Development of a high-throughput screen and its use in the discovery of Streptococcus pneumoniae immunoglobulin A1 protease inhibitors

Article abstract of DOI:10.1021/ja404180x

Streptococcus pneumoniae relies on a number of virulence factors, including immunoglobulin A1 protease (IgA1P), a Zn²⁺ metalloprotease produced on the extracellular surface of the bacteria, to promote pathogenic colonization. IgA1P exhibits a unique function, in that it catalyzes the proteolysis of human IgA1 at its hinge region to leave the bacterial cell surface masked by IgA1 Fab, enabling the bacteria to evade the host's immune system and adhere to host epithelial cells to promote colonization. Thus, S. pneumoniae IgA1P has emerged as a promising antibacterial target; however, the lack of an appropriate screening assay has limited the investigation of this metalloprotease virulence factor. Relying on electrostatics-mediated AuNP aggregation, we have designed a promising high-throughput colorimetric assay for IgA1P. By using this assay, we have uncovered inhibitors of the enzyme that should be useful in deciphering its role in pneumococcal colonization and virulence.

Full text of DOI:10.1021/ja404180x

Communication
pubs.acs.org/JACS
Development of a High-Throughput Screen and Its Use in the
Discovery of Streptococcus pneumoniae Immunoglobulin A1
Protease Inhibitors
Amanda L. Garner,† Jessica L. Fullagar,‡ Joshua A. Day,‡ Seth M. Cohen,‡ and Kim D. Janda*,†
†Departments of Chemistry and Immunology and Microbial Science, The Skaggs Institute for Chemical Biology, and The Worm
Institute for Research and Medicine, The Scripps Research Institute, and ‡Department of Chemistry and Biochemistry, University of
California, San Diego, La Jolla, California 92037, United States
S
* Supporting Information
ABSTRACT: Streptococcus pneumoniae relies on a number
of virulence factors, including immunoglobulin A1
protease (IgA1P), a Zn2+ metalloprotease produced on
the extracellular surface of the bacteria, to promote
pathogenic colonization. IgA1P exhibits a unique function,
in that it catalyzes the proteolysis of human IgA1 at its
hinge region to leave the bacterial cell surface masked by
IgA1 Fab, enabling the bacteria to evade the host’s
immune system and adhere to host epithelial cells to
Figure 1. IgA1 cleavage by IgA1Ps. (a) IgA1P-catalyzed degradation of
promote colonization. Thus, S. pneumoniae IgA1P has
IgA1 (black = heavy chain, gray = light chain). (b) Sequence of amino
emerged as a promising antibacterial target; however, the
acids in the hinge region of the α chain of IgA1 Fab and proposed site
of cleavage by S. pneumoniae IgA1P.
lack of an appropriate screening assay has limited the
investigation of this metalloprotease virulence factor.
Relying on electrostatics-mediated AuNP aggregation, we
Interestingly, in the case of S. pneumoniae IgA1P, upon
cleavage of IgA1, the bacterium retains the Fab fragments on
its surface to mask its capsular polysaccharide, the immuno-
dominant surface antigen on pneumococci.9 This event
promotes, rather than inhibits, adherence to epithelial cells
via electrostatic interactions between the positively charged Fab
molecules coating the bacteria and negatively charged plasma
membrane of the host’s cells, thereby enhancing infection.9 In
addition to promoting adherence, S. pneumoniae IgA1P has also
been linked to pneumococcal virulence in murine models of
pneumonia10,11 and sepsis12 via mutation analyses; yet, there
exists a disconnect with the targets of this protease between
rodents and humans, as currently the enzyme is only known to
cleave human5 and primate13 IgA1. Thus, it may be likely that
IgA1P engages additional host substrates to promote
virulence,10 but this remains to be determined.
A number of factors have precluded the investigation of S.
pneumoniae IgA1P as an antibacterial target. In addition to the
lack of an appropriate animal model,5 no high-throughput-
amenable screening assays have been reported for IgA1P
activity. ELISA-based tests have been described;14,15 however,
they have never been utilized for small molecule inhibitor
screening. Moreover, these methods are not applicable to such
efforts, as they require several expensive antibody reagents and
a separate reaction mixture transfer step.14,15 With respect to
examining inhibition of IgA1P activity, only Western blot-
based16 and autoradiographic17 assays have been reported. As a
have designed a promising high-throughput colorimetric
assay for IgA1P. By using this assay, we have uncovered
inhibitors of the enzyme that should be useful in
deciphering its role in pneumococcal colonization and
virulence.
