PH responsive Janus-like supramolecular fusion proteins for functional protein delivery

Article abstract of DOI:10.1021/ja4084122

A facile, noncovalent solid-phase immobilization platform is described to assemble Janus-like supramolecular fusion proteins that are responsive to external stimuli. A chemically postmodified transporter protein, DHSA, is fused with (imino)biotinylated cargo proteins via an avidin adaptor with a high degree of spatial control. Notably, the derived heterofusion proteins are able to cross cellular membranes, dissociate at acidic pH due to the iminobiotin linker and preserve the enzymatic activity of the cargo proteins β-galactosidase and the enzymatic subunit of Clostridium botulinum C2 toxin. The mix-and-match strategy described herein opens unique opportunities to access macromolecular architectures of high structural definition and biological activity, thus complementing protein ligation and recombinant protein expression techniques.

Full text of DOI:10.1021/ja4084122

Communication
pubs.acs.org/JACS
pH Responsive Janus-like Supramolecular Fusion Proteins for
Functional Protein Delivery
Seah Ling Kuan,∥,† David Y. W. Ng,∥,†,‡ Yuzhou Wu,†,‡ Christina Fortsch,§ Holger Barth,§
̈
Mikheil Doroshenko,‡ Kaloian Koynov,‡ Christoph Meier,† and Tanja Weil*,†
†Institute of Organic Chemistry III, Macromolecular Chemistry, University of Ulm, Albert-Einstein-Allee 11, D-89081 Ulm, Germany
‡Max Planck Institute for Polymer Research Ackermannweg 10, D-55128 Mainz, Germany
§Institute of Pharmacology and Toxicology, University of Ulm Medical Center, Albert-Einstein-Allee 11, D-89081 Ulm, Germany
S
* Supporting Information
through the production of monovalent or divalent streptavidin
proteins via genetic engineering,14,15 the applications are still
ABSTRACT: A facile, noncovalent solid-phase immobili-
hampered by lower binding constants15 and a lack of spatial
zation platform is described to assemble Janus-like
supramolecular fusion proteins that are responsive to
external stimuli. A chemically postmodified transporter
protein, DHSA, is fused with (imino)biotinylated cargo
proteins via an avidin adaptor with a high degree of spatial
control. Notably, the derived heterofusion proteins are
able to cross cellular membranes, dissociate at acidic pH
due to the iminobiotin linker and preserve the enzymatic
activity of the cargo proteins β-galactosidase and the
enzymatic subunit of Clostridium botulinum C2 toxin. The
mix-and-match strategy described herein opens unique
opportunities to access macromolecular architectures of
high structural definition and biological activity, thus
complementing protein ligation and recombinant protein
expression techniques.
control. A noncovalent strategy to rapidly self-assemble the
protein components with a high degree of control over
composition and spatial arrangement is therefore highly
attractive for the systematic construction of asymmetric protein
motifs in a “built-to-order” fashion.
Iminobiotin, an imine analogue of biotin, interacts non-
covalently with avidin, and its resultant binding affinity is
dependent on the extent of protonation of the imine group
(Figure 1a, basic pH, Kd∼10−11 M; acidic pH, Kd ∼10−3 M).16
he development of a versatile approach to derive stimuli
Tresponsive, Janus-like fusion proteins holds great promise
in the realization of precise multifunctional platforms that
utilize the tumor microenvironment as a trigger, e.g., acidic1 or
reducing2 milieus, for the delivery and controlled release of
functional proteins. Genetic fusion is to date the major route to
access these chimeras as exemplified by antibody-enzyme
conjugates.3,4 However, controlled degradation of the recombi-
nant fusion protein proceeds mainly via enzymatic cleavage,
and it is not possible to introduce pH or light sensitive break
points via genetic engineering. Chemical conjugation strategies
are valuable alternatives to engineer fusion proteins, since they
allow the combination of native and chemically postmodified
proteins, thereby expanding the repertoire of proteins nature
offers.5−7 Furthermore, chemical approaches allow small
molecules to be programmed into the protein chimeras8 for
controlled disintegration of the fusion proteins. In particular,
noncovalent approaches are of emerging interest, since mild
conditions can be applied to retain bioactivity of the fusion
construct to impart in vivo functionality.8−11 Avidin−biotin
technology has been long since employed as a molecular
adaptor for heterofunctional bioconjugates with extensive use in
cancer pretargeting.12,13 Nevertheless, precise control of the
resultant stoichiometry and distinct orientation of the conjugate
is limited. Despite attempts to improve stoichiometric control
Figure 1. Synthesis of the DHSA-Av dimer. (a) Protonation of the
imine group in iminobiotin at acidic pH. (b) Biotin-PEO-maleimide
linker that is conjugated to the SH-group of protein 1 (DHSA). (c)
Schematic overview for the toposelective solid phase immobilization to
assemble DHSA-Av dimers.
