β-Lactamase triggered visual detection of bacteria using cephalosporin functionalized biomaterials

Article abstract of DOI:10.1039/d0cc04088f

We report the conjugation of a chromogenic cephalosporin β-lactamase (βL) substrate to polymers and integration into biomaterials for facile, visual βL detection. Identification of these bacterial enzymes, which are a leading cause of antibiotic resistance, is critical in the treatment of infectious diseases. The βL substrate polymer conjugate undergoes a clear to deep yellow color change upon incubation with common pathogenic Gram-positive and Gram-negative bacteria species. We have demonstrated the feasibility of formulating hydrogels with the βL substrate covalently tethered to a poly(ethylene glycol) (PEG) polymer matrix, exhibiting a visible color change in the presence of βLs. This approach has the potential to be used in diagnostic biomaterials for point-of-care detection of βL-producing bacteria, helping combat the spread of drug resistant microbes.

Full text of DOI:10.1039/d0cc04088f

ChemComm
View Article Online
View Journal
COMMUNICATION
b-Lactamase triggered visual detection of bacteria
using cephalosporin functionalized biomaterials†
Cite this:DOI: 10.1039/d0cc04088f
a
Dahlia Alkekhia, ‡a Hannah Safford,‡a Shashank Shukla,
Russel Hopsonb and
a
Received 11th June 2020,
Accepted 4th August 2020
Anita Shukla
*
DOI: 10.1039/d0cc04088f
rsc.li/chemcomm
We report the conjugation of a chromogenic cephalosporin suitable for use without specialized instrumentation and
b-lactamase (bL) substrate to polymers and integration into bioma- trained personnel.5 Recognizing the limitations of these meth-
terials for facile, visual bL detection. Identification of these bacterial ods, recent efforts in point-of-care infection diagnostics exhi-
enzymes, which are a leading cause of antibiotic resistance, is biting visual detection of bacteria, have led to the development
critical in the treatment of infectious diseases. The bL substrate of paper strips that change color due to electrostatic interac-
polymer conjugate undergoes a clear to deep yellow color change tions with bacteria6 and wound dressings that change color in
upon incubation with common pathogenic Gram-positive and the presence of lipase-producing bacteria.7 However, the sti-
Gram-negative bacteria species. We have demonstrated the feasi- muli causing color change in these materials are generally non-
bility of formulating hydrogels with the bL substrate covalently specific to bacteria, limiting their use. In this work, we focused
tethered to a poly(ethylene glycol) (PEG) polymer matrix, exhibiting instead on visually detecting bLs, which are only produced by
a visible color change in the presence of bLs. This approach has the bacteria and in particular, antibiotic resistant bacteria. Non-
potential to be used in diagnostic biomaterials for point-of-care colorimetric bL hydrolysis indicators include luminescent ruthe-
detection of bL-producing bacteria, helping combat the spread of nium (i),8 fluorescent,9,10 chemiluminescent,11 and fluorescence
drug resistant microbes.
resonance energy transfer12 probes, again requiring the use of
external equipment. Alternatively, chromogenic bL substrates
Infectious diseases are among the leading causes of death have been utilized for a visual indication in solution, on agar,
worldwide. Treatment of infections is becoming increasingly or on discs, including commercially available CENTA,13,14
difficult due to the spread of antibiotic resistant microbes. PADAC,15,16 and nitrocefin.4
According to a 2019 report by the United States Centers for
Our goal was to synthesize a chromogenic bL substrate
Disease Control and Prevention, resistant microbes cause over that can be readily conjugated to a range of polymers, providing
2.8 million illnesses and 35 000 deaths annually in the US building blocks for infection diagnostic biomaterials (Scheme 1).
alone.1 Production of b-lactamases (bLs) is one of the primary These biomaterials could be used in several different scenarios
mechanisms of resistance to b-lactam antibiotics, especially (e.g., contaminated surgical equipment and hospital surfaces,
for Gram-negative bacteria.2 bLs are enzymes that cleave the which are linked to increased risk of healthcare-associated
b-lactam ring and inactivate b-lactam antibiotics, which make up infections,17 or superficial bacterial infections), allowing for
65% of all prescribed intravenous antibiotics in the US,3 including real-time in situ monitoring of bL-producing bacteria via a color
penicillins and cephalosporins.4 The facile detection of bL- change. Inspired by the structure of the chromogenic bL
producing bacteria has the potential to improve prevention and substrate, CENTA,13,14 we synthesized a bL substrate by mod-
treatment and reduce the spread of antibiotic resistant infections. ifying a cephalosporin derivative, 7-amino-3-chloromethyl-3-
Conventional bacteria detection methods, including culture cephem-4-carboxylic acid p-methoxybenzyl ester (ACLE), with
and DNA amplification, although highly sensitive are often not a chromophore, 4-nitrobenzenethiol (NBT), at the 30-position,
yielding compound 1 (Scheme 1 and Section S3, Fig. S1, ESI†).
a School of Engineering, Center for Biomedical Engineering, Institute for Molecular
and Nanoscale Innovation, Brown University, Providence, RI 02912, USA.
E-mail: anita_shukla@brown.edu
b Department of Chemistry, Brown University, Providence, RI 02912, USA
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
d0cc04088f
In this substrate, the primary amine at the 70-position of
the cephem nucleus is readily modifiable and can be used to
conjugate the substrate to compounds bearing any one of an
abundance of amine reactive groups. Here, we modified the
bL substrate with a maleimide moiety, allowing facile conjuga-
tion to a range of molecules and macromolecules via modular
‡ These authors contributed equally to this work.
Chem. Commun.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Communication
ChemComm
Scheme 1 Synthetic scheme for bL substrate and substrate–polymer conjugates and illustration of bL detecting hydrogel. ACLE: 7-amino-3-chloromethyl-3-
cephem-4-carboxylic acid p-methoxybenzyl ester; NBT: 4-nitrobenzenethiol; TEA: triethylamine; NMM: 4-methylmorpholine; DCM: dichloromethane; TFA:
trifluoroacetic acid; bLs: b-lactamases; HATU: 1-[Bis(dimethylamino)methylene]-1H-1,2,3-triazolo[4,5-b]pyridinium 3-oxid hexafluorophosphate; DIPEA: N,N-
diisopropylethylamine; DMSO: dimethyl sulfoxide; DMF: dimethylformamide; PB: sodium phosphate buffer; PBS: phosphate buffered saline; mPEG-SH:
methoxy-poly(ethylene glycol)-thiol; 4-arm-PEG-SH: thiol modified 4-arm-PEG; mal-PEG-mal: bis-maleimide-PEG.
click-chemistry under mild conditions.18,19 The p-methoxybenzyl bL-BC, bL-PA, and bL-EC, respectively, for the concentration of
protecting group prevents modification of the carboxyl on the 2 tested (Fig. S5, ESI†); these values are similar to the lowest
cephem core, which is a key enzyme-substrate recognition site enzyme concentration leading to a color change discernible by
for many bLs,20,21 during chemical reactions and polymer eye (Fig. 1a). Utilizing the Michaelis–Menten model, we com-
functionalization. Covalent conjugation of the bL substrate to pared the enzyme kinetics of substrate 2 to CENTA and nitro-
biomaterial forming polymers enables the substrate to remain cefin (Fig. S6, ESI†), summarized in Table 1. The kinetic
tethered to the biomaterial backbone preventing its premature parameters obtained for CENTA and nitrocefin with bL-BC were
release (via diffusion, for example).
within a similar range of values previously reported with bLs
Prior to modification with polymers, 1 was deprotected to derived from B. cereus,9,13 as expected. The kcat/KM values for 2
yield 2 (Scheme 1 and Section S4, Fig. S2, ESI†) in order to were slightly below those of CENTA and nitrocefin for bL-BC
characterize the response of the substrate to bLs. 2 was tested and bL-PA. With bL-EC, an extremely high enzyme concen-
with bLs produced by Bacillus cereus (bL-BC), Pseudomonas tration was required for substrate hydrolysis yielding a negli-
aeruginosa (bL-PA), and Enterobacter cloacae (bL-EC). bL-BC gible kcat, indicating low turnover number. The differences
was selected due to its previous thorough characterization,2,22 observed in these kinetic parameters between the bLs tested
while bL-PA and bL-EC were examined due to the clinical are likely related to variation in substrate specificity for differ-
relevance of the pathogenic, Gram-negative bacteria that produce ent bLs.21 As a preliminary investigation of bL specificity,
them.2,23,24 Substrate 2 was found to exhibit a bL concentration- substrate 2, nitrocefin, and CENTA were also incubated with
dependent color change from clear to yellow when mixed with bLs bacteria-produced collagenases; no color change was observed
from any of the three bacterial species (Fig. 1a).
(Fig. 1d).
To examine potential antibacterial effects of the substrate,
Ultraviolet-visible (UV-vis) spectroscopy was used to monitor
the kinetics of substrate 2 cleavage. As the b-lactam ring is which could limit its utility in bacteria detection, the minimum
hydrolyzed by the bL, its characteristic absorbance at 260 nm inhibitory concentration (MIC) of 2 was measured against
decreases, and the chromophore is expelled (Scheme 1; 20), different bacterial species (Table S1, ESI†). Compound 2 exhib-
causing a clear red shift in the maximum absorbance (lmax
)
ited antibacterial effects (MIC = 8 mg mLꢀ1) against Staphylo-
from 345 to 410 nm (Fig. S3, ESI†). As expected, the absorbance coccus aureus 25923, which is a non-bL-producing strain,25,26
at lmax is enzyme concentration-dependent (Fig. 1b and Fig. S4, and S. aureus 29213, which produces low levels of bLs,26,27
ESI†). Fig. 1c shows the absorbance at 410 nm versus bL but showed no antibacterial effect at concentrations up to
concentration. The limit of detection was determined to be 128 mg mLꢀ1 against other bacteria investigated (methicillin-
4.6, 0.66, and 0.11 units (U) mLꢀ1 (B0.05, 0.02, 12.86 mM) for resistant S. aureus (MRSA), P. aeruginosa, E. cloacae, and B. cereus).
Chem. Commun.
This journal is © The Royal Society of Chemistry 2020

View Article Online
ChemComm
Communication
nanoparticles, and coatings given its relative biocompatiblity and
tunability.28,29 We first functionalized 1 with a maleimide
utilizing 3-maleimidopropionic acid, yielding 3 (Scheme 1
and Section S9, Fig. S7–S13, ESI†). 3 was then conjugated to a
linear, short chain methoxy-PEG-thiol (mPEG-SH) (1.7 kDa) via
Michael-type addition forming conjugate 4 (Scheme 1 and
1
Section S10, ESI†). H-NMR showed the disappearance of the
maleimide protons at 7.00 ppm and the appearance of a
new proton signal at 4.02 ppm in 4, indicating formation of
the thiol-maleimide adduct (Fig. S14, ESI†). Additional protons
on the adduct, which would likely appear in the region of
2.4–3.2 ppm,30 were obscured by the PEG repeat unit protons.
Matrix-assisted laser desorption ionization-time of flight
(MALDI-TOF) mass spectrometry also confirmed PEG conjuga-
tion, demonstrating an increase in the molecular weight of 4
compared to unmodified mPEG-SH (Fig. S15, ESI†). 4 and
unmodified mPEG-SH were also compared using size exclusion
chromatography (SEC), where mPEG-SH and 4 had comparable
elution times (measured via a refractive index detector; data not
shown) and only 4 had an absorbance signal at 345 nm (Fig. S16,
ESI†), indicating successful conjugation. In the presence of bL-BC,
4 changed color from clear to a bright yellow, accompanied by a
decrease in absorbance at 260 nm and the expected red shift in
l
max (Fig. S17, ESI†), confirming that 4 remains responsive to bLs.
Fig. 1 (a) Substrate 2 (544 mM) incubated with bL-BC, bL-PA, or bL-EC
(U mLꢀ1) for 45 minutes at 37 1C. (b) Absorbance spectra of 2 incubated
with bL-BC (U mLꢀ1). (c) Absorbance at 410 nm versus bL concentration
for solutions in (a). (d) Images of CENTA, nitrocefin and 2 (all at 544 mM)
incubated in 1ꢁ PBS with 1 mM CaCl2 alone or with 1 mg mLꢀ1 collage-
nases or with 30 U mLꢀ1 of bL-PA in PBS for 0.5, 2, or 5 hours at 37 1C.
Note, 1 U is defined as hydrolyzing 1.0 mmole of benzylpenicillin per min at
pH 7.0 at 25 1C.
A decrease in the 345 nm absorbance signal of 4 after incubation
with bL was also observed via SEC (Fig. S16, ESI†).
Upon conjugation to the hydrophilic PEG, aqueous solubi-
lity of the substrate was greatly improved facilitating detection
of bL-producing bacteria in in vitro cultures at higher substrate
concentrations. We incubated conjugate 4 at increasing con-
centrations (up to 2028 mM) with bL-producing strains of
P. aeruginosa,12 E. cloacae,12,31 and B. cereus,8 as well as a
These results are comparable to CENTA, nitrocefin, and PADAC, non-pathogenic, non-bL-producing E. coli (DH5-a).8,31,32 As
which have been shown to lack antibacterial activity against most shown in Fig. 2a, the conjugate changed color when incubated
Gram-negative species but do exhibit activity against S. aureus and with bL-producing bacteria, even at the lowest concentrations
some Streptococci.14,15
of 4 tested (B127–254 mM), while no color change was observed
After characterizing substrate 2 and its response to bLs, for E. coli DH5-a. For B. cereus, at the highest concentration of 4
we investigated its responsiveness following conjugation to (45.8 times the highest concentration of 2 investigated for
poly(ethylene glycol) (PEG). PEG was chosen as our model antibacterial effects), no visible color change was achieved prior
polymer due to its extensive use in biomaterials such as hydrogels, to growth inhibition (suggested by solution clearing).
Table 1 Kinetic parameters of substrate 2, CENTA, and nitrocefin hydrolysis by bLs
Enzyme
Substrate
[E]a (mM)
KM (mM)
kcat (sꢀ1
)
kcat/KM (mMꢀ1 sꢀ1
)
bL from B. cereus
Substrate 2
CENTAb
0.32
0.32
0.32
866.80 ꢂ 141.40
345.30 ꢂ 36.08
450.70 ꢂ 69.55
4.99 ꢂ 0.57
16.50 ꢂ 0.88
49.67 ꢂ 4.35
0.0058 ꢂ 0.0011
0.048 ꢂ 0.0056
0.11 ꢂ 0.012
Nitrocefinc
bL from P. aeruginosa
Substrate 2
CENTA
Nitrocefin
1.19
0.47
0.47
58.01 ꢂ 9.54
12.63 ꢂ 6.84
178.80 ꢂ 32.71
0.66 ꢂ 0.02
0.42 ꢂ 0.02
5.14 ꢂ 0.37
0.011 ꢂ 0.002
0.033 ꢂ 0.018
0.029 ꢂ 0.0056
d
d
bL from E. cloacae
Substrate 2
CENTA
Nitrocefin
116.9
1.17
0.58
864.6 ꢂ 123.4
68.88 ꢂ 11.83
315.20 ꢂ 26.75
—
—
1.48 ꢂ 0.062
0.021 ꢂ 0.0038
9.04 ꢂ 0.38
0.029 ꢂ 0.0027
Enzyme concentration. Previously reported: (1) [E] = 0.014 mM, kM = 330 mM, kcat = 50 sꢀ1, kcat/kM = 0.15 mMꢀ1 sꢀ1 13 (2) [E] = 0.001 mM, kM = 135.9
;
a
b
ꢂ 16.4 mM, kcat = 8.9 sꢀ1, kcat/kM = 0.06 mMꢀ1 sꢀ1 9 c
.
Previously reported: (1) [E] = 0.014 mM, kM = 70 mM, kcat = 45 sꢀ1, kcat/kM = 0.64 mMꢀ1 sꢀ1 13 (2) [E]
;
= 0.001 mM, kM = 8.1 ꢂ 1.0 mM, kcat = 14.9 sꢀ1, kcat/kM = 1.8 mMꢀ1 sꢀ1 9 d
.
—: negligible.
Chem. Commun.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Communication
ChemComm
the National Institute of General Medical Sciences of the National
Institutes of Health [grant number P20GM103430].
Conflicts of interest
There are no conflicts to declare.
References
1 Centers for Disease Control and Prevention, Antibiotic resistance
threats in the United States, 2019.
2 K. Bush, Antimicrob. Agents Chemother., 2018, 62, e01076-18.
3 K. Bush and P. A. Bradford, Cold Spring Harbor Perspect. Med., 2016,
6, a025247.
4 D. M. Livermore and D. F. J. Brown, J. Antimicrob. Chemother., 2001,
48, 59–64.
Fig. 2 (a) Conjugate 4 incubated with bL-producing (P. aeruginosa,
E. cloacae, B. cereus) or non-bL-producing bacteria (E. coli), or in media
for 18–24 hours at 37 1C. (b) Representative images of 20% (w/v) b-lactam-
PEG and non-responsive PEG hydrogels incubated in 1ꢁ PBS with and
without 400 U mLꢀ1 of bL-BC for o10 minutes.
´
5 L. Varadi, J. L. Luo, D. E. Hibbs, J. D. Perry, R. J. Anderson, S. Orenga
and P. W. Groundwater, Chem. Soc. Rev., 2017, 46, 4818–4832.
6 B. Creran, X. Li, B. Duncan, C. S. Kim, D. F. Moyano and
V. M. Rotello, ACS Appl. Mater. Interfaces, 2014, 6, 19525–19530.
7 H. Singh, W. Li, M. R. Kazemian, R. Yang, C. Yang, S. Logsetty and
S. Liu, ACS Appl. Bio Mater., 2019, 2, 2028–2036.
8 Q. Shao and B. Xing, Chem. Commun., 2012, 48, 1739–1741.
9 R. Cain, C. J. Schofield, J. Brem, C. W. G. Fishwick, S. S. van Berkel,
R. J. Owens, J. Spencer, A. M. Rydzik, R. Salimraj and A. Verma,
J. Med. Chem., 2013, 56, 6945–6953.
Having demonstrated successful response to bLs in bacterial
suspensions using PEG modified 2, we investigated the for-
mation of a macroscale bL-responsive chromogenic biomaterial
using 3. As shown in Scheme 1, 3 was reacted with thiol
functionalized 4-arm-PEG to decorate on average one of the
polymer arms with the bL substrate (5). Subsequently, a bis-
maleimide-PEG cross-linker was used to cross-link the hydrogel 10 H. Xie, H. Hassounah, J. Rao, J. D. Cirillo, Y. Cheng, E. A. Graviss,
Y. Kong and P. Sule, Angew. Chem., Int. Ed., 2014, 53, 9360–9364.
network. Control non-b-lactam hydrogels were formed similarly
11 S. Maity, X. Wang, S. Das, M. He, L. W. Riley and N. Murthy, Chem.
omitting the addition of 3. An increased swelling was observed
Commun., 2020, 56, 3516–3519.
in the b-lactam-PEG hydrogels compared to control PEG hydrogels 12 H. L. Chan, L. Lyu, J. Aw, W. Zhang, J. Li, H.-H. H. Yang, H. Hayashi,
S. Chiba and B. Xing, ACS Chem. Biol., 2018, 13, 1890–1896.
13 C. Bebrone, C. Moali, F. Mahy, S. Rival, J. D. Docquier, G. M. Rossolini,
(Fig. 2b) likely resulting from increased network defects due to
incomplete cross-linking at sites where 3 was conjugated.33
J. Fastrez, R. F. Pratt, J.-M. Frere and M. Galleni, Antimicrob. Agents
Finally, when both hydrogels were incubated with bL-BC, only
b-lactam containing hydrogels showed a rapid and distinct color
change from clear to yellow (Fig. 2b), indicating that these
materials successfully maintain bL responsiveness.
In summary, we have demonstrated a modular approach to
covalently attach a chromogenic bL substrate to polymers
which exhibit a color change specifically in the presence of
bLs. To the best of our knowledge, this is the first report of
polymer-bL substrate conjugates for direct and facile colori-
metric detection of bL-producing bacteria involved in several
difficult to treat infections. While we have demonstrated the
feasibility of conjugating this substrate to PEG and PEG deri-
Chemother., 2001, 45, 1868–1871.
14 R. N. Jones, H. W. Wilson, W. J. Novick, A. L. Barry and C. Thornsberry,
J. Clin. Microbiol., 1982, 15, 954–958.
15 R. N. Jones, H. W. Wilson and W. J. Novick, J. Clin. Microbiol., 1982,
15, 677–683.
16 S. Kobayashi, S. Arai, S. Hayashi and T. Sakaguchi, Antimicrob.
Agents Chemother., 1988, 32, 1040–1045.
17 R. F. Chemaly, S. Simmons, C. Dale, S. S. Ghantoji, M. Rodriguez, J. Gubb,
J. Stachowiak and M. Stibich, Ther. Adv. Infect. Dis., 2014, 2, 79–90.
18 M. P. Lutolf and J. A. Hubbell, Biomacromolecules, 2003, 4, 713–722.
19 A. Gandini, Prog. Polym. Sci., 2013, 38, 1–29.
´
20 F. Morandi, E. Caselli, S. Morandi, P. J. Focia, J. Blazquez,
B. K. Shoichet and F. Prati, J. Am. Chem. Soc., 2003, 125, 685–695.
21 K. Bush and P. A. Bradford, Nat. Rev. Microbiol., 2019, 17, 295–306.
22 J. Imsande, F. D. Gillin, R. J. Tanis and A. G. Atherly, J. Biol. Chem.,
1970, 245, 2205–2212.
vatives, this versatile substrate has the potential to be readily 23 K. M. Kazmierczak, S. Rabine, M. Hackel, R. E. McLaughlin, D. J.
Biedenbach, S. K. Bouchillon, D. F. Sahm and P. A. Bradford,
Antimicrob. Agents Chemother., 2016, 60, 1067–1078.
modified by a broad range of additional polymers. These
conjugates could potentially be used for the detection of
`
24 A. Davin-Regli and J.-M. Pages, Front. Microbiol., 2015, 6, 392.
bacteria in a range of environments (e.g., medical devices, 25 S. Khan, U. W. Sallum, X. Zheng, G. J. Nau and T. Hasan, BMC
Microbiol., 2014, 14, 84.
26 S. Lee, K. T. Kwon, H.-I. Kim, H. H. Chang, J.-M. Lee, P. G. Choe,
household surfaces) and may ultimately be incorporated into
multifunctional biomaterials for the enhanced prevention,
W. B. Park, N. J. Kim, M.-D. Oh, D. Y. Song and S.-W. Kim, Microb.
detection, and treatment of bacterial infections.
This work was supported by the Office of Naval Research
[grant number N00014-17-1-2120]. The authors thank Dr Tun-Li
Drug Resist., 2014, 20, 568–574.
27 D. S. Kernodle, P. A. McGraw, C. W. Stratton and A. B. Kaiser,
Antimicrob. Agents Chemother., 1990, 34, 420–425.
28 J. Zhu, Biomaterials, 2010, 31, 4639–4656.
Shen at the Mass Spectrometry Facility at Brown University for 29 I. W. Hamley, Biomacromolecules, 2014, 15, 1543–1559.
30 B. H. Northrop, S. H. Frayne and U. Choudhary, Polym. Chem., 2015,
his technical support, Dr Vicki Colvin at Brown University for
the use of her PerkinElmer Lambda 950 spectrophotometer,
6, 3415–3430.
31 J. Aw, F. Widjaja, Y. Ding, J. Mu, Y. Liang and B. Xing, Chem. Commun.,
and Dr Chao Yu at Brown University for support with illustrations.
Research was made possible by the use of the MALDI-TOF MS
available through the Rhode Island Institutional Development
Award (IDeA) Network of Biomedical Research Excellence from
2017, 53, 3330–3333.
`
32 L. Boschi, P. S. Mercuri, M. L. Riccio, G. Amicosante, M. Galleni, J. M. Frere
and G. M. Rossolini, Antimicrob. Agents Chemother., 2000, 44, 1538–1543.
33 J. Kim, Y. P. Kong, S. M. Niedzielski, R. K. Singh, A. J. Putnam and
A. Shikanov, Soft Matter, 2016, 12, 2076–2085.
Chem. Commun.
This journal is © The Royal Society of Chemistry 2020