Streptococcus pneumoniae is a commensal colonizer of the upper
respiratory tract of humans, where it inhabits the mucosal
surfaces lining the nasopharynx; however, the bacteria can
spread from the upper respiratory tract to the sterile spaces of
the lower respiratory tract, ear, or bloodstream to cause
pneumococcal diseases such as pneumonia, otitis media, sepsis,
and meningitis.1,2 To promote pathogenic colonization, S.
pneumoniae relies on a combination of virulence factors and
immune modulators; in particular, the bacteria use up to four
surface-presented, extracellular zinc metalloproteases to evade
the host immune system and promote virulence.1,2 Of these
metalloproteases, S. pneumoniae immunoglobulin A1 protease
(IgA1P) has emerged as the most promising antibacterial
target.3,4
IgA1 proteases are a class of extracellular bacterial enzymes
that act as post-proline endopeptidases and specifically cleave
human IgA1 (Figure 1),5 the predominant form of Ig in
mucosal tissues and secretions.6 Cleavage occurs in the unique
hinge region of the α chain of IgA1 to yield intact Fabα and Fcα
fragments,6−8 leaving IgA1P-producing organisms coated with
antibody fragments lacking Fc domains, thus enabling the
bacteria to evade recognition by Fc receptors and complement
to prevent production of an inflammatory response.2
Received: April 26, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja404180x | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
result, no S. pneumoniae IgA1P-specific inhibitors have been
disclosed.16 Thus, in order to identify IgA1P-specific inhibitors
with the capacity to elucidate the role of S. pneumoniae IgA1P
in pneumococcal colonization and virulence, we sought to
design a user-friendly, potentially high-throughput screening
assay for IgA1P.
To test the applicability of the designed AuNP assay for S.
pneumoniae IgA1P, D39 strain S. pneumoniae was grown to
stationary phase (OD600 0.4−0.5) in Todd−Hewitt broth
supplemented with 0.5% yeast extract.3 Bacterial supernatant
was then obtained via centrifugation and used as the source of
S. pneumoniae IgA1P. Although production of recombinant S.
pneumoniae IgA1P has been reported, the activity and stability
of this recombinant enzyme are poor.21 For the assay, diluted
IgA1P-containing S. pneumoniae supernatant (2× in 50 mM
Tris-HCl, pH 7.0)3 was incubated with human secretory IgA
(0.15 mg/mL) at 37 °C for 18 h.22 After a freeze−thaw cycle to
disrupt protein−antibody interactions, AuNPs (20 nm) were
added, and absorbance signals A650 and A520 were then
measured. As Figure 2b shows, our assay design was successful,
and under these conditions a 2.3-fold raw ratiometric
absorbance enhancement was observed over samples not
treated with IgA1. These positive results were further verified
via specificity and knockout studies, and in the presence of
IgA2, which is not cleaved by IgA1P,5 and when using
supernatant from an IgA1P-knockout S. pneumoniae strain
(P1285),22 no ratiometric absorbance signal was observed over
background (Figure 2b). Western blot analysis was performed
as previously described16,22 for each of these samples to
monitor IgA1 structural integrity and confirm cleavage status.
Positive correlation was observed between our results using the
AuNP assay and Western blot findings, demonstrating the
accuracy of our assay for monitoring secretory IgA cleavage by
IgA1P (see Figure S4 in the Supporting Information (SI)).
Given these successful results, we next characterized the
IgA1P-catalyzed IgA1 cleavage reaction using our AuNP assay.
We first examined the time dependence of IgA1 cleavage by
IgA1P. As Figure 3a shows, the proteolysis of IgA1 was linear
As inspiration for our assay design, we chose to take
advantage of the fact that, following IgA1P-mediated cleavage,
positively charged Fab fragments are produced.9 Thus, we
reasoned that if a pro-colorimetric, negatively charged reagent
could selectively bind to the positively charged Fab molecules
to produce a measurable signal, then a simple, colorimetric
high-throughput assay for IgA1P activity would be established.
