The serine protease of pic mediates a dose-dependent mucolytic activity after binding to sugar constituents of the mucin substrate
The serine protease motif of Pic mediates a dose-dependent mucolytic activity
after binding to sugar constituents of the mucin substrate
Javier Gutie´rrez-Jime´nez, Ivonne Arciniega, Fernando Navarro-Garcı´a*
Department of Cell Biology, Centro de Investigacio´n y de Estudios Avanzados (CINVESTAV-IPN), Zacatenco, Ap. Postal 14-740, 07000 Me´xico, DF, Mexico
a r t i c l e i n f o
Received 23 January 2008
Received in revised form 14 April 2008
Accepted 18 April 2008
Available online 2 May 2008
a b s t r a c t
The pic gene is harbored on the chromosomes of three important pathogens: enteroaggregative
Escherichia coli (EAEC), uropathogenic E. coli (UPEC), and Shigella ﬂexneri. Since Pic is secreted into the
intestinal lumen during EAEC infection, we sought to identify intestinal–mucosal substrates for Pic. Pic
did not damage epithelial cells, cleave fodrin, or degrade host defense proteins embedded in the mucus
layer (sIgA, lactoferrin and lysozyme). However, by using a solid–phase assay to evaluate the mucinolytic
activity of EAEC Pic, we documented a speciﬁc, dose-dependent mucinolytic activity. A serine protease
inhibitor and an enzymatically inactive variant of Pic were used to show that the Pic serine protease
motif is required for mucinolytic activity. Pic binds mucin, and this binding was blocked in competition
assays using monosaccharide constituents of the oligosaccharide side chains of mucin. Moreover, Pic
mucinolytic activity decreased when sialic acid was removed from mucin. Thus, Pic is a mucinase with
lectin-like activity that can be related to its reported hemagglutinin activity. Our results suggest that
EAEC may secrete Pic into the intestinal lumen as a strategy for penetrating the gel-like mucus layer
during EAEC colonization.
Ó 2008 Elsevier Ltd. All rights reserved.
Enteroaggregative Escherichia coli (EAEC) is associated with
persistent pediatric diarrhea and belong to diarrheagenic patho-
types recognized in E. coli . This pathotype is deﬁned by its
aggregating pattern of adherence to HEp-2 cells and shows two
prominent pathogenic features: (i) formation of a thick mucus
blanket on the intestinal mucosa and (ii) mucosal damage by
cytotoxins [2,3]. Another recognized enteric bacterial pathogen is
Shigella ﬂexneri, an intracellular pathogen that causes bacillary
dysentery by invasion and tissue damage of colonic epithelium. The
stools are generally watery, purulent, bloody, and mucoid .
These enteric pathogens share the pic/set chromosomal locus,
which in S. ﬂexneri is found on the she pathogenicity island . This
locus encodes a 109 kDa protein named Pic (Protein involved in
colonization), and a protein homologue named PicU has recently
been identiﬁed in the CFT073 strain of uropathogenic E. coli (UPEC)
. Pic is one of two autotransporter proteins secreted by EAEC that
belong to the serine protease autotransporter of Enterobacteriaceae
(SPATE) family . The full contribution of these proteins to
pathogenesis remains elusive: no universal contribution has been
suggested and different activities have been reported for various
members of the SPATE family [8,9].
Unlike plasmid-encoded toxin (Pet), the other autotransporter
secreted by EAEC, Pic has no enterotoxic or cytotoxic activities
[8,10]. However, the cytotoxic effect of Pic has only been tested on
HEp-2 cells with a single concentration of toxin . Several
biological roles for Pic have been reported, including hemaggluti-
nation, degradation of complement protein, and mucinolytic
activity against bovine submaxillary gland mucin (BSM) and mouse
intestinal mucus . PicU has also shown an in vitro mucinolytic
activity against BSM . However, another group was unable to
identify a mucinase activity associated with PicU or the EAEC
homologue . Thus, the biological role of Pic in these pathogens
is unknown. However, the possibility of a mucinolytic activity for
Pic suggests an effective strategy for penetrating the intestinal
mucus layer and/or obtaining a carbon source for the metabolic
process. In this work, we further studied this possibility. We tested
different intestinal mucosa substrates for Pic and using a solid–
phase assay to evaluate mucinolytic activity of Pic, we found
a speciﬁc, dose-dependent mucinolytic activity for EAEC Pic. This
activity depended on its serine protease motif and on Pic binding to
monosaccharide constituents of the oligosaccharide side chains of
* Corresponding author. Tel.: þ52 5 5747 3990; fax: þ52 5 5747 3393.
E-mail address: firstname.lastname@example.org (F. Navarro-Garcı´a).
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/micpath
0882-4010/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved.
Microbial Pathogenesis 45 (2008) 115–123
2.1. Interaction of Pic with epithelial cells
Some SPATE members (i.e. Pet, EspC, Sat, and EspP) display
cytotoxic activities against mammalian cells [3,8,9,12]. To
determine whether Pic is cytotoxic to intestinal epithelial cells,
HT29-C1 cells were incubated with various doses of Pic for up to
12 h. Potential cytotoxic effects were then detected by Giemsa
staining to visualize the morphology of intoxicated cells. As a neg-
ative control, cells were treated with concentrated supernatants
from non-toxigenic HB101 (pACYC184). As a positive control, cells
were treated for 4 h with culture supernatant containing Pet pro-
tein at 37 mg/ml (355 nM). Cells in the negative control exhibited
a normal cell morphology (Fig. 1A), whereas cell rounding and
detachment from the glass substratum was observed for cells in the
positive control (Fig. 1B). HT29-C1 cells treated with culture
supernatant containing 364 nM Pic exhibited normal cell mor-
phology, indicating that Pic did not cause a cytotoxic effect (Fig.1C).