The pH responsive iminobiotin−avidin interaction has been
employed mostly for protein affinity purification17 and, more
recently, in the synthesis of layer-by-layer nanoparticles that
exhibit a pH-sensitive outer stealth layer.18 We hypothesize that
adopting the iminobiotin−avidin technology will allow us to
fulfill two criteria: (1) The combination with solid phase
synthesis will lead to the construction of heterofusion proteins
with high degree of spatial control. (2) Programming the pH
Received: August 14, 2013
Published: October 25, 2013
© 2013 American Chemical Society
17254
dx.doi.org/10.1021/ja4084122 | J. Am. Chem. Soc. 2013, 135, 17254−17257

Journal of the American Chemical Society
Communication
trigger, iminobiotin, into the cargo protein will allow controlled
release and preservation of bioactivity in vivo, which opens
access to stimuli responsive release of the protein cargo inside
acidic cancer cells,1,19 thereby serving as an attractive option to
existing release strategies.20 Herein, the first examples of pH
responsive Janus-like supramolecular fusion proteins based on
the iminobiotin−avidin linkage are reported. Five different
asymmetric supramolecular fusion proteins with opposite
“faces” are assembled, i.e., a biotinylated chemically post-
modified transporter protein at one site and a biotinylated or
pH responsive iminobiotinylated protein cargo at the opposite
site. The release and preservation of enzymatic activity of the
assembled supramolecular fusion proteins are also investigated.
Dendronized human serum albumin (DHSA)21 is selected as
a chemically postmodified transporter protein because of its
enhanced cellular uptake and low cytotoxicity and its potential
to deliver drug molecules into cells.21 As the first step to
construct the fusion protein, biotin is tethered to the single
unpaired sulfhydryl of the Cys34 residue on HSA via a PEO-
maleimide linker (Figure 1b, Chart S1, Supporting Information
(SI)). Dendronization of HSA is accomplished by conjugation
of ethynyl-polyamidoamine dendrons to azido-functionalized
HSA via the Huisgen 1,3-dipolar cycloaddition as reported
previously (Figure S1 (SI)).21 Avidin (Av) is first immobilized
onto iminobiotin beads to mask one hemisphere of Av,
followed by assembly of biotinylated DHSA onto the
toposelective Av (Figure 1c). The DHSA-Av heterodimer is
isolated by subsequent cleavage from the beads at acidic pH
(Figure 1c) and finally purified by cation exchange chromatog-
raphy. The resultant DHSA-Av construct possesses a vacant site
at the opposite hemisphere, formerly occupied by the solid
phase, for the conjugation of the “cargo” protein entity (Figure
2). The purified DHSA-Av (Figure S3 (SI)) is characterized by
Figure 3. Characterization of the supramolecular fusion proteins. (a)
Fluorescence polarization of DHSA-Av and DHSA-Av-HSA against
protein standards. *Molecular weights of the protein standards as
reported or calculated from MALDI-ToF (b) AFM image (top)
showing DHSA-Av heterodimers (H = 8 nm) on mica surface and
hydrodynamic radii of DHSA-Av determined by FCS measurements
(bottom). (c) AFM image (top) and height profile (bottom) of
DHSA-Av-(iminobiotin)-βGal (3) trimer on mica surface. The
respective structures of the constructs are depicted above the AFM
images. Shapes that are expected for the projections of the
heterofusion proteins cross section are indicated in the AFM images.
homogeneity of the resultant fusion construct and the statistical
analysis of the height profile and aspect ratios of the
macromolecules indicate a population composing of 89%
dimers with one Av per DHSA molecule (Figures 3b, S8 (SI)).
Noteworthy, only one protein is coupled to each site for steric
reasons, and therefore, no higher order assemblies with three or
four proteins are detected. These findings indicate the efficiency
of the proposed method for the construction of defined protein
conjugates in a controlled fashion. The membrane translocation
of Av is shown to be mediated by the appended DHSA, with a
6-fold increase in uptake (Figure S11 (SI)).
Three different types of cargo proteins, namely, HSA, the
enzymatic subunit of Clostridium botulinum C2 toxin (C2I) and
β-galactosidase (βGal) have been applied to construct the
supramolecular fusion proteins. All three cargo proteins, HSA,
C2I and βGal, have been statistically labeled with biotin or
iminobiotin by reaction with the respective NHS esters (Figure
2a, Chart S1b (SI)).
Thereafter, they are assembled with DHSA-Av by incubation
(Figure 2b) to yield the resultant DHSA-Av-Cargo complexes
(1−5), which are used for subsequent characterization and
biological evaluation. The formation of the fusion DHSA-Av-
Cargo complexes is determined by fluorescence polarization or
AFM. The formation of the protein trimers DHSA-Av-(biotin)-
HSA (1) is apparent from the size increase (MW ∼ 230 kDa)
detected in fluorescence polarization (Figure 3a). AFM
characterization is applied to the DHSA-Av-(iminobiotin)-
βGal (3), since it allows better differentiation of the protein
entities due to the greater size difference of βGal to the protein
components in DHSA-Av. The AFM image shows the
formation of a heterotrimer (Figure 3c). The heights of
DHSA, Av and βGal in the heterotrimeric structure are
determined to be 6, 6.5, and 8 nm, respectively. The heights of
DHSA and βGal are consistent with the sizes of the individual
entities. Av has a higher than expected height of 6.5 nm, as it is
supported by the two larger proteins (Figure 3c).
Figure 2. Assembly of DHSA-Av-Cargo complexes. (a) Iminobiotin
(pH-cleavable) and biotin (non-pH-cleavable) NHS esters conjugated
to amine groups of the cargo proteins HSA, βGal, and C2I. (b)
Assembly of DHSA-Av with the (imino)biotinylated cargo proteins
HSA, βGal, or C2I.
the SDS-PAGE, MALDI-ToF and fluorescence polarization
(Figures 3a, S4, S5 (SI)), which corroborates with the
formation of the fusion construct. Further characterization is
carried out using atomic force microscopy (AFM) and
fluorescence correlation spectroscopy (FCS) to determine the
size and the homogeneity of the resultant construct. The mean
hydrodynamic radius of DHSA-Av determined by FCS is 7.4
nm (Figures 3b, S6 (SI)) supports the formation of the fusion
construct, which is consistent with AFM data (8 nm, Figure S8
(SI)). AFM allows not only the estimation of the complex sizes
but also provides information about sample homogeneity and
the physical arrangement of the constituent components in the
observed multiprotein complexes.22,23 Estimation of the
The pH responsiveness of the avidin-iminobiotin interaction
is verified by a competitive binding assay using 4′-
hydroxyazobenzene-2-carboxylic acid, which has a lower
17255
dx.doi.org/10.1021/ja4084122 | J. Am. Chem. Soc. 2013, 135, 17254−17257

Journal of the American Chemical Society
Communication
binding affinity for Av than iminobiotin (Figure 4a).1 In the
following evaluation, pH-induced release of the cargo protein
Clostridium botulinum C2 toxin, C2I, is evaluated in vitro. C2I
contributes to the cytotoxic activity of the toxin,25,26 and it has
been selected to investigate the pH-triggered release of bacterial
protein toxins from the fusion DHSA-Av carrier. In addition,
the preservation of the cytotoxicity after release is of interest for
the induction of apoptotic cell deaths.27 The cytotoxic activity
of C2I has been assessed by the observation of toxin-specific
cell-rounding due to C2I-catalyzed ADP-ribosylation of actin in
the cytosol.28 Both biotin-C2I and iminobiotin-C2I cargo
proteins are conjugated to DHSA-Av at basic pH, and unbound
C2I is removed by ultrafiltration with a cutoff of 100 kDa.
Thereafter, the thus-assembled DHSA-Av-(biotin)-C2I (4) and
DHSA-Av-(iminobiotin)-C2I (5) are incubated for 30 min at
pH 4 to release the C2I cargo. DHSA-Av (∼158 kDa) and the
released cargo (∼50 kDa) are then separated by ultrafiltration
with a 100 kDa cutoff (Figure 4d). The flow-through and
supernatant fractions obtained are subsequently analyzed by
Western blotting (Figure 4d). Only C2I conjugated via the
iminobiotin linker is detected in the flow-through fraction by a
specific C2I-antibody in the Western blot analysis (Figure 4d).
For (4), no C2I is detected in the flow-through. The flow-
through fractions of carrier-released iminobiotin-C2I is then
analyzed for the presence and the preservation of the toxin
activity (Figure 5a).
Figure 4. pH-induced disassembly of DHSA-Av-Cargo complexes. (a)
Extracellular competitive binding assay with 4′-hydroxyazobenzene-2-
carboxylic acid against iminobiotin-Av showing pH responsiveness. (b)
Confocal images for the intracellular pH release study of DHSA-Av
containing biotin-HSA (1) and iminobiotin-HSA (2). Scale bar is 20
μm. (c) The normalized single cell fluorescence intensity shows
enhanced quenching in the donor emission channel due to energy
transfer in (1). Data are expressed as mean
SEM, N = 5 (SI).
Statistical significance is calculated using student’s t test (*p < 0.05, **
p < 0.005). (d) Schematic overview for the pH release experiment of
DHSA-Av containing biotin-C2I (4) and iminobiotin-C2I (5) and the
corresponding Western blot analysis. The supernatant (S) and
flowthrough (F) fractions after incubation of (4) and (5) at pH 4
and subsequent ultrafiltration (100 kDa cutoff) are collected and
analyzed by Western blotting with an antibody against C2I.
HSA from the complexes (1) and (2) has been investigated by
Forster resonance energy transfer (FRET). FRET allows
̈
detecting the proximity between two fluorophores up to 10
nm distances.24 It is thus possible to evaluate the release of the
HSA cargo by FCS either at the single molecule level in vitro or
in living cells directly by fluorescence imaging. The DHSA-Av
carrier is labeled with Atto-425 as FRET donor and biotin- or
iminobiotin-HSA cargos are labeled with Atto-520 as FRET
acceptor. FCS studies in the emission channel of the donor
(Figure S7 (SI)) reveal similar extent of quenching for the
biotin-HSA containing complex (1) and iminobiotin-HSA
containing complex (2) at pH 9. However, at pH 4, (2)
exhibits significantly diminished quenching (about 3-fold less
than (1)) indicating cargo release. In addition, the fusion
constructs, (1) and (2), have been evaluated for HSA release in
living cells by fluorescence imaging.
Confocal microscopy studies reveal that both complexes (1)
and (2) are taken up by A549 cells after incubation overnight
regardless of the different linkers used (Figure S12 (SI)). The
fluorescence images and intensities of A549 cancer cells
incubated with (1) and (2) are recorded with excitation at
458 nm and emission at two channels: 463−509 nm (donor)
and 520−658 nm (acceptor), (Figures 4b,c, S13 (SI)). The
confocal studies reveal enhanced quenching in the donor
emission channel due to energy transfer in (1) compared with
(2), further substantiating the release of the iminobiotinylated
HSA but not the biotinylated HSA in A549 cells.
Figure 5. Protein activity assays. (a) The presence of catalytically
active C2I protein in the flow-through fractions (F) has been analyzed
in the cell rounding assay. HeLa cells are incubated for 24 h at 37 °C
with the flow-through fractions of carrier-released iminobiotin-C2I or
biotin-C2I. The transport component of C2 toxin, C2IIa, has been
added to mediate the transport of C2I into the cytosol of the cells. For
control cells are left untreated or treated with C2IIa and C2I for
positive control. (b) Microscopic images of A549 cells treated with
100 nM of (left) iminobiotin-βGal and (right) DHSA-Av-(iminobio-
tin)-βGal (3), followed by treatment with the substrate X-gal.
The transport component of C2 toxin, C2IIa, is then added
together with the flow-through fractions into cultured cells to
mediate the transport of C2I into the cytosol of the cells to
assess whether it still displays functional activity. Notably, cell
rounding is observed for the flow-through fractions of carrier-
released iminobiotin-C2I but not that of biotin-C2I (negative
control). Taken together, the results show that only
iminobiotin-C2I, but not biotin-C2I, is released from the
DHSA-Av carrier under acidic conditions. Moreover, the
carrier-released iminobiotin-C2I retains activity to induce cell
rounding in HeLa cells.
Thereafter, cellular studies have been accomplished to assess
the enzymatic activity of cargo proteins inside cells. βGal is a
tetrameric membrane-impermeable protein that catalyzes the
hydrolysis of β-glycosidic bonds. DHSA-Av nanocarrier plat-
form is applied for the delivery of βGal into A549 cancer cells
to directly study the retention of enzymatic activity of the cargo
The retention of activity is a key concern in protein delivery
and the enzymatic activity provides a stringent test. Thus, the
pH-induced disassembly of the enzymatic subunit of the
17256
dx.doi.org/10.1021/ja4084122 | J. Am. Chem. Soc. 2013, 135, 17254−17257

Journal of the American Chemical Society
Communication
Notes
after cell uptake. The enzymatic activity of βGal can be detected
by a colorimetric assay through the incubation of the treated
cells for 24 h with 5-bromo-4-chloro-3-indolyl-β-D-galactopyr-
anoside (X-gal), a substrate for βGal. Cellular delivery of βGal
is accomplished by assembling iminobiotinylated βGal
(iminobiotin-βGal) and DHSA-Av. Confocal microscopy
shows that (3) is internalized into A549 cells (Figure S14
(SI)). The enzymatic activity of βGal after membrane
transduction is investigated by incubating A549 cells that are
pretreated with (3) for a further 24 h with X-gal. Upon
enzymatic hydrolysis of the substrate X-gal to galactose and 5-
bromo-4-chloro-3-hydroxyindole by βGal, the color changes
from colorless to blue, thus demonstrating the successful
intracellular delivery of the protein. No color change is
observed when cells were treated with βGal alone (Figure
5b). Our findings clearly indicate that the catalytic activity of
βGal is preserved upon internalization.
In summary, we have demonstrated a versatile, supra-
molecular approach to derive defined heteroassembly Janus-
like fusion proteins. A high degree of spatial control is achieved
using solid phase immobilization based on the pH responsive
iminobiotin−avidin interaction. Following this approach, five
precision Janus-like fusion conjugates have been prepared: Two
noncleavable, biotin-containing cargoes and three pH-cleavable,
iminobiotin-containing cargoes, which are the first examples of
pH responsive Janus-like fusion proteins based on the
iminobiotin−avidin interaction. Our strategy allows the non-
covalent conjugation of both native and chemically post-
modified proteins, which grants access to greater diversity and
expands the toolbox nature offers, as exemplified by the
dendronized transporter protein DHSA. Notably, we have
demonstrated (a) through FRET studies and in vitro assays that
the iminobiotin moiety can serve as a pH trigger for the release
of the cargo proteins, iminobiotin-HSA and -C2I, (b) that the
DHSA-Av carrier can transport the cargo proteins into cells and
the cargo proteins are released in the acidic tumor environ-
ment, as exemplified by the HSA derivatives, and (c) that the
catalytic activity of the carrier-released iminobiotin-C2I and the
iminobiotin-βGal that is internalized into cells is preserved.
One could envision programming fusion proteins combining
cell- or tissue-selectivity and intracellular release at acidic pH
found in the tumor microenvironment, which is highly
attractive to achieve targeted delivery and controlled release
of functional proteins, e.g., catalytic component of toxins. The
mix-and-match strategy described opens unique opportunities
to access macromolecular biohybrid architectures of high
structural definition and biological activity and in this way
complements protein ligation and recombinant protein
expression techniques.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors are grateful to the financial support from the
BMBF Project Biotechnologie 2020+, ERC Synergy Grant
319130-BioQ and the German Research Foundation (DFG
Grant BA 2087/2-2 to H.B. and SFB 625 to T.W.). S.L.K. is
grateful to the Alexander von Humboldt Foundation for
providing a research fellowship.
REFERENCES
■
(1) Tannock, I. F.; Rotin, D. Cancer Res. 1989, 49, 4373−4384.
(2) Li, C.; Hu, J.; Liu, S. Soft Matter 2012, 8, 7096−7102.
(3) Andrady, C.; Sharma, S. K.; Chester, K. A. Immunotherapy 2011,
3, 193−211.
(4) Huang, C. Curr. Opin. Biotechnol. 2009, 20, 692−699.
(5) Hudak, J. E.; Barfield, R. M.; de Gregory, W., H.; Grob, P.;
Nogales, E.; Bertozzi, C. R.; Rabuka, D.; de Hart, G. W. Angew. Chem.,
Int. Ed. 2012, 51, 4161−4165.
(6) Hutchins, B. M.; Kazane, S. A.; Staflin, K.; Forsyth, J. S.; Felding-
Habermann, B.; Smider, V. V.; Schultz, P. G. Chem Biol. 2011, 18,
299−303.
(7) Schoffelen, S.; Beekwilder, J.; Debets, M. F.; Bosch, D.; van Hest,
J. C. Bioconjugate Chem. 2013, 24, 987−996.
(8) Gangar, A.; Fegan, A.; Kumarapperuma, S. C.; Wagner, C. R. J.
Am. Chem. Soc. 2012, 134, 2895−2897.
(9) Komatsu, T.; Terada, H.; Kobayashi, N. Chem.Eur. J. 2011, 17,
1849−1854.
(10) Frokjaer, S.; Otzen, D. E. Nat. Rev. Drug Discovery 2005, 4, 298−
306.
(11) Padilla, J. E.; Colovos, C.; Yeates, T. O. Proc. Natl. Acad. Sci. U.
S. A. 2001, 98, 2217−2221.
(12) Sakahara, H.; Saga, T. Adv. Drug Delivery Rev. 1999, 37, 89−101.
(13) Liu, G.; Hnatowich, D. J. Bioconjugate Chem. 2008, 19, 2095−
2104.
(14) Aslan, F. M.; Yu, Y.; Mohr, S. C.; Cantor, C. R. Proc. Natl. Acad.
Sci. U. S. A. 2005, 102, 8507−8512.
(15) Wu, S.-C.; Wong, S.-L. J. Biol. Chem. 2005, 280, 23225−23231.
(16) Fudem-Goldin, B.; Orr, G. A. In Avidin Biotin Technology:
Methods in Enzymology; Wilchek, M., Bayer, E. A., Eds.; Academic
Press: San Diego, 1990.
(17) Cao, Y.; Bai, G.; Zhang, L.; Bai, F.; Yang, W. Artif. Cells, Blood
Substitutes, Immobilization Biotechnol. 2006, 34, 487−500.
(18) Poon, Z.; Chang, D.; Zhao, X.; Hammond, P. T. ACS Nano
2011, 5, 4284−4292.
(19) Wojtkowiak, J. W.; Rothberg, J. M.; Kumar, V.; Schramm, K. J.;
Haller, E.; Proemsey, J. B.; Lloyd, M. C.; Sloane, B. F.; Gillies, R. J.
Cancer Res. 2012, 72, 3938−3947.
(20) Kale, A. A.; Torchilin, V. P. Bioconjugate Chem. 2007, 18, 363−
370.
(21) Kuan, S. L.; Stockle, B.; Reichenwallner, J.; Ng, D. Y. W.; Wu,
Y.; Doroshenko, M.; Koynov, K.; Hinderberger, D.; Mullen, K.; Weil,
T. Biomacromolecules 2013, 14, 367−376.
(22) Muller, D. J.; Janovjak, H.; Lehto, T.; Kuerschner, L.; Anderson,
K. Prog. Biophys. Mol. Biol. 2002, 79, 1−43.
(23) Gaczynska, M.; Osmulski, P. A. Curr. Opin. Colloid Interface Sci.
2008, 13, 351−367.
(24) Kenworthy, A. K. Methods 2001, 24, 289−296.
(25) Aktories, K.; Barmann, M.; Ohishi, I.; Tsuyama, S.; Jakobs, K.;
̈
̈
ASSOCIATED CONTENT
̈
■
S
* Supporting Information
Supporting figures (Scheme 1, Chart S1, Figures S1−15) and
methods. This material is available free of charge via the
Internet at http://pubs.acs.org.
̈
Habermann, H. E. Nature 1986, 322, 390−392.
(26) Aktories, K.; Barth, H. Eur. J. Cell Biol. 2011, 90, 944−950.
AUTHOR INFORMATION
■
(27) Heine, K.; Pust, S.; Enzenmuller, S.; Barth, H. Infect. Immun.
̈
2009, 76, 4600−4608.
(28) Barth, H.; Blocker, D.; Behlke, J.; Bergsma-Schutter, W.;
Brisson, A.; Benz, R.; Aktories, K. J. Biol. Chem. 2000, 275, 18704−
18711.
Corresponding Author
̈
tanja.weil@uni-ulm.de
Author Contributions
∥S.L. Kuan and D.Y.W. Ng contributed equally.
17257
dx.doi.org/10.1021/ja4084122 | J. Am. Chem. Soc. 2013, 135, 17254−17257