Fortunately, negatively charged gold nanoparticles (AuNPs)
have previously been employed as a means to interrogate
protein kinase and phosphatase activity in vitro and in cell lysate
using a similar electrostatics-based principle.18,19 Briefly, the
charge differential between a phosphorylated and non-
phosphorylated arginine- or lysine-containing, kinase- or
phosphatase-recognized peptide substrate was monitored via
aggregation of citrate-modified AuNPs,18,19 where dispersed
AuNPs exhibit a deep red color while aggregated AuNPs are
blue, based on a shift in the wavelength of the surface plasmon
resonance due to increased particle size.20 Taking a similar
approach, in the absence of IgA1P, AuNPs would be dispersed
with no noticeable color change; however, in the presence of
IgA1P, the enzyme would cleave human IgA1 to yield positively
charged Fab fragments, which would electrostatically interact
with negatively charged, citrate-adsorbed AuNPs to induce
AuNP aggregation and a color change from red to blue, thereby
establishing a colorimetric screening assay for IgA1P activity
(Figure 2a).
Figure 3. Characterization of AuNP assay for IgA1P-mediated IgA1
cleavage: A650/520 = normalized ratiometric absorbance signal with
respect to the negative control (wt−). (a) Time dependence. (b)
Dependence on the concentration of secretory IgA. (c) Dependence
on the volume of D39 strain S. pneumoniae supernatant containing
IgA1P. (d) Z′ factor.
Figure 2. AuNP-based colorimetric IgA1P assay. (a) Assay design
(black = heavy chain, gray = light chain). (b) Proof-of-concept data: wt
+ = D39 strain S. pneumoniae supernatant treated with secretory IgA;
wt− = D39 strain S. pneumoniae supernatant without secretory IgA;
wt/IgA2 = D39 strain S. pneumoniae supernatant treated with IgA2;
KO/IgA1 = P1285 strain S. pneumoniae supernatant treated with
secretory IgA; A650/520 = raw ratiometric absorbance signal.
with time for ∼5 h. The assay was also linear with respect to the
concentration of secretory IgA (up to ∼0.05 mg/mL) added to
the wells (Figure 3b). Moreover, as Figure 3c shows, the assay
exhibited saturation kinetics with respect to the volume of S.
pneumoniae supernatant. In addition to reaction character-
ization, we also assessed the suitability of this AuNP assay for
B
dx.doi.org/10.1021/ja404180x | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
high-throughput screening. The signal-to-noise ratio (S/N) was
22.7, the signal-to-background ratio (S/B) was 3.8, and the Z′
factor23 was 0.60, indicating the potential of our AuNP assay for
high-throughput screening, as assays with Z′ values between 0.5
and 1.0 are typically regarded as excellent assays (Figure 3d).
Using our AuNP assay for IgA1P activity, we initiated a
screening campaign to identify small molecule inhibitors of S.
pneumoniae IgA1P. Libraries of substituted pyridines and
naphthalenes,24,25 natural products (Enzo Natural Product
Library), and metal-chelating fragments (CFL1.1)26,27 were
analyzed. These efforts resulted in the identification of three
natural products, remerine (1), norharmane (2), and berberine
(3) (Figure 4a; see Figure S1 for IC50 curves), and one family
Table 1. Structures and IC50 Values for Select 3,4-HOPTO
Analogue Hits
Figure 4. Identified inhibitors of S. pneumoniae IgA1P. (a) Natural
product hits. (b) 3,4-HOPTO-based hits.
of chelator fragments, 3-hydroxy-1-alkyl-2-methylpyridine-
4(1H)-thiones (3,4-HOPTO, compounds 4−6, Figure 4b; see
Figures S2 and S3 for the activity of other CFL1.1 members),
exhibiting dose-dependent inhibition of S. pneumoniae IgA1P.