Since cytotoxicity could be dose dependent, HT29-C1 cells were
incubated with 546, 728, 910, 1366 and 1821 nM of Pic for 12 h at
37 C (Fig. 1D-F). None of these doses produced a cytotoxic effect.
Pic was also unable to produce a cytotoxic effect in another epi-
thelial cell line, HEp-2 cells (data not shown).
Confocal microscopy was used to further study possible
interactions between Pic and epithelial cells. After incubation with
910 nM Pic, HEp-2 cells were immunostained with anti-Pic anti-
bodies and rhodamine-labeled phalloidin to visualize the actin
cytoskeleton. HEp-2 cells were treated with 355 nM Pet protein as
a positive control in parallel experiments. Consistent with previous
results [3,13], Pet-treated cells showed cytoskeleton contraction
(Fig. 1G), loss of actin stress ﬁbers (Fig. 1G), and an intracellular
distribution of Pet (Fig. 1H-I). In contrast, a higher concentration of
Pic (910 nM) was unable to produce any of the Pet-related effects:
Pic did not damage the actin cytoskeleton (Fig. 1J) and was not
present in HEp-2 cells (Fig. 1K). We were also unable to detect Pic
binding to the cell membrane (Fig. 1L). Identical results were
obtained with an even higher 1.8 mM concentration of Pic (data not
Fodrin cleavage is involved with the cytotoxic effects elicited by
Pet and EspC. To explore whether Pic is able to cleave fodrin, we
performed in vitro degradation experiments using a recombinant,
109 kDa GST–fodrin . Previous work has shown that GST–fodrin
is a substrate of Pet and EspC [12,15]. GST–fodrin was exposed to Pic
at several times (0, 30 and 180 min) and then resolved by SDS-
PAGE. At time zero, fodrin and Pic were detected as a band of
109 kDa: both proteins have the same molecular weights and run at
the same position in SDS-PAGE gels (Fig. 2A). After 30 and 180 min
of incubation, no breakdown products of GST–fodrin were detected
(Fig. 2A). The band from time zero which corresponds to GST–
fodrin and Pic was identical to the bands from samples taken after
30 and 180 min of incubation. The same result, indicating no Pic-
mediated cleavage of GST–fodrin, was obtained after a 6 h
incubation (data not shown). In contrast, Pet and EspC were able to
cleave GST–fodrin in 60 min to generate different breakdown
products on SDS-PAGE (Fig. 2A). Pet and EspC were also able to bind
fodrin as determined by overlay analyses, whereas Pic was unable
to bind fodrin in the same assay (Fig. 2B). All these data suggest that
Pic, unlike Pet and EspC [3,12], is unable to interact with and
damage epithelial cells of the intestinal mucosa.
2.2. The serine protease motif of Pic is involved
in the dose-dependent mucinolytic activity
High molecular weight molecules such as mucin (2 Â 106
are not resolved by SDS-PAGE because they are unable to enter the
gel. Thus, as an alternative method to analyze the interaction of Pic
with mucin, we employed a colorimetric technique  and stan-
dardized the degradation assay by using commercial BSM and
Proteinase K (PK). BSM was coupled to biotin, adjusted to different
concentrations (0.1, 0.2, 0.4 and 0.8 mg/ml), adsorbed to microtiter
plates, and exposed to 1 or 10 mg of PK per well. BSM-coated wells
were incubated with PBS as a negative control. The assay showed
that BSM was efﬁciently degraded by 1 mg of PK since this protease
treatment dramatically reduced the mucin-dependent absorbance
of the treated wells (Fig. 3A). Reduction of the mucin-dependent
absorbance was related to PK concentration, as 10 mg of PK reduced
absorbance values more than 1 mg of PK (Fig. 3A).
To evaluate the mucinolytic activity of Pic, we used commercial
hog stomach mucin (HSM) as well as BSM. Both substrates were
coupled to biotin and adsorbed to microtiter plates as mentioned
above. Degradation assays showed that Pic at 1 or 10 mg per well
was able to degrade HSM in a dose-dependent fashion (Fig. 3B).
Thus, 1 mg of Pic degraded around 32% of the HSM sample, while
10 mg of Pic degraded around 56% of the HSM. Similar results were
obtained when BSM was used as the substrate (data not shown).
These data indicate that Pic is able to degrade mucin in a quanti-
tative, dose-dependent manner.
Since the serine protease motif harbored by SPATE proteins is
crucial for their biological activities , we analyzed the role of this
motif on the mucinolytic activity of Pic. Pic was preincubated for
30 min with the serine protease inhibitor phenylmethylsulphonyl
ﬂuoride (PMSF) and then added at 1 or 10 mg per well to HSM-
coated plates. Absorbance values obtained with PMSF-treated Pic
were similar to those obtained when HSM was incubated with PBS
(Fig. 3C). This demonstrated that PMSF-treated Pic was unable to
degrade HSM at either 1 or 10 mg of Pic per well. To further assess
the role of the serine protease motif on the mucinolytic activity of
Pic, site-directed mutagenesis was used to convert its predicted
catalytic serine residue (S258) to isoleucine. This mutated protein,
Pic S258I, was used at 1 or 10 mg per well in the HSM degradation
assay. As with PMSF-treated Pic, absorbance values obtained with
Pic S258I were similar to those values obtained when HSM was
incubated with PBS (Fig. 3D). These data demonstrate that the Pic
serine protease motif is crucial to cleave mucin, since the muci-
nolytic activity is abolished by blocking this motif with an inhibitor
or by a mutation.
2.3. Interaction of Pic with proteins immersed in the mucus layer
We analyzed whether Pic displays proteolytic activity against
proteins embedded in the intestinal mucus such as lysozyme,
lactoferrin and secretory IgA (sIgA). Each potential substrate was
incubated with Pic for 0, 0.5, 3, or 4 h before degradation products
were analyzed by SDS-PAGE. Pic and lysozyme were resolved at
time zero as full-length proteins of 109 and 14 kDa, respectively.
Lysozyme was also resolved as a full-length protein without deg-
radation products after 0.5, 3, and 4 h of incubation, indicating
that Pic is unable to cleave lysozyme in vitro (Fig. 4A). When
apolactoferrin was incubated with Pic, SDS-PAGE analysis showed
at time zero two bands of 109 and 80 kDa for Pic and apolacto-
ferrin, respectively. Both protein bands were preserved without
degradation products after 0.5, 3, and 4 h of incubation, demon-
strating that Pic was also unable to cleave apolactoferrin (Fig. 4B).
Since sIgA is a heteromeric protein, its interaction with Pic was
analyzed by SDS-PAGE under reducing and non-reducing condi-
tions. In both cases, Pic and sIgA proteins were incubated for 1, 4
and 8 h as previously described . Under non-reducing condi-
tions (no 2-b-mercaptoethanol and heating), SDS-PAGE showed
Pic as a full-length protein band of 109 kDa while sIgA was
observed as a high molecular weight protein (around 400 kDa)
which hardly entered into the separating gel. This proﬁle was
J. Gutie´rrez-Jime´nez et al. / Microbial Pathogenesis 45 (2008) 115–123116
similar after 1, 4, and 8 h of incubation, indicating no degradation
of sIgA by Pic (Fig. 4C). Under reducing conditions, SDS-PAGE
showed Pic as a single band of 109 kDa and sIgA as three bands
representing the secretory component (70 kDa), the heavy chains
(52 kDa), and the light chains (23 kDa). None of these sIgA bands
were degraded by Pic at any of the incubation times tested
(Fig. 4D). All these data indicate that Pic is unable to cleave pro-
teins embedded in the mucus layer such as lysozyme, lactoferrin
and sIgA. Furthermore, the data demonstrate that there is
substrate speciﬁcity for the serine protease motif of Pic.
2.4. Role of monosaccharide components of mucin
in the interaction with Pic
The ability of Pic to bind to bovine mucin was examined in a solid-
phase assay using mucin-coated microtiter wells . GST-coated
Fig. 1. Pic does not cause cytotoxicity and is not internalized into cells. (A–F) Pic does not induce cytotoxic effects. HT29-C1 cells were incubated with different doses of Pic, 355 nM
Pet, or left without treatment. After incubation, HT29-C1 cells were ﬁxed, permeabilized and stained with Giemsa. Untreated cells (A). Cells treated with Pet as a positive control (B).
Cells treated with puriﬁed Pic at 364 (C), 728 (D), 1366 (E) and 1821 nM (F). (G–L) Pic is not internalized into epithelial cells. HEp-2 cells were incubated with Pic (910 nM), Pet
(355 nM) proteins or left without treatment. After incubation, the cells were ﬁxed, permeabilized, and stained with rhodamine–phalloidin (red; G,J) and anti-Pic (K) or anti-Pet (H)
antibodies followed by a secondary ﬂuorescein-labeled goat anti-rabbit antibody (green). Slides were observed under confocal microscopy. (G–I) Pet-treated cells. (J–L) Pic-treated
cells. Merged images are shown in panels I and L. (For interpretation of the references to colour in this ﬁgure legend, the reader is referred to the web version of this article.)
J. Gutie´rrez-Jime´nez et al. / Microbial Pathogenesis 45 (2008) 115–123 117
microtiter wells were used as a negative control. The binding of Pic to
bovine mucin was higher than that to GST, as revealed by anti-Pic or
anti-GST antibodies, respectively (Fig. 5A). To determine if any of the
individual sugar components of mucin were important for Pic
binding, the solid-phase assay was repeated with three mono-
saccharide constituents of the oligosaccharide side chains of mucin:
N-acetylglucosamine (GlcNAc), N-acetylgalactosamine (GalNAc), and
N-acetylneuraminic acid (sialic acid, NANA). Pic bound to all the three
Fig. 2. Pic does not interact with fodrin. (A) Degradation assay. Puriﬁed GST–fodrin (1 mg) was incubated with 1 mg Pic, Pet or EspC proteins at different times and then separated by
SDS-PAGE. The degradation products are indicated with a closed circle for Pet and with an asterisk for EspC. (B) Overlay assay. Pic (4 mg), Pet (4 mg), or EspC (4 mg) were separated by
SDS-PAGE and transferred to nitrocellulose membranes. The membrane containing these proteins was incubated with GST–fodrin (5 mg/ml) and the afﬁnity reaction was revealed
by using anti-GST antibodies and AP-labeled anti-mouse antibodies.
Fig. 3. Pic mucinolytic activity is a dose- and serine protease motif-dependent event. Microplate wells were coated with biotin-labeled mucin, and then treated for 4 h at 37 C with
1 or 10 mg per well of either Proteinase K used as positive control (A), Pic (B), Pic pretreated for 30 min with the serine protease inhibitor (PMSF) (C), or the Pic S258I serine protease
motif mutant (D). After incubation, the undigested labeled mucin was detected with streptavidin–peroxidase. Enzymatic activity was determined with different concentrations of
HSM by plotting absorbance readings on a graph generated from a standard curve of labeled mucin in PBS.
J. Gutie´rrez-Jime´nez et al. / Microbial Pathogenesis 45 (2008) 115–123118
monosaccharides, showing a higher afﬁnity for GalNAc than for
GlcNAc or sialic acid. However, Pic had a higher afﬁnity for mucin
than for any of the individual monosaccharides (Fig. 5A).
To further analyze the interaction of Pic with these mono-
saccharides, GlcNAc, GalNAc and sialic acid were used in inhibition
experiments. Pretreating Pic with GlcNAc reduced Pic binding to
bovine mucin by 70% compared to the untreated Pic control
(Fig. 5B), indicating that GlcNAc is a mucin constituent recognized
by Pic. GalNAc decreased Pic binding to mucin by 34% compared to
the untreated control. However, when Pic was pretreated with
sialic acid, the percentage of Pic binding to mucin increased by 35%
compared to the untreated control (Fig. 5B). This suggested that
sialic acid may act as a bridge between Pic and mucin, thereby
enhancing Pic binding. Interestingly, when Pic was preincubated
with mucin and then added to monosaccharide-coated wells,
mucin was able to block Pic binding to all three sugars with equal
and essentially complete efﬁciency (around 97% inhibition as
compared with the untreated Pic) (Fig. 5C).
Our data suggested that Pic binding to the sugar residues of
mucin could be important for Pic-mediated degradation of the
protein component of mucin. To test this hypothesis, we used
neuraminidase to remove sialic acid from BSM. A colorimetric
assay  demonstrated that the amount of sialic acid in com-
mercial BSM decreased from 0.602 mM to 0.419 mM after treat-
ment with 0.06 U/ml of neuraminidase. Both untreated and
neuraminidase-treated mucins were coupled to biotin for use in
the solid-phase degradation assay. Pic (9 nM) degraded both
mucins at concentrations of 0.1, 0.2, 0.4 and 0.8 mg mucin/ml
(Fig. 5D). However, we observed a signiﬁcant decrease of Pic
activity against neuraminidase-treated mucin in comparison with
untreated mucin (Fig. 5D). The maximum difference was
observed at 0.2 mg/ml of mucin, where Pic degraded 50% of the
untreated mucin versus 17% of the neuraminidase-treated mucin
(Fig. 5D). These data indicate that the presence of sugars on
mucin constitutes an important factor for Pic binding and its
Fig. 4. Pic is unable to cleave lysozyme, apolactoferrin or sIgA. Puriﬁed Pic protein (1 mg) was incubated with 1 mg lysozyme, apolactoferrin, or sIgA at different times and then
separated by SDS-PAGE at 12%, 10%, and 8%, respectively. (A) Pic–lysozyme interaction for 0.5, 3, or 4 h. (B) Pic–lactoferrin interaction for 0.5, 3, or 4 h. (C-D) Pic–sIgA interaction for
1, 4, or 8 h; gels (8% and 12%) were run under non-reducing (C) and reducing (D) conditions.
J. Gutie´rrez-Jime´nez et al. / Microbial Pathogenesis 45 (2008) 115–123 119
Epidemiological studies have shown that the pic gene is associ-
ated with diarrheogenic strains of EAEC [20,21]. Since the Pic protein
is secreted into the intestinal lumen during EAEC infection, the aim of
this work was to identify intestinal mucosa substrates for Pic. Here,
we show that Pic does not directly damage epithelial cells. Pic does
not cleave the actin-binding protein fodrin, and it does not cleave
host defense proteins such as sIgA, lactoferrin, or lysozyme that are
embedded in the mucus layer. Instead, Pic binds and cleaves mucin.
Pic binds to the monosaccharide constituents of the mucin oligo-
saccharide side chains, and its serine protease motif is required for
mucin cleavage. Secretion of a thick mucus layer over the intestinal
epithelium is a hallmark of EAEC infection; Pic mucinolytic activity
could explain why the thick mucus does not inhibit EAEC infection.
Dutta et al. previously showed that 20 mg/ml of Pic is not cyto-
toxic to HEp-2 cells . Here, toxin dose response experiments
performed after 4–12 h of intoxication conclusively demonstrated
that Pic is not cytotoxic to HT29-C1 cells, an intestinal epithelial cell
line. Confocal microscopy further showed that, at even highest dose
and incubation time used, Pic (i) does not bind to the plasma
membrane of epithelial cells; (ii) is not internalized into epithelial
cells; and (iii) does not produce cytoskeletal damage in epithelial
cells. Unlike the SPATEs Pet and EspC, Pic did not bind and cleave
fodrin. This observation supports the data reported by Dutta et al.
 but is contrary to the data reported by Parham et al. . The
discrepancies may be attributed to different sources of fodrin/
spectrin: we used a consistent source of recombinant aII spectrin,
whereas the other authors used puriﬁed erythrocyte spectrin
which can vary in substrate quality and thus produce unclear
results. Together, our data suggest that Pic is unable to interact with
and damage the epithelial cells of the intestinal mucosa.
Pic has been reported to be a serine protease with mucinase
activity that is involved with multiple aspects of EAEC pathogenesis
. We decided to further explore its capacity to degrade host
defense proteins that are embedded in the mucus layer. Pic and the
IgA1 protease autotransporter of Neisseria spp. both possess a con-
sensus serine protease motif . However, Pic was unable to
cleave sIgA, an important immunoglobulin in the mucosal immu-
nity against adherent pathogens . Pic was also unable to cleave
lactoferrin, a molecule with inhibitory effects on bioﬁlm
development, bacterial adhesion, and colonization . Further-
more, lactoferrin is a serine protease which has been reported to
proteolytically cleave and inactivate three autotransporter coloni-
zation factors: the IgA protease precursor protein; Hap, the non-
pilus adhesin ; and Aae from A. actinomycetemcomitans .
Finally, Pic was unable to cleave lysozyme, an antimicrobial protein
which speciﬁcally attacks and destroys peptidoglycan .
Fig. 5. Interaction of Pic with mucin. (A) Binding of Pic to mucin and monosaccharide constituents of mucin. Wells were coated with mucin, exposed to the indicated mono-
saccharides or GST, and then blocked with BSA. The wells were subsequently incubated with increasing concentrations of Pic for 1 h at 37 C. Pic or GST binding was detected by
using anti-Pic or anti-GST antibodies followed of a secondary antibody labeled with peroxidase. (B) Effect of monosaccharides on the binding of Pic to mucin. Pic was preincubated
with the indicated sugars for 30 min at 37 C and then added to mucin-coated wells. Pic binding was detected as indicated in panel A. (C) Effect of mucin on the binding of Pic to the
monosaccharide constituents of mucin. Pic was preincubated with mucin for 30 min at 37 C and then added to wells coated with the indicated sugars. Pic binding was detected as
indicated in the panel A. (D) Effect of removing sialic acid from mucin on the Pic mucinolytic activity. Untreated and neuraminidase-treated mucins were coupled to biotin. The
degradation assay on solid phase was performed using Pic at 1 mg per well as indicated in Fig. 3 legend.
J. Gutie´rrez-Jime´nez et al. / Microbial Pathogenesis 45 (2008) 115–123120
Studies of Pic as a mucinase have been complicated by the
technologies used to demonstrate its mucinolytic activity. In the
ﬁrst study on Pic, SDS-PAGE analysis was used to monitor Pic
mucinolytic activity against crude mouse mucus . As the crude
mucus was probably unable to enter the separating gel, the
resulting data was unclear. The use of Sephacryl S-400 column
chromatography of untreated crude mouse mucus versus mucus
treated with Pic and measuring the optical density of PAS staining
fractions was even more complicated . Another methodology
used by various authors [6–8] detects zones of mucin lysis on BSM
in agarose plates after treatment with Pic (24 h at 37 C) and
staining with amido black. Unfortunately, this is a rudimentary
technique with no quantitative measurement after long
incubations. It also produces inconsistent results. For example, with
this methodology Pic showed mucinolytic activity against BSM and
mouse crude mucin but did not show mucinolytic activity against
HSM . Heimer et al. , in contrast, were unable to identify
a mucinase activity associated with Pic from either EAEC of UPEC.
Here, we have shown that Pic is indeed a mucinase with mucino-
lytic activity against both BSM and HSM. Use of the microplate
assay with immobilized biotin-labeled mucin as substrate and a 4 h
37 C treatment with Pic allowed us to (i) clearly demonstrate the
mucinolytic activity of Pic; (ii) establish that this effect is a dose-
dependent event (from substrate or enzyme); and (iii) quantify Pic
activity (i.e. percentages of mucin degradation at different Pic
concentrations). We also demonstrated with the microplate assay
that Pic mucinolytic activity depends on its serine protease motif
since the use of a serine protease inhibitor (i.e. PMSF) or a serine
protease motif mutant (Pic S258I) completely blocked mucinolytic
activity against either BSM or HSM. The microplate assay used here
could also resolve a controversy regarding the possible mucinolytic
activity of the SPATE Tsh [11,27]. Collectively, our data indicate that
Pic is functionally similar to EpeA, a SPATE from enterohemorrhagic
E. coli which has mucinolytic activity but is not associated with
a cytopathic effect on epithelial cells .
Mucin, the principal glycoprotein of mucous secretion, plays an
important role in protection against microbial invasion because of
its heavy glycosylation and its ability to form a gel. However, enteric
pathogens have developed various strategies (motility, adhesion,
and proteolytic cleavage of mucin) for penetrating this protective
gel-like layer . Many enteric pathogens are able to bind mucin
as an initial event towards colonization , but few pathogens
express proteins that exhibit features of both a lectin and a pro-
tease. A soluble Zn-metalloprotease, hemagglutinin/protease
(HapA) , exhibits both of these features and proteolytically
degrades several physiologically important host proteins including
mucin . Here, we show that Pic is a mucinase that displays
a lectin-like activity by binding to monosaccharide constituents of
the oligosaccharide side chains of mucin. This dual function might
be related to other activities attributed to Pic, such as its ability to
inactivate complement and its role in hemagglutination.
Our collective data indicate that Pic, which is secreted by three
important pathogens (EAEC, UPEC, and S. ﬂexneri), is a mucinase
with a lectin-like activity. The lectin-like activity of Pic can be
related to its previously reported hemagglutinin activity which is
also found in its homologue Tsh from avian pathogenic E. coli [7,33].
Finally, this work suggests an important role for Pic in penetrating
the protective gel-like mucus layer during EAEC (as well as UPEC
and S. ﬂexneri) colonization.
4. Material and methods
4.1. Strains and plasmids
As previously described , the minimal Pic clone HB101(pPic1)
was used to obtain Pic protein. Supernatant proteins from the non-
toxigenic HB101 (pACYC184) were used as a negative control. The
strains were maintained on L-agar or L-broth containing 10 mg/ml
tetracycline. Clone 18531, representing bp 2531–4689 of human aII
spectrin, was cloned into the inducible bacterial expression vector
pGEX-3X. The strains were maintained on L-agar or L-broth con-
taining 100 mg/ml ampicillin. Clone 18531 was kindly provided by
Stabatch and Morrow .
4.2. Antibodies and recombinant protein preparation
Broth cultures from HB101(pPic) were incubated overnight at
37 C and then centrifuged at 7000 Â g for 15 min. The culture
supernatant was ﬁltered throughout 0.22 mm cellulose acetate
membrane ﬁlters (Corning, Cambridge, MA), concentrated 100-fold
in an ultrafree centrifugal ﬁlter device with a 100 kDa cut-off
(Millipore, Bedford, MA), ﬁlter-sterilized again and stored at À20 C
for up to 3 months.
Rabbit anti-Pic polyclonal antibodies were elicited by excising
proteins from polyacrylamide gels and injecting the gel slices
subcutaneously into rabbits in two doses, 2 weeks apart. The
antibody responses and speciﬁcity were determined by immuno-
blotting and the gamma fractions from the antisera were obtained.
GST–fodrin was prepared as previously described . Brieﬂy,
overnight bacterial culture from clone 18531 expressed in BL21 was
induced with IPTG. The supernatant of the lysate was afﬁnity
absorbed on glutathione–Sepharose beads at 4 C and washed with
1 mM DTT. The bound peptide was eluted during 10 min at room
temperature in the same buffer containing PBS, 1 mM DTT and
10 mM reduced glutathione.
4.3. Culture and tissue culture assays
HEp-2 cells were propagated in humidiﬁed 5% CO2–95% air at
37 C in Dulbecco’s modiﬁed Eagle’s medium (DMEM) supple-
mented with 5% fetal bovine serum (Hyclone), 1% non-essential
amino acids, 5 mM L-glutamine, penicillin (100 units/ml), and
streptomycin (100 mg/ml). For experimental use, subconﬂuent HEp-
2 cells were resuspended with EDTA–trypsin, plated into eight-well
LabTek slides (VWR), and allowed to grow to 70% conﬂuence. The
HT29-C1 clone, obtained from Daniel Louvard (Institut Pasteur,
Paris, France), was grown in DMEM supplemented with 10% fetal
bovine serum, 44 mM sodium bicarbonate, 10 mg of human trans-
ferrin (Sigma, St. Louis, Mo.), 50 IU of streptomycin, and 50 mg/ml of
penicillin. HT29-C1 cells were grown at 37 C in humidiﬁed 10%
CO2–90% air; medium was changed 6 days/week. For experimental
use, HT29-C1 cells were prepared in eight-well LabTek slides and
grown to 70% conﬂuence (about 3 days).
For all experiments, Pic was diluted directly into tissue culture
medium (without antibiotics or serum) and added to the target
cells at a ﬁnal volume of 250 ml per well (eight-well LabTek slides).
Following the speciﬁed incubation times in humidiﬁed atmosphere
of 5% CO2–95% air at 37 C, the medium was aspirated, cells were
washed twice with PBS and processed by means of the following
methods: (i) Giemsa staining. Cells were ﬁxed with 70% methanol
and stained with 10% Giemsa (Sigma). Slides were analyzed at
a 100Â magniﬁcation by standard bright ﬁeld light microscopy .
(ii) Immunostaining. Cells were ﬁxed with 2% formalin–PBS,
washed, permeabilized by adding 0.1% Triton X-100 in PBS, and
stained with 0.05 mg/ml of tetramethyl rhodamine isothiocyanate
(TRITC) phalloidin and with a rabbit anti-Pic polyclonal antibody as
previously described , followed by an anti-rabbit ﬂuorescein-
labeled antibody. Slides were mounted on Gelvatol, covered with
glass coverslips, and examined under a Leica TCS SP2 confocal
J. Gutie´rrez-Jime´nez et al. / Microbial Pathogenesis 45 (2008) 115–123 121
4.4. Overlay assay
Overlay assays were performed as previously reported .
Brieﬂy, about 2 mg of protein (Pet, Pic or EspC) were separated by
SDS-PAGE and transferred to a nitrocellulose membrane. Individual
BSA-blocked strips were then incubated for 1 h with 5 mg/ml
GST–fodrin. Strips were incubated with anti-GST mouse polyclonal
antiserum (1:1000) for 1 h, followed of a goat anti-mouse AP-
conjugated antibody (1:500) for 1 h. Binding was detected using
1-Step NBT/BCIP substrate.
4.5. Degradation assay
Two micrograms of GST–fodrin, or lactoferrin, lysozyme and
secretory IgA (Sigma) were mixed with an equal volume of 2Â
digestion buffer containing Pic or Pet. Reactions were carried out at
30 C at several times and stopped by the addition of 4 Â SDS
sample buffer. All samples were analyzed by SDS-PAGE; the gels
were stained with Coomassie blue .
4.6. Site-directed mutagenesis
Site-directed mutagenesis was performed using the Quik-
ChangeÔ Kit from Stratagene as speciﬁed by the manufacturer. The
template used for construction of the site-directed mutant
(pPicS258I) was pPic1, as we had previously done to produce Pet
S260I . The primers used for this purpose were
50-GAGCCCCTGGGGATATTGGTTCTCCTTTGT-30 and 50-ACAAAGGA-
GAACCAATATCCCCCAGGGGCTC-30. The primers encompassed res-
idues 1743–1772 of the pic sequence (AF097644) but encoded a T
instead of a G at nucleotide 1758, thereby substituting an isoleucine
for the serine at residue 258. Following mutagenesis, the S258I
mutation was conﬁrmed by double-strand sequencing of the area
encompassing the serine protease active site.
4.7. Detection and quantiﬁcation of mucinolytic activity
The presence of mucinolytic activity was quantiﬁed in a micro-
plate assay using immobilized biotin-labeled mucin as substrate
. Brieﬂy, microplate wells were coated with biotin-labeled
mucin, and concentrated supernatants containing Pic were added
and incubated for 4 h at 37 C in sodium acetate buffer. The un-
digested labeled mucin was detected with streptavidin–peroxidase.
Enzymatic activity was determined by reporting absorbance read-
ings on a plot obtained with a standard curve of labeled mucin.
Activity was expressed as the percentage of decrease in absorbance
compared with that from control wells devoid of enzyme.
4.8. Pic binding assays
4.8.1. Immobilized mucin and monosaccharide assays
A modiﬁed solid-phase mucin binding assay was performed
with BSM or monosaccharides (GlcNAc, GalNAc, and NANA from
Sigma) as previously described . Brieﬂy, the bovine mucin (in
HEPES) and 50 mM monosaccharides (in 1 M NaOH) were added to
microtiter plate wells. Pic was added to the wells at various con-
centrations. Reaction was detected by using a polyclonal anti-Pic
antibody (1:700), followed by incubation with a secondary anti-
body (HRP-coupled anti-rabbit). The color was developed with
o-phenylene-diamine, the reaction was stopped with 2 N H2SO4,
and the color was read at 490 nm.
4.8.2. Inhibition of Pic binding to mucin or monosaccharides
Three monosaccharide constituents of the oligosaccharide
side chains of mucin, N-acetylglucosamine (GlcNAc), N-
acetylgalactosamine (GalNAc) and N-acetylneuraminic acid (sialic
acid, NANA), were used in the solid-phase assay. Microtiter plate
wells were coated with bovine mucin as described above. Solutions
of GlcNAc and GalNAc (100 mM) in PBS or sialic acid (50 mM) in
10 mM Tris–HCl were mixed with Pic (2.8 nM). Pic in PBS served as
the control. Samples from Pic-sugar suspension were added to
mucin-coated wells. The mucin adherence assay was continued as
described above. To determine if mucin was able to inhibit Pic
binding to monosaccharide, Pic was preincubated with mucin and
the wells were coated with the monosaccharides.
We thank Ken Teter for the invaluable critical review of the
manuscript and Lucia Chavez-Duen˜as for technical help. This work
was supported by grants from Consejo Nacional de Ciencia y
Tecnologı´a de Me´xico (CONACYT, 30004M and C02-44660) to FNG.
JGJ was supported by a scholarship from Promep (SEP).
 Kaper JB, Nataro JP, Mobley HL. Pathogenic Escherichia coli. Nat Rev Microbiol
 Tzipori S, Montanaro J, Robins-Browne RM, Vial P, Gibson R, Levine MM.
Studies with enteroaggregative Escherichia coli in the gnotobiotic piglet
gastroenteritis model. Infect Immun 1992;60:5302–6.
 Navarro-Garcia F, Sears C, Eslava C, Cravioto A, Nataro JP. Cytoskeletal effects
induced by Pet, the serine protease enterotoxin of enteroaggregative
Escherichia coli. Infect Immun 1999;67:2184–92.
 Bennish ML. Potentially lethal complications of shigellosis. Rev Infect Dis 1991;
 Al-Hasani K, Rajakumar K, Bulach D, Robins-Browne R, Adler B, Sakellaris H.
Genetic organization of the she pathogenicity island in Shigella ﬂexneri 2a.
Microb Pathog 2001;30:1–8.
 Parham NJ, Srinivasan U, Desvaux M, Foxman B, Marrs CF, Henderson IR. PicU,
a second serine protease autotransporter of uropathogenic Escherichia coli.
FEMS Microbiol Lett 2004;230:73–83.
 Henderson IR, Czeczulin J, Eslava C, Noriega F, Nataro JP. Characterization of
pic, a secreted protease of Shigella ﬂexneri and enteroaggregative Escherichia
coli. Infect Immun 1999;67:5587–96.
 Dutta PR, Cappello R, Navarro-Garcia F, Nataro JP. Functional comparison of
serine protease autotransporters of Enterobacteriaceae. Infect Immun 2002;70:
 Henderson IR, Navarro-Garcia F, Desvaux M, Fernandez RC, Ala’Aldeen D. Type
V protein pathway: the autotransporter story. Microbiol Mol Biol Rev 2004;68:
 Navarro-Garcia F, Eslava C, Villaseca JM, Lopez-Revilla R, Czeczulin JR,
Srinivas S, et al. In vitro effects of a high-molecular-weight heat-labile
enterotoxin from enteroaggregative Escherichia coli. Infect Immun 1998;66:
 Heimer SR, Rasko DA, Lockatell CV, Johnson DE, Mobley HL. Autotransporter
genes pic and tsh are associated with Escherichia coli strains that cause acute
pyelonephritis and are expressed during urinary tract infection. Infect Immun
 Navarro-Garcia F, Canizalez-Roman A, Sui BQ, Nataro JP, Azamar Y. The serine
protease motif of EspC from enteropathogenic Escherichia coli produces
epithelial damage by a mechanism different from that of Pet toxin from
enteroaggregative E. coli. Infect Immun 2004;72:3609–21.
 Navarro-Garcia F, Canizalez-Roman A, Luna J, Sears C, Nataro JP. Plasmid-
encoded toxin of enteroaggregative Escherichia coli is internalized by epithelial
cells. Infect Immun 2001;69:1053–60.
 Stabach PR, Cianci CD, Glantz SB, Zhang Z, Morrow JS. Site-directed
mutagenesis of alpha II spectrin at codon 1175 modulates its mu-calpain
susceptibility. Biochemistry 1997;36:57–65.
 Canizalez-Roman A, Navarro-Garcia F. Fodrin CaM-binding domain cleavage
by Pet from enteroaggregative Escherichia coli leads to actin cytoskeletal
disruption. Mol Microbiol 2003;48:947–58.
 Colina AR, Aumont F, Deslauriers N, Belhumeur P, de Repentigny L. Evidence
for degradation of gastrointestinal mucin by Candida albicans secretory
aspartyl proteinase. Infect Immun 1996;64:4514–9.
 Koomey JM, Gill RE, Falkow S. Genetic and biochemical analysis of gonococcal
IgA1 protease: cloning in Escherichia coli and construction of mutants of
gonococci that fail to produce the activity. Proc Natl Acad Sci U S A 1982;79:
 Ryan PA, Pancholi V, Fischetti VA. Group A streptococci bind to mucin and
human pharyngeal cells through sialic acid-containing receptors. Infect
 Jourdian GW, Dean L, Roseman S. The sialic acids. XI. A periodate–resorcinol
method for the quantitative estimation of free sialic acids and their glycosides.
J Biol Chem 1971;246:430–5.
J. Gutie´rrez-Jime´nez et al. / Microbial Pathogenesis 45 (2008) 115–123122
 Moon JY, Park JH, Kim YB. Molecular epidemiological characteristics of virulence
factors on enteroaggregative E. coli. FEMS Microbiol Lett 2005;253:215–20.
 Piva IC, Pereira AL, Ferraz LR, Silva RS, Vieira AC, Blanco JE, et al. Virulence
markers of enteroaggregative Escherichia coli isolated from children and adults
with diarrhea in Brasilia, Brazil. J Clin Microbiol 2003;41:1827–32.
 Jose J, Jahnig F, Meyer TF. Common structural features of IgA1 protease-like
outer membrane protein autotransporters. Mol Microbiol 1995;18:378–80.
 Orsi N. The antimicrobial activity of lactoferrin: current status and
perspectives. Biometals 2004;17:189–96.
 Plaut AG, Qiu J, St Geme 3rd JW. Human lactoferrin proteolytic activity:
analysis of the cleaved region in the IgA protease of Haemophilus inﬂuenzae.
Vaccine 2000;19(Suppl. 1):S148–52.
 Rose JE, Meyer DH, Fives-Taylor PM. Aae, an autotransporter involved in
adhesion of Actinobacillus actinomycetemcomitans to epithelial cells. Infect
 Chan DI, Prenner EJ, Vogel HJ. Tryptophan- and arginine-rich antimicrobial
peptides: structures and mechanisms of action. Biochim Biophys Acta 2006;
 Stathopoulos C, Provence DL, Curtiss 3rd R. Characterization of the avian
pathogenic Escherichia coli hemagglutinin Tsh, a member of the
immunoglobulin A protease-type family of autotransporters. Infect Immun
 Leyton DL, Sloan J, Hill RE, Doughty S, Hartland EL. Transfer region of pO113
from enterohemorrhagic Escherichia coli: similarity with R64 and
identiﬁcation of a novel plasmid-encoded autotransporter, EpeA. Infect
 Lehker MW, Sweeney D. Trichomonad invasion of the mucous layer requires
adhesins, mucinases, and motility. Sex Transm Infect 1999;75:231–8.
 Mantle M, Husar SD. Binding of Yersinia enterocolitica to puriﬁed, native small
intestinal mucins from rabbits and humans involves interactions with the
mucin carbohydrate moiety. Infect Immun 1994;62:1219–27.
 Hase CC, Finkelstein RA. Cloning and nucleotide sequence of the Vibrio
cholerae hemagglutinin/protease (HA/protease) gene and construction of an
HA/protease-negative strain. J Bacteriol 1991;173:3311–7.
 Finkelstein RA, Boesman-Finkelstein M, Holt P. Vibrio cholerae hemagglutinin/
lectin/protease hydrolyzes ﬁbronectin and ovomucin: F.M. Burnet revisited.
Proc Natl Acad Sci U S A 1983;80:1092–5.
 Provence DL, Curtiss 3rd R. Isolation and characterization of a gene involved in
hemagglutination by an avian pathogenic Escherichia coli strain. Infect Immun
J. Gutie´rrez-Jime´nez et al. / Microbial Pathogenesis 45 (2008) 115–123 123