Cell News 01/2017
9
NIKON YOUNG SCIENTIST AWARD 2017
ubiquitinome upon
Salmonella
infection, we established a quan-
titative proteomics platform for the detection of endogenous
ubiquitinated proteins (Fig. 1a). Importantly, we combined the
use of different workflows to characterize ubiquitinated sub-
strates both on the level of their modified lysine residues as well
as attached polyubiquitin chains. To globally map ubiquitination
sites, we performed stable isotope of amino acids in cell culture
(SILAC)-coupled diGly proteomics experiments. Developed in
recent years this methodology is based on the immunoprecipita-
tion of diGly-remnant containing peptides, which directly result
from tryptic digestion of ubiquitinated proteins (Ordureau et al.,
2015). Using this approach, we were able to map multiple thou-
sand novel ubiquitination events derived from both host and
pathogen including previously uncharacterized modifications of
secreted bacterial effectors and outer-membrane proteins. Unbi-
ased gene ontology enrichment analysis of this data set revealed
several significantly over-represented biological processes
including "regulation of actin cytoskeleton" and "regulation of
apoptosis". These categories nicely corroborate previous findings
that
Salmonella
infection impacts inflammation and cell archi-
tecture respectively (Fig. 1b).
The host inflammatory response to
Salmonella
requires the formation of linear ubiquitin chains
Observed changes in the ubiquitination patterns revealed a
number of important insights including the identity of multiple
pathogen-stimulated inflammatory pathways establishing a role
for linear ubiquitin chains in bacteria triggered inflammation.
Salmonella
infection as such is a complex stimulus expected
to trigger multiple different inflammatory pathways unlike the
treatment of cells with a single cytokine.
Salmonella
is there-
fore able to stimulate a variety of both extra- and intracellular
pattern recognition receptors via its surface antigens. Moreover,
live
Salmonella
impacts cellular inflammatory signaling addi-
tionally via the secretion of metabolites and the translocation
of virulence factors (Gaudet et al., 2015).
Salmonella
-induced
gastroenteritis for example requires several virulence factors,
which contribute to the induction of inflammation (Bruno et
al., 2009). The analysis of the ubiquitinome upon
Salmonel-
la
infection of epithelial cells revealed the clear presence of
an inflammation signature within the data (Fig. 1b). Robustly
increased ubiquitination was observed for multiple regulators of
the NF-
K
B pathway. Importantly, we detected highly increased
ubiquitination sites throughout all subunits of LUBAC (RNF31/
HOIP, HOIL, SHARPIN). These data suggested that LUBAC activity
might be stimulated upon
Salmonella
infection. In order to test
this hypothesis, we immunoprecipitated linear polyubiquitinated
proteins upon exposure of cells to
Salmonella
under denaturing
conditions using a previously characterized M1-Ub specific anti-
body (Matsumoto et al., 2012). Our experiments revealed an in-
crease in linear ubiquitin chains within 15 minutes of infection
implying that LUBAC is stimulated upon bacterial exposure (Fig.
2a,b). Moreover,
Salmonella
-driven M1-ubiquitination was ac-
companied with the downstream activation of NF-
K
B signaling.
As previous reports indicated that
Salmonella
effector proteins
SopE, SopE2 and SopB contribute to bacterial NF-
K
B stimulation
(Bruno et al., 2009), we tested whether these effectors have
an impact on LUBAC activity. Indeed, a triple knockout strain
deficient in SopE/SopE2/SopB demonstrated an impaired ability
to induce both linear ubiquitination as well as downstream NF-
K
B activation (Fig. 2b). Notably, treatment of cells with the HOIP
inhibitor Gliotoxin significantly impaired the
Salmonella
induced
pro-inflammatory response supporting the notion that LUBAC
activity is required for these events (Fig. 2c). Using linear Ub
specific immunoprecipitation in combination with SILAC based
quantitative proteomics we were able to identify the complete
set of
Salmonella
-induced LUBAC substrates (Fig. 2d). These
included proteins involved in multiple NF-
K
B activating path-
ways. We identified
Salmonella
induced linear ubiquitination of
IRAK1 and IRAK4, events previously linked to the activation of
TLR signaling (Emmerich et al., 2013). Infection also induced the
M1-ubiquitination of RIPK2 indicative of NOD stimulation (Fiil
et al., 2013). These findings are consistent with the ability of
bacterial products to stimulate these PRRs, as LPS and Flagellin
can stimulate TLR4 and TLR5 respectively and bacterial peptido-
glycan is sensed via NOD1/2.
Bacterial manipulation of the host ubiquitin system
- discovering host targets of the
Salmonella
E3 ligase
SopA
Although prokaryotes lack the canonical Ub-proteasome system,
a wide range of bacterial pathogens acquired strategies to
exploit the host Ub machinery in order to support their own life
cycle and to counteract host immune defense programs. One
example is the injection of bacterial encoded E3 ligases into the
host cytosol to modify specific targets and to trigger diverse
cellular responses (Maculins et al., 2016). Work over the past
years has identified a variety of bacterial E3 ligase effectors and
began to elucidate their impact on host proteins and signaling
cascades. Among these, the ligase effector SopA from
Salmo-
nella
typhimurium constitutes a very striking example of so
called molecular mimicry. Despite the lack of obvious sequence
homology to eukaryotic HECT E3s, SopA adopts a HECT-like
architecture consisting of N- and C-lobe connected by a flexible
helical element (Diao et al., 2008). While these previous studies
have characterized the biochemical activity of SopA and im-
plicated it in the induction of the enteritis phenotype, the host
targets and molecular functions of this virulence factor have
remained unknown. In our studies, we established the use of
ubiquitin proteomics for the identification of SopA substrates in
pathogen-infected cells (Fig. 3a). These experiments interesting-
ly revealed human TRIM56 and TRIM65 as
bona fide
substrates