The 3,4-HOPTO scaffold is amenable to medicinal chemistry,
as previously demonstrated in the discovery of a potent
inhibitor of another Zn2+ metalloprotease virulence factor, LasB
of Pseudomonas aeruginosa.28 Thus, we chose to focus our initial
efforts on the 3,4-HOPTO scaffold for the development of
more advanced lead compounds.
In order to identify 3,4-HOPTO derivatives with improved in
vitro potency against S. pneumoniae IgA1P, a sublibrary of
compounds based on fragments 4 and 5 was analyzed.26,28 3,4-
HOPTO analogues 8 were constructed via a one-step
condensation of a primary amine with thiomaltol 7 (Table
1)26,28 and initially tested for inhibition at 100 μM. From this
sublibrary and additional structure−activity relationship studies,
one compound was discovered with >200-fold improvement in
potency, compound 8a with an IC50 value of 4.30 μM (Table 1;
see Table S1 for the activity of other sublibrary members). With
this inhibitor in hand, we then confirmed its inhibitory activity
via Western blot analysis as described above, and complete
inhibition was observed with 25 μM compound 8a (see Figure
S4). These results provide additional evidence that libraries of
metal-chelating fragments can successfully be used to identify
first-in-class inhibitors of bacterial Zn2+ metalloprotease
virulence factors.28,29 Future synthetic efforts will focus on
the preparation of additional analogues to identify compounds
that not only have enhanced affinity for IgA1P but are also
highly specific for this virulence factor, as 3,4-HOPTO
derivatives including 8a have been shown to be active against
other metalloenzymes (see Figure S5 for the metalloenzyme
specificity of 8a).26,27,29
In summary, we have described the first user-friendly,
colorimetric screening assay for IgA1P from S. pneumoniae.
By utilizing the enzyme’s ability to modulate the surface charge
of the bacteria, we have designed an electrostatics-based assay
in which the resulting Fab fragments are able to induce the
aggregation of citrate-adsorbed, negatively charged AuNPs to
produce a colorimetric absorbance signal. Importantly, this
assay approach offers many advantages over the current state-
of-the-art for IgA1Ps, including amenability to high-throughput
screening for the identification of small molecule inhibitors and
elimination of the requirement of harmful or expensive
reagents. Although the overall charges of IgA1 Fab fragments
produced by other IgA1Ps have yet to be measured, the assay
platform described herein may be amenable to monitoring the
activity of these related proteases, such as the IgA1Ps produced
by oral Streptococcal species, Haemophilus influenzae, and
Neisseria meningitidis.6 Finally, and most significantly, using
our AuNP assay for S. pneumoniae IgA1P, we have uncovered
the first inhibitors of the enzyme, and these compounds should
enable the investigation of this pneumococcal virulence factor
as a potential antibacterial target.
C
dx.doi.org/10.1021/ja404180x | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
(24) Fujimori, T.; Wirsching, P.; Janda, K. D. J. Comb. Chem. 2003, 5,
625.
(25) Xu, Y.; Shi, J.; Yamamoto, N.; Moss, J. A.; Vogt, P. K.; Janda, K.
D. Bioorg. Med. Chem. 2006, 14, 2660.
(26) Agrawal, A.; Johnson, S. L.; Jacobsen, J. A.; Miller, M. T.; Chen,
L.; Pellecchia, M.; Cohen, S. M. ChemMedChem 2010, 5, 195.
(27) Jacobsen, J. A.; Fullager, J. L.; Miller, M. T.; Cohen, S. M. J. Med.
Chem. 2011, 54, 590.
(28) Garner, A. L.; Struss, A. K.; Fullager, J. L.; Agrawal, A.; Moreno,
A. Y.; Cohen, S. M.; Janda, K. D. ACS Med. Chem. Lett. 2012, 3, 668.
(29) Fullager, J. L.; Garner, A. L.; Struss, A. K.; Day, J. A.; Martin, D.
P.; Yu, J.; Cai, X.; Janda, K. D.; Cohen, S. M. Chem. Commun. 2013,
49, 3197.
ASSOCIATED CONTENT
■
S
* Supporting Information
General methods and materials, IgA1P preparation, assay
protocols, and supplemental figures. This material is available
free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
kdjanda@scripps.edu
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We dedicate this paper to Prof. Robert M. Williams on the
occasion of his 60th birthday. This work was supported by the
NIH (R01 AI077644 to K.D.J. and R01 GM098435 to S.M.C.).
