli lepB gene as a tool for cellular testing of antimicrobial com-
pounds that inhibit signal peptidase I in vitro. Antimicrob Agents
Chemother 2002; 46: 3549 54.
20 Ollinger J, OMalley T, Ahn J et al. Inhibition of the sole type I signal pep-
tidase of Mycobacterium tuberculosis is bactericidal under replicating and
nonreplicating conditions. J Bacteriol 2012; 194: 2614 9.
21 Parsons YN, Panagea S, Smart CH et al. Use of subtractive hybridization
to identify a diagnostic probe for a cystic fibrosis epidemic strain of
Pseudomonas aeruginosa. J Clin Microbiol 2002; 40: 4607 11.
22 Smart CH, Walshaw MJ, Hart CA et al. Use of suppression subtractive
hybridization to examine the accessory genome of the Liverpool cystic
fibrosis epidemic strain of Pseudomonas aeruginosa. J Med Microbiol
2006; 55: 67788.
23 Qiu D, Damron FH, Mima T et al. PBAD-based shuttle vectors for func-
tional analysis of toxic and highly regulated genes in Pseudomonas and
Burkholderia spp. and other bacteria. Appl Environ Microbiol 2008; 74:
7422 6.
24 Rao S, Bockstael K, Nath S et al. Enzymatic investigation of the
Staphylococcus aureus type I signal peptidase SpsBimplications for the
search for novel antibiotics. FEBS J 2009; 276: 3222 34.
25 Yoshimura F, Nikaido H. Permeability of Pseudomonas aeruginosa outer
membrane to hydrophilic solutes. J Bacteriol 1982; 152: 63642.
26 Tsubery H, Ofek I, Cohen S et al. Structure-function studies of poly-
myxin B nonapeptide: implications to sensitization of gram-negative bac-
teria. J Med Chem 2000; 43: 3085 92.
27 Li XZ, Zhang L, Poole K. Interplay between the MexA-MexB-OprM multi-
drug efflux system and the outer membrane barrier in the multiple
8 of 8
antibiotic resistance of Pseudomonas aeruginosa. J Antimicrob
Chemother 2000; 45: 4336.
28 Wareham DW, Gordon NC, Hornsey M. In vitro activity of teicoplanin
combined with colistin versus multidrug-resistant strains of Acinetobacter
baumannii. J Antimicrob Chemother 2011; 66: 104751.
29 Lee HJ, Bergen PJ, Bulitta JB et al. Synergistic activity of colistin and
rifampin combination against multidrug-resistant Acinetobacter bauman-
nii in an in vitro pharmacokinetic/pharmacodynamic model. Antimicrob
Agents Chemother 2013; 57: 3738 45.
30 Timurkaynak F, Can F, Azap OK et al. In vitro activities of non-traditional
antimicrobials alone or in combination against multidrug-resistant strains
of Pseudomonas aeruginosa and Acinetobacter baumannii isolated from
intensive care units. Int J Antimicrob Agents 2006; 27: 224 8.
31 Hornsey M, Longshaw C, Phee L et al. In vitro activity of telavancin in
combination with colistin versus Gram-negative bacterial pathogens.
Antimicrob Agents Chemother 2012; 56: 3080 5.
Downloaded from http://jac.oxfordjournals.org/ at Aston University on August 22, 2014
32 Smith PA, Roberts TC, Romesberg FE. Broad-spectrum antibiotic activity
of the arylomycin natural products is masked by natural target mutations.
Chem Biol 2010; 17: 1223 31.
33 Luke I, Handford JI, Palmer T et al. Proteolytic processing of Escherichia
coli twin-arginine signal peptides by LepB. Arch Microbiol 2009; 191: 91925.
34 Jacobs MA, Alwood A, Thaipisuttikul I et al. Comprehensive transposon
mutant library of Pseudomonas aeruginosa. Proc Natl Acad Sci USA 2003;
100: 14339 44.
35 Whiley RA, Sheikh NP, Mushtaq N et al. Differential potentiation of the
virulence of the Pseudomonas aeruginosa cystic fibrosis liverpool epidemic
strain by oral commensal streptococci. J Infect Dis 2014; 209: 769 80.
36 Hornsey M, Phee L, Stubbings W et al. In vitro activity of the novel
monosulfactam BAL30072 alone and in combination with meropenem
versus a diverse collection of important Gram-negative pathogens. Int J
Antimicrob Agents 2013; 42: 3436.
37 Gordon NC, Png K, Wareham DW. Potent synergy and sustained
bactericidal activity of a vancomycin-colistin combination versus multi-
drug-resistant strains of Acinetobacter baumannii. Antimicrob Agents
Chemother 2010; 54: 5316 22.
38 Hornsey M, Ellington MJ, Doumith M et al. AdeABC-mediated efflux
and tigecycline MICs for epidemic clones of Acinetobacter baumannii.
J Antimicrob Chemother 2010; 65: 1589 93.
39 Betts JW, Phee LM, Woodford N et al. Activity of colistin in combination
with tigecycline or rifampicin against multidrug-resistant Stenotrophomonas
maltophilia. Eur J Clin Microbiol Infect Dis 2014; doi:10.1007/s10096-014-
2101-3.
40 Le J, Castanheira M, Burgess DS et al. Clonal dissemination of Klebsiella
pneumoniae carbapenemase KPC-3 in Long Beach, California. J Clin
Microbiol 2010; 48: 6235.
41 Lammertyn E, Van Mellaert L, Meyen E et al. Molecular and functional
characterization of type I signal peptidase from Legionella pneumophila.
Microbiology 2004; 150: 1475 83.