Supplementary
Information
Identification of Secreted Bacterial Proteins
by Non
-
C
anonical Amino Acid Tagging
Authors:
Alborz Mahdavi
a,b
, Janek Szychowski
a
, John T. Ngo
a
, Michael J. Sweredoski
c
, Robert
L.J
. Graham
c
, Sonja Hess
c
, Olaf Schneewind
d
, Sarkis
K. Mazmanian
e
, David A. Tirrell
a,1
Affiliations:
a
Division of Chemistry and Chemical Engineering, California Institute of Technology,
Pasadena, CA 91125, USA.
b
Department of Bioengineering, California Institute of Technology, Pasadena, CA 91125, USA.
c
Proteome Exploration Laboratory, Beckman Institute, California Institute of Technology,
Pasadena, CA 91125, USA.
d
Department of Microbiology, University of Chicago, Chicago, IL 60637, USA.
e
Division of Biology
and Biological Engineering
, California Inst
itute of Technology, CA 91125,
USA.
Corresponding author:
David A. Tirrell
Division of Chemistry and Chemical Engineering
California Institute of Technology
1200 East California Blvd. MC 210
-
41
Pasadena, CA, 91125
Email: tirrell@caltech.edu
Tel:
626
-
395
-
3140
Fax: 626
-
568
-
8743
Figure S1.
Synthesis of Anl, the Acid
-
Cleavable Biotin
-
Alkyne Affinity Enrichment Tag and Copper
-
Catalyzed Click Reaction
Synthesis of Anl by Copper
-
Catalyzed Diazo T
ransfer
.
The synthesis of Anl was based on a
previously
published protocol
(
1
); the
starting material was
Boc
-
lysin
.
Briefly,
5.27 g (81.1
mmol) of sodium azide was treated with 2.7 ml (16
mmol) of distilled triflic anhydride in 13 ml
of water for 2 hours. The triflic azide product was extracted with 10 ml
methylene chloride and
added dropwise to a flask containing Boc
-
Lys
-
OH (2 g, 8.1 mmol), K
2
CO
3
(1.68 g, 12.2 mmol) and
CuSO
4
(20 mg, 0.08 mmol) in 26 ml of water and 250 ml of methanol. After 20 hours at room
temperature the product was extracted with ethyl
acetate, redissolved in methylene chloride
and purified by silica gel chromatography. After Boc deprotection with hydrochloric acid, the
final product was purified by cation exchange chromatography.
Copper
-
Catalyzed Click R
eaction
.
The copper
-
catalyzed
azide
-
alkyne coupling reaction, and the
synthesis of the requisite THPTA ligand were performed as described previously
(2)
.
Labeling
was carried out at room temperature for 2 hours, at a final concentration of
0.2
-
0.5 mg/ml of
HeLa lysate proteins, 0.1 mM copper sulfate, 0.5 mM THPTA ligand, 5 mM sodium ascorbate, 5
mM aminoguanidine and 100 μM alkyne probe. EDTA
-
free protease inhibitor (Roche) was
added to
all reactions
and lysates.
1.
Link AJ, Vink MK, &
Tirrell DA (2007) Preparation of the functionalizable methionine surrogate
azidohomoalanine via copper
-
catalyzed diazo transfer.
Nat. Protoc.
2(8):1879
-
1883.
2.
Hong V, Presolski SI, Ma C, & Finn MG (2009) Analysis and optimization of copper
-
catalyzed az
ide
-
alkyne cycloaddition for bioconjugation.
Angew. Chem. Int. Ed.
48(52):9879
-
9883.
A
cid
-
cleavable
biotin
-
alkyne
affinity
enrichment
tag
N
-
(2
-
(((hex
-
5
-
yn
-
1
-
yloxy)diphenylsilyl)oxy)
-
2
-
methylpropyl)
-
1
-
(5
-
((3a
S
,4
S
,6a
R
)
-
2
-
oxohexahydro
-
1
H
-
thieno[3,4
-
d]imida
zol
-
4
-
yl)pentanamido)
-
3,6,9,12
-
tetraoxapentadecan
-
15
-
amide.
Synthesis of the acid
cleavable alkyne tag was based on our recent work on the analogous acid
-
sensitive
dialkoxydiphenylsilane biotin
-
azide tag (
45
), with the exception that
hex
-
5
-
yn
-
1
-
ol was used
in
place of 6
-
azidohexanol.
1
H NMR (500 MHz, CDCl
3
),
δ
(ppm):
7.72
-
7.50 (m, 4H), 7.40 (m, 6H),
7.06
-
6.95 (m, 1H), 6.66 (s, 1H), 6.50 (t,
J =
5.65, 5.65 Hz, 1H), 5.80 (s, 1H), 4.53
-
4.37 (m, 1H),
4.34
-
4.19 (m, 1H), 3.73 (td,
J =
19.58, 6.04, 6.04 Hz, 4H), 3.
65
-
3.48 (m, 14H), 3.48
-
3.35 (m, 2H),
3.35
-
3.22 (m, 1H), 3.11 (dd,
J =
11.73, 7.06 Hz, 1H), 2.85 (td,
J =
16.69, 8.51, 8.51 Hz, 1H), 2.72
(t,
J =
10.43, 10.43 Hz, 2H), 2.41 (t,
J =
6.01, 6.01 Hz, 2H), 2.27
-
2.05 (m, 4H), 1.95 (dd,
J =
3.09,
2.17 Hz, 1H), 1.8
1
-
1.52 (m, 8H), 1.50
-
1.33 (m, 2H), 1.23 (d,
J =
22.65 Hz, 6H)
;
13
C NMR (126 MHz,
CDCl
3
),
δ
(ppm): 173.5, 171.2, 164.1, 134.8 (m, 6C), 130.2 (m, 2C), 127.9 (m, 4C), 84.3, 75.7, 70.4
(m, 8C), 68.6, 67.3, 62.6, 61.8, 60.2, 55.7, 50.4, 40.5, 39.1, 37.1, 35.9,
31.3, 28.3, 28.0, 27.6,
25.6, 24.8, 18.1;
m/z
calculated for C
43
H
67
N
7
O
9
SSi [M+H]
+
: 840.4163, MS found: 863.6 (M+Na).
HRMS found: 863.4228.
NMR spectra of acid
-
cleavable biotin
-
alkyne tag (top,
1
H NMR spectrum; bottom,
13
C NMR
spectrum)
Figure S2.
Mass shifts at methionine positions as a result of Anl incorporation and addition of cleavable
tag.
Mass modifications associated with replacement of Met by Anl, with reaction of Anl with
the biotin
-
alkyne affinity enrichment tag, and with cleavage of the
tag by formic acid are
represented. These three mass modifications are included as variable modifications
for tandem
mass spectrometry.
Figure S3.
Sequence of the
NLL
-
MetRS
expression
construct.
DNA sequence of the NheI
-
flanked
expression cassette with
NLL
-
MetRS under control of the
E. coli
MetG promoter and followed
by its natural
E. coli
transcriptional termination signal. The MetRS sequence is underlined.
Mutations L13N, Y260L, H301L are highlighted in green.
GCTAGC
TCTAGAGACGTCCGGCCGGAGCTCCACCGCGGTGGCGGCCGCTCTAGAGTCACTTACTTAACATTTTCCCATTTGGTACTATCTAACC
CCTTTTCACTATTAAGAAGTAATGCCTACT
ATGAC
TCAAGTCGCGAAGAAAATTCTGGTGACGTGCGCA
AAC
CCGTACGCTAACGGCTCAATC
CACCTCGGCCATATGCTGGAGCACATCCAGGCTGATGTCTGGGTCCGTTACCAGCGAATGCGCGGCCACGAGGTCAACTTCATCTGCGCCGA
CGATGCCCACGGTACACCGATCATGCTGAAAGCTCAGCAGCTTGGTATCACCCCGGAGCAGATGATTGGCGAAATGAGTCAGGAGCATCAG
ACTGATTTCGCAGGC
TTTAACATCAGCTATGACAACTATCACTCGACGCACAGCGAAGAGAACCGCCAGTTGTCAGAACTTATCTACTCTCGCC
TGAAAGAAAACGGTTTTATTAAAAACCGCACCATCTCTCAGCTGTACGATCCGGAAAAAGGCATGTTCCTGCCGGACCGTTTTGTGAAAGGCA
CCTGCCCGAAATGTAAATCCCCGGATCAATACGGCGATAACTGCGAAGTCTGCGGCGCGACCTACAGCCCGACTGAACTGATCG
AGCCGAAA
TCGGTGGTTTCTGGCGCTACGCCGGTAATGCGTGATTCTGAACACTTCTTCTTTGATCTGCCCTCTTTCAGCGAAATGTTGCAGGCATGGACCC
GCAGCGGTGCGTTGCAGGAGCAGGTGGCAAATAAAATGCAGGAGTGGTTTGAATCTGGCCTGCAACAGTGGGATATCTCCCGCGACGCCCC
TTACTTCGGTTTTGAAATTCCGAACGCGCCGGGCAAATATTTCTACGTCTGGCTGGACGCACC
GATTGGC
CTG
ATGGGTTCTTTCAAGAATCTG
TGCGACAAGCGCGGCGACAGCGTAAGCTTCGATGAATACTGGAAGAAAGACTCCACCGCCGAGCTGTACCACTTCATCGGTAAAGATATTGT
TTACTTC
CTG
AGCCTGTTCTGGCCTGCCATGCTGGAAGGCAGCAACTTCCGCAAGCCGTCCAACCTGTTTGTTCATGGCTATGTGACGGTGAA
CGGCGCAAAGATGTCCAAGTCTCGCGGCACCTTTATTAAA
GCCAGCACCTGGCTGAATCATTTTGACGCAGACAGCCTGCGTTACTACTACAC
TGCGAAACTCTCTTCGCGCATTGATGATATCGATCTCAACCTGGAAGATTTCGTTCAGCGTGTGAATGCCGATATCGTTAACAAAGTGGTTAA
CCTGGCCTCCCGTAATGCGGGCTTTATCAACAAGCGTTTTGACGGCGTGCTGGCAAGCGAACTGGCTGACCCGCAGTTGTACAAAACCTTCAC
TGATGCCGCTGAAGTGA
TTGGTGAAGCGTGGGAAAGCCGTGAATTTGGTAAAGCCGTGCGCGAAATCATGGCGCTGGCTGATCTGGCTAAC
CGCTATGTCGATGAACAGGCTCCGTGGGTGGTGGCGAAACAGGAAGGCCGCGATGCCGACCTGCAGGCAATTTGCTCAATGGGCATCAACC
TGTTCCGCGTGCTGATGACTTACCTGAAGCCGGTACTGCCGAAACTGACCGAGCGTGCAGAAGCATTCCTCAATACGGAACTGACCTGGGA
T
GGTATCCAGCAACCGCTGCTGGGCCACAAAGTGAATCCGTTCAAGGCGCTGTATAACCGCATCGATATGAGGCAGGTTGAAGCACTGGTGG
AAGCCTCTAAATGAGAAGTAAAAGCCGCTGCCGCGCCGGTAACTGGCCCGCTGGCAGATGATCCGATTCAGGAAACCATCACCTTTGACGAC
TTCGCTAAAGTTGACCTGCGCGTGGCGCTGATTGAAAACGCAGAGTTTGTTGAAGGTTCTGACAAACTGCTG
CGCCTGACGCTGGATCTCGG
CGGTGAAAAACGCAATGTCTTCTCCGGTATTCGTTCTGCTTACCCGGATCCGCAGGCACTGATTGGTCGTCACACCATTATGGTGGCTAACCT
GGCACCACGTAAAATGCGCTTCGGTATCTCTGAAGGCATGGTGATGGCTGCCGGTCCTGGCGGGAAAGATATTTTCCTGCTAAGCCCGGATG
CCGGTGCTAAACCGGGTCATCAGGTGAAATAA
TCCCCCTTCAAGGCGCTGC
ATCGACAGCCTTTTGCTTTATAAATTCCTAAAGTTGTTTTCTT
GCGATTTTGTCTCTCTCTAACCCGCATAAATACTGGTAGCATCTGCATTCAACTGGATAAAATTACAGGGATGCAGAATGAGACACTTTATCTA
TCAGGACGAAAAATCACATAAATTCAGGGCAGTTGAGCAACAGGGAAACGAGTTGCATATCAGTTGGGGAAAAGTTGGCACCAAAGGCAAA
GCCAGATAAAAAGTTTTTCAGATGCTGC
GGCAGCGGCAAAAGCGGAGCCCGACCTCGAGGGGGGGCCCGGTACCCGGCCGGACGTCTCTAG
A
GCTAGC
Figure S4.
Alkyne functionalized fluorescent dyes for visualization of Anl labeled proteins.
Alkyne
functionalized tetramethylrhodamine (TAMRA) or Alexa Fluor 488 dyes are appended to
Anl
residues by copper catalyzed click reaction.
Figure S5.
HeLa cell morphology after labeling and infection by various
Y. enterocolitica
strains.
HeLa
cells were incubated with
Y. enterocolitica
at multiplicity of infection of 100 in the presence of 1
mM Met or 1 mM Anl. Infections with secretion wild type
Yersinia
resulted in a rounded
morphology, whereas infections with secretion mutant
Yersinia
did not change HeLa cell
morphology.
Magnified ve
rsions are shown in the insets to the right of each image.
Figure S6.
Identifying infection conditions to i
ncreas
e
efficiency of Yop injection into HeLa cells.
Optimal
conditions for Anl labeling of injected Yops were identified by investigating Anl concentrations
of 0.5, 1 and 2 mM, as well as the timing of T3SS activation.
A
.
Anl incorporation was detected
by labeling with TAMRA alkyne dye and in
-
gel fluoresce
nce detection. Activation of T3SS
machinery was induced by pre
-
incubation of
Yersinia
at 37 °C for various times (0
-
60 min)
before the start of infection. The labeling control (HeLa only) did not include any
Yersinia
cells.
A pre
-
incubation step at 37 °C f
or 120 min was included to determine if longer pre
-
incubation
would result in more labeling of injected proteins. Higher Anl concentrations consistently
yielded higher levels of labeling with TAMRA dye, and longer pre
-
incubation at 37 °C resulted in
more e
fficient injection. Secretion of Yops under LCR conditions (M) is included as a basis of
comparison for the expected position of Yops
B
.
Colloidal blue staining of the same gels shows
equal protein loading across different conditions. Pre
-
incubation of
Yer
sinia
at 37 °C increases
the efficiency of injection of labeled T3SS substrates.
C.
T3SS secretion competent
Y.
enterocolitica
were pre
-
incubated at 37 °C for 1
-
5 hours before start of infection of HeLa cells.
Infections were carried out in media containin
g 0.5, 1, 2 or 4 mM Anl.
D.
Colloidal blue dye
staining of the same samples shows equal loading of HeLa lysates.
H: HeLa lysates after
infection with secretion WT
Yersinia
. M: Media secreted Yops under LCR conditions.
Positions
corresponding to different
Yops are
indicated to the right of the image.
Figure S7.
Selective lysis of HeLa cells in the presence of intact
Y. enterocolitica.
Digitonin was used as a
detergent to disrupt the HeLa cell membrane but not the bacterial envelope. Co
-
cultures of
HeLa and
Yersinia
cells were lysed with various concentrations of digitonin in PBS. Briefly, 6
-
well
plates containing 1.5 x 10
6
HeLa cells and
Yersinia
at MOI of 100 were lysed for 20 min at room
temperature. After lysis of HeLa cells,
Yersinia
cells were removed by centrifugation at 15000
rcf and the remaining supernatant proteins were precipitated with chloroform/methanol. The
pr
oteins were resu
s
pended in 1% SDS and electrophoresed on a 12% polyacrylamide gel and
stained with colloidal blue. Efficient lysis of HeLa cells was observed with digitonin
concentrations as low as 0.025% v/w. In the absence of HeLa cells and presence of
Y
ersinia
, no
protein could be observed after incubation with digitonin, indicating lack of
Yersinia
lysis.
Western blot analysis with antibody specific for
Yersinia
RpoA confirmed the lack of bacterial
lysis. A small amount of non
-
specific binding was obser
ved with this antibody under all
conditions, including those with no bacteria.
Yersinia
lysis with 2% SDS served as a positive
control for RpoA labeling. Western blot analysis with an antibody specific to α
-
tubulin was used
as a loading control to show the
extent of HeLa cell lysis. Western blots probed with α
-
tubulin
antibody confirmed that 0.05% digitonin is sufficient for HeLa cell lysis. Digitonin concentrations
are given as weight per volume of PBS.
Figure S8.
C
ontrol
conditions
for
detection
of
labeled
proteins in absence of Anl incorporation.
The
images in this section are labeling controls for the i
nfection studies shown in Fig. 3
D,
E
of the
main text. The images were obtained from infections in 1 mM Anl with secretion wild type
Y.
enterocolitica
lacking the NLL
-
MetRS. Anl residues were labeled with alkyne
-
functionalized
AlexaFluor 488; the resulting fluorescence is shown in green. AlexaFluor 633 conjugated to
wheat germ agglutinin (WGA) labels membranes of HeLa cells; the correspond
ing fluorescence
is shown in red. In the absence of NLL
-
MetRS, Anl is not incorporated into
Yersinia
proteins;
therefore there is no labeling with alkyne
-
functionalized AlexaFluor 488.
The wild
-
type
MetRS
does not activate Anl to any appreciable
extent
;
however the NLL
-
MetRS
does activate Met. We have previously measured the kinetics
of amino acid activation by
NLL
-
MetRS
;
k
cat
/K
m
is 350±70 M
-
1
s
-
1
for Met
and 410±80 M
-
1
s
-
1
for Anl
.
See Ref. 6 of the main text
for measure
ment
of
the
catalytic efficiency of MetRS variants.
As shown in
F
igure 3
A
lane 5
of
the main text
,
labeling in absence of NLL
-
MetRS is negligible
.
This is also evident in the images
above
as bacteria lacking the NLL
-
MetRS are not labeled
, even in the presence of Anl
.
The
Optimem medi
um
contains
about 100
μM Met
;
infections
we
re performed in 1
mM Anl,
Figure S9.
Directed mass spectrometry search for gel
-
extracted peptides derived from Yops secreted
under LCR conditions.
A directed mass spectrometric search was used to determine the Met
positions occupied by Anl in each Yop. Yops were secreted under LCR conditions in the
presence of 1 mM Anl in 19
-
amino acid M9 medium supplemented with 25 μM Met and lacking
calcium chlori
de. Secreted proteins were precipitated by chloroform/methanol treatment,
resuspended in 1% SDS in PBS, electrophoresed on a 12% bis
-
tris acrylamide gel and stained
with colloidal blue. The resulting bands were cut out, and in
-
gel tryptic digestion was
per
formed for each gel fragment. To help identify peptides corresponding to each Yop, an m/z
inclusion list for expected peptide ions was generated taking into account
up to two
missed
cleavages by trypsin as well as fixed carbamidomethyl modification of cyst
eine and variable
oxidation of Met.
R
eplacement of Met by Anl was included as a variable modification. This
inclusion list was used to perform directed LC
-
MS/MS experiments on an LTQ
-
FT Ultra mass
spectrometer (Thermo Fisher Scientific) equipped with a nan
oelectrospray ion source (Thermo
Fisher Scientific) connected to an EASY
-
nLC II (Thermo Fisher Scientific).
A.
Image of t
he
cut out
gel fragments with the total number of identified spectra
shown for
each fragment. The corresponding Yop positions are indicated next to the image
.
B.
The total
number of Met positions
based on the protein sequence
for each Yop
and the molecular weight
is listed.
C.
The number of identified spectra for each Yop
is shown for each extracted gel
section, the order of the gel slices correspond to decreasing molecular weight of protein.
Figure S10.
Detection of r
esidue
-
specific incorporation of Anl in Yops.
Directed mass spectrometry was
performed based on an expected peptide ion list (see
Fig.
S
9
for details) from gel
-
extracted
Yops
from proteins secreted under LCR conditions
An example of this analysis for YopH is
illustrated here
showing the sequence cover
age obtained in presence or absence of Anl labeling
A
.
Highlighted
sequences
depict coverage by LC
-
MS/MS; modified Met residues are highlighted
in green
indicating oxidation of methionine in the “+Methionine” condition and incorporation
of Anl in the “+A
zidonorleucine” condition
. Small differences were observed between the
sequence coverages obtained for Met
-
and Anl
-
treated samples.
B
.
The Met positions at which
Anl was observed are shown for YopH
as a function of residue position along the sequence
.
Black
boxes are positions at which Anl was observed, and the frequencies at which Met and Anl
were observed are
shown
for each position.
C.
A
representative spectrum from an LC
-
MS/MS
analysis on a precursor ion of the M
+23
TQQVGLGDGIM
+16(ox)
ADM
+23
YTLTIR pep
tide is shown,
with y
-
ions shown in blue and b
-
ions shown in red. A mass increment of 16 amu indicates that
the Met at position 328 is oxidized. Anl is incorporated at positions 307 and 331, as indicated by
mass shifts of 23 amu each. This mass shift corre
sponds to replacement of Met by Anl (
Fig S2
).
A similar analysis was performed for each Yop.
Figure S11.
Identifying the pattern of r
esidue
-
specific replacement of methionine by azidonorleucine in
Yops
as a function of protein molecular weight and numbe
r of methionines
.
Directed mass
spectrometry was used to determine the total number of Met positions at which Anl
incorporation was observed.
A.
The number of Met positions as a function of molecular weight
for each Yop is plotted
to show the number of
methionines contained in sequence of each Yop
.
B.
The experimentally observed number of Anl residues (replacement of methionine) for each
Yop as identified by MS
.
C
.
The sequence coverage observed
by MS
for each Yop
.
Sequence
coverage
is represented in yel
low, and the position of each Met is represented as a circle.
Black
circles represent detection of Anl at a given Met position. These results were obtained from
secretion of Yops under LCR conditions. The extent of labeling and observed sequence coverage
m
ay differ from those of Yops injected in HeLa cell infection.
Figure S12.
Detection of biotin after click reaction of alkyne
-
biotin probe to proteins containing Anl
.
HeLa
infections and secretion of Yops under LCR conditions were performed as
described in the
methods section of the main text.
Proteins were treated with the biotin
-
alkyne affinity tag
(structure
4
in the main text)
and biotinylation was detected by western blot
with a streptavidin
Alexa
fluor
-
633
conjugate antibody. Image bel
ow s
hows biotin labeling in injected HeLa cell
lysates as well as proteins secreted under LCR conditions
. Biotinylated
horseradish peroxidase
(HRP)
was used as positive control for labeling.
Figure S13.
Affinity enrichment of
injected
Y. enterocolitica
proteins
from infected HeLa cell lysates
.
A
.
Infected HeLa cell lysates were treated with probe
4
and affinity purified on a streptavidin
column. Flow
-
through (FT) and elution (EL) fractions were stained with colloidal blue.
B
.
Western blot detection of Yo
pE and mammalian α
-
tubulin in the same fractions
as those
represented in
A
.
YopE was enriched from HeLa
infections with
T3SS
-
Wt
Yersinia
. Lack of α
-
tubulin
in the elution fractions
confirmed efficient removal of HeLa cell proteins.
Figure S14.
Comparison of
MS/MS spectra from Yops secreted under LCR conditions
and
Yops injected
into HeLa cells.
A.
Tandem mass
spectr
um
corresponding to a peptide from
YopD from
proteins
secreted under
LCR conditions
.
B.
Tandem mass spectrum
identified
from
infected
HeLa cell
lysate
s and corresponding to the same peptide sequence as in
A
.
Both samples were gel
extracted and desalted on C18 ziptip columns. A doubly charged m/z 701.86 amu peptide ion
corresponding to the sequence “EKEVNASIAANEK” gave a precurso
r mass of 1401.71
(theoretical value 1401.7172). The y
-
and b
-
ions are color coded blue and red, respectively,
along with the matching peptide sequence on top of each spectrum. Spectra were obtained on
a
nanoL
C
-
Thermo Fisher Scientific LTQ
-
FT spectrometer
with the peptide mass range set to 200
-
1700 amu.