We are grateful to Prof. Jeffrey Weiser and Mikhail Shchepetov
(University of Pennsylvania) for providing D39 and IgA1P KO
strain S. pneumoniae and helpful bacterial culture advice. We
also thank Dr. Major Gooyit for assay verification studies.
REFERENCES
■
(1) Bogaert, D.; de Groot, R.; Hermans, P. W. M. Lancet Infect. Dis.
2004, 4, 144.
(2) Kadioglu, A.; Weiser, J. N.; Paton, J. C.; Andrew, P. W. Nat. Rev.
Microbiol. 2008, 6, 288.
(3) Wani, J. H.; Gilbert, J. V.; Plaut, A. G.; Weiser, J. N. Infect. Immun.
1996, 64, 3967.
(4) Poulsen, K.; Reinholdt, J.; Kilian, M. Infect. Immun. 1996, 64,
3957.
(5) Mulks, M. H.; Kornfeld, S. J.; Plaut, A. G. J. Infect. Dis. 1980, 141,
450.
(6) Kilian, M.; Reinholdt, J.; Lomholt, H.; Poulsen, K.; Frandsen, E.
V. G. APMIS 1996, 104, 321.
(7) Batten, M. R.; Senior, B. W.; Kilian, M.; Woof, J. M. Infect.
Immun. 2003, 71, 1462.
(8) Chintalacharuvu, K. R.; Chuang, P. D.; Dragoman, A.; Fernandez,
C. Z.; Qiu, J.; Plaut, A. G.; Trinh, K. R.; Gala, F. A.; Morrison, S. L.
Infect. Immun. 2003, 71, 2563.
(9) Weiser, J. N.; Bae, D.; Fasching, C.; Scamurra, R. W.; Ratner, A.
J.; Janoff, E. N. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 4215.
(10) Polissi, A.; Pontiggia, A.; Feger, G.; Altieri, M.; Mottl, H.;
Ferrari, L.; Simon, D. Infect. Immun. 1998, 66, 5620.
(11) Chiavolini, D.; Memmi, G.; Maggi, T.; Iannelli, F.; Pozzi, G.;
Oggioni, M. R. BMC Microbiol. 2003, 3, 14.
(12) Hava, D. L.; Camilli, A. Mol. Microbiol. 2002, 45, 1389.
(13) Qiu, J.; Brackee, G. P.; Plaut, A. G. Infect. Immun. 1996, 64, 933.
(14) Reinholdt, J.; Kilian, M. J. Immunol. Methods 1983, 63, 367.
(15) Reinholdt, J. J. Immunol. Methods 1996, 191, 39.
(16) Walker, S. G.; Carnu, O. I.; Tuter, G.; Ryan, M. E. FEMS
Immunol. Med. Microbiol. 2006, 48, 218.
(17) Bachovchin, W. W.; Plaut, A. G.; Flentke, G. R.; Lynch, M.;
Kettner, C. A. J. Biol. Chem. 1990, 265, 3738.
(18) Choi, Y.; Ho, N.-H.; Tung, C.-H. Angew. Chem., Int. Ed. 2007,
46, 707.
(19) Oishi, J.; Asami, Y.; Mori, T.; Kang, J.-H.; Tanabe, M.; Niidome,
T.; Katayama, Y. ChemBioChem 2007, 8, 875.
(20) Elghanian, R.; Storhoff, J. J.; Mucic, R. C.; Letsinger, R. L.;
Mirkin, C. A. Science 1997, 277, 1078.
(21) Romanello, V.; Marcacci, M.; Molin, F. D.; Moschioni, M.;
Censini, S.; Covacci, A.; Baritussio, A. G.; Montecucco, C.; Tonello, F.
Protein Exp. Pur. 2006, 45, 142.
(22) Bender, M. H.; Weiser, J. N. Mol. Microbiol. 2006, 61, 526.
(23) Zhang, J.; Chung, T. D. Y.; Oldenburg, K. R. J. Biomol. Screen.
1999, 4, 67.
D
dx.doi.org/10.1021/ja404180x